diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,65844 @@
+{
+  "best_global_step": 1600,
+  "best_metric": 0.11357858031988144,
+  "best_model_checkpoint": "saves/lora/gemma-3-1b-it/train_cola_1744902669/checkpoint-1600",
+  "epoch": 83.16008316008316,
+  "eval_steps": 200,
+  "global_step": 40000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.010395010395010396,
+      "grad_norm": 45.610389709472656,
+      "learning_rate": 4.999999876629946e-05,
+      "loss": 6.9949,
+      "num_input_tokens_seen": 3808,
+      "step": 5
+    },
+    {
+      "epoch": 0.02079002079002079,
+      "grad_norm": 8.579319953918457,
+      "learning_rate": 4.999999375439123e-05,
+      "loss": 1.6699,
+      "num_input_tokens_seen": 7744,
+      "step": 10
+    },
+    {
+      "epoch": 0.031185031185031187,
+      "grad_norm": 6.6488823890686035,
+      "learning_rate": 4.9999984887169785e-05,
+      "loss": 0.9866,
+      "num_input_tokens_seen": 11648,
+      "step": 15
+    },
+    {
+      "epoch": 0.04158004158004158,
+      "grad_norm": 15.529796600341797,
+      "learning_rate": 4.9999972164636506e-05,
+      "loss": 0.681,
+      "num_input_tokens_seen": 15648,
+      "step": 20
+    },
+    {
+      "epoch": 0.05197505197505198,
+      "grad_norm": 9.921183586120605,
+      "learning_rate": 4.999995558679334e-05,
+      "loss": 0.252,
+      "num_input_tokens_seen": 19552,
+      "step": 25
+    },
+    {
+      "epoch": 0.062370062370062374,
+      "grad_norm": 2.4148218631744385,
+      "learning_rate": 4.999993515364287e-05,
+      "loss": 0.2042,
+      "num_input_tokens_seen": 23392,
+      "step": 30
+    },
+    {
+      "epoch": 0.07276507276507277,
+      "grad_norm": 3.0714080333709717,
+      "learning_rate": 4.999991086518822e-05,
+      "loss": 0.1813,
+      "num_input_tokens_seen": 27328,
+      "step": 35
+    },
+    {
+      "epoch": 0.08316008316008316,
+      "grad_norm": 1.055045485496521,
+      "learning_rate": 4.999988272143315e-05,
+      "loss": 0.169,
+      "num_input_tokens_seen": 31296,
+      "step": 40
+    },
+    {
+      "epoch": 0.09355509355509356,
+      "grad_norm": 1.0464402437210083,
+      "learning_rate": 4.999985072238199e-05,
+      "loss": 0.1731,
+      "num_input_tokens_seen": 35168,
+      "step": 45
+    },
+    {
+      "epoch": 0.10395010395010396,
+      "grad_norm": 2.519477605819702,
+      "learning_rate": 4.999981486803969e-05,
+      "loss": 0.1956,
+      "num_input_tokens_seen": 39008,
+      "step": 50
+    },
+    {
+      "epoch": 0.11434511434511435,
+      "grad_norm": 9.787467002868652,
+      "learning_rate": 4.999977515841176e-05,
+      "loss": 0.24,
+      "num_input_tokens_seen": 42944,
+      "step": 55
+    },
+    {
+      "epoch": 0.12474012474012475,
+      "grad_norm": 3.136486053466797,
+      "learning_rate": 4.9999731593504344e-05,
+      "loss": 0.174,
+      "num_input_tokens_seen": 46880,
+      "step": 60
+    },
+    {
+      "epoch": 0.13513513513513514,
+      "grad_norm": 4.001842498779297,
+      "learning_rate": 4.999968417332415e-05,
+      "loss": 0.1963,
+      "num_input_tokens_seen": 50752,
+      "step": 65
+    },
+    {
+      "epoch": 0.14553014553014554,
+      "grad_norm": 2.3512396812438965,
+      "learning_rate": 4.999963289787848e-05,
+      "loss": 0.1973,
+      "num_input_tokens_seen": 54592,
+      "step": 70
+    },
+    {
+      "epoch": 0.15592515592515593,
+      "grad_norm": 1.4661264419555664,
+      "learning_rate": 4.999957776717526e-05,
+      "loss": 0.1624,
+      "num_input_tokens_seen": 58624,
+      "step": 75
+    },
+    {
+      "epoch": 0.16632016632016633,
+      "grad_norm": 2.0171024799346924,
+      "learning_rate": 4.9999518781222984e-05,
+      "loss": 0.1541,
+      "num_input_tokens_seen": 62656,
+      "step": 80
+    },
+    {
+      "epoch": 0.17671517671517672,
+      "grad_norm": 5.369823932647705,
+      "learning_rate": 4.9999455940030746e-05,
+      "loss": 0.1912,
+      "num_input_tokens_seen": 66464,
+      "step": 85
+    },
+    {
+      "epoch": 0.18711018711018712,
+      "grad_norm": 4.731435775756836,
+      "learning_rate": 4.999938924360824e-05,
+      "loss": 0.1803,
+      "num_input_tokens_seen": 70304,
+      "step": 90
+    },
+    {
+      "epoch": 0.19750519750519752,
+      "grad_norm": 2.901169538497925,
+      "learning_rate": 4.999931869196575e-05,
+      "loss": 0.1642,
+      "num_input_tokens_seen": 74176,
+      "step": 95
+    },
+    {
+      "epoch": 0.2079002079002079,
+      "grad_norm": 2.0974717140197754,
+      "learning_rate": 4.999924428511416e-05,
+      "loss": 0.1736,
+      "num_input_tokens_seen": 78208,
+      "step": 100
+    },
+    {
+      "epoch": 0.2182952182952183,
+      "grad_norm": 3.789848804473877,
+      "learning_rate": 4.999916602306494e-05,
+      "loss": 0.1743,
+      "num_input_tokens_seen": 82048,
+      "step": 105
+    },
+    {
+      "epoch": 0.2286902286902287,
+      "grad_norm": 0.6832077503204346,
+      "learning_rate": 4.999908390583016e-05,
+      "loss": 0.1563,
+      "num_input_tokens_seen": 86112,
+      "step": 110
+    },
+    {
+      "epoch": 0.2390852390852391,
+      "grad_norm": 6.4458746910095215,
+      "learning_rate": 4.999899793342247e-05,
+      "loss": 0.279,
+      "num_input_tokens_seen": 90016,
+      "step": 115
+    },
+    {
+      "epoch": 0.2494802494802495,
+      "grad_norm": 9.534346580505371,
+      "learning_rate": 4.999890810585516e-05,
+      "loss": 0.2002,
+      "num_input_tokens_seen": 93888,
+      "step": 120
+    },
+    {
+      "epoch": 0.2598752598752599,
+      "grad_norm": 3.762131452560425,
+      "learning_rate": 4.999881442314206e-05,
+      "loss": 0.1738,
+      "num_input_tokens_seen": 97984,
+      "step": 125
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 3.6249701976776123,
+      "learning_rate": 4.9998716885297617e-05,
+      "loss": 0.2456,
+      "num_input_tokens_seen": 101856,
+      "step": 130
+    },
+    {
+      "epoch": 0.2806652806652807,
+      "grad_norm": 2.8140504360198975,
+      "learning_rate": 4.999861549233688e-05,
+      "loss": 0.1876,
+      "num_input_tokens_seen": 105920,
+      "step": 135
+    },
+    {
+      "epoch": 0.2910602910602911,
+      "grad_norm": 0.7389298677444458,
+      "learning_rate": 4.999851024427548e-05,
+      "loss": 0.1907,
+      "num_input_tokens_seen": 109824,
+      "step": 140
+    },
+    {
+      "epoch": 0.30145530145530147,
+      "grad_norm": 1.1445457935333252,
+      "learning_rate": 4.999840114112965e-05,
+      "loss": 0.1432,
+      "num_input_tokens_seen": 113856,
+      "step": 145
+    },
+    {
+      "epoch": 0.31185031185031187,
+      "grad_norm": 2.4285688400268555,
+      "learning_rate": 4.999828818291621e-05,
+      "loss": 0.1255,
+      "num_input_tokens_seen": 117792,
+      "step": 150
+    },
+    {
+      "epoch": 0.32224532224532226,
+      "grad_norm": 2.173661470413208,
+      "learning_rate": 4.999817136965259e-05,
+      "loss": 0.1833,
+      "num_input_tokens_seen": 121856,
+      "step": 155
+    },
+    {
+      "epoch": 0.33264033264033266,
+      "grad_norm": 4.439704895019531,
+      "learning_rate": 4.9998050701356794e-05,
+      "loss": 0.1777,
+      "num_input_tokens_seen": 125856,
+      "step": 160
+    },
+    {
+      "epoch": 0.34303534303534305,
+      "grad_norm": 1.2934468984603882,
+      "learning_rate": 4.999792617804744e-05,
+      "loss": 0.1704,
+      "num_input_tokens_seen": 129728,
+      "step": 165
+    },
+    {
+      "epoch": 0.35343035343035345,
+      "grad_norm": 4.36632776260376,
+      "learning_rate": 4.9997797799743724e-05,
+      "loss": 0.1544,
+      "num_input_tokens_seen": 133536,
+      "step": 170
+    },
+    {
+      "epoch": 0.36382536382536385,
+      "grad_norm": 1.4044604301452637,
+      "learning_rate": 4.999766556646545e-05,
+      "loss": 0.1623,
+      "num_input_tokens_seen": 137504,
+      "step": 175
+    },
+    {
+      "epoch": 0.37422037422037424,
+      "grad_norm": 4.0994367599487305,
+      "learning_rate": 4.9997529478232996e-05,
+      "loss": 0.1324,
+      "num_input_tokens_seen": 141408,
+      "step": 180
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.5019556283950806,
+      "learning_rate": 4.9997389535067365e-05,
+      "loss": 0.1835,
+      "num_input_tokens_seen": 145248,
+      "step": 185
+    },
+    {
+      "epoch": 0.39501039501039503,
+      "grad_norm": 2.546299695968628,
+      "learning_rate": 4.999724573699012e-05,
+      "loss": 0.166,
+      "num_input_tokens_seen": 149120,
+      "step": 190
+    },
+    {
+      "epoch": 0.40540540540540543,
+      "grad_norm": 1.030928134918213,
+      "learning_rate": 4.9997098084023457e-05,
+      "loss": 0.1688,
+      "num_input_tokens_seen": 152864,
+      "step": 195
+    },
+    {
+      "epoch": 0.4158004158004158,
+      "grad_norm": 0.547962486743927,
+      "learning_rate": 4.999694657619013e-05,
+      "loss": 0.1486,
+      "num_input_tokens_seen": 156832,
+      "step": 200
+    },
+    {
+      "epoch": 0.4158004158004158,
+      "eval_loss": 0.13875435292720795,
+      "eval_runtime": 11.7395,
+      "eval_samples_per_second": 72.916,
+      "eval_steps_per_second": 18.229,
+      "num_input_tokens_seen": 156832,
+      "step": 200
+    },
+    {
+      "epoch": 0.4261954261954262,
+      "grad_norm": 5.252303123474121,
+      "learning_rate": 4.999679121351352e-05,
+      "loss": 0.1371,
+      "num_input_tokens_seen": 160832,
+      "step": 205
+    },
+    {
+      "epoch": 0.4365904365904366,
+      "grad_norm": 2.6132638454437256,
+      "learning_rate": 4.9996631996017565e-05,
+      "loss": 0.1328,
+      "num_input_tokens_seen": 164640,
+      "step": 210
+    },
+    {
+      "epoch": 0.446985446985447,
+      "grad_norm": 1.0061838626861572,
+      "learning_rate": 4.9996468923726835e-05,
+      "loss": 0.1539,
+      "num_input_tokens_seen": 168672,
+      "step": 215
+    },
+    {
+      "epoch": 0.4573804573804574,
+      "grad_norm": 2.0787553787231445,
+      "learning_rate": 4.999630199666647e-05,
+      "loss": 0.163,
+      "num_input_tokens_seen": 172736,
+      "step": 220
+    },
+    {
+      "epoch": 0.4677754677754678,
+      "grad_norm": 1.5094966888427734,
+      "learning_rate": 4.999613121486222e-05,
+      "loss": 0.1574,
+      "num_input_tokens_seen": 176832,
+      "step": 225
+    },
+    {
+      "epoch": 0.4781704781704782,
+      "grad_norm": 1.9534474611282349,
+      "learning_rate": 4.999595657834041e-05,
+      "loss": 0.1611,
+      "num_input_tokens_seen": 180704,
+      "step": 230
+    },
+    {
+      "epoch": 0.4885654885654886,
+      "grad_norm": 0.8851607441902161,
+      "learning_rate": 4.999577808712798e-05,
+      "loss": 0.1505,
+      "num_input_tokens_seen": 184512,
+      "step": 235
+    },
+    {
+      "epoch": 0.498960498960499,
+      "grad_norm": 2.314516305923462,
+      "learning_rate": 4.999559574125244e-05,
+      "loss": 0.1625,
+      "num_input_tokens_seen": 188352,
+      "step": 240
+    },
+    {
+      "epoch": 0.5093555093555093,
+      "grad_norm": 0.9527634382247925,
+      "learning_rate": 4.9995409540741934e-05,
+      "loss": 0.1596,
+      "num_input_tokens_seen": 192320,
+      "step": 245
+    },
+    {
+      "epoch": 0.5197505197505198,
+      "grad_norm": 3.8342361450195312,
+      "learning_rate": 4.999521948562516e-05,
+      "loss": 0.1619,
+      "num_input_tokens_seen": 196224,
+      "step": 250
+    },
+    {
+      "epoch": 0.5301455301455301,
+      "grad_norm": 3.920595407485962,
+      "learning_rate": 4.999502557593143e-05,
+      "loss": 0.1635,
+      "num_input_tokens_seen": 200192,
+      "step": 255
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 3.2861549854278564,
+      "learning_rate": 4.999482781169066e-05,
+      "loss": 0.1776,
+      "num_input_tokens_seen": 203904,
+      "step": 260
+    },
+    {
+      "epoch": 0.5509355509355509,
+      "grad_norm": 0.7414162158966064,
+      "learning_rate": 4.9994626192933324e-05,
+      "loss": 0.1656,
+      "num_input_tokens_seen": 207776,
+      "step": 265
+    },
+    {
+      "epoch": 0.5613305613305614,
+      "grad_norm": 4.784208297729492,
+      "learning_rate": 4.999442071969054e-05,
+      "loss": 0.1736,
+      "num_input_tokens_seen": 211808,
+      "step": 270
+    },
+    {
+      "epoch": 0.5717255717255717,
+      "grad_norm": 3.9673733711242676,
+      "learning_rate": 4.999421139199397e-05,
+      "loss": 0.1675,
+      "num_input_tokens_seen": 215712,
+      "step": 275
+    },
+    {
+      "epoch": 0.5821205821205822,
+      "grad_norm": 0.617060124874115,
+      "learning_rate": 4.999399820987592e-05,
+      "loss": 0.1217,
+      "num_input_tokens_seen": 219680,
+      "step": 280
+    },
+    {
+      "epoch": 0.5925155925155925,
+      "grad_norm": 3.072108745574951,
+      "learning_rate": 4.999378117336924e-05,
+      "loss": 0.1478,
+      "num_input_tokens_seen": 223584,
+      "step": 285
+    },
+    {
+      "epoch": 0.6029106029106029,
+      "grad_norm": 3.0659289360046387,
+      "learning_rate": 4.9993560282507415e-05,
+      "loss": 0.1281,
+      "num_input_tokens_seen": 227456,
+      "step": 290
+    },
+    {
+      "epoch": 0.6133056133056133,
+      "grad_norm": 1.9073257446289062,
+      "learning_rate": 4.9993335537324495e-05,
+      "loss": 0.0978,
+      "num_input_tokens_seen": 231296,
+      "step": 295
+    },
+    {
+      "epoch": 0.6237006237006237,
+      "grad_norm": 1.2062827348709106,
+      "learning_rate": 4.999310693785516e-05,
+      "loss": 0.1687,
+      "num_input_tokens_seen": 235040,
+      "step": 300
+    },
+    {
+      "epoch": 0.6340956340956341,
+      "grad_norm": 1.744913935661316,
+      "learning_rate": 4.9992874484134653e-05,
+      "loss": 0.1617,
+      "num_input_tokens_seen": 239136,
+      "step": 305
+    },
+    {
+      "epoch": 0.6444906444906445,
+      "grad_norm": 4.2405524253845215,
+      "learning_rate": 4.999263817619882e-05,
+      "loss": 0.1158,
+      "num_input_tokens_seen": 243008,
+      "step": 310
+    },
+    {
+      "epoch": 0.6548856548856549,
+      "grad_norm": 1.1594067811965942,
+      "learning_rate": 4.9992398014084105e-05,
+      "loss": 0.1574,
+      "num_input_tokens_seen": 247072,
+      "step": 315
+    },
+    {
+      "epoch": 0.6652806652806653,
+      "grad_norm": 6.583249568939209,
+      "learning_rate": 4.999215399782754e-05,
+      "loss": 0.1714,
+      "num_input_tokens_seen": 250880,
+      "step": 320
+    },
+    {
+      "epoch": 0.6756756756756757,
+      "grad_norm": 2.282191753387451,
+      "learning_rate": 4.999190612746675e-05,
+      "loss": 0.1565,
+      "num_input_tokens_seen": 255008,
+      "step": 325
+    },
+    {
+      "epoch": 0.6860706860706861,
+      "grad_norm": 1.101054072380066,
+      "learning_rate": 4.999165440303998e-05,
+      "loss": 0.1544,
+      "num_input_tokens_seen": 258944,
+      "step": 330
+    },
+    {
+      "epoch": 0.6964656964656964,
+      "grad_norm": 1.9154175519943237,
+      "learning_rate": 4.999139882458603e-05,
+      "loss": 0.1571,
+      "num_input_tokens_seen": 262880,
+      "step": 335
+    },
+    {
+      "epoch": 0.7068607068607069,
+      "grad_norm": 2.6875693798065186,
+      "learning_rate": 4.9991139392144314e-05,
+      "loss": 0.1371,
+      "num_input_tokens_seen": 266752,
+      "step": 340
+    },
+    {
+      "epoch": 0.7172557172557172,
+      "grad_norm": 1.0691272020339966,
+      "learning_rate": 4.999087610575485e-05,
+      "loss": 0.1259,
+      "num_input_tokens_seen": 270688,
+      "step": 345
+    },
+    {
+      "epoch": 0.7276507276507277,
+      "grad_norm": 4.2721710205078125,
+      "learning_rate": 4.999060896545824e-05,
+      "loss": 0.1486,
+      "num_input_tokens_seen": 274592,
+      "step": 350
+    },
+    {
+      "epoch": 0.738045738045738,
+      "grad_norm": 0.6534197330474854,
+      "learning_rate": 4.999033797129568e-05,
+      "loss": 0.1607,
+      "num_input_tokens_seen": 278368,
+      "step": 355
+    },
+    {
+      "epoch": 0.7484407484407485,
+      "grad_norm": 1.2525132894515991,
+      "learning_rate": 4.999006312330894e-05,
+      "loss": 0.1547,
+      "num_input_tokens_seen": 282272,
+      "step": 360
+    },
+    {
+      "epoch": 0.7588357588357588,
+      "grad_norm": 2.6068663597106934,
+      "learning_rate": 4.998978442154043e-05,
+      "loss": 0.1328,
+      "num_input_tokens_seen": 286208,
+      "step": 365
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 1.2813936471939087,
+      "learning_rate": 4.9989501866033125e-05,
+      "loss": 0.1566,
+      "num_input_tokens_seen": 289984,
+      "step": 370
+    },
+    {
+      "epoch": 0.7796257796257796,
+      "grad_norm": 1.0099918842315674,
+      "learning_rate": 4.998921545683059e-05,
+      "loss": 0.1376,
+      "num_input_tokens_seen": 293760,
+      "step": 375
+    },
+    {
+      "epoch": 0.7900207900207901,
+      "grad_norm": 1.1336792707443237,
+      "learning_rate": 4.9988925193976996e-05,
+      "loss": 0.1443,
+      "num_input_tokens_seen": 297664,
+      "step": 380
+    },
+    {
+      "epoch": 0.8004158004158004,
+      "grad_norm": 0.8535900712013245,
+      "learning_rate": 4.998863107751711e-05,
+      "loss": 0.1262,
+      "num_input_tokens_seen": 301536,
+      "step": 385
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 2.385103464126587,
+      "learning_rate": 4.998833310749629e-05,
+      "loss": 0.1342,
+      "num_input_tokens_seen": 305440,
+      "step": 390
+    },
+    {
+      "epoch": 0.8212058212058212,
+      "grad_norm": 0.7003737092018127,
+      "learning_rate": 4.998803128396047e-05,
+      "loss": 0.1905,
+      "num_input_tokens_seen": 309344,
+      "step": 395
+    },
+    {
+      "epoch": 0.8316008316008316,
+      "grad_norm": 3.849493980407715,
+      "learning_rate": 4.9987725606956215e-05,
+      "loss": 0.1359,
+      "num_input_tokens_seen": 313248,
+      "step": 400
+    },
+    {
+      "epoch": 0.8316008316008316,
+      "eval_loss": 0.1436040997505188,
+      "eval_runtime": 11.7236,
+      "eval_samples_per_second": 73.015,
+      "eval_steps_per_second": 18.254,
+      "num_input_tokens_seen": 313248,
+      "step": 400
+    },
+    {
+      "epoch": 0.841995841995842,
+      "grad_norm": 0.7200118899345398,
+      "learning_rate": 4.998741607653066e-05,
+      "loss": 0.1541,
+      "num_input_tokens_seen": 317120,
+      "step": 405
+    },
+    {
+      "epoch": 0.8523908523908524,
+      "grad_norm": 3.5762441158294678,
+      "learning_rate": 4.9987102692731523e-05,
+      "loss": 0.1778,
+      "num_input_tokens_seen": 321152,
+      "step": 410
+    },
+    {
+      "epoch": 0.8627858627858628,
+      "grad_norm": 1.3075878620147705,
+      "learning_rate": 4.9986785455607157e-05,
+      "loss": 0.1334,
+      "num_input_tokens_seen": 324992,
+      "step": 415
+    },
+    {
+      "epoch": 0.8731808731808732,
+      "grad_norm": 3.804544448852539,
+      "learning_rate": 4.9986464365206456e-05,
+      "loss": 0.1407,
+      "num_input_tokens_seen": 328928,
+      "step": 420
+    },
+    {
+      "epoch": 0.8835758835758836,
+      "grad_norm": 1.8650304079055786,
+      "learning_rate": 4.9986139421578956e-05,
+      "loss": 0.1343,
+      "num_input_tokens_seen": 332736,
+      "step": 425
+    },
+    {
+      "epoch": 0.893970893970894,
+      "grad_norm": 3.979048013687134,
+      "learning_rate": 4.998581062477477e-05,
+      "loss": 0.1681,
+      "num_input_tokens_seen": 336736,
+      "step": 430
+    },
+    {
+      "epoch": 0.9043659043659044,
+      "grad_norm": 3.6140568256378174,
+      "learning_rate": 4.998547797484458e-05,
+      "loss": 0.1169,
+      "num_input_tokens_seen": 340608,
+      "step": 435
+    },
+    {
+      "epoch": 0.9147609147609148,
+      "grad_norm": 2.1027162075042725,
+      "learning_rate": 4.9985141471839706e-05,
+      "loss": 0.1588,
+      "num_input_tokens_seen": 344480,
+      "step": 440
+    },
+    {
+      "epoch": 0.9251559251559252,
+      "grad_norm": 2.0764429569244385,
+      "learning_rate": 4.998480111581203e-05,
+      "loss": 0.1482,
+      "num_input_tokens_seen": 348384,
+      "step": 445
+    },
+    {
+      "epoch": 0.9355509355509356,
+      "grad_norm": 3.590261459350586,
+      "learning_rate": 4.998445690681405e-05,
+      "loss": 0.1582,
+      "num_input_tokens_seen": 352224,
+      "step": 450
+    },
+    {
+      "epoch": 0.9459459459459459,
+      "grad_norm": 3.2917912006378174,
+      "learning_rate": 4.9984108844898834e-05,
+      "loss": 0.1462,
+      "num_input_tokens_seen": 356320,
+      "step": 455
+    },
+    {
+      "epoch": 0.9563409563409564,
+      "grad_norm": 2.2158613204956055,
+      "learning_rate": 4.9983756930120076e-05,
+      "loss": 0.1199,
+      "num_input_tokens_seen": 360160,
+      "step": 460
+    },
+    {
+      "epoch": 0.9667359667359667,
+      "grad_norm": 1.4454841613769531,
+      "learning_rate": 4.9983401162532025e-05,
+      "loss": 0.1076,
+      "num_input_tokens_seen": 364032,
+      "step": 465
+    },
+    {
+      "epoch": 0.9771309771309772,
+      "grad_norm": 2.1778509616851807,
+      "learning_rate": 4.998304154218955e-05,
+      "loss": 0.1179,
+      "num_input_tokens_seen": 367904,
+      "step": 470
+    },
+    {
+      "epoch": 0.9875259875259875,
+      "grad_norm": 0.8267582654953003,
+      "learning_rate": 4.998267806914812e-05,
+      "loss": 0.1486,
+      "num_input_tokens_seen": 371808,
+      "step": 475
+    },
+    {
+      "epoch": 0.997920997920998,
+      "grad_norm": 4.604156494140625,
+      "learning_rate": 4.998231074346378e-05,
+      "loss": 0.1419,
+      "num_input_tokens_seen": 375776,
+      "step": 480
+    },
+    {
+      "epoch": 1.0083160083160083,
+      "grad_norm": 3.7127249240875244,
+      "learning_rate": 4.998193956519317e-05,
+      "loss": 0.1366,
+      "num_input_tokens_seen": 379760,
+      "step": 485
+    },
+    {
+      "epoch": 1.0187110187110187,
+      "grad_norm": 1.430838942527771,
+      "learning_rate": 4.9981564534393545e-05,
+      "loss": 0.1281,
+      "num_input_tokens_seen": 383632,
+      "step": 490
+    },
+    {
+      "epoch": 1.0291060291060292,
+      "grad_norm": 2.004713773727417,
+      "learning_rate": 4.998118565112272e-05,
+      "loss": 0.1098,
+      "num_input_tokens_seen": 387568,
+      "step": 495
+    },
+    {
+      "epoch": 1.0395010395010396,
+      "grad_norm": 3.3961589336395264,
+      "learning_rate": 4.998080291543914e-05,
+      "loss": 0.1416,
+      "num_input_tokens_seen": 391440,
+      "step": 500
+    },
+    {
+      "epoch": 1.04989604989605,
+      "grad_norm": 1.1622871160507202,
+      "learning_rate": 4.9980416327401826e-05,
+      "loss": 0.153,
+      "num_input_tokens_seen": 395312,
+      "step": 505
+    },
+    {
+      "epoch": 1.0602910602910602,
+      "grad_norm": 3.8734970092773438,
+      "learning_rate": 4.998002588707038e-05,
+      "loss": 0.1317,
+      "num_input_tokens_seen": 399344,
+      "step": 510
+    },
+    {
+      "epoch": 1.0706860706860706,
+      "grad_norm": 1.2433947324752808,
+      "learning_rate": 4.997963159450503e-05,
+      "loss": 0.1554,
+      "num_input_tokens_seen": 403184,
+      "step": 515
+    },
+    {
+      "epoch": 1.0810810810810811,
+      "grad_norm": 4.193972587585449,
+      "learning_rate": 4.9979233449766575e-05,
+      "loss": 0.1385,
+      "num_input_tokens_seen": 407184,
+      "step": 520
+    },
+    {
+      "epoch": 1.0914760914760915,
+      "grad_norm": 1.7979377508163452,
+      "learning_rate": 4.997883145291641e-05,
+      "loss": 0.1299,
+      "num_input_tokens_seen": 410896,
+      "step": 525
+    },
+    {
+      "epoch": 1.1018711018711018,
+      "grad_norm": 1.2090972661972046,
+      "learning_rate": 4.9978425604016536e-05,
+      "loss": 0.1497,
+      "num_input_tokens_seen": 414832,
+      "step": 530
+    },
+    {
+      "epoch": 1.1122661122661124,
+      "grad_norm": 3.0061264038085938,
+      "learning_rate": 4.9978015903129536e-05,
+      "loss": 0.1276,
+      "num_input_tokens_seen": 418736,
+      "step": 535
+    },
+    {
+      "epoch": 1.1226611226611227,
+      "grad_norm": 4.9195170402526855,
+      "learning_rate": 4.997760235031859e-05,
+      "loss": 0.1517,
+      "num_input_tokens_seen": 422608,
+      "step": 540
+    },
+    {
+      "epoch": 1.133056133056133,
+      "grad_norm": 1.9237570762634277,
+      "learning_rate": 4.9977184945647473e-05,
+      "loss": 0.1584,
+      "num_input_tokens_seen": 426448,
+      "step": 545
+    },
+    {
+      "epoch": 1.1434511434511434,
+      "grad_norm": 3.3735451698303223,
+      "learning_rate": 4.997676368918055e-05,
+      "loss": 0.1565,
+      "num_input_tokens_seen": 430416,
+      "step": 550
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.747287929058075,
+      "learning_rate": 4.9976338580982794e-05,
+      "loss": 0.1282,
+      "num_input_tokens_seen": 434384,
+      "step": 555
+    },
+    {
+      "epoch": 1.1642411642411643,
+      "grad_norm": 3.0234885215759277,
+      "learning_rate": 4.9975909621119755e-05,
+      "loss": 0.1298,
+      "num_input_tokens_seen": 438448,
+      "step": 560
+    },
+    {
+      "epoch": 1.1746361746361746,
+      "grad_norm": 1.4938101768493652,
+      "learning_rate": 4.997547680965758e-05,
+      "loss": 0.1268,
+      "num_input_tokens_seen": 442288,
+      "step": 565
+    },
+    {
+      "epoch": 1.185031185031185,
+      "grad_norm": 1.1884517669677734,
+      "learning_rate": 4.997504014666302e-05,
+      "loss": 0.1538,
+      "num_input_tokens_seen": 446192,
+      "step": 570
+    },
+    {
+      "epoch": 1.1954261954261955,
+      "grad_norm": 1.5035388469696045,
+      "learning_rate": 4.997459963220342e-05,
+      "loss": 0.1618,
+      "num_input_tokens_seen": 450064,
+      "step": 575
+    },
+    {
+      "epoch": 1.2058212058212059,
+      "grad_norm": 4.883091449737549,
+      "learning_rate": 4.997415526634671e-05,
+      "loss": 0.1614,
+      "num_input_tokens_seen": 454032,
+      "step": 580
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 1.4672636985778809,
+      "learning_rate": 4.99737070491614e-05,
+      "loss": 0.1269,
+      "num_input_tokens_seen": 457872,
+      "step": 585
+    },
+    {
+      "epoch": 1.2266112266112266,
+      "grad_norm": 3.8486790657043457,
+      "learning_rate": 4.997325498071663e-05,
+      "loss": 0.1187,
+      "num_input_tokens_seen": 461712,
+      "step": 590
+    },
+    {
+      "epoch": 1.237006237006237,
+      "grad_norm": 2.066298484802246,
+      "learning_rate": 4.997279906108211e-05,
+      "loss": 0.1404,
+      "num_input_tokens_seen": 465616,
+      "step": 595
+    },
+    {
+      "epoch": 1.2474012474012475,
+      "grad_norm": 2.1758906841278076,
+      "learning_rate": 4.9972339290328155e-05,
+      "loss": 0.1519,
+      "num_input_tokens_seen": 469520,
+      "step": 600
+    },
+    {
+      "epoch": 1.2474012474012475,
+      "eval_loss": 0.13767853379249573,
+      "eval_runtime": 11.7508,
+      "eval_samples_per_second": 72.846,
+      "eval_steps_per_second": 18.212,
+      "num_input_tokens_seen": 469520,
+      "step": 600
+    },
+    {
+      "epoch": 1.2577962577962578,
+      "grad_norm": 0.9251719117164612,
+      "learning_rate": 4.9971875668525646e-05,
+      "loss": 0.1021,
+      "num_input_tokens_seen": 473264,
+      "step": 605
+    },
+    {
+      "epoch": 1.2681912681912682,
+      "grad_norm": 2.634211301803589,
+      "learning_rate": 4.997140819574609e-05,
+      "loss": 0.1254,
+      "num_input_tokens_seen": 477072,
+      "step": 610
+    },
+    {
+      "epoch": 1.2785862785862787,
+      "grad_norm": 1.127885341644287,
+      "learning_rate": 4.997093687206159e-05,
+      "loss": 0.1028,
+      "num_input_tokens_seen": 481104,
+      "step": 615
+    },
+    {
+      "epoch": 1.288981288981289,
+      "grad_norm": 1.2469841241836548,
+      "learning_rate": 4.997046169754482e-05,
+      "loss": 0.1068,
+      "num_input_tokens_seen": 484944,
+      "step": 620
+    },
+    {
+      "epoch": 1.2993762993762994,
+      "grad_norm": 1.1840534210205078,
+      "learning_rate": 4.996998267226905e-05,
+      "loss": 0.1325,
+      "num_input_tokens_seen": 488912,
+      "step": 625
+    },
+    {
+      "epoch": 1.3097713097713097,
+      "grad_norm": 3.291097640991211,
+      "learning_rate": 4.996949979630817e-05,
+      "loss": 0.1872,
+      "num_input_tokens_seen": 492720,
+      "step": 630
+    },
+    {
+      "epoch": 1.32016632016632,
+      "grad_norm": 1.6606706380844116,
+      "learning_rate": 4.996901306973663e-05,
+      "loss": 0.153,
+      "num_input_tokens_seen": 496848,
+      "step": 635
+    },
+    {
+      "epoch": 1.3305613305613306,
+      "grad_norm": 1.9148038625717163,
+      "learning_rate": 4.996852249262949e-05,
+      "loss": 0.1627,
+      "num_input_tokens_seen": 500752,
+      "step": 640
+    },
+    {
+      "epoch": 1.340956340956341,
+      "grad_norm": 1.4481282234191895,
+      "learning_rate": 4.996802806506241e-05,
+      "loss": 0.1378,
+      "num_input_tokens_seen": 504624,
+      "step": 645
+    },
+    {
+      "epoch": 1.3513513513513513,
+      "grad_norm": 1.5727436542510986,
+      "learning_rate": 4.996752978711164e-05,
+      "loss": 0.1367,
+      "num_input_tokens_seen": 508624,
+      "step": 650
+    },
+    {
+      "epoch": 1.3617463617463619,
+      "grad_norm": 0.8288475871086121,
+      "learning_rate": 4.996702765885401e-05,
+      "loss": 0.1202,
+      "num_input_tokens_seen": 512432,
+      "step": 655
+    },
+    {
+      "epoch": 1.3721413721413722,
+      "grad_norm": 4.659414291381836,
+      "learning_rate": 4.9966521680366964e-05,
+      "loss": 0.1357,
+      "num_input_tokens_seen": 516176,
+      "step": 660
+    },
+    {
+      "epoch": 1.3825363825363826,
+      "grad_norm": 2.0323102474212646,
+      "learning_rate": 4.9966011851728524e-05,
+      "loss": 0.1505,
+      "num_input_tokens_seen": 520080,
+      "step": 665
+    },
+    {
+      "epoch": 1.392931392931393,
+      "grad_norm": 3.893711566925049,
+      "learning_rate": 4.996549817301731e-05,
+      "loss": 0.1966,
+      "num_input_tokens_seen": 523888,
+      "step": 670
+    },
+    {
+      "epoch": 1.4033264033264032,
+      "grad_norm": 2.2676761150360107,
+      "learning_rate": 4.9964980644312544e-05,
+      "loss": 0.1212,
+      "num_input_tokens_seen": 527728,
+      "step": 675
+    },
+    {
+      "epoch": 1.4137214137214138,
+      "grad_norm": 2.547551155090332,
+      "learning_rate": 4.996445926569403e-05,
+      "loss": 0.1034,
+      "num_input_tokens_seen": 531600,
+      "step": 680
+    },
+    {
+      "epoch": 1.4241164241164241,
+      "grad_norm": 2.3840415477752686,
+      "learning_rate": 4.996393403724218e-05,
+      "loss": 0.1296,
+      "num_input_tokens_seen": 535408,
+      "step": 685
+    },
+    {
+      "epoch": 1.4345114345114345,
+      "grad_norm": 2.9187703132629395,
+      "learning_rate": 4.9963404959037985e-05,
+      "loss": 0.1071,
+      "num_input_tokens_seen": 539216,
+      "step": 690
+    },
+    {
+      "epoch": 1.444906444906445,
+      "grad_norm": 4.831179141998291,
+      "learning_rate": 4.996287203116303e-05,
+      "loss": 0.1368,
+      "num_input_tokens_seen": 543248,
+      "step": 695
+    },
+    {
+      "epoch": 1.4553014553014554,
+      "grad_norm": 2.326526641845703,
+      "learning_rate": 4.996233525369951e-05,
+      "loss": 0.1253,
+      "num_input_tokens_seen": 547184,
+      "step": 700
+    },
+    {
+      "epoch": 1.4656964656964657,
+      "grad_norm": 1.006280541419983,
+      "learning_rate": 4.99617946267302e-05,
+      "loss": 0.152,
+      "num_input_tokens_seen": 551248,
+      "step": 705
+    },
+    {
+      "epoch": 1.476091476091476,
+      "grad_norm": 0.7572613954544067,
+      "learning_rate": 4.996125015033846e-05,
+      "loss": 0.1067,
+      "num_input_tokens_seen": 555088,
+      "step": 710
+    },
+    {
+      "epoch": 1.4864864864864864,
+      "grad_norm": 2.819971799850464,
+      "learning_rate": 4.996070182460827e-05,
+      "loss": 0.1384,
+      "num_input_tokens_seen": 558992,
+      "step": 715
+    },
+    {
+      "epoch": 1.496881496881497,
+      "grad_norm": 0.9361287951469421,
+      "learning_rate": 4.996014964962418e-05,
+      "loss": 0.1187,
+      "num_input_tokens_seen": 562896,
+      "step": 720
+    },
+    {
+      "epoch": 1.5072765072765073,
+      "grad_norm": 1.1760891675949097,
+      "learning_rate": 4.9959593625471344e-05,
+      "loss": 0.1467,
+      "num_input_tokens_seen": 566864,
+      "step": 725
+    },
+    {
+      "epoch": 1.5176715176715176,
+      "grad_norm": 1.9068094491958618,
+      "learning_rate": 4.995903375223552e-05,
+      "loss": 0.1293,
+      "num_input_tokens_seen": 570832,
+      "step": 730
+    },
+    {
+      "epoch": 1.5280665280665282,
+      "grad_norm": 2.162119150161743,
+      "learning_rate": 4.995847003000302e-05,
+      "loss": 0.1443,
+      "num_input_tokens_seen": 574704,
+      "step": 735
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.81267911195755,
+      "learning_rate": 4.9957902458860804e-05,
+      "loss": 0.1641,
+      "num_input_tokens_seen": 578672,
+      "step": 740
+    },
+    {
+      "epoch": 1.5488565488565489,
+      "grad_norm": 1.3887544870376587,
+      "learning_rate": 4.995733103889639e-05,
+      "loss": 0.1293,
+      "num_input_tokens_seen": 582512,
+      "step": 745
+    },
+    {
+      "epoch": 1.5592515592515592,
+      "grad_norm": 2.874913454055786,
+      "learning_rate": 4.99567557701979e-05,
+      "loss": 0.1559,
+      "num_input_tokens_seen": 586480,
+      "step": 750
+    },
+    {
+      "epoch": 1.5696465696465696,
+      "grad_norm": 1.4742555618286133,
+      "learning_rate": 4.995617665285403e-05,
+      "loss": 0.1235,
+      "num_input_tokens_seen": 590448,
+      "step": 755
+    },
+    {
+      "epoch": 1.5800415800415801,
+      "grad_norm": 1.0371143817901611,
+      "learning_rate": 4.99555936869541e-05,
+      "loss": 0.1501,
+      "num_input_tokens_seen": 594352,
+      "step": 760
+    },
+    {
+      "epoch": 1.5904365904365905,
+      "grad_norm": 0.8540729880332947,
+      "learning_rate": 4.995500687258803e-05,
+      "loss": 0.1427,
+      "num_input_tokens_seen": 598192,
+      "step": 765
+    },
+    {
+      "epoch": 1.6008316008316008,
+      "grad_norm": 1.2054797410964966,
+      "learning_rate": 4.995441620984628e-05,
+      "loss": 0.1509,
+      "num_input_tokens_seen": 602000,
+      "step": 770
+    },
+    {
+      "epoch": 1.6112266112266114,
+      "grad_norm": 3.248786449432373,
+      "learning_rate": 4.995382169881996e-05,
+      "loss": 0.1275,
+      "num_input_tokens_seen": 605968,
+      "step": 775
+    },
+    {
+      "epoch": 1.6216216216216215,
+      "grad_norm": 2.303069591522217,
+      "learning_rate": 4.9953223339600755e-05,
+      "loss": 0.133,
+      "num_input_tokens_seen": 609968,
+      "step": 780
+    },
+    {
+      "epoch": 1.632016632016632,
+      "grad_norm": 1.5103806257247925,
+      "learning_rate": 4.995262113228091e-05,
+      "loss": 0.0977,
+      "num_input_tokens_seen": 613840,
+      "step": 785
+    },
+    {
+      "epoch": 1.6424116424116424,
+      "grad_norm": 1.1814939975738525,
+      "learning_rate": 4.995201507695332e-05,
+      "loss": 0.1057,
+      "num_input_tokens_seen": 617584,
+      "step": 790
+    },
+    {
+      "epoch": 1.6528066528066527,
+      "grad_norm": 1.565474510192871,
+      "learning_rate": 4.995140517371144e-05,
+      "loss": 0.1079,
+      "num_input_tokens_seen": 621488,
+      "step": 795
+    },
+    {
+      "epoch": 1.6632016632016633,
+      "grad_norm": 2.3072304725646973,
+      "learning_rate": 4.995079142264932e-05,
+      "loss": 0.1529,
+      "num_input_tokens_seen": 625360,
+      "step": 800
+    },
+    {
+      "epoch": 1.6632016632016633,
+      "eval_loss": 0.12164340168237686,
+      "eval_runtime": 11.7477,
+      "eval_samples_per_second": 72.866,
+      "eval_steps_per_second": 18.216,
+      "num_input_tokens_seen": 625360,
+      "step": 800
+    },
+    {
+      "epoch": 1.6735966735966736,
+      "grad_norm": 1.0892314910888672,
+      "learning_rate": 4.995017382386162e-05,
+      "loss": 0.0882,
+      "num_input_tokens_seen": 629296,
+      "step": 805
+    },
+    {
+      "epoch": 1.683991683991684,
+      "grad_norm": 1.4066522121429443,
+      "learning_rate": 4.994955237744356e-05,
+      "loss": 0.0971,
+      "num_input_tokens_seen": 633232,
+      "step": 810
+    },
+    {
+      "epoch": 1.6943866943866945,
+      "grad_norm": 1.5949170589447021,
+      "learning_rate": 4.994892708349101e-05,
+      "loss": 0.15,
+      "num_input_tokens_seen": 637136,
+      "step": 815
+    },
+    {
+      "epoch": 1.7047817047817047,
+      "grad_norm": 2.1417219638824463,
+      "learning_rate": 4.994829794210035e-05,
+      "loss": 0.099,
+      "num_input_tokens_seen": 641072,
+      "step": 820
+    },
+    {
+      "epoch": 1.7151767151767152,
+      "grad_norm": 3.5216476917266846,
+      "learning_rate": 4.994766495336864e-05,
+      "loss": 0.1572,
+      "num_input_tokens_seen": 644944,
+      "step": 825
+    },
+    {
+      "epoch": 1.7255717255717256,
+      "grad_norm": 1.4649467468261719,
+      "learning_rate": 4.994702811739348e-05,
+      "loss": 0.1093,
+      "num_input_tokens_seen": 648816,
+      "step": 830
+    },
+    {
+      "epoch": 1.735966735966736,
+      "grad_norm": 1.3467360734939575,
+      "learning_rate": 4.994638743427308e-05,
+      "loss": 0.111,
+      "num_input_tokens_seen": 652784,
+      "step": 835
+    },
+    {
+      "epoch": 1.7463617463617465,
+      "grad_norm": 0.8305129408836365,
+      "learning_rate": 4.994574290410624e-05,
+      "loss": 0.103,
+      "num_input_tokens_seen": 656592,
+      "step": 840
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 1.0096068382263184,
+      "learning_rate": 4.9945094526992364e-05,
+      "loss": 0.1025,
+      "num_input_tokens_seen": 660432,
+      "step": 845
+    },
+    {
+      "epoch": 1.7671517671517671,
+      "grad_norm": 3.591475486755371,
+      "learning_rate": 4.994444230303142e-05,
+      "loss": 0.1101,
+      "num_input_tokens_seen": 664272,
+      "step": 850
+    },
+    {
+      "epoch": 1.7775467775467777,
+      "grad_norm": 3.790574073791504,
+      "learning_rate": 4.994378623232402e-05,
+      "loss": 0.1497,
+      "num_input_tokens_seen": 668016,
+      "step": 855
+    },
+    {
+      "epoch": 1.7879417879417878,
+      "grad_norm": 2.151954174041748,
+      "learning_rate": 4.99431263149713e-05,
+      "loss": 0.1539,
+      "num_input_tokens_seen": 671952,
+      "step": 860
+    },
+    {
+      "epoch": 1.7983367983367984,
+      "grad_norm": 2.990436553955078,
+      "learning_rate": 4.9942462551075056e-05,
+      "loss": 0.1765,
+      "num_input_tokens_seen": 675824,
+      "step": 865
+    },
+    {
+      "epoch": 1.8087318087318087,
+      "grad_norm": 2.7082676887512207,
+      "learning_rate": 4.994179494073764e-05,
+      "loss": 0.1363,
+      "num_input_tokens_seen": 679728,
+      "step": 870
+    },
+    {
+      "epoch": 1.819126819126819,
+      "grad_norm": 0.878171980381012,
+      "learning_rate": 4.9941123484062e-05,
+      "loss": 0.1026,
+      "num_input_tokens_seen": 683664,
+      "step": 875
+    },
+    {
+      "epoch": 1.8295218295218296,
+      "grad_norm": 0.8126267790794373,
+      "learning_rate": 4.99404481811517e-05,
+      "loss": 0.137,
+      "num_input_tokens_seen": 687760,
+      "step": 880
+    },
+    {
+      "epoch": 1.83991683991684,
+      "grad_norm": 2.768674850463867,
+      "learning_rate": 4.9939769032110864e-05,
+      "loss": 0.1317,
+      "num_input_tokens_seen": 691760,
+      "step": 885
+    },
+    {
+      "epoch": 1.8503118503118503,
+      "grad_norm": 0.9904656410217285,
+      "learning_rate": 4.993908603704423e-05,
+      "loss": 0.114,
+      "num_input_tokens_seen": 695504,
+      "step": 890
+    },
+    {
+      "epoch": 1.8607068607068609,
+      "grad_norm": 3.476804733276367,
+      "learning_rate": 4.9938399196057126e-05,
+      "loss": 0.1366,
+      "num_input_tokens_seen": 699440,
+      "step": 895
+    },
+    {
+      "epoch": 1.871101871101871,
+      "grad_norm": 1.2326109409332275,
+      "learning_rate": 4.993770850925547e-05,
+      "loss": 0.1258,
+      "num_input_tokens_seen": 703408,
+      "step": 900
+    },
+    {
+      "epoch": 1.8814968814968815,
+      "grad_norm": 0.9935004711151123,
+      "learning_rate": 4.993701397674577e-05,
+      "loss": 0.1374,
+      "num_input_tokens_seen": 707216,
+      "step": 905
+    },
+    {
+      "epoch": 1.8918918918918919,
+      "grad_norm": 0.4536069929599762,
+      "learning_rate": 4.993631559863515e-05,
+      "loss": 0.0987,
+      "num_input_tokens_seen": 711216,
+      "step": 910
+    },
+    {
+      "epoch": 1.9022869022869022,
+      "grad_norm": 1.0992226600646973,
+      "learning_rate": 4.9935613375031283e-05,
+      "loss": 0.1563,
+      "num_input_tokens_seen": 715088,
+      "step": 915
+    },
+    {
+      "epoch": 1.9126819126819128,
+      "grad_norm": 2.7366180419921875,
+      "learning_rate": 4.993490730604248e-05,
+      "loss": 0.1237,
+      "num_input_tokens_seen": 719248,
+      "step": 920
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 1.9953835010528564,
+      "learning_rate": 4.993419739177761e-05,
+      "loss": 0.1262,
+      "num_input_tokens_seen": 723216,
+      "step": 925
+    },
+    {
+      "epoch": 1.9334719334719335,
+      "grad_norm": 2.5190019607543945,
+      "learning_rate": 4.9933483632346164e-05,
+      "loss": 0.1339,
+      "num_input_tokens_seen": 727120,
+      "step": 930
+    },
+    {
+      "epoch": 1.943866943866944,
+      "grad_norm": 2.8206253051757812,
+      "learning_rate": 4.993276602785821e-05,
+      "loss": 0.1335,
+      "num_input_tokens_seen": 731184,
+      "step": 935
+    },
+    {
+      "epoch": 1.9542619542619541,
+      "grad_norm": 2.7601120471954346,
+      "learning_rate": 4.993204457842441e-05,
+      "loss": 0.1189,
+      "num_input_tokens_seen": 735152,
+      "step": 940
+    },
+    {
+      "epoch": 1.9646569646569647,
+      "grad_norm": 2.222433090209961,
+      "learning_rate": 4.993131928415602e-05,
+      "loss": 0.1294,
+      "num_input_tokens_seen": 738992,
+      "step": 945
+    },
+    {
+      "epoch": 1.975051975051975,
+      "grad_norm": 3.0745630264282227,
+      "learning_rate": 4.993059014516489e-05,
+      "loss": 0.1218,
+      "num_input_tokens_seen": 742864,
+      "step": 950
+    },
+    {
+      "epoch": 1.9854469854469854,
+      "grad_norm": 2.4255318641662598,
+      "learning_rate": 4.9929857161563464e-05,
+      "loss": 0.1479,
+      "num_input_tokens_seen": 746928,
+      "step": 955
+    },
+    {
+      "epoch": 1.995841995841996,
+      "grad_norm": 2.229947090148926,
+      "learning_rate": 4.992912033346477e-05,
+      "loss": 0.0944,
+      "num_input_tokens_seen": 750864,
+      "step": 960
+    },
+    {
+      "epoch": 2.006237006237006,
+      "grad_norm": 1.2301019430160522,
+      "learning_rate": 4.992837966098245e-05,
+      "loss": 0.1204,
+      "num_input_tokens_seen": 754624,
+      "step": 965
+    },
+    {
+      "epoch": 2.0166320166320166,
+      "grad_norm": 1.3781042098999023,
+      "learning_rate": 4.992763514423071e-05,
+      "loss": 0.1033,
+      "num_input_tokens_seen": 758560,
+      "step": 970
+    },
+    {
+      "epoch": 2.027027027027027,
+      "grad_norm": 1.373133897781372,
+      "learning_rate": 4.992688678332437e-05,
+      "loss": 0.1395,
+      "num_input_tokens_seen": 762560,
+      "step": 975
+    },
+    {
+      "epoch": 2.0374220374220373,
+      "grad_norm": 1.2036548852920532,
+      "learning_rate": 4.992613457837884e-05,
+      "loss": 0.0719,
+      "num_input_tokens_seen": 766528,
+      "step": 980
+    },
+    {
+      "epoch": 2.047817047817048,
+      "grad_norm": 3.631380796432495,
+      "learning_rate": 4.992537852951011e-05,
+      "loss": 0.1184,
+      "num_input_tokens_seen": 770528,
+      "step": 985
+    },
+    {
+      "epoch": 2.0582120582120584,
+      "grad_norm": 1.7038533687591553,
+      "learning_rate": 4.9924618636834785e-05,
+      "loss": 0.1072,
+      "num_input_tokens_seen": 774560,
+      "step": 990
+    },
+    {
+      "epoch": 2.0686070686070686,
+      "grad_norm": 0.945583701133728,
+      "learning_rate": 4.9923854900470046e-05,
+      "loss": 0.0634,
+      "num_input_tokens_seen": 778368,
+      "step": 995
+    },
+    {
+      "epoch": 2.079002079002079,
+      "grad_norm": 2.235551118850708,
+      "learning_rate": 4.992308732053367e-05,
+      "loss": 0.1103,
+      "num_input_tokens_seen": 782304,
+      "step": 1000
+    },
+    {
+      "epoch": 2.079002079002079,
+      "eval_loss": 0.14475902915000916,
+      "eval_runtime": 11.7375,
+      "eval_samples_per_second": 72.928,
+      "eval_steps_per_second": 18.232,
+      "num_input_tokens_seen": 782304,
+      "step": 1000
+    },
+    {
+      "epoch": 2.0893970893970892,
+      "grad_norm": 2.069002151489258,
+      "learning_rate": 4.992231589714402e-05,
+      "loss": 0.1353,
+      "num_input_tokens_seen": 786176,
+      "step": 1005
+    },
+    {
+      "epoch": 2.0997920997921,
+      "grad_norm": 2.173896312713623,
+      "learning_rate": 4.992154063042007e-05,
+      "loss": 0.1735,
+      "num_input_tokens_seen": 790080,
+      "step": 1010
+    },
+    {
+      "epoch": 2.1101871101871104,
+      "grad_norm": 2.7577319145202637,
+      "learning_rate": 4.992076152048136e-05,
+      "loss": 0.1356,
+      "num_input_tokens_seen": 794016,
+      "step": 1015
+    },
+    {
+      "epoch": 2.1205821205821205,
+      "grad_norm": 1.0546536445617676,
+      "learning_rate": 4.991997856744807e-05,
+      "loss": 0.1085,
+      "num_input_tokens_seen": 797984,
+      "step": 1020
+    },
+    {
+      "epoch": 2.130977130977131,
+      "grad_norm": 3.098329782485962,
+      "learning_rate": 4.9919191771440905e-05,
+      "loss": 0.1139,
+      "num_input_tokens_seen": 801984,
+      "step": 1025
+    },
+    {
+      "epoch": 2.141372141372141,
+      "grad_norm": 0.8387546539306641,
+      "learning_rate": 4.991840113258122e-05,
+      "loss": 0.0855,
+      "num_input_tokens_seen": 805888,
+      "step": 1030
+    },
+    {
+      "epoch": 2.1517671517671517,
+      "grad_norm": 1.1556094884872437,
+      "learning_rate": 4.9917606650990933e-05,
+      "loss": 0.1028,
+      "num_input_tokens_seen": 809792,
+      "step": 1035
+    },
+    {
+      "epoch": 2.1621621621621623,
+      "grad_norm": 1.502564787864685,
+      "learning_rate": 4.9916808326792566e-05,
+      "loss": 0.0852,
+      "num_input_tokens_seen": 813728,
+      "step": 1040
+    },
+    {
+      "epoch": 2.1725571725571724,
+      "grad_norm": 1.8533755540847778,
+      "learning_rate": 4.9916006160109235e-05,
+      "loss": 0.1157,
+      "num_input_tokens_seen": 817760,
+      "step": 1045
+    },
+    {
+      "epoch": 2.182952182952183,
+      "grad_norm": 1.1098006963729858,
+      "learning_rate": 4.991520015106464e-05,
+      "loss": 0.1095,
+      "num_input_tokens_seen": 821760,
+      "step": 1050
+    },
+    {
+      "epoch": 2.1933471933471935,
+      "grad_norm": 1.9087544679641724,
+      "learning_rate": 4.991439029978308e-05,
+      "loss": 0.1146,
+      "num_input_tokens_seen": 825792,
+      "step": 1055
+    },
+    {
+      "epoch": 2.2037422037422036,
+      "grad_norm": 0.931681215763092,
+      "learning_rate": 4.9913576606389434e-05,
+      "loss": 0.1163,
+      "num_input_tokens_seen": 829696,
+      "step": 1060
+    },
+    {
+      "epoch": 2.214137214137214,
+      "grad_norm": 1.3091639280319214,
+      "learning_rate": 4.991275907100919e-05,
+      "loss": 0.1016,
+      "num_input_tokens_seen": 833664,
+      "step": 1065
+    },
+    {
+      "epoch": 2.2245322245322248,
+      "grad_norm": 1.331113338470459,
+      "learning_rate": 4.9911937693768434e-05,
+      "loss": 0.1323,
+      "num_input_tokens_seen": 837568,
+      "step": 1070
+    },
+    {
+      "epoch": 2.234927234927235,
+      "grad_norm": 1.1060495376586914,
+      "learning_rate": 4.991111247479382e-05,
+      "loss": 0.0911,
+      "num_input_tokens_seen": 841312,
+      "step": 1075
+    },
+    {
+      "epoch": 2.2453222453222454,
+      "grad_norm": 1.1464821100234985,
+      "learning_rate": 4.9910283414212605e-05,
+      "loss": 0.1134,
+      "num_input_tokens_seen": 845056,
+      "step": 1080
+    },
+    {
+      "epoch": 2.2557172557172556,
+      "grad_norm": 0.9112980365753174,
+      "learning_rate": 4.990945051215265e-05,
+      "loss": 0.0903,
+      "num_input_tokens_seen": 848928,
+      "step": 1085
+    },
+    {
+      "epoch": 2.266112266112266,
+      "grad_norm": 0.5969815254211426,
+      "learning_rate": 4.99086137687424e-05,
+      "loss": 0.091,
+      "num_input_tokens_seen": 852800,
+      "step": 1090
+    },
+    {
+      "epoch": 2.2765072765072767,
+      "grad_norm": 2.181849718093872,
+      "learning_rate": 4.9907773184110874e-05,
+      "loss": 0.1399,
+      "num_input_tokens_seen": 856736,
+      "step": 1095
+    },
+    {
+      "epoch": 2.286902286902287,
+      "grad_norm": 1.6436723470687866,
+      "learning_rate": 4.9906928758387715e-05,
+      "loss": 0.1233,
+      "num_input_tokens_seen": 860608,
+      "step": 1100
+    },
+    {
+      "epoch": 2.2972972972972974,
+      "grad_norm": 1.9354290962219238,
+      "learning_rate": 4.9906080491703146e-05,
+      "loss": 0.1289,
+      "num_input_tokens_seen": 864480,
+      "step": 1105
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 2.9185686111450195,
+      "learning_rate": 4.990522838418797e-05,
+      "loss": 0.1097,
+      "num_input_tokens_seen": 868352,
+      "step": 1110
+    },
+    {
+      "epoch": 2.318087318087318,
+      "grad_norm": 0.6285694241523743,
+      "learning_rate": 4.9904372435973604e-05,
+      "loss": 0.1107,
+      "num_input_tokens_seen": 872320,
+      "step": 1115
+    },
+    {
+      "epoch": 2.3284823284823286,
+      "grad_norm": 2.5175721645355225,
+      "learning_rate": 4.990351264719203e-05,
+      "loss": 0.1157,
+      "num_input_tokens_seen": 876256,
+      "step": 1120
+    },
+    {
+      "epoch": 2.3388773388773387,
+      "grad_norm": 3.2958950996398926,
+      "learning_rate": 4.990264901797586e-05,
+      "loss": 0.1534,
+      "num_input_tokens_seen": 880256,
+      "step": 1125
+    },
+    {
+      "epoch": 2.3492723492723493,
+      "grad_norm": 0.6458130478858948,
+      "learning_rate": 4.990178154845826e-05,
+      "loss": 0.0812,
+      "num_input_tokens_seen": 884064,
+      "step": 1130
+    },
+    {
+      "epoch": 2.35966735966736,
+      "grad_norm": 1.0521550178527832,
+      "learning_rate": 4.9900910238773014e-05,
+      "loss": 0.0844,
+      "num_input_tokens_seen": 887968,
+      "step": 1135
+    },
+    {
+      "epoch": 2.37006237006237,
+      "grad_norm": 1.694340705871582,
+      "learning_rate": 4.990003508905448e-05,
+      "loss": 0.1126,
+      "num_input_tokens_seen": 891936,
+      "step": 1140
+    },
+    {
+      "epoch": 2.3804573804573805,
+      "grad_norm": 2.6967995166778564,
+      "learning_rate": 4.989915609943763e-05,
+      "loss": 0.1048,
+      "num_input_tokens_seen": 895776,
+      "step": 1145
+    },
+    {
+      "epoch": 2.390852390852391,
+      "grad_norm": 2.0080816745758057,
+      "learning_rate": 4.9898273270058e-05,
+      "loss": 0.1318,
+      "num_input_tokens_seen": 899584,
+      "step": 1150
+    },
+    {
+      "epoch": 2.401247401247401,
+      "grad_norm": 1.6507608890533447,
+      "learning_rate": 4.989738660105174e-05,
+      "loss": 0.0971,
+      "num_input_tokens_seen": 903392,
+      "step": 1155
+    },
+    {
+      "epoch": 2.4116424116424118,
+      "grad_norm": 3.3507463932037354,
+      "learning_rate": 4.989649609255559e-05,
+      "loss": 0.081,
+      "num_input_tokens_seen": 907392,
+      "step": 1160
+    },
+    {
+      "epoch": 2.422037422037422,
+      "grad_norm": 1.0209333896636963,
+      "learning_rate": 4.989560174470687e-05,
+      "loss": 0.1063,
+      "num_input_tokens_seen": 911296,
+      "step": 1165
+    },
+    {
+      "epoch": 2.4324324324324325,
+      "grad_norm": 0.7667717933654785,
+      "learning_rate": 4.989470355764351e-05,
+      "loss": 0.0718,
+      "num_input_tokens_seen": 915232,
+      "step": 1170
+    },
+    {
+      "epoch": 2.442827442827443,
+      "grad_norm": 1.201318383216858,
+      "learning_rate": 4.9893801531504e-05,
+      "loss": 0.0989,
+      "num_input_tokens_seen": 919072,
+      "step": 1175
+    },
+    {
+      "epoch": 2.453222453222453,
+      "grad_norm": 1.8414338827133179,
+      "learning_rate": 4.9892895666427475e-05,
+      "loss": 0.0905,
+      "num_input_tokens_seen": 922976,
+      "step": 1180
+    },
+    {
+      "epoch": 2.4636174636174637,
+      "grad_norm": 1.5301529169082642,
+      "learning_rate": 4.9891985962553606e-05,
+      "loss": 0.1146,
+      "num_input_tokens_seen": 926976,
+      "step": 1185
+    },
+    {
+      "epoch": 2.474012474012474,
+      "grad_norm": 2.180487632751465,
+      "learning_rate": 4.989107242002269e-05,
+      "loss": 0.1394,
+      "num_input_tokens_seen": 930720,
+      "step": 1190
+    },
+    {
+      "epoch": 2.4844074844074844,
+      "grad_norm": 1.1226143836975098,
+      "learning_rate": 4.989015503897561e-05,
+      "loss": 0.0817,
+      "num_input_tokens_seen": 934880,
+      "step": 1195
+    },
+    {
+      "epoch": 2.494802494802495,
+      "grad_norm": 3.743725538253784,
+      "learning_rate": 4.988923381955383e-05,
+      "loss": 0.0885,
+      "num_input_tokens_seen": 938560,
+      "step": 1200
+    },
+    {
+      "epoch": 2.494802494802495,
+      "eval_loss": 0.1171746477484703,
+      "eval_runtime": 11.7393,
+      "eval_samples_per_second": 72.918,
+      "eval_steps_per_second": 18.229,
+      "num_input_tokens_seen": 938560,
+      "step": 1200
+    },
+    {
+      "epoch": 2.505197505197505,
+      "grad_norm": 2.137660026550293,
+      "learning_rate": 4.988830876189942e-05,
+      "loss": 0.1045,
+      "num_input_tokens_seen": 942528,
+      "step": 1205
+    },
+    {
+      "epoch": 2.5155925155925156,
+      "grad_norm": 1.3337303400039673,
+      "learning_rate": 4.988737986615503e-05,
+      "loss": 0.1051,
+      "num_input_tokens_seen": 946496,
+      "step": 1210
+    },
+    {
+      "epoch": 2.525987525987526,
+      "grad_norm": 2.2751457691192627,
+      "learning_rate": 4.988644713246391e-05,
+      "loss": 0.0993,
+      "num_input_tokens_seen": 950400,
+      "step": 1215
+    },
+    {
+      "epoch": 2.5363825363825363,
+      "grad_norm": 2.268911600112915,
+      "learning_rate": 4.988551056096991e-05,
+      "loss": 0.0996,
+      "num_input_tokens_seen": 954272,
+      "step": 1220
+    },
+    {
+      "epoch": 2.546777546777547,
+      "grad_norm": 2.1609456539154053,
+      "learning_rate": 4.988457015181743e-05,
+      "loss": 0.0772,
+      "num_input_tokens_seen": 958176,
+      "step": 1225
+    },
+    {
+      "epoch": 2.5571725571725574,
+      "grad_norm": 0.9419129490852356,
+      "learning_rate": 4.988362590515153e-05,
+      "loss": 0.0955,
+      "num_input_tokens_seen": 962112,
+      "step": 1230
+    },
+    {
+      "epoch": 2.5675675675675675,
+      "grad_norm": 1.764098882675171,
+      "learning_rate": 4.9882677821117805e-05,
+      "loss": 0.1139,
+      "num_input_tokens_seen": 965952,
+      "step": 1235
+    },
+    {
+      "epoch": 2.577962577962578,
+      "grad_norm": 1.0912017822265625,
+      "learning_rate": 4.988172589986246e-05,
+      "loss": 0.1086,
+      "num_input_tokens_seen": 969920,
+      "step": 1240
+    },
+    {
+      "epoch": 2.5883575883575882,
+      "grad_norm": 2.8189423084259033,
+      "learning_rate": 4.9880770141532304e-05,
+      "loss": 0.1118,
+      "num_input_tokens_seen": 973792,
+      "step": 1245
+    },
+    {
+      "epoch": 2.598752598752599,
+      "grad_norm": 2.8706629276275635,
+      "learning_rate": 4.987981054627472e-05,
+      "loss": 0.1199,
+      "num_input_tokens_seen": 977792,
+      "step": 1250
+    },
+    {
+      "epoch": 2.609147609147609,
+      "grad_norm": 1.5601402521133423,
+      "learning_rate": 4.987884711423769e-05,
+      "loss": 0.1104,
+      "num_input_tokens_seen": 981696,
+      "step": 1255
+    },
+    {
+      "epoch": 2.6195426195426195,
+      "grad_norm": 0.966006875038147,
+      "learning_rate": 4.9877879845569784e-05,
+      "loss": 0.1177,
+      "num_input_tokens_seen": 985504,
+      "step": 1260
+    },
+    {
+      "epoch": 2.62993762993763,
+      "grad_norm": 0.4674605131149292,
+      "learning_rate": 4.9876908740420175e-05,
+      "loss": 0.0911,
+      "num_input_tokens_seen": 989472,
+      "step": 1265
+    },
+    {
+      "epoch": 2.64033264033264,
+      "grad_norm": 4.481584548950195,
+      "learning_rate": 4.987593379893861e-05,
+      "loss": 0.1083,
+      "num_input_tokens_seen": 993376,
+      "step": 1270
+    },
+    {
+      "epoch": 2.6507276507276507,
+      "grad_norm": 1.3552768230438232,
+      "learning_rate": 4.987495502127545e-05,
+      "loss": 0.0819,
+      "num_input_tokens_seen": 997312,
+      "step": 1275
+    },
+    {
+      "epoch": 2.6611226611226613,
+      "grad_norm": 0.8347285985946655,
+      "learning_rate": 4.987397240758162e-05,
+      "loss": 0.1218,
+      "num_input_tokens_seen": 1001120,
+      "step": 1280
+    },
+    {
+      "epoch": 2.6715176715176714,
+      "grad_norm": 0.7202677130699158,
+      "learning_rate": 4.9872985958008664e-05,
+      "loss": 0.0739,
+      "num_input_tokens_seen": 1004896,
+      "step": 1285
+    },
+    {
+      "epoch": 2.681912681912682,
+      "grad_norm": 1.9180819988250732,
+      "learning_rate": 4.987199567270871e-05,
+      "loss": 0.1084,
+      "num_input_tokens_seen": 1008864,
+      "step": 1290
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 3.202082872390747,
+      "learning_rate": 4.9871001551834444e-05,
+      "loss": 0.0717,
+      "num_input_tokens_seen": 1012608,
+      "step": 1295
+    },
+    {
+      "epoch": 2.7027027027027026,
+      "grad_norm": 1.826272964477539,
+      "learning_rate": 4.98700035955392e-05,
+      "loss": 0.0626,
+      "num_input_tokens_seen": 1016416,
+      "step": 1300
+    },
+    {
+      "epoch": 2.713097713097713,
+      "grad_norm": 0.8048884272575378,
+      "learning_rate": 4.986900180397686e-05,
+      "loss": 0.1074,
+      "num_input_tokens_seen": 1020256,
+      "step": 1305
+    },
+    {
+      "epoch": 2.7234927234927238,
+      "grad_norm": 1.5013339519500732,
+      "learning_rate": 4.9867996177301926e-05,
+      "loss": 0.0827,
+      "num_input_tokens_seen": 1024064,
+      "step": 1310
+    },
+    {
+      "epoch": 2.733887733887734,
+      "grad_norm": 2.3943207263946533,
+      "learning_rate": 4.9866986715669464e-05,
+      "loss": 0.129,
+      "num_input_tokens_seen": 1028096,
+      "step": 1315
+    },
+    {
+      "epoch": 2.7442827442827444,
+      "grad_norm": 1.6364383697509766,
+      "learning_rate": 4.9865973419235155e-05,
+      "loss": 0.1232,
+      "num_input_tokens_seen": 1032000,
+      "step": 1320
+    },
+    {
+      "epoch": 2.7546777546777546,
+      "grad_norm": 1.7162449359893799,
+      "learning_rate": 4.986495628815526e-05,
+      "loss": 0.0991,
+      "num_input_tokens_seen": 1035968,
+      "step": 1325
+    },
+    {
+      "epoch": 2.765072765072765,
+      "grad_norm": 0.7651729583740234,
+      "learning_rate": 4.986393532258663e-05,
+      "loss": 0.1105,
+      "num_input_tokens_seen": 1039808,
+      "step": 1330
+    },
+    {
+      "epoch": 2.7754677754677752,
+      "grad_norm": 2.252964973449707,
+      "learning_rate": 4.986291052268671e-05,
+      "loss": 0.1029,
+      "num_input_tokens_seen": 1043584,
+      "step": 1335
+    },
+    {
+      "epoch": 2.785862785862786,
+      "grad_norm": 0.9601475596427917,
+      "learning_rate": 4.986188188861355e-05,
+      "loss": 0.1004,
+      "num_input_tokens_seen": 1047520,
+      "step": 1340
+    },
+    {
+      "epoch": 2.7962577962577964,
+      "grad_norm": 1.4805073738098145,
+      "learning_rate": 4.9860849420525766e-05,
+      "loss": 0.0877,
+      "num_input_tokens_seen": 1051392,
+      "step": 1345
+    },
+    {
+      "epoch": 2.8066528066528065,
+      "grad_norm": 1.6725808382034302,
+      "learning_rate": 4.9859813118582575e-05,
+      "loss": 0.1195,
+      "num_input_tokens_seen": 1055200,
+      "step": 1350
+    },
+    {
+      "epoch": 2.817047817047817,
+      "grad_norm": 1.4402563571929932,
+      "learning_rate": 4.98587729829438e-05,
+      "loss": 0.1028,
+      "num_input_tokens_seen": 1059040,
+      "step": 1355
+    },
+    {
+      "epoch": 2.8274428274428276,
+      "grad_norm": 1.5252577066421509,
+      "learning_rate": 4.985772901376983e-05,
+      "loss": 0.0869,
+      "num_input_tokens_seen": 1062944,
+      "step": 1360
+    },
+    {
+      "epoch": 2.8378378378378377,
+      "grad_norm": 1.146884560585022,
+      "learning_rate": 4.9856681211221666e-05,
+      "loss": 0.0984,
+      "num_input_tokens_seen": 1066816,
+      "step": 1365
+    },
+    {
+      "epoch": 2.8482328482328483,
+      "grad_norm": 3.3587729930877686,
+      "learning_rate": 4.985562957546089e-05,
+      "loss": 0.109,
+      "num_input_tokens_seen": 1070560,
+      "step": 1370
+    },
+    {
+      "epoch": 2.858627858627859,
+      "grad_norm": 1.71546471118927,
+      "learning_rate": 4.9854574106649686e-05,
+      "loss": 0.1179,
+      "num_input_tokens_seen": 1074560,
+      "step": 1375
+    },
+    {
+      "epoch": 2.869022869022869,
+      "grad_norm": 0.6439744830131531,
+      "learning_rate": 4.985351480495081e-05,
+      "loss": 0.0674,
+      "num_input_tokens_seen": 1078560,
+      "step": 1380
+    },
+    {
+      "epoch": 2.8794178794178795,
+      "grad_norm": 1.6250063180923462,
+      "learning_rate": 4.985245167052762e-05,
+      "loss": 0.115,
+      "num_input_tokens_seen": 1082368,
+      "step": 1385
+    },
+    {
+      "epoch": 2.88981288981289,
+      "grad_norm": 2.2628583908081055,
+      "learning_rate": 4.9851384703544066e-05,
+      "loss": 0.0979,
+      "num_input_tokens_seen": 1086208,
+      "step": 1390
+    },
+    {
+      "epoch": 2.9002079002079,
+      "grad_norm": 2.03609299659729,
+      "learning_rate": 4.985031390416469e-05,
+      "loss": 0.0993,
+      "num_input_tokens_seen": 1090208,
+      "step": 1395
+    },
+    {
+      "epoch": 2.9106029106029108,
+      "grad_norm": 2.0300698280334473,
+      "learning_rate": 4.984923927255461e-05,
+      "loss": 0.1365,
+      "num_input_tokens_seen": 1094144,
+      "step": 1400
+    },
+    {
+      "epoch": 2.9106029106029108,
+      "eval_loss": 0.11820784211158752,
+      "eval_runtime": 11.7461,
+      "eval_samples_per_second": 72.875,
+      "eval_steps_per_second": 18.219,
+      "num_input_tokens_seen": 1094144,
+      "step": 1400
+    },
+    {
+      "epoch": 2.920997920997921,
+      "grad_norm": 1.6313543319702148,
+      "learning_rate": 4.984816080887958e-05,
+      "loss": 0.0968,
+      "num_input_tokens_seen": 1098016,
+      "step": 1405
+    },
+    {
+      "epoch": 2.9313929313929314,
+      "grad_norm": 1.4707082509994507,
+      "learning_rate": 4.9847078513305875e-05,
+      "loss": 0.095,
+      "num_input_tokens_seen": 1102016,
+      "step": 1410
+    },
+    {
+      "epoch": 2.9417879417879416,
+      "grad_norm": 1.2573354244232178,
+      "learning_rate": 4.984599238600043e-05,
+      "loss": 0.0816,
+      "num_input_tokens_seen": 1105952,
+      "step": 1415
+    },
+    {
+      "epoch": 2.952182952182952,
+      "grad_norm": 1.1725883483886719,
+      "learning_rate": 4.9844902427130716e-05,
+      "loss": 0.0827,
+      "num_input_tokens_seen": 1109888,
+      "step": 1420
+    },
+    {
+      "epoch": 2.9625779625779627,
+      "grad_norm": 2.862562656402588,
+      "learning_rate": 4.984380863686482e-05,
+      "loss": 0.1078,
+      "num_input_tokens_seen": 1113984,
+      "step": 1425
+    },
+    {
+      "epoch": 2.972972972972973,
+      "grad_norm": 3.826087236404419,
+      "learning_rate": 4.984271101537143e-05,
+      "loss": 0.0972,
+      "num_input_tokens_seen": 1118048,
+      "step": 1430
+    },
+    {
+      "epoch": 2.9833679833679834,
+      "grad_norm": 2.891716718673706,
+      "learning_rate": 4.9841609562819816e-05,
+      "loss": 0.1578,
+      "num_input_tokens_seen": 1122016,
+      "step": 1435
+    },
+    {
+      "epoch": 2.993762993762994,
+      "grad_norm": 1.6001083850860596,
+      "learning_rate": 4.984050427937983e-05,
+      "loss": 0.0622,
+      "num_input_tokens_seen": 1125792,
+      "step": 1440
+    },
+    {
+      "epoch": 3.004158004158004,
+      "grad_norm": 2.01772403717041,
+      "learning_rate": 4.983939516522191e-05,
+      "loss": 0.1137,
+      "num_input_tokens_seen": 1129680,
+      "step": 1445
+    },
+    {
+      "epoch": 3.0145530145530146,
+      "grad_norm": 1.5638049840927124,
+      "learning_rate": 4.983828222051711e-05,
+      "loss": 0.0954,
+      "num_input_tokens_seen": 1133680,
+      "step": 1450
+    },
+    {
+      "epoch": 3.024948024948025,
+      "grad_norm": 2.426571846008301,
+      "learning_rate": 4.983716544543705e-05,
+      "loss": 0.0767,
+      "num_input_tokens_seen": 1137584,
+      "step": 1455
+    },
+    {
+      "epoch": 3.0353430353430353,
+      "grad_norm": 1.5635465383529663,
+      "learning_rate": 4.983604484015395e-05,
+      "loss": 0.1064,
+      "num_input_tokens_seen": 1141616,
+      "step": 1460
+    },
+    {
+      "epoch": 3.045738045738046,
+      "grad_norm": 1.7077943086624146,
+      "learning_rate": 4.983492040484064e-05,
+      "loss": 0.0517,
+      "num_input_tokens_seen": 1145456,
+      "step": 1465
+    },
+    {
+      "epoch": 3.056133056133056,
+      "grad_norm": 5.166476249694824,
+      "learning_rate": 4.98337921396705e-05,
+      "loss": 0.075,
+      "num_input_tokens_seen": 1149392,
+      "step": 1470
+    },
+    {
+      "epoch": 3.0665280665280665,
+      "grad_norm": 3.9282407760620117,
+      "learning_rate": 4.983266004481753e-05,
+      "loss": 0.0903,
+      "num_input_tokens_seen": 1153360,
+      "step": 1475
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 1.5178333520889282,
+      "learning_rate": 4.9831524120456316e-05,
+      "loss": 0.102,
+      "num_input_tokens_seen": 1157296,
+      "step": 1480
+    },
+    {
+      "epoch": 3.087318087318087,
+      "grad_norm": 3.058875560760498,
+      "learning_rate": 4.9830384366762026e-05,
+      "loss": 0.0878,
+      "num_input_tokens_seen": 1161328,
+      "step": 1485
+    },
+    {
+      "epoch": 3.0977130977130978,
+      "grad_norm": 1.6579962968826294,
+      "learning_rate": 4.9829240783910436e-05,
+      "loss": 0.1021,
+      "num_input_tokens_seen": 1165264,
+      "step": 1490
+    },
+    {
+      "epoch": 3.108108108108108,
+      "grad_norm": 1.4158744812011719,
+      "learning_rate": 4.982809337207789e-05,
+      "loss": 0.0571,
+      "num_input_tokens_seen": 1169200,
+      "step": 1495
+    },
+    {
+      "epoch": 3.1185031185031185,
+      "grad_norm": 3.4372661113739014,
+      "learning_rate": 4.9826942131441337e-05,
+      "loss": 0.1224,
+      "num_input_tokens_seen": 1173104,
+      "step": 1500
+    },
+    {
+      "epoch": 3.128898128898129,
+      "grad_norm": 1.2253714799880981,
+      "learning_rate": 4.9825787062178315e-05,
+      "loss": 0.0586,
+      "num_input_tokens_seen": 1176944,
+      "step": 1505
+    },
+    {
+      "epoch": 3.139293139293139,
+      "grad_norm": 1.2087267637252808,
+      "learning_rate": 4.9824628164466945e-05,
+      "loss": 0.055,
+      "num_input_tokens_seen": 1180816,
+      "step": 1510
+    },
+    {
+      "epoch": 3.1496881496881497,
+      "grad_norm": 2.8129379749298096,
+      "learning_rate": 4.982346543848595e-05,
+      "loss": 0.0445,
+      "num_input_tokens_seen": 1184752,
+      "step": 1515
+    },
+    {
+      "epoch": 3.1600831600831603,
+      "grad_norm": 2.8411333560943604,
+      "learning_rate": 4.9822298884414626e-05,
+      "loss": 0.0765,
+      "num_input_tokens_seen": 1188528,
+      "step": 1520
+    },
+    {
+      "epoch": 3.1704781704781704,
+      "grad_norm": 1.0603411197662354,
+      "learning_rate": 4.982112850243288e-05,
+      "loss": 0.127,
+      "num_input_tokens_seen": 1192496,
+      "step": 1525
+    },
+    {
+      "epoch": 3.180873180873181,
+      "grad_norm": 1.9706228971481323,
+      "learning_rate": 4.98199542927212e-05,
+      "loss": 0.056,
+      "num_input_tokens_seen": 1196432,
+      "step": 1530
+    },
+    {
+      "epoch": 3.1912681912681915,
+      "grad_norm": 1.8726189136505127,
+      "learning_rate": 4.981877625546066e-05,
+      "loss": 0.0677,
+      "num_input_tokens_seen": 1200272,
+      "step": 1535
+    },
+    {
+      "epoch": 3.2016632016632016,
+      "grad_norm": 2.195997953414917,
+      "learning_rate": 4.981759439083293e-05,
+      "loss": 0.127,
+      "num_input_tokens_seen": 1204144,
+      "step": 1540
+    },
+    {
+      "epoch": 3.212058212058212,
+      "grad_norm": 1.1727287769317627,
+      "learning_rate": 4.981640869902027e-05,
+      "loss": 0.0771,
+      "num_input_tokens_seen": 1208048,
+      "step": 1545
+    },
+    {
+      "epoch": 3.2224532224532223,
+      "grad_norm": 0.4863424599170685,
+      "learning_rate": 4.9815219180205517e-05,
+      "loss": 0.0621,
+      "num_input_tokens_seen": 1211984,
+      "step": 1550
+    },
+    {
+      "epoch": 3.232848232848233,
+      "grad_norm": 1.3246047496795654,
+      "learning_rate": 4.9814025834572126e-05,
+      "loss": 0.0644,
+      "num_input_tokens_seen": 1215792,
+      "step": 1555
+    },
+    {
+      "epoch": 3.2432432432432434,
+      "grad_norm": 0.7587156891822815,
+      "learning_rate": 4.981282866230411e-05,
+      "loss": 0.0907,
+      "num_input_tokens_seen": 1219568,
+      "step": 1560
+    },
+    {
+      "epoch": 3.2536382536382535,
+      "grad_norm": 3.0291249752044678,
+      "learning_rate": 4.981162766358611e-05,
+      "loss": 0.074,
+      "num_input_tokens_seen": 1223472,
+      "step": 1565
+    },
+    {
+      "epoch": 3.264033264033264,
+      "grad_norm": 2.5555543899536133,
+      "learning_rate": 4.9810422838603316e-05,
+      "loss": 0.0899,
+      "num_input_tokens_seen": 1227184,
+      "step": 1570
+    },
+    {
+      "epoch": 3.274428274428274,
+      "grad_norm": 0.9471989274024963,
+      "learning_rate": 4.9809214187541533e-05,
+      "loss": 0.0834,
+      "num_input_tokens_seen": 1231056,
+      "step": 1575
+    },
+    {
+      "epoch": 3.284823284823285,
+      "grad_norm": 1.1901923418045044,
+      "learning_rate": 4.980800171058715e-05,
+      "loss": 0.0854,
+      "num_input_tokens_seen": 1235056,
+      "step": 1580
+    },
+    {
+      "epoch": 3.2952182952182953,
+      "grad_norm": 1.0427956581115723,
+      "learning_rate": 4.980678540792715e-05,
+      "loss": 0.0808,
+      "num_input_tokens_seen": 1239024,
+      "step": 1585
+    },
+    {
+      "epoch": 3.3056133056133055,
+      "grad_norm": 0.6777945756912231,
+      "learning_rate": 4.980556527974909e-05,
+      "loss": 0.0741,
+      "num_input_tokens_seen": 1242768,
+      "step": 1590
+    },
+    {
+      "epoch": 3.316008316008316,
+      "grad_norm": 1.6756219863891602,
+      "learning_rate": 4.980434132624114e-05,
+      "loss": 0.0893,
+      "num_input_tokens_seen": 1246608,
+      "step": 1595
+    },
+    {
+      "epoch": 3.3264033264033266,
+      "grad_norm": 2.8532302379608154,
+      "learning_rate": 4.980311354759205e-05,
+      "loss": 0.0822,
+      "num_input_tokens_seen": 1250544,
+      "step": 1600
+    },
+    {
+      "epoch": 3.3264033264033266,
+      "eval_loss": 0.11357858031988144,
+      "eval_runtime": 11.7298,
+      "eval_samples_per_second": 72.977,
+      "eval_steps_per_second": 18.244,
+      "num_input_tokens_seen": 1250544,
+      "step": 1600
+    },
+    {
+      "epoch": 3.3367983367983367,
+      "grad_norm": 1.828748106956482,
+      "learning_rate": 4.980188194399116e-05,
+      "loss": 0.0982,
+      "num_input_tokens_seen": 1254416,
+      "step": 1605
+    },
+    {
+      "epoch": 3.3471933471933473,
+      "grad_norm": 1.1847913265228271,
+      "learning_rate": 4.9800646515628384e-05,
+      "loss": 0.0719,
+      "num_input_tokens_seen": 1258288,
+      "step": 1610
+    },
+    {
+      "epoch": 3.357588357588358,
+      "grad_norm": 1.1204500198364258,
+      "learning_rate": 4.979940726269426e-05,
+      "loss": 0.1133,
+      "num_input_tokens_seen": 1262256,
+      "step": 1615
+    },
+    {
+      "epoch": 3.367983367983368,
+      "grad_norm": 1.3574473857879639,
+      "learning_rate": 4.979816418537988e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 1266064,
+      "step": 1620
+    },
+    {
+      "epoch": 3.3783783783783785,
+      "grad_norm": 2.9061965942382812,
+      "learning_rate": 4.979691728387696e-05,
+      "loss": 0.0411,
+      "num_input_tokens_seen": 1269936,
+      "step": 1625
+    },
+    {
+      "epoch": 3.3887733887733886,
+      "grad_norm": 2.5860774517059326,
+      "learning_rate": 4.979566655837776e-05,
+      "loss": 0.0638,
+      "num_input_tokens_seen": 1273776,
+      "step": 1630
+    },
+    {
+      "epoch": 3.399168399168399,
+      "grad_norm": 3.4192051887512207,
+      "learning_rate": 4.9794412009075184e-05,
+      "loss": 0.1024,
+      "num_input_tokens_seen": 1277840,
+      "step": 1635
+    },
+    {
+      "epoch": 3.4095634095634098,
+      "grad_norm": 2.2081363201141357,
+      "learning_rate": 4.979315363616269e-05,
+      "loss": 0.1114,
+      "num_input_tokens_seen": 1281872,
+      "step": 1640
+    },
+    {
+      "epoch": 3.41995841995842,
+      "grad_norm": 0.8070342540740967,
+      "learning_rate": 4.979189143983434e-05,
+      "loss": 0.054,
+      "num_input_tokens_seen": 1285776,
+      "step": 1645
+    },
+    {
+      "epoch": 3.4303534303534304,
+      "grad_norm": 0.9763333797454834,
+      "learning_rate": 4.979062542028478e-05,
+      "loss": 0.0958,
+      "num_input_tokens_seen": 1289744,
+      "step": 1650
+    },
+    {
+      "epoch": 3.4407484407484406,
+      "grad_norm": 3.5090322494506836,
+      "learning_rate": 4.978935557770923e-05,
+      "loss": 0.0875,
+      "num_input_tokens_seen": 1293584,
+      "step": 1655
+    },
+    {
+      "epoch": 3.451143451143451,
+      "grad_norm": 2.8172214031219482,
+      "learning_rate": 4.978808191230353e-05,
+      "loss": 0.0729,
+      "num_input_tokens_seen": 1297360,
+      "step": 1660
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 2.3486151695251465,
+      "learning_rate": 4.9786804424264085e-05,
+      "loss": 0.0888,
+      "num_input_tokens_seen": 1301264,
+      "step": 1665
+    },
+    {
+      "epoch": 3.471933471933472,
+      "grad_norm": 1.4960966110229492,
+      "learning_rate": 4.978552311378792e-05,
+      "loss": 0.0772,
+      "num_input_tokens_seen": 1305136,
+      "step": 1670
+    },
+    {
+      "epoch": 3.4823284823284824,
+      "grad_norm": 0.9146766662597656,
+      "learning_rate": 4.978423798107261e-05,
+      "loss": 0.0738,
+      "num_input_tokens_seen": 1308944,
+      "step": 1675
+    },
+    {
+      "epoch": 3.492723492723493,
+      "grad_norm": 0.6939899921417236,
+      "learning_rate": 4.978294902631635e-05,
+      "loss": 0.0965,
+      "num_input_tokens_seen": 1312816,
+      "step": 1680
+    },
+    {
+      "epoch": 3.503118503118503,
+      "grad_norm": 2.356868028640747,
+      "learning_rate": 4.9781656249717914e-05,
+      "loss": 0.0771,
+      "num_input_tokens_seen": 1316752,
+      "step": 1685
+    },
+    {
+      "epoch": 3.5135135135135136,
+      "grad_norm": 2.858529567718506,
+      "learning_rate": 4.9780359651476645e-05,
+      "loss": 0.07,
+      "num_input_tokens_seen": 1320688,
+      "step": 1690
+    },
+    {
+      "epoch": 3.523908523908524,
+      "grad_norm": 1.0865362882614136,
+      "learning_rate": 4.977905923179251e-05,
+      "loss": 0.0908,
+      "num_input_tokens_seen": 1324688,
+      "step": 1695
+    },
+    {
+      "epoch": 3.5343035343035343,
+      "grad_norm": 1.955013394355774,
+      "learning_rate": 4.977775499086606e-05,
+      "loss": 0.051,
+      "num_input_tokens_seen": 1328720,
+      "step": 1700
+    },
+    {
+      "epoch": 3.544698544698545,
+      "grad_norm": 2.425605535507202,
+      "learning_rate": 4.97764469288984e-05,
+      "loss": 0.0763,
+      "num_input_tokens_seen": 1332688,
+      "step": 1705
+    },
+    {
+      "epoch": 3.555093555093555,
+      "grad_norm": 1.3733640909194946,
+      "learning_rate": 4.977513504609127e-05,
+      "loss": 0.0931,
+      "num_input_tokens_seen": 1336624,
+      "step": 1710
+    },
+    {
+      "epoch": 3.5654885654885655,
+      "grad_norm": 1.6114391088485718,
+      "learning_rate": 4.9773819342646965e-05,
+      "loss": 0.0665,
+      "num_input_tokens_seen": 1340528,
+      "step": 1715
+    },
+    {
+      "epoch": 3.5758835758835756,
+      "grad_norm": 1.6096720695495605,
+      "learning_rate": 4.97724998187684e-05,
+      "loss": 0.0938,
+      "num_input_tokens_seen": 1344400,
+      "step": 1720
+    },
+    {
+      "epoch": 3.586278586278586,
+      "grad_norm": 1.325593113899231,
+      "learning_rate": 4.9771176474659045e-05,
+      "loss": 0.0623,
+      "num_input_tokens_seen": 1348432,
+      "step": 1725
+    },
+    {
+      "epoch": 3.5966735966735968,
+      "grad_norm": 4.302254676818848,
+      "learning_rate": 4.976984931052299e-05,
+      "loss": 0.0723,
+      "num_input_tokens_seen": 1352368,
+      "step": 1730
+    },
+    {
+      "epoch": 3.607068607068607,
+      "grad_norm": 1.5492842197418213,
+      "learning_rate": 4.976851832656489e-05,
+      "loss": 0.0582,
+      "num_input_tokens_seen": 1356144,
+      "step": 1735
+    },
+    {
+      "epoch": 3.6174636174636174,
+      "grad_norm": 1.6570332050323486,
+      "learning_rate": 4.9767183522990004e-05,
+      "loss": 0.0927,
+      "num_input_tokens_seen": 1360144,
+      "step": 1740
+    },
+    {
+      "epoch": 3.627858627858628,
+      "grad_norm": 1.7081284523010254,
+      "learning_rate": 4.9765844900004176e-05,
+      "loss": 0.0889,
+      "num_input_tokens_seen": 1364016,
+      "step": 1745
+    },
+    {
+      "epoch": 3.638253638253638,
+      "grad_norm": 2.4115359783172607,
+      "learning_rate": 4.9764502457813834e-05,
+      "loss": 0.0762,
+      "num_input_tokens_seen": 1368048,
+      "step": 1750
+    },
+    {
+      "epoch": 3.6486486486486487,
+      "grad_norm": 3.0755650997161865,
+      "learning_rate": 4.9763156196626005e-05,
+      "loss": 0.0537,
+      "num_input_tokens_seen": 1371888,
+      "step": 1755
+    },
+    {
+      "epoch": 3.6590436590436592,
+      "grad_norm": 1.9900604486465454,
+      "learning_rate": 4.97618061166483e-05,
+      "loss": 0.0629,
+      "num_input_tokens_seen": 1375792,
+      "step": 1760
+    },
+    {
+      "epoch": 3.6694386694386694,
+      "grad_norm": 2.3767342567443848,
+      "learning_rate": 4.9760452218088915e-05,
+      "loss": 0.0614,
+      "num_input_tokens_seen": 1379600,
+      "step": 1765
+    },
+    {
+      "epoch": 3.67983367983368,
+      "grad_norm": 3.414614677429199,
+      "learning_rate": 4.975909450115663e-05,
+      "loss": 0.0922,
+      "num_input_tokens_seen": 1383600,
+      "step": 1770
+    },
+    {
+      "epoch": 3.6902286902286905,
+      "grad_norm": 1.8949253559112549,
+      "learning_rate": 4.975773296606084e-05,
+      "loss": 0.0789,
+      "num_input_tokens_seen": 1387632,
+      "step": 1775
+    },
+    {
+      "epoch": 3.7006237006237006,
+      "grad_norm": 0.9566705822944641,
+      "learning_rate": 4.97563676130115e-05,
+      "loss": 0.0722,
+      "num_input_tokens_seen": 1391632,
+      "step": 1780
+    },
+    {
+      "epoch": 3.711018711018711,
+      "grad_norm": 1.3604012727737427,
+      "learning_rate": 4.9754998442219166e-05,
+      "loss": 0.0515,
+      "num_input_tokens_seen": 1395696,
+      "step": 1785
+    },
+    {
+      "epoch": 3.7214137214137213,
+      "grad_norm": 1.59651780128479,
+      "learning_rate": 4.9753625453894984e-05,
+      "loss": 0.0839,
+      "num_input_tokens_seen": 1399504,
+      "step": 1790
+    },
+    {
+      "epoch": 3.731808731808732,
+      "grad_norm": 1.4416794776916504,
+      "learning_rate": 4.975224864825068e-05,
+      "loss": 0.0823,
+      "num_input_tokens_seen": 1403504,
+      "step": 1795
+    },
+    {
+      "epoch": 3.742203742203742,
+      "grad_norm": 1.2233071327209473,
+      "learning_rate": 4.9750868025498576e-05,
+      "loss": 0.0472,
+      "num_input_tokens_seen": 1407440,
+      "step": 1800
+    },
+    {
+      "epoch": 3.742203742203742,
+      "eval_loss": 0.11633815616369247,
+      "eval_runtime": 11.7158,
+      "eval_samples_per_second": 73.064,
+      "eval_steps_per_second": 18.266,
+      "num_input_tokens_seen": 1407440,
+      "step": 1800
+    },
+    {
+      "epoch": 3.7525987525987525,
+      "grad_norm": 3.1031711101531982,
+      "learning_rate": 4.974948358585158e-05,
+      "loss": 0.1074,
+      "num_input_tokens_seen": 1411248,
+      "step": 1805
+    },
+    {
+      "epoch": 3.762993762993763,
+      "grad_norm": 1.5377775430679321,
+      "learning_rate": 4.9748095329523205e-05,
+      "loss": 0.0552,
+      "num_input_tokens_seen": 1415088,
+      "step": 1810
+    },
+    {
+      "epoch": 3.773388773388773,
+      "grad_norm": 3.0684947967529297,
+      "learning_rate": 4.974670325672752e-05,
+      "loss": 0.0748,
+      "num_input_tokens_seen": 1419056,
+      "step": 1815
+    },
+    {
+      "epoch": 3.7837837837837838,
+      "grad_norm": 1.5123759508132935,
+      "learning_rate": 4.974530736767921e-05,
+      "loss": 0.049,
+      "num_input_tokens_seen": 1422992,
+      "step": 1820
+    },
+    {
+      "epoch": 3.7941787941787943,
+      "grad_norm": 2.1821107864379883,
+      "learning_rate": 4.9743907662593524e-05,
+      "loss": 0.0596,
+      "num_input_tokens_seen": 1426896,
+      "step": 1825
+    },
+    {
+      "epoch": 3.8045738045738045,
+      "grad_norm": 3.80794358253479,
+      "learning_rate": 4.974250414168633e-05,
+      "loss": 0.0773,
+      "num_input_tokens_seen": 1430704,
+      "step": 1830
+    },
+    {
+      "epoch": 3.814968814968815,
+      "grad_norm": 2.418774366378784,
+      "learning_rate": 4.974109680517407e-05,
+      "loss": 0.0991,
+      "num_input_tokens_seen": 1434768,
+      "step": 1835
+    },
+    {
+      "epoch": 3.8253638253638256,
+      "grad_norm": 2.061624765396118,
+      "learning_rate": 4.973968565327376e-05,
+      "loss": 0.1144,
+      "num_input_tokens_seen": 1438672,
+      "step": 1840
+    },
+    {
+      "epoch": 3.8357588357588357,
+      "grad_norm": 1.1311324834823608,
+      "learning_rate": 4.973827068620303e-05,
+      "loss": 0.0693,
+      "num_input_tokens_seen": 1442640,
+      "step": 1845
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 1.1184659004211426,
+      "learning_rate": 4.973685190418008e-05,
+      "loss": 0.0482,
+      "num_input_tokens_seen": 1446512,
+      "step": 1850
+    },
+    {
+      "epoch": 3.856548856548857,
+      "grad_norm": 2.1937286853790283,
+      "learning_rate": 4.97354293074237e-05,
+      "loss": 0.0851,
+      "num_input_tokens_seen": 1450384,
+      "step": 1855
+    },
+    {
+      "epoch": 3.866943866943867,
+      "grad_norm": 2.4927890300750732,
+      "learning_rate": 4.9734002896153276e-05,
+      "loss": 0.1014,
+      "num_input_tokens_seen": 1454160,
+      "step": 1860
+    },
+    {
+      "epoch": 3.8773388773388775,
+      "grad_norm": 1.9283006191253662,
+      "learning_rate": 4.973257267058877e-05,
+      "loss": 0.1342,
+      "num_input_tokens_seen": 1458096,
+      "step": 1865
+    },
+    {
+      "epoch": 3.8877338877338876,
+      "grad_norm": 1.1847726106643677,
+      "learning_rate": 4.973113863095076e-05,
+      "loss": 0.0856,
+      "num_input_tokens_seen": 1462032,
+      "step": 1870
+    },
+    {
+      "epoch": 3.898128898128898,
+      "grad_norm": 2.2974116802215576,
+      "learning_rate": 4.9729700777460384e-05,
+      "loss": 0.0854,
+      "num_input_tokens_seen": 1466064,
+      "step": 1875
+    },
+    {
+      "epoch": 3.9085239085239083,
+      "grad_norm": 0.9547013640403748,
+      "learning_rate": 4.972825911033937e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 1469840,
+      "step": 1880
+    },
+    {
+      "epoch": 3.918918918918919,
+      "grad_norm": 1.494032382965088,
+      "learning_rate": 4.9726813629810056e-05,
+      "loss": 0.0694,
+      "num_input_tokens_seen": 1473680,
+      "step": 1885
+    },
+    {
+      "epoch": 3.9293139293139294,
+      "grad_norm": 1.777726411819458,
+      "learning_rate": 4.9725364336095326e-05,
+      "loss": 0.0528,
+      "num_input_tokens_seen": 1477616,
+      "step": 1890
+    },
+    {
+      "epoch": 3.9397089397089395,
+      "grad_norm": 3.392860174179077,
+      "learning_rate": 4.972391122941871e-05,
+      "loss": 0.0723,
+      "num_input_tokens_seen": 1481424,
+      "step": 1895
+    },
+    {
+      "epoch": 3.95010395010395,
+      "grad_norm": 0.4901174306869507,
+      "learning_rate": 4.972245431000428e-05,
+      "loss": 0.0253,
+      "num_input_tokens_seen": 1485392,
+      "step": 1900
+    },
+    {
+      "epoch": 3.9604989604989607,
+      "grad_norm": 3.200289726257324,
+      "learning_rate": 4.972099357807671e-05,
+      "loss": 0.0912,
+      "num_input_tokens_seen": 1489296,
+      "step": 1905
+    },
+    {
+      "epoch": 3.970893970893971,
+      "grad_norm": 2.250957489013672,
+      "learning_rate": 4.971952903386127e-05,
+      "loss": 0.0686,
+      "num_input_tokens_seen": 1493104,
+      "step": 1910
+    },
+    {
+      "epoch": 3.9812889812889813,
+      "grad_norm": 4.5319671630859375,
+      "learning_rate": 4.971806067758381e-05,
+      "loss": 0.0899,
+      "num_input_tokens_seen": 1497072,
+      "step": 1915
+    },
+    {
+      "epoch": 3.991683991683992,
+      "grad_norm": 1.0679068565368652,
+      "learning_rate": 4.971658850947076e-05,
+      "loss": 0.0972,
+      "num_input_tokens_seen": 1501072,
+      "step": 1920
+    },
+    {
+      "epoch": 4.002079002079002,
+      "grad_norm": 0.6308817863464355,
+      "learning_rate": 4.9715112529749165e-05,
+      "loss": 0.0446,
+      "num_input_tokens_seen": 1505080,
+      "step": 1925
+    },
+    {
+      "epoch": 4.012474012474012,
+      "grad_norm": 0.9566804766654968,
+      "learning_rate": 4.9713632738646624e-05,
+      "loss": 0.04,
+      "num_input_tokens_seen": 1509080,
+      "step": 1930
+    },
+    {
+      "epoch": 4.022869022869023,
+      "grad_norm": 3.9046921730041504,
+      "learning_rate": 4.971214913639134e-05,
+      "loss": 0.064,
+      "num_input_tokens_seen": 1512760,
+      "step": 1935
+    },
+    {
+      "epoch": 4.033264033264033,
+      "grad_norm": 0.3602249324321747,
+      "learning_rate": 4.9710661723212104e-05,
+      "loss": 0.0481,
+      "num_input_tokens_seen": 1516664,
+      "step": 1940
+    },
+    {
+      "epoch": 4.043659043659043,
+      "grad_norm": 1.8709810972213745,
+      "learning_rate": 4.9709170499338295e-05,
+      "loss": 0.0873,
+      "num_input_tokens_seen": 1520504,
+      "step": 1945
+    },
+    {
+      "epoch": 4.054054054054054,
+      "grad_norm": 3.663151264190674,
+      "learning_rate": 4.9707675464999895e-05,
+      "loss": 0.0394,
+      "num_input_tokens_seen": 1524536,
+      "step": 1950
+    },
+    {
+      "epoch": 4.0644490644490645,
+      "grad_norm": 2.4017012119293213,
+      "learning_rate": 4.970617662042743e-05,
+      "loss": 0.0308,
+      "num_input_tokens_seen": 1528280,
+      "step": 1955
+    },
+    {
+      "epoch": 4.074844074844075,
+      "grad_norm": 0.3524039685726166,
+      "learning_rate": 4.970467396585206e-05,
+      "loss": 0.0589,
+      "num_input_tokens_seen": 1532184,
+      "step": 1960
+    },
+    {
+      "epoch": 4.085239085239086,
+      "grad_norm": 3.385021924972534,
+      "learning_rate": 4.97031675015055e-05,
+      "loss": 0.1137,
+      "num_input_tokens_seen": 1536152,
+      "step": 1965
+    },
+    {
+      "epoch": 4.095634095634096,
+      "grad_norm": 3.036262273788452,
+      "learning_rate": 4.9701657227620075e-05,
+      "loss": 0.0771,
+      "num_input_tokens_seen": 1540120,
+      "step": 1970
+    },
+    {
+      "epoch": 4.106029106029106,
+      "grad_norm": 1.1490225791931152,
+      "learning_rate": 4.9700143144428685e-05,
+      "loss": 0.0468,
+      "num_input_tokens_seen": 1543960,
+      "step": 1975
+    },
+    {
+      "epoch": 4.116424116424117,
+      "grad_norm": 0.8724983334541321,
+      "learning_rate": 4.969862525216482e-05,
+      "loss": 0.0388,
+      "num_input_tokens_seen": 1547800,
+      "step": 1980
+    },
+    {
+      "epoch": 4.126819126819127,
+      "grad_norm": 1.9951585531234741,
+      "learning_rate": 4.9697103551062556e-05,
+      "loss": 0.0737,
+      "num_input_tokens_seen": 1551608,
+      "step": 1985
+    },
+    {
+      "epoch": 4.137214137214137,
+      "grad_norm": 2.2757885456085205,
+      "learning_rate": 4.9695578041356565e-05,
+      "loss": 0.0923,
+      "num_input_tokens_seen": 1555512,
+      "step": 1990
+    },
+    {
+      "epoch": 4.147609147609147,
+      "grad_norm": 1.2507476806640625,
+      "learning_rate": 4.969404872328209e-05,
+      "loss": 0.0601,
+      "num_input_tokens_seen": 1559448,
+      "step": 1995
+    },
+    {
+      "epoch": 4.158004158004158,
+      "grad_norm": 0.7527347803115845,
+      "learning_rate": 4.969251559707498e-05,
+      "loss": 0.0569,
+      "num_input_tokens_seen": 1563512,
+      "step": 2000
+    },
+    {
+      "epoch": 4.158004158004158,
+      "eval_loss": 0.11810138821601868,
+      "eval_runtime": 11.7447,
+      "eval_samples_per_second": 72.884,
+      "eval_steps_per_second": 18.221,
+      "num_input_tokens_seen": 1563512,
+      "step": 2000
+    },
+    {
+      "epoch": 4.168399168399168,
+      "grad_norm": 4.866814136505127,
+      "learning_rate": 4.9690978662971674e-05,
+      "loss": 0.07,
+      "num_input_tokens_seen": 1567384,
+      "step": 2005
+    },
+    {
+      "epoch": 4.1787941787941785,
+      "grad_norm": 0.218743234872818,
+      "learning_rate": 4.968943792120916e-05,
+      "loss": 0.0402,
+      "num_input_tokens_seen": 1571384,
+      "step": 2010
+    },
+    {
+      "epoch": 4.1891891891891895,
+      "grad_norm": 1.4916744232177734,
+      "learning_rate": 4.9687893372025046e-05,
+      "loss": 0.0289,
+      "num_input_tokens_seen": 1575160,
+      "step": 2015
+    },
+    {
+      "epoch": 4.1995841995842,
+      "grad_norm": 1.104396104812622,
+      "learning_rate": 4.9686345015657535e-05,
+      "loss": 0.0515,
+      "num_input_tokens_seen": 1579224,
+      "step": 2020
+    },
+    {
+      "epoch": 4.20997920997921,
+      "grad_norm": 3.1168956756591797,
+      "learning_rate": 4.968479285234538e-05,
+      "loss": 0.0491,
+      "num_input_tokens_seen": 1583128,
+      "step": 2025
+    },
+    {
+      "epoch": 4.220374220374221,
+      "grad_norm": 3.9578585624694824,
+      "learning_rate": 4.9683236882327974e-05,
+      "loss": 0.0465,
+      "num_input_tokens_seen": 1587000,
+      "step": 2030
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 1.0281132459640503,
+      "learning_rate": 4.968167710584526e-05,
+      "loss": 0.0443,
+      "num_input_tokens_seen": 1590808,
+      "step": 2035
+    },
+    {
+      "epoch": 4.241164241164241,
+      "grad_norm": 0.7010374069213867,
+      "learning_rate": 4.968011352313775e-05,
+      "loss": 0.0171,
+      "num_input_tokens_seen": 1594712,
+      "step": 2040
+    },
+    {
+      "epoch": 4.251559251559252,
+      "grad_norm": 4.028834819793701,
+      "learning_rate": 4.967854613444659e-05,
+      "loss": 0.0198,
+      "num_input_tokens_seen": 1598584,
+      "step": 2045
+    },
+    {
+      "epoch": 4.261954261954262,
+      "grad_norm": 2.7524290084838867,
+      "learning_rate": 4.967697494001349e-05,
+      "loss": 0.0587,
+      "num_input_tokens_seen": 1602520,
+      "step": 2050
+    },
+    {
+      "epoch": 4.272349272349272,
+      "grad_norm": 1.0941822528839111,
+      "learning_rate": 4.9675399940080736e-05,
+      "loss": 0.0731,
+      "num_input_tokens_seen": 1606264,
+      "step": 2055
+    },
+    {
+      "epoch": 4.282744282744282,
+      "grad_norm": 3.586545705795288,
+      "learning_rate": 4.9673821134891226e-05,
+      "loss": 0.0415,
+      "num_input_tokens_seen": 1610136,
+      "step": 2060
+    },
+    {
+      "epoch": 4.293139293139293,
+      "grad_norm": 2.5060245990753174,
+      "learning_rate": 4.967223852468842e-05,
+      "loss": 0.0571,
+      "num_input_tokens_seen": 1613912,
+      "step": 2065
+    },
+    {
+      "epoch": 4.303534303534303,
+      "grad_norm": 2.5483553409576416,
+      "learning_rate": 4.967065210971639e-05,
+      "loss": 0.0369,
+      "num_input_tokens_seen": 1617784,
+      "step": 2070
+    },
+    {
+      "epoch": 4.313929313929314,
+      "grad_norm": 1.8607265949249268,
+      "learning_rate": 4.966906189021977e-05,
+      "loss": 0.0575,
+      "num_input_tokens_seen": 1621688,
+      "step": 2075
+    },
+    {
+      "epoch": 4.324324324324325,
+      "grad_norm": 2.057943105697632,
+      "learning_rate": 4.966746786644379e-05,
+      "loss": 0.0243,
+      "num_input_tokens_seen": 1625560,
+      "step": 2080
+    },
+    {
+      "epoch": 4.334719334719335,
+      "grad_norm": 2.101684808731079,
+      "learning_rate": 4.966587003863429e-05,
+      "loss": 0.0489,
+      "num_input_tokens_seen": 1629432,
+      "step": 2085
+    },
+    {
+      "epoch": 4.345114345114345,
+      "grad_norm": 3.4833898544311523,
+      "learning_rate": 4.966426840703765e-05,
+      "loss": 0.0706,
+      "num_input_tokens_seen": 1633208,
+      "step": 2090
+    },
+    {
+      "epoch": 4.355509355509356,
+      "grad_norm": 0.8471901416778564,
+      "learning_rate": 4.9662662971900875e-05,
+      "loss": 0.0837,
+      "num_input_tokens_seen": 1637048,
+      "step": 2095
+    },
+    {
+      "epoch": 4.365904365904366,
+      "grad_norm": 0.24271565675735474,
+      "learning_rate": 4.9661053733471534e-05,
+      "loss": 0.05,
+      "num_input_tokens_seen": 1640824,
+      "step": 2100
+    },
+    {
+      "epoch": 4.376299376299376,
+      "grad_norm": 0.7736706733703613,
+      "learning_rate": 4.965944069199781e-05,
+      "loss": 0.0511,
+      "num_input_tokens_seen": 1644792,
+      "step": 2105
+    },
+    {
+      "epoch": 4.386694386694387,
+      "grad_norm": 0.3205808401107788,
+      "learning_rate": 4.965782384772842e-05,
+      "loss": 0.0662,
+      "num_input_tokens_seen": 1648824,
+      "step": 2110
+    },
+    {
+      "epoch": 4.397089397089397,
+      "grad_norm": 1.7462029457092285,
+      "learning_rate": 4.9656203200912734e-05,
+      "loss": 0.0983,
+      "num_input_tokens_seen": 1652664,
+      "step": 2115
+    },
+    {
+      "epoch": 4.407484407484407,
+      "grad_norm": 1.9608981609344482,
+      "learning_rate": 4.965457875180067e-05,
+      "loss": 0.0552,
+      "num_input_tokens_seen": 1656536,
+      "step": 2120
+    },
+    {
+      "epoch": 4.417879417879418,
+      "grad_norm": 0.42880797386169434,
+      "learning_rate": 4.9652950500642724e-05,
+      "loss": 0.0358,
+      "num_input_tokens_seen": 1660408,
+      "step": 2125
+    },
+    {
+      "epoch": 4.428274428274428,
+      "grad_norm": 1.0406259298324585,
+      "learning_rate": 4.965131844769001e-05,
+      "loss": 0.0342,
+      "num_input_tokens_seen": 1664376,
+      "step": 2130
+    },
+    {
+      "epoch": 4.4386694386694385,
+      "grad_norm": 0.9198408126831055,
+      "learning_rate": 4.96496825931942e-05,
+      "loss": 0.0314,
+      "num_input_tokens_seen": 1668344,
+      "step": 2135
+    },
+    {
+      "epoch": 4.4490644490644495,
+      "grad_norm": 1.3925645351409912,
+      "learning_rate": 4.9648042937407566e-05,
+      "loss": 0.0374,
+      "num_input_tokens_seen": 1672248,
+      "step": 2140
+    },
+    {
+      "epoch": 4.45945945945946,
+      "grad_norm": 2.7168407440185547,
+      "learning_rate": 4.964639948058297e-05,
+      "loss": 0.0672,
+      "num_input_tokens_seen": 1676248,
+      "step": 2145
+    },
+    {
+      "epoch": 4.46985446985447,
+      "grad_norm": 1.549904704093933,
+      "learning_rate": 4.9644752222973846e-05,
+      "loss": 0.0728,
+      "num_input_tokens_seen": 1680216,
+      "step": 2150
+    },
+    {
+      "epoch": 4.48024948024948,
+      "grad_norm": 1.85928475856781,
+      "learning_rate": 4.964310116483422e-05,
+      "loss": 0.0648,
+      "num_input_tokens_seen": 1684152,
+      "step": 2155
+    },
+    {
+      "epoch": 4.490644490644491,
+      "grad_norm": 1.6178593635559082,
+      "learning_rate": 4.964144630641872e-05,
+      "loss": 0.0466,
+      "num_input_tokens_seen": 1688056,
+      "step": 2160
+    },
+    {
+      "epoch": 4.501039501039501,
+      "grad_norm": 3.4581243991851807,
+      "learning_rate": 4.9639787647982525e-05,
+      "loss": 0.0837,
+      "num_input_tokens_seen": 1691992,
+      "step": 2165
+    },
+    {
+      "epoch": 4.511434511434511,
+      "grad_norm": 2.093769073486328,
+      "learning_rate": 4.963812518978143e-05,
+      "loss": 0.0892,
+      "num_input_tokens_seen": 1695896,
+      "step": 2170
+    },
+    {
+      "epoch": 4.521829521829522,
+      "grad_norm": 0.5826231837272644,
+      "learning_rate": 4.963645893207182e-05,
+      "loss": 0.0365,
+      "num_input_tokens_seen": 1699736,
+      "step": 2175
+    },
+    {
+      "epoch": 4.532224532224532,
+      "grad_norm": 2.888047933578491,
+      "learning_rate": 4.963478887511063e-05,
+      "loss": 0.0646,
+      "num_input_tokens_seen": 1703768,
+      "step": 2180
+    },
+    {
+      "epoch": 4.542619542619542,
+      "grad_norm": 3.088243007659912,
+      "learning_rate": 4.963311501915542e-05,
+      "loss": 0.0298,
+      "num_input_tokens_seen": 1707576,
+      "step": 2185
+    },
+    {
+      "epoch": 4.553014553014553,
+      "grad_norm": 1.7297030687332153,
+      "learning_rate": 4.963143736446432e-05,
+      "loss": 0.0516,
+      "num_input_tokens_seen": 1711416,
+      "step": 2190
+    },
+    {
+      "epoch": 4.5634095634095635,
+      "grad_norm": 3.3710927963256836,
+      "learning_rate": 4.962975591129603e-05,
+      "loss": 0.0995,
+      "num_input_tokens_seen": 1715224,
+      "step": 2195
+    },
+    {
+      "epoch": 4.573804573804574,
+      "grad_norm": 2.557830333709717,
+      "learning_rate": 4.962807065990986e-05,
+      "loss": 0.0738,
+      "num_input_tokens_seen": 1719064,
+      "step": 2200
+    },
+    {
+      "epoch": 4.573804573804574,
+      "eval_loss": 0.13142085075378418,
+      "eval_runtime": 11.736,
+      "eval_samples_per_second": 72.938,
+      "eval_steps_per_second": 18.235,
+      "num_input_tokens_seen": 1719064,
+      "step": 2200
+    },
+    {
+      "epoch": 4.584199584199585,
+      "grad_norm": 2.711688280105591,
+      "learning_rate": 4.9626381610565714e-05,
+      "loss": 0.0796,
+      "num_input_tokens_seen": 1723000,
+      "step": 2205
+    },
+    {
+      "epoch": 4.594594594594595,
+      "grad_norm": 1.23819899559021,
+      "learning_rate": 4.9624688763524043e-05,
+      "loss": 0.0516,
+      "num_input_tokens_seen": 1726936,
+      "step": 2210
+    },
+    {
+      "epoch": 4.604989604989605,
+      "grad_norm": 2.1595842838287354,
+      "learning_rate": 4.962299211904591e-05,
+      "loss": 0.0357,
+      "num_input_tokens_seen": 1730808,
+      "step": 2215
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 3.259979486465454,
+      "learning_rate": 4.962129167739296e-05,
+      "loss": 0.0306,
+      "num_input_tokens_seen": 1734616,
+      "step": 2220
+    },
+    {
+      "epoch": 4.625779625779626,
+      "grad_norm": 0.6946565508842468,
+      "learning_rate": 4.961958743882742e-05,
+      "loss": 0.0821,
+      "num_input_tokens_seen": 1738360,
+      "step": 2225
+    },
+    {
+      "epoch": 4.636174636174636,
+      "grad_norm": 2.5261144638061523,
+      "learning_rate": 4.961787940361211e-05,
+      "loss": 0.1022,
+      "num_input_tokens_seen": 1742136,
+      "step": 2230
+    },
+    {
+      "epoch": 4.646569646569646,
+      "grad_norm": 0.5816161036491394,
+      "learning_rate": 4.961616757201043e-05,
+      "loss": 0.051,
+      "num_input_tokens_seen": 1746168,
+      "step": 2235
+    },
+    {
+      "epoch": 4.656964656964657,
+      "grad_norm": 1.3163511753082275,
+      "learning_rate": 4.961445194428637e-05,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 1749912,
+      "step": 2240
+    },
+    {
+      "epoch": 4.667359667359667,
+      "grad_norm": 2.1269617080688477,
+      "learning_rate": 4.9612732520704486e-05,
+      "loss": 0.0726,
+      "num_input_tokens_seen": 1753784,
+      "step": 2245
+    },
+    {
+      "epoch": 4.6777546777546775,
+      "grad_norm": 1.235656976699829,
+      "learning_rate": 4.961100930152994e-05,
+      "loss": 0.0372,
+      "num_input_tokens_seen": 1757752,
+      "step": 2250
+    },
+    {
+      "epoch": 4.6881496881496885,
+      "grad_norm": 1.5069435834884644,
+      "learning_rate": 4.960928228702849e-05,
+      "loss": 0.0833,
+      "num_input_tokens_seen": 1761752,
+      "step": 2255
+    },
+    {
+      "epoch": 4.698544698544699,
+      "grad_norm": 3.161324977874756,
+      "learning_rate": 4.960755147746645e-05,
+      "loss": 0.061,
+      "num_input_tokens_seen": 1765688,
+      "step": 2260
+    },
+    {
+      "epoch": 4.708939708939709,
+      "grad_norm": 1.693697214126587,
+      "learning_rate": 4.9605816873110736e-05,
+      "loss": 0.0423,
+      "num_input_tokens_seen": 1769720,
+      "step": 2265
+    },
+    {
+      "epoch": 4.71933471933472,
+      "grad_norm": 1.6283940076828003,
+      "learning_rate": 4.960407847422883e-05,
+      "loss": 0.0506,
+      "num_input_tokens_seen": 1773720,
+      "step": 2270
+    },
+    {
+      "epoch": 4.72972972972973,
+      "grad_norm": 2.0996880531311035,
+      "learning_rate": 4.960233628108885e-05,
+      "loss": 0.1257,
+      "num_input_tokens_seen": 1777496,
+      "step": 2275
+    },
+    {
+      "epoch": 4.74012474012474,
+      "grad_norm": 1.4215102195739746,
+      "learning_rate": 4.960059029395942e-05,
+      "loss": 0.0645,
+      "num_input_tokens_seen": 1781496,
+      "step": 2280
+    },
+    {
+      "epoch": 4.75051975051975,
+      "grad_norm": 1.9666770696640015,
+      "learning_rate": 4.959884051310983e-05,
+      "loss": 0.0685,
+      "num_input_tokens_seen": 1785368,
+      "step": 2285
+    },
+    {
+      "epoch": 4.760914760914761,
+      "grad_norm": 3.107551336288452,
+      "learning_rate": 4.959708693880991e-05,
+      "loss": 0.0565,
+      "num_input_tokens_seen": 1789272,
+      "step": 2290
+    },
+    {
+      "epoch": 4.771309771309771,
+      "grad_norm": 2.7823517322540283,
+      "learning_rate": 4.9595329571330074e-05,
+      "loss": 0.0509,
+      "num_input_tokens_seen": 1793304,
+      "step": 2295
+    },
+    {
+      "epoch": 4.781704781704782,
+      "grad_norm": 1.9672950506210327,
+      "learning_rate": 4.9593568410941326e-05,
+      "loss": 0.0309,
+      "num_input_tokens_seen": 1797304,
+      "step": 2300
+    },
+    {
+      "epoch": 4.792099792099792,
+      "grad_norm": 1.538166880607605,
+      "learning_rate": 4.959180345791528e-05,
+      "loss": 0.0482,
+      "num_input_tokens_seen": 1801400,
+      "step": 2305
+    },
+    {
+      "epoch": 4.802494802494802,
+      "grad_norm": 2.5043864250183105,
+      "learning_rate": 4.9590034712524086e-05,
+      "loss": 0.0497,
+      "num_input_tokens_seen": 1805336,
+      "step": 2310
+    },
+    {
+      "epoch": 4.8128898128898125,
+      "grad_norm": 1.6404062509536743,
+      "learning_rate": 4.958826217504053e-05,
+      "loss": 0.0273,
+      "num_input_tokens_seen": 1809272,
+      "step": 2315
+    },
+    {
+      "epoch": 4.8232848232848236,
+      "grad_norm": 1.53585684299469,
+      "learning_rate": 4.958648584573795e-05,
+      "loss": 0.028,
+      "num_input_tokens_seen": 1813272,
+      "step": 2320
+    },
+    {
+      "epoch": 4.833679833679834,
+      "grad_norm": 5.274715423583984,
+      "learning_rate": 4.958470572489028e-05,
+      "loss": 0.0883,
+      "num_input_tokens_seen": 1817176,
+      "step": 2325
+    },
+    {
+      "epoch": 4.844074844074844,
+      "grad_norm": 3.1148996353149414,
+      "learning_rate": 4.958292181277203e-05,
+      "loss": 0.1076,
+      "num_input_tokens_seen": 1821144,
+      "step": 2330
+    },
+    {
+      "epoch": 4.854469854469855,
+      "grad_norm": 0.7819375395774841,
+      "learning_rate": 4.958113410965832e-05,
+      "loss": 0.0608,
+      "num_input_tokens_seen": 1825016,
+      "step": 2335
+    },
+    {
+      "epoch": 4.864864864864865,
+      "grad_norm": 0.5736908912658691,
+      "learning_rate": 4.957934261582481e-05,
+      "loss": 0.0248,
+      "num_input_tokens_seen": 1828792,
+      "step": 2340
+    },
+    {
+      "epoch": 4.875259875259875,
+      "grad_norm": 2.000171184539795,
+      "learning_rate": 4.95775473315478e-05,
+      "loss": 0.0317,
+      "num_input_tokens_seen": 1832696,
+      "step": 2345
+    },
+    {
+      "epoch": 4.885654885654886,
+      "grad_norm": 3.2827367782592773,
+      "learning_rate": 4.9575748257104124e-05,
+      "loss": 0.0497,
+      "num_input_tokens_seen": 1836472,
+      "step": 2350
+    },
+    {
+      "epoch": 4.896049896049896,
+      "grad_norm": 0.2923671305179596,
+      "learning_rate": 4.9573945392771224e-05,
+      "loss": 0.036,
+      "num_input_tokens_seen": 1840440,
+      "step": 2355
+    },
+    {
+      "epoch": 4.906444906444906,
+      "grad_norm": 1.1491643190383911,
+      "learning_rate": 4.9572138738827134e-05,
+      "loss": 0.0707,
+      "num_input_tokens_seen": 1844376,
+      "step": 2360
+    },
+    {
+      "epoch": 4.916839916839917,
+      "grad_norm": 1.9995085000991821,
+      "learning_rate": 4.957032829555046e-05,
+      "loss": 0.0408,
+      "num_input_tokens_seen": 1848280,
+      "step": 2365
+    },
+    {
+      "epoch": 4.927234927234927,
+      "grad_norm": 2.202824115753174,
+      "learning_rate": 4.956851406322039e-05,
+      "loss": 0.0802,
+      "num_input_tokens_seen": 1852216,
+      "step": 2370
+    },
+    {
+      "epoch": 4.9376299376299375,
+      "grad_norm": 2.6806259155273438,
+      "learning_rate": 4.9566696042116704e-05,
+      "loss": 0.0585,
+      "num_input_tokens_seen": 1856088,
+      "step": 2375
+    },
+    {
+      "epoch": 4.948024948024948,
+      "grad_norm": 1.9189386367797852,
+      "learning_rate": 4.9564874232519766e-05,
+      "loss": 0.0278,
+      "num_input_tokens_seen": 1859896,
+      "step": 2380
+    },
+    {
+      "epoch": 4.958419958419959,
+      "grad_norm": 2.128568649291992,
+      "learning_rate": 4.9563048634710516e-05,
+      "loss": 0.0236,
+      "num_input_tokens_seen": 1863800,
+      "step": 2385
+    },
+    {
+      "epoch": 4.968814968814969,
+      "grad_norm": 2.9501397609710693,
+      "learning_rate": 4.956121924897049e-05,
+      "loss": 0.027,
+      "num_input_tokens_seen": 1867640,
+      "step": 2390
+    },
+    {
+      "epoch": 4.979209979209979,
+      "grad_norm": 3.494018077850342,
+      "learning_rate": 4.955938607558181e-05,
+      "loss": 0.0695,
+      "num_input_tokens_seen": 1871512,
+      "step": 2395
+    },
+    {
+      "epoch": 4.98960498960499,
+      "grad_norm": 1.0756707191467285,
+      "learning_rate": 4.955754911482715e-05,
+      "loss": 0.0523,
+      "num_input_tokens_seen": 1875384,
+      "step": 2400
+    },
+    {
+      "epoch": 4.98960498960499,
+      "eval_loss": 0.15491805970668793,
+      "eval_runtime": 11.7272,
+      "eval_samples_per_second": 72.993,
+      "eval_steps_per_second": 18.248,
+      "num_input_tokens_seen": 1875384,
+      "step": 2400
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.2396276295185089,
+      "learning_rate": 4.9555708366989804e-05,
+      "loss": 0.0632,
+      "num_input_tokens_seen": 1879248,
+      "step": 2405
+    },
+    {
+      "epoch": 5.01039501039501,
+      "grad_norm": 1.227403163909912,
+      "learning_rate": 4.9553863832353655e-05,
+      "loss": 0.0419,
+      "num_input_tokens_seen": 1883248,
+      "step": 2410
+    },
+    {
+      "epoch": 5.020790020790021,
+      "grad_norm": 2.9086012840270996,
+      "learning_rate": 4.955201551120313e-05,
+      "loss": 0.0234,
+      "num_input_tokens_seen": 1887120,
+      "step": 2415
+    },
+    {
+      "epoch": 5.031185031185031,
+      "grad_norm": 3.0317583084106445,
+      "learning_rate": 4.955016340382328e-05,
+      "loss": 0.0584,
+      "num_input_tokens_seen": 1891056,
+      "step": 2420
+    },
+    {
+      "epoch": 5.041580041580041,
+      "grad_norm": 0.8113784193992615,
+      "learning_rate": 4.954830751049972e-05,
+      "loss": 0.0301,
+      "num_input_tokens_seen": 1895024,
+      "step": 2425
+    },
+    {
+      "epoch": 5.051975051975052,
+      "grad_norm": 0.6022761464118958,
+      "learning_rate": 4.954644783151864e-05,
+      "loss": 0.0329,
+      "num_input_tokens_seen": 1899152,
+      "step": 2430
+    },
+    {
+      "epoch": 5.0623700623700625,
+      "grad_norm": 1.7706804275512695,
+      "learning_rate": 4.954458436716684e-05,
+      "loss": 0.0201,
+      "num_input_tokens_seen": 1903152,
+      "step": 2435
+    },
+    {
+      "epoch": 5.072765072765073,
+      "grad_norm": 1.0601933002471924,
+      "learning_rate": 4.954271711773168e-05,
+      "loss": 0.046,
+      "num_input_tokens_seen": 1907088,
+      "step": 2440
+    },
+    {
+      "epoch": 5.083160083160083,
+      "grad_norm": 1.2593992948532104,
+      "learning_rate": 4.9540846083501115e-05,
+      "loss": 0.0333,
+      "num_input_tokens_seen": 1910992,
+      "step": 2445
+    },
+    {
+      "epoch": 5.093555093555094,
+      "grad_norm": 4.84083366394043,
+      "learning_rate": 4.953897126476369e-05,
+      "loss": 0.045,
+      "num_input_tokens_seen": 1914864,
+      "step": 2450
+    },
+    {
+      "epoch": 5.103950103950104,
+      "grad_norm": 0.2864864766597748,
+      "learning_rate": 4.9537092661808514e-05,
+      "loss": 0.0406,
+      "num_input_tokens_seen": 1918640,
+      "step": 2455
+    },
+    {
+      "epoch": 5.114345114345114,
+      "grad_norm": 4.088659763336182,
+      "learning_rate": 4.9535210274925306e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 1922608,
+      "step": 2460
+    },
+    {
+      "epoch": 5.124740124740125,
+      "grad_norm": 0.11531715095043182,
+      "learning_rate": 4.953332410440435e-05,
+      "loss": 0.0391,
+      "num_input_tokens_seen": 1926608,
+      "step": 2465
+    },
+    {
+      "epoch": 5.135135135135135,
+      "grad_norm": 1.648222804069519,
+      "learning_rate": 4.9531434150536496e-05,
+      "loss": 0.0395,
+      "num_input_tokens_seen": 1930512,
+      "step": 2470
+    },
+    {
+      "epoch": 5.145530145530145,
+      "grad_norm": 2.2682533264160156,
+      "learning_rate": 4.952954041361322e-05,
+      "loss": 0.0302,
+      "num_input_tokens_seen": 1934416,
+      "step": 2475
+    },
+    {
+      "epoch": 5.155925155925156,
+      "grad_norm": 0.4399753212928772,
+      "learning_rate": 4.952764289392655e-05,
+      "loss": 0.0367,
+      "num_input_tokens_seen": 1938352,
+      "step": 2480
+    },
+    {
+      "epoch": 5.166320166320166,
+      "grad_norm": 0.3311882019042969,
+      "learning_rate": 4.952574159176912e-05,
+      "loss": 0.008,
+      "num_input_tokens_seen": 1942192,
+      "step": 2485
+    },
+    {
+      "epoch": 5.1767151767151764,
+      "grad_norm": 1.0251822471618652,
+      "learning_rate": 4.952383650743413e-05,
+      "loss": 0.0162,
+      "num_input_tokens_seen": 1946192,
+      "step": 2490
+    },
+    {
+      "epoch": 5.1871101871101875,
+      "grad_norm": 4.446981430053711,
+      "learning_rate": 4.952192764121536e-05,
+      "loss": 0.0663,
+      "num_input_tokens_seen": 1950000,
+      "step": 2495
+    },
+    {
+      "epoch": 5.197505197505198,
+      "grad_norm": 1.2716466188430786,
+      "learning_rate": 4.9520014993407185e-05,
+      "loss": 0.025,
+      "num_input_tokens_seen": 1953808,
+      "step": 2500
+    },
+    {
+      "epoch": 5.207900207900208,
+      "grad_norm": 1.912648320198059,
+      "learning_rate": 4.951809856430456e-05,
+      "loss": 0.033,
+      "num_input_tokens_seen": 1957680,
+      "step": 2505
+    },
+    {
+      "epoch": 5.218295218295219,
+      "grad_norm": 0.6633192896842957,
+      "learning_rate": 4.951617835420303e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 1961552,
+      "step": 2510
+    },
+    {
+      "epoch": 5.228690228690229,
+      "grad_norm": 0.8859670758247375,
+      "learning_rate": 4.951425436339869e-05,
+      "loss": 0.0456,
+      "num_input_tokens_seen": 1965456,
+      "step": 2515
+    },
+    {
+      "epoch": 5.239085239085239,
+      "grad_norm": 1.9564597606658936,
+      "learning_rate": 4.9512326592188274e-05,
+      "loss": 0.0649,
+      "num_input_tokens_seen": 1969232,
+      "step": 2520
+    },
+    {
+      "epoch": 5.24948024948025,
+      "grad_norm": 4.266429901123047,
+      "learning_rate": 4.9510395040869054e-05,
+      "loss": 0.0304,
+      "num_input_tokens_seen": 1973040,
+      "step": 2525
+    },
+    {
+      "epoch": 5.25987525987526,
+      "grad_norm": 3.4240670204162598,
+      "learning_rate": 4.9508459709738905e-05,
+      "loss": 0.048,
+      "num_input_tokens_seen": 1976944,
+      "step": 2530
+    },
+    {
+      "epoch": 5.27027027027027,
+      "grad_norm": 1.707228183746338,
+      "learning_rate": 4.950652059909627e-05,
+      "loss": 0.0444,
+      "num_input_tokens_seen": 1980848,
+      "step": 2535
+    },
+    {
+      "epoch": 5.28066528066528,
+      "grad_norm": 0.08145427703857422,
+      "learning_rate": 4.95045777092402e-05,
+      "loss": 0.0487,
+      "num_input_tokens_seen": 1984752,
+      "step": 2540
+    },
+    {
+      "epoch": 5.291060291060291,
+      "grad_norm": 7.088893413543701,
+      "learning_rate": 4.950263104047031e-05,
+      "loss": 0.0726,
+      "num_input_tokens_seen": 1988592,
+      "step": 2545
+    },
+    {
+      "epoch": 5.301455301455301,
+      "grad_norm": 1.9855923652648926,
+      "learning_rate": 4.9500680593086775e-05,
+      "loss": 0.0633,
+      "num_input_tokens_seen": 1992656,
+      "step": 2550
+    },
+    {
+      "epoch": 5.3118503118503115,
+      "grad_norm": 0.3952619135379791,
+      "learning_rate": 4.94987263673904e-05,
+      "loss": 0.0476,
+      "num_input_tokens_seen": 1996784,
+      "step": 2555
+    },
+    {
+      "epoch": 5.3222453222453225,
+      "grad_norm": 7.177328109741211,
+      "learning_rate": 4.949676836368256e-05,
+      "loss": 0.0417,
+      "num_input_tokens_seen": 2000592,
+      "step": 2560
+    },
+    {
+      "epoch": 5.332640332640333,
+      "grad_norm": 0.17644605040550232,
+      "learning_rate": 4.949480658226518e-05,
+      "loss": 0.0484,
+      "num_input_tokens_seen": 2004464,
+      "step": 2565
+    },
+    {
+      "epoch": 5.343035343035343,
+      "grad_norm": 0.43695083260536194,
+      "learning_rate": 4.949284102344082e-05,
+      "loss": 0.0218,
+      "num_input_tokens_seen": 2008368,
+      "step": 2570
+    },
+    {
+      "epoch": 5.353430353430354,
+      "grad_norm": 0.747542142868042,
+      "learning_rate": 4.9490871687512565e-05,
+      "loss": 0.0226,
+      "num_input_tokens_seen": 2012208,
+      "step": 2575
+    },
+    {
+      "epoch": 5.363825363825364,
+      "grad_norm": 6.063629627227783,
+      "learning_rate": 4.948889857478413e-05,
+      "loss": 0.0386,
+      "num_input_tokens_seen": 2015952,
+      "step": 2580
+    },
+    {
+      "epoch": 5.374220374220374,
+      "grad_norm": 1.7891842126846313,
+      "learning_rate": 4.948692168555978e-05,
+      "loss": 0.0311,
+      "num_input_tokens_seen": 2019856,
+      "step": 2585
+    },
+    {
+      "epoch": 5.384615384615385,
+      "grad_norm": 0.41442546248435974,
+      "learning_rate": 4.94849410201444e-05,
+      "loss": 0.0313,
+      "num_input_tokens_seen": 2023792,
+      "step": 2590
+    },
+    {
+      "epoch": 5.395010395010395,
+      "grad_norm": 4.142585754394531,
+      "learning_rate": 4.948295657884341e-05,
+      "loss": 0.0688,
+      "num_input_tokens_seen": 2027696,
+      "step": 2595
+    },
+    {
+      "epoch": 5.405405405405405,
+      "grad_norm": 6.4825053215026855,
+      "learning_rate": 4.9480968361962835e-05,
+      "loss": 0.0289,
+      "num_input_tokens_seen": 2031440,
+      "step": 2600
+    },
+    {
+      "epoch": 5.405405405405405,
+      "eval_loss": 0.15861836075782776,
+      "eval_runtime": 11.7337,
+      "eval_samples_per_second": 72.952,
+      "eval_steps_per_second": 18.238,
+      "num_input_tokens_seen": 2031440,
+      "step": 2600
+    },
+    {
+      "epoch": 5.415800415800415,
+      "grad_norm": 6.522375583648682,
+      "learning_rate": 4.9478976369809305e-05,
+      "loss": 0.0315,
+      "num_input_tokens_seen": 2035344,
+      "step": 2605
+    },
+    {
+      "epoch": 5.426195426195426,
+      "grad_norm": 0.16328278183937073,
+      "learning_rate": 4.947698060268999e-05,
+      "loss": 0.0482,
+      "num_input_tokens_seen": 2039344,
+      "step": 2610
+    },
+    {
+      "epoch": 5.4365904365904365,
+      "grad_norm": 1.5238748788833618,
+      "learning_rate": 4.9474981060912665e-05,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 2043408,
+      "step": 2615
+    },
+    {
+      "epoch": 5.446985446985447,
+      "grad_norm": 8.83666706085205,
+      "learning_rate": 4.94729777447857e-05,
+      "loss": 0.0419,
+      "num_input_tokens_seen": 2047472,
+      "step": 2620
+    },
+    {
+      "epoch": 5.457380457380458,
+      "grad_norm": 1.1469340324401855,
+      "learning_rate": 4.947097065461801e-05,
+      "loss": 0.0718,
+      "num_input_tokens_seen": 2051376,
+      "step": 2625
+    },
+    {
+      "epoch": 5.467775467775468,
+      "grad_norm": 0.401373028755188,
+      "learning_rate": 4.9468959790719125e-05,
+      "loss": 0.0164,
+      "num_input_tokens_seen": 2055184,
+      "step": 2630
+    },
+    {
+      "epoch": 5.478170478170478,
+      "grad_norm": 0.7865161895751953,
+      "learning_rate": 4.9466945153399146e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 2059056,
+      "step": 2635
+    },
+    {
+      "epoch": 5.488565488565489,
+      "grad_norm": 1.1036028861999512,
+      "learning_rate": 4.9464926742968755e-05,
+      "loss": 0.0372,
+      "num_input_tokens_seen": 2062928,
+      "step": 2640
+    },
+    {
+      "epoch": 5.498960498960499,
+      "grad_norm": 0.9749707579612732,
+      "learning_rate": 4.946290455973921e-05,
+      "loss": 0.0154,
+      "num_input_tokens_seen": 2066992,
+      "step": 2645
+    },
+    {
+      "epoch": 5.509355509355509,
+      "grad_norm": 0.3660118281841278,
+      "learning_rate": 4.9460878604022365e-05,
+      "loss": 0.0199,
+      "num_input_tokens_seen": 2071056,
+      "step": 2650
+    },
+    {
+      "epoch": 5.51975051975052,
+      "grad_norm": 2.2302656173706055,
+      "learning_rate": 4.945884887613065e-05,
+      "loss": 0.0601,
+      "num_input_tokens_seen": 2074992,
+      "step": 2655
+    },
+    {
+      "epoch": 5.53014553014553,
+      "grad_norm": 1.2271231412887573,
+      "learning_rate": 4.9456815376377055e-05,
+      "loss": 0.03,
+      "num_input_tokens_seen": 2078864,
+      "step": 2660
+    },
+    {
+      "epoch": 5.54054054054054,
+      "grad_norm": 4.041025638580322,
+      "learning_rate": 4.9454778105075195e-05,
+      "loss": 0.0553,
+      "num_input_tokens_seen": 2082736,
+      "step": 2665
+    },
+    {
+      "epoch": 5.5509355509355505,
+      "grad_norm": 2.408539295196533,
+      "learning_rate": 4.945273706253924e-05,
+      "loss": 0.0529,
+      "num_input_tokens_seen": 2086608,
+      "step": 2670
+    },
+    {
+      "epoch": 5.5613305613305615,
+      "grad_norm": 1.8175849914550781,
+      "learning_rate": 4.9450692249083925e-05,
+      "loss": 0.0498,
+      "num_input_tokens_seen": 2090704,
+      "step": 2675
+    },
+    {
+      "epoch": 5.571725571725572,
+      "grad_norm": 0.7717705368995667,
+      "learning_rate": 4.9448643665024605e-05,
+      "loss": 0.0583,
+      "num_input_tokens_seen": 2094672,
+      "step": 2680
+    },
+    {
+      "epoch": 5.582120582120583,
+      "grad_norm": 2.8313677310943604,
+      "learning_rate": 4.944659131067719e-05,
+      "loss": 0.0334,
+      "num_input_tokens_seen": 2098576,
+      "step": 2685
+    },
+    {
+      "epoch": 5.592515592515593,
+      "grad_norm": 0.5528305172920227,
+      "learning_rate": 4.944453518635818e-05,
+      "loss": 0.0372,
+      "num_input_tokens_seen": 2102544,
+      "step": 2690
+    },
+    {
+      "epoch": 5.602910602910603,
+      "grad_norm": 0.4642880856990814,
+      "learning_rate": 4.944247529238465e-05,
+      "loss": 0.0474,
+      "num_input_tokens_seen": 2106352,
+      "step": 2695
+    },
+    {
+      "epoch": 5.613305613305613,
+      "grad_norm": 1.076480746269226,
+      "learning_rate": 4.944041162907427e-05,
+      "loss": 0.0541,
+      "num_input_tokens_seen": 2110320,
+      "step": 2700
+    },
+    {
+      "epoch": 5.623700623700624,
+      "grad_norm": 0.9065725803375244,
+      "learning_rate": 4.943834419674529e-05,
+      "loss": 0.0616,
+      "num_input_tokens_seen": 2114128,
+      "step": 2705
+    },
+    {
+      "epoch": 5.634095634095634,
+      "grad_norm": 2.137449264526367,
+      "learning_rate": 4.9436272995716506e-05,
+      "loss": 0.0433,
+      "num_input_tokens_seen": 2117936,
+      "step": 2710
+    },
+    {
+      "epoch": 5.644490644490644,
+      "grad_norm": 1.7382537126541138,
+      "learning_rate": 4.943419802630735e-05,
+      "loss": 0.0756,
+      "num_input_tokens_seen": 2121904,
+      "step": 2715
+    },
+    {
+      "epoch": 5.654885654885655,
+      "grad_norm": 0.1500329077243805,
+      "learning_rate": 4.94321192888378e-05,
+      "loss": 0.0363,
+      "num_input_tokens_seen": 2125808,
+      "step": 2720
+    },
+    {
+      "epoch": 5.665280665280665,
+      "grad_norm": 0.18378840386867523,
+      "learning_rate": 4.943003678362842e-05,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 2129744,
+      "step": 2725
+    },
+    {
+      "epoch": 5.675675675675675,
+      "grad_norm": 2.1144022941589355,
+      "learning_rate": 4.942795051100036e-05,
+      "loss": 0.0447,
+      "num_input_tokens_seen": 2133648,
+      "step": 2730
+    },
+    {
+      "epoch": 5.686070686070686,
+      "grad_norm": 1.4472310543060303,
+      "learning_rate": 4.942586047127536e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 2137584,
+      "step": 2735
+    },
+    {
+      "epoch": 5.696465696465697,
+      "grad_norm": 2.9420251846313477,
+      "learning_rate": 4.942376666477571e-05,
+      "loss": 0.0204,
+      "num_input_tokens_seen": 2141520,
+      "step": 2740
+    },
+    {
+      "epoch": 5.706860706860707,
+      "grad_norm": 6.842442035675049,
+      "learning_rate": 4.9421669091824304e-05,
+      "loss": 0.027,
+      "num_input_tokens_seen": 2145360,
+      "step": 2745
+    },
+    {
+      "epoch": 5.717255717255718,
+      "grad_norm": 4.913285732269287,
+      "learning_rate": 4.9419567752744634e-05,
+      "loss": 0.0493,
+      "num_input_tokens_seen": 2149232,
+      "step": 2750
+    },
+    {
+      "epoch": 5.727650727650728,
+      "grad_norm": 4.254665374755859,
+      "learning_rate": 4.941746264786074e-05,
+      "loss": 0.0336,
+      "num_input_tokens_seen": 2153104,
+      "step": 2755
+    },
+    {
+      "epoch": 5.738045738045738,
+      "grad_norm": 3.5793089866638184,
+      "learning_rate": 4.9415353777497254e-05,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 2156944,
+      "step": 2760
+    },
+    {
+      "epoch": 5.748440748440748,
+      "grad_norm": 0.15350663661956787,
+      "learning_rate": 4.9413241141979394e-05,
+      "loss": 0.018,
+      "num_input_tokens_seen": 2160880,
+      "step": 2765
+    },
+    {
+      "epoch": 5.758835758835759,
+      "grad_norm": 7.260200023651123,
+      "learning_rate": 4.9411124741632956e-05,
+      "loss": 0.0741,
+      "num_input_tokens_seen": 2164688,
+      "step": 2770
+    },
+    {
+      "epoch": 5.769230769230769,
+      "grad_norm": 0.4035978317260742,
+      "learning_rate": 4.940900457678431e-05,
+      "loss": 0.0156,
+      "num_input_tokens_seen": 2168624,
+      "step": 2775
+    },
+    {
+      "epoch": 5.779625779625779,
+      "grad_norm": 1.878625512123108,
+      "learning_rate": 4.9406880647760425e-05,
+      "loss": 0.058,
+      "num_input_tokens_seen": 2172432,
+      "step": 2780
+    },
+    {
+      "epoch": 5.79002079002079,
+      "grad_norm": 1.7479143142700195,
+      "learning_rate": 4.9404752954888824e-05,
+      "loss": 0.0155,
+      "num_input_tokens_seen": 2176240,
+      "step": 2785
+    },
+    {
+      "epoch": 5.8004158004158,
+      "grad_norm": 0.2874299883842468,
+      "learning_rate": 4.940262149849762e-05,
+      "loss": 0.0313,
+      "num_input_tokens_seen": 2180112,
+      "step": 2790
+    },
+    {
+      "epoch": 5.8108108108108105,
+      "grad_norm": 0.8156207203865051,
+      "learning_rate": 4.9400486278915526e-05,
+      "loss": 0.0347,
+      "num_input_tokens_seen": 2184080,
+      "step": 2795
+    },
+    {
+      "epoch": 5.8212058212058215,
+      "grad_norm": 0.40522682666778564,
+      "learning_rate": 4.939834729647181e-05,
+      "loss": 0.0208,
+      "num_input_tokens_seen": 2187952,
+      "step": 2800
+    },
+    {
+      "epoch": 5.8212058212058215,
+      "eval_loss": 0.16124659776687622,
+      "eval_runtime": 11.7412,
+      "eval_samples_per_second": 72.905,
+      "eval_steps_per_second": 18.226,
+      "num_input_tokens_seen": 2187952,
+      "step": 2800
+    },
+    {
+      "epoch": 5.831600831600832,
+      "grad_norm": 2.628782272338867,
+      "learning_rate": 4.9396204551496326e-05,
+      "loss": 0.0082,
+      "num_input_tokens_seen": 2191792,
+      "step": 2805
+    },
+    {
+      "epoch": 5.841995841995842,
+      "grad_norm": 0.06309270113706589,
+      "learning_rate": 4.939405804431952e-05,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 2195696,
+      "step": 2810
+    },
+    {
+      "epoch": 5.852390852390853,
+      "grad_norm": 3.3200418949127197,
+      "learning_rate": 4.9391907775272414e-05,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 2199504,
+      "step": 2815
+    },
+    {
+      "epoch": 5.862785862785863,
+      "grad_norm": 3.297598123550415,
+      "learning_rate": 4.9389753744686604e-05,
+      "loss": 0.028,
+      "num_input_tokens_seen": 2203312,
+      "step": 2820
+    },
+    {
+      "epoch": 5.873180873180873,
+      "grad_norm": 1.8152594566345215,
+      "learning_rate": 4.938759595289426e-05,
+      "loss": 0.0317,
+      "num_input_tokens_seen": 2207216,
+      "step": 2825
+    },
+    {
+      "epoch": 5.883575883575883,
+      "grad_norm": 0.030661484226584435,
+      "learning_rate": 4.938543440022815e-05,
+      "loss": 0.0792,
+      "num_input_tokens_seen": 2210960,
+      "step": 2830
+    },
+    {
+      "epoch": 5.893970893970894,
+      "grad_norm": 0.08718180656433105,
+      "learning_rate": 4.938326908702161e-05,
+      "loss": 0.0199,
+      "num_input_tokens_seen": 2214896,
+      "step": 2835
+    },
+    {
+      "epoch": 5.904365904365904,
+      "grad_norm": 0.891739010810852,
+      "learning_rate": 4.9381100013608554e-05,
+      "loss": 0.0393,
+      "num_input_tokens_seen": 2219024,
+      "step": 2840
+    },
+    {
+      "epoch": 5.914760914760915,
+      "grad_norm": 2.6833608150482178,
+      "learning_rate": 4.9378927180323485e-05,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 2222864,
+      "step": 2845
+    },
+    {
+      "epoch": 5.925155925155925,
+      "grad_norm": 0.4111284017562866,
+      "learning_rate": 4.937675058750148e-05,
+      "loss": 0.0387,
+      "num_input_tokens_seen": 2226800,
+      "step": 2850
+    },
+    {
+      "epoch": 5.9355509355509355,
+      "grad_norm": 0.19528307020664215,
+      "learning_rate": 4.937457023547819e-05,
+      "loss": 0.0504,
+      "num_input_tokens_seen": 2230704,
+      "step": 2855
+    },
+    {
+      "epoch": 5.945945945945946,
+      "grad_norm": 1.0128448009490967,
+      "learning_rate": 4.9372386124589876e-05,
+      "loss": 0.0863,
+      "num_input_tokens_seen": 2234672,
+      "step": 2860
+    },
+    {
+      "epoch": 5.956340956340957,
+      "grad_norm": 1.5820633172988892,
+      "learning_rate": 4.937019825517333e-05,
+      "loss": 0.0374,
+      "num_input_tokens_seen": 2238736,
+      "step": 2865
+    },
+    {
+      "epoch": 5.966735966735967,
+      "grad_norm": 6.139444351196289,
+      "learning_rate": 4.9368006627565954e-05,
+      "loss": 0.0147,
+      "num_input_tokens_seen": 2242608,
+      "step": 2870
+    },
+    {
+      "epoch": 5.977130977130977,
+      "grad_norm": 2.956362724304199,
+      "learning_rate": 4.936581124210573e-05,
+      "loss": 0.0515,
+      "num_input_tokens_seen": 2246576,
+      "step": 2875
+    },
+    {
+      "epoch": 5.987525987525988,
+      "grad_norm": 3.3747339248657227,
+      "learning_rate": 4.9363612099131216e-05,
+      "loss": 0.0743,
+      "num_input_tokens_seen": 2250448,
+      "step": 2880
+    },
+    {
+      "epoch": 5.997920997920998,
+      "grad_norm": 1.451723575592041,
+      "learning_rate": 4.936140919898155e-05,
+      "loss": 0.0531,
+      "num_input_tokens_seen": 2254320,
+      "step": 2885
+    },
+    {
+      "epoch": 6.008316008316008,
+      "grad_norm": 1.6113674640655518,
+      "learning_rate": 4.9359202541996426e-05,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 2258272,
+      "step": 2890
+    },
+    {
+      "epoch": 6.018711018711019,
+      "grad_norm": 0.25212323665618896,
+      "learning_rate": 4.935699212851616e-05,
+      "loss": 0.0224,
+      "num_input_tokens_seen": 2262176,
+      "step": 2895
+    },
+    {
+      "epoch": 6.029106029106029,
+      "grad_norm": 0.9302971959114075,
+      "learning_rate": 4.935477795888162e-05,
+      "loss": 0.0401,
+      "num_input_tokens_seen": 2266112,
+      "step": 2900
+    },
+    {
+      "epoch": 6.039501039501039,
+      "grad_norm": 1.7203342914581299,
+      "learning_rate": 4.935256003343426e-05,
+      "loss": 0.0085,
+      "num_input_tokens_seen": 2269984,
+      "step": 2905
+    },
+    {
+      "epoch": 6.04989604989605,
+      "grad_norm": 0.07220667600631714,
+      "learning_rate": 4.93503383525161e-05,
+      "loss": 0.02,
+      "num_input_tokens_seen": 2273792,
+      "step": 2910
+    },
+    {
+      "epoch": 6.0602910602910605,
+      "grad_norm": 0.1504482924938202,
+      "learning_rate": 4.934811291646977e-05,
+      "loss": 0.026,
+      "num_input_tokens_seen": 2277856,
+      "step": 2915
+    },
+    {
+      "epoch": 6.070686070686071,
+      "grad_norm": 1.9064278602600098,
+      "learning_rate": 4.934588372563845e-05,
+      "loss": 0.0157,
+      "num_input_tokens_seen": 2281600,
+      "step": 2920
+    },
+    {
+      "epoch": 6.081081081081081,
+      "grad_norm": 0.6857333779335022,
+      "learning_rate": 4.93436507803659e-05,
+      "loss": 0.0089,
+      "num_input_tokens_seen": 2285376,
+      "step": 2925
+    },
+    {
+      "epoch": 6.091476091476092,
+      "grad_norm": 0.06444158405065536,
+      "learning_rate": 4.934141408099649e-05,
+      "loss": 0.0609,
+      "num_input_tokens_seen": 2289312,
+      "step": 2930
+    },
+    {
+      "epoch": 6.101871101871102,
+      "grad_norm": 0.2495444267988205,
+      "learning_rate": 4.9339173627875135e-05,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 2293216,
+      "step": 2935
+    },
+    {
+      "epoch": 6.112266112266112,
+      "grad_norm": 0.3173748552799225,
+      "learning_rate": 4.9336929421347335e-05,
+      "loss": 0.0431,
+      "num_input_tokens_seen": 2297376,
+      "step": 2940
+    },
+    {
+      "epoch": 6.122661122661123,
+      "grad_norm": 2.9207396507263184,
+      "learning_rate": 4.933468146175918e-05,
+      "loss": 0.0251,
+      "num_input_tokens_seen": 2301440,
+      "step": 2945
+    },
+    {
+      "epoch": 6.133056133056133,
+      "grad_norm": 0.14804820716381073,
+      "learning_rate": 4.933242974945734e-05,
+      "loss": 0.0395,
+      "num_input_tokens_seen": 2305312,
+      "step": 2950
+    },
+    {
+      "epoch": 6.143451143451143,
+      "grad_norm": 3.9064009189605713,
+      "learning_rate": 4.933017428478906e-05,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 2309280,
+      "step": 2955
+    },
+    {
+      "epoch": 6.153846153846154,
+      "grad_norm": 5.434149265289307,
+      "learning_rate": 4.932791506810214e-05,
+      "loss": 0.007,
+      "num_input_tokens_seen": 2313120,
+      "step": 2960
+    },
+    {
+      "epoch": 6.164241164241164,
+      "grad_norm": 0.968045711517334,
+      "learning_rate": 4.932565209974499e-05,
+      "loss": 0.0166,
+      "num_input_tokens_seen": 2317184,
+      "step": 2965
+    },
+    {
+      "epoch": 6.174636174636174,
+      "grad_norm": 5.382054805755615,
+      "learning_rate": 4.93233853800666e-05,
+      "loss": 0.0353,
+      "num_input_tokens_seen": 2321184,
+      "step": 2970
+    },
+    {
+      "epoch": 6.185031185031185,
+      "grad_norm": 1.5844146013259888,
+      "learning_rate": 4.932111490941651e-05,
+      "loss": 0.0153,
+      "num_input_tokens_seen": 2324928,
+      "step": 2975
+    },
+    {
+      "epoch": 6.1954261954261955,
+      "grad_norm": 5.043550491333008,
+      "learning_rate": 4.9318840688144876e-05,
+      "loss": 0.0214,
+      "num_input_tokens_seen": 2328960,
+      "step": 2980
+    },
+    {
+      "epoch": 6.205821205821206,
+      "grad_norm": 0.07053793966770172,
+      "learning_rate": 4.9316562716602387e-05,
+      "loss": 0.0251,
+      "num_input_tokens_seen": 2332928,
+      "step": 2985
+    },
+    {
+      "epoch": 6.216216216216216,
+      "grad_norm": 0.036285389214754105,
+      "learning_rate": 4.9314280995140346e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 2337056,
+      "step": 2990
+    },
+    {
+      "epoch": 6.226611226611227,
+      "grad_norm": 0.07890532165765762,
+      "learning_rate": 4.931199552411063e-05,
+      "loss": 0.0424,
+      "num_input_tokens_seen": 2340928,
+      "step": 2995
+    },
+    {
+      "epoch": 6.237006237006237,
+      "grad_norm": 0.019469575956463814,
+      "learning_rate": 4.930970630386568e-05,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 2344864,
+      "step": 3000
+    },
+    {
+      "epoch": 6.237006237006237,
+      "eval_loss": 0.1914750337600708,
+      "eval_runtime": 11.7091,
+      "eval_samples_per_second": 73.106,
+      "eval_steps_per_second": 18.276,
+      "num_input_tokens_seen": 2344864,
+      "step": 3000
+    },
+    {
+      "epoch": 6.247401247401247,
+      "grad_norm": 5.998438358306885,
+      "learning_rate": 4.9307413334758524e-05,
+      "loss": 0.0118,
+      "num_input_tokens_seen": 2348864,
+      "step": 3005
+    },
+    {
+      "epoch": 6.257796257796258,
+      "grad_norm": 7.5015106201171875,
+      "learning_rate": 4.930511661714276e-05,
+      "loss": 0.0458,
+      "num_input_tokens_seen": 2352704,
+      "step": 3010
+    },
+    {
+      "epoch": 6.268191268191268,
+      "grad_norm": 4.160377025604248,
+      "learning_rate": 4.9302816151372576e-05,
+      "loss": 0.0729,
+      "num_input_tokens_seen": 2356576,
+      "step": 3015
+    },
+    {
+      "epoch": 6.278586278586278,
+      "grad_norm": 4.9023847579956055,
+      "learning_rate": 4.930051193780274e-05,
+      "loss": 0.0522,
+      "num_input_tokens_seen": 2360512,
+      "step": 3020
+    },
+    {
+      "epoch": 6.288981288981289,
+      "grad_norm": 0.046715669333934784,
+      "learning_rate": 4.929820397678858e-05,
+      "loss": 0.0131,
+      "num_input_tokens_seen": 2364352,
+      "step": 3025
+    },
+    {
+      "epoch": 6.299376299376299,
+      "grad_norm": 0.27996915578842163,
+      "learning_rate": 4.9295892268686015e-05,
+      "loss": 0.0176,
+      "num_input_tokens_seen": 2368160,
+      "step": 3030
+    },
+    {
+      "epoch": 6.3097713097713095,
+      "grad_norm": 0.49580928683280945,
+      "learning_rate": 4.9293576813851536e-05,
+      "loss": 0.0404,
+      "num_input_tokens_seen": 2371904,
+      "step": 3035
+    },
+    {
+      "epoch": 6.3201663201663205,
+      "grad_norm": 0.9389116168022156,
+      "learning_rate": 4.929125761264223e-05,
+      "loss": 0.0767,
+      "num_input_tokens_seen": 2375712,
+      "step": 3040
+    },
+    {
+      "epoch": 6.330561330561331,
+      "grad_norm": 2.3065052032470703,
+      "learning_rate": 4.928893466541573e-05,
+      "loss": 0.076,
+      "num_input_tokens_seen": 2379680,
+      "step": 3045
+    },
+    {
+      "epoch": 6.340956340956341,
+      "grad_norm": 0.08882429450750351,
+      "learning_rate": 4.928660797253027e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 2383488,
+      "step": 3050
+    },
+    {
+      "epoch": 6.351351351351352,
+      "grad_norm": 0.08732178062200546,
+      "learning_rate": 4.928427753434467e-05,
+      "loss": 0.0065,
+      "num_input_tokens_seen": 2387392,
+      "step": 3055
+    },
+    {
+      "epoch": 6.361746361746362,
+      "grad_norm": 0.07492464780807495,
+      "learning_rate": 4.9281943351218286e-05,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 2391296,
+      "step": 3060
+    },
+    {
+      "epoch": 6.372141372141372,
+      "grad_norm": 0.906123697757721,
+      "learning_rate": 4.9279605423511095e-05,
+      "loss": 0.0324,
+      "num_input_tokens_seen": 2395200,
+      "step": 3065
+    },
+    {
+      "epoch": 6.382536382536383,
+      "grad_norm": 3.6891300678253174,
+      "learning_rate": 4.927726375158363e-05,
+      "loss": 0.0284,
+      "num_input_tokens_seen": 2399040,
+      "step": 3070
+    },
+    {
+      "epoch": 6.392931392931393,
+      "grad_norm": 0.24633274972438812,
+      "learning_rate": 4.9274918335797004e-05,
+      "loss": 0.0098,
+      "num_input_tokens_seen": 2402944,
+      "step": 3075
+    },
+    {
+      "epoch": 6.403326403326403,
+      "grad_norm": 0.20260301232337952,
+      "learning_rate": 4.927256917651292e-05,
+      "loss": 0.0215,
+      "num_input_tokens_seen": 2406816,
+      "step": 3080
+    },
+    {
+      "epoch": 6.413721413721413,
+      "grad_norm": 0.305388867855072,
+      "learning_rate": 4.927021627409364e-05,
+      "loss": 0.0481,
+      "num_input_tokens_seen": 2410656,
+      "step": 3085
+    },
+    {
+      "epoch": 6.424116424116424,
+      "grad_norm": 3.982872724533081,
+      "learning_rate": 4.9267859628902005e-05,
+      "loss": 0.0249,
+      "num_input_tokens_seen": 2414496,
+      "step": 3090
+    },
+    {
+      "epoch": 6.4345114345114345,
+      "grad_norm": 0.09157679975032806,
+      "learning_rate": 4.9265499241301454e-05,
+      "loss": 0.0195,
+      "num_input_tokens_seen": 2418272,
+      "step": 3095
+    },
+    {
+      "epoch": 6.444906444906445,
+      "grad_norm": 0.07051081210374832,
+      "learning_rate": 4.926313511165598e-05,
+      "loss": 0.0311,
+      "num_input_tokens_seen": 2422208,
+      "step": 3100
+    },
+    {
+      "epoch": 6.455301455301456,
+      "grad_norm": 3.4662933349609375,
+      "learning_rate": 4.926076724033016e-05,
+      "loss": 0.0461,
+      "num_input_tokens_seen": 2426016,
+      "step": 3105
+    },
+    {
+      "epoch": 6.465696465696466,
+      "grad_norm": 0.21878628432750702,
+      "learning_rate": 4.9258395627689146e-05,
+      "loss": 0.0352,
+      "num_input_tokens_seen": 2429920,
+      "step": 3110
+    },
+    {
+      "epoch": 6.476091476091476,
+      "grad_norm": 2.438678503036499,
+      "learning_rate": 4.925602027409868e-05,
+      "loss": 0.0325,
+      "num_input_tokens_seen": 2434016,
+      "step": 3115
+    },
+    {
+      "epoch": 6.486486486486487,
+      "grad_norm": 0.25928011536598206,
+      "learning_rate": 4.925364117992507e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 2437952,
+      "step": 3120
+    },
+    {
+      "epoch": 6.496881496881497,
+      "grad_norm": 0.09472639858722687,
+      "learning_rate": 4.92512583455352e-05,
+      "loss": 0.0413,
+      "num_input_tokens_seen": 2441696,
+      "step": 3125
+    },
+    {
+      "epoch": 6.507276507276507,
+      "grad_norm": 0.025656303390860558,
+      "learning_rate": 4.9248871771296536e-05,
+      "loss": 0.0127,
+      "num_input_tokens_seen": 2445696,
+      "step": 3130
+    },
+    {
+      "epoch": 6.517671517671518,
+      "grad_norm": 0.7066136598587036,
+      "learning_rate": 4.924648145757711e-05,
+      "loss": 0.007,
+      "num_input_tokens_seen": 2449632,
+      "step": 3135
+    },
+    {
+      "epoch": 6.528066528066528,
+      "grad_norm": 3.153752565383911,
+      "learning_rate": 4.924408740474554e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 2453600,
+      "step": 3140
+    },
+    {
+      "epoch": 6.538461538461538,
+      "grad_norm": 0.05097276717424393,
+      "learning_rate": 4.924168961317103e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 2457600,
+      "step": 3145
+    },
+    {
+      "epoch": 6.548856548856548,
+      "grad_norm": 0.021507594734430313,
+      "learning_rate": 4.9239288083223334e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 2461504,
+      "step": 3150
+    },
+    {
+      "epoch": 6.5592515592515594,
+      "grad_norm": 2.3601343631744385,
+      "learning_rate": 4.9236882815272803e-05,
+      "loss": 0.0265,
+      "num_input_tokens_seen": 2465376,
+      "step": 3155
+    },
+    {
+      "epoch": 6.56964656964657,
+      "grad_norm": 0.10477852821350098,
+      "learning_rate": 4.9234473809690365e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 2469280,
+      "step": 3160
+    },
+    {
+      "epoch": 6.58004158004158,
+      "grad_norm": 1.7486748695373535,
+      "learning_rate": 4.923206106684752e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 2473184,
+      "step": 3165
+    },
+    {
+      "epoch": 6.590436590436591,
+      "grad_norm": 0.053241755813360214,
+      "learning_rate": 4.922964458711634e-05,
+      "loss": 0.0268,
+      "num_input_tokens_seen": 2477024,
+      "step": 3170
+    },
+    {
+      "epoch": 6.600831600831601,
+      "grad_norm": 0.04432320222258568,
+      "learning_rate": 4.9227224370869474e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 2480896,
+      "step": 3175
+    },
+    {
+      "epoch": 6.611226611226611,
+      "grad_norm": 4.2861857414245605,
+      "learning_rate": 4.9224800418480155e-05,
+      "loss": 0.0191,
+      "num_input_tokens_seen": 2484672,
+      "step": 3180
+    },
+    {
+      "epoch": 6.621621621621622,
+      "grad_norm": 0.1452285647392273,
+      "learning_rate": 4.9222372730322176e-05,
+      "loss": 0.0856,
+      "num_input_tokens_seen": 2488576,
+      "step": 3185
+    },
+    {
+      "epoch": 6.632016632016632,
+      "grad_norm": 0.07084045559167862,
+      "learning_rate": 4.921994130676993e-05,
+      "loss": 0.0428,
+      "num_input_tokens_seen": 2492544,
+      "step": 3190
+    },
+    {
+      "epoch": 6.642411642411642,
+      "grad_norm": 3.1588563919067383,
+      "learning_rate": 4.9217506148198366e-05,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 2496448,
+      "step": 3195
+    },
+    {
+      "epoch": 6.652806652806653,
+      "grad_norm": 1.1454668045043945,
+      "learning_rate": 4.921506725498302e-05,
+      "loss": 0.0309,
+      "num_input_tokens_seen": 2500448,
+      "step": 3200
+    },
+    {
+      "epoch": 6.652806652806653,
+      "eval_loss": 0.21970447897911072,
+      "eval_runtime": 11.7744,
+      "eval_samples_per_second": 72.7,
+      "eval_steps_per_second": 18.175,
+      "num_input_tokens_seen": 2500448,
+      "step": 3200
+    },
+    {
+      "epoch": 6.663201663201663,
+      "grad_norm": 1.1405205726623535,
+      "learning_rate": 4.9212624627499994e-05,
+      "loss": 0.0182,
+      "num_input_tokens_seen": 2504320,
+      "step": 3205
+    },
+    {
+      "epoch": 6.673596673596673,
+      "grad_norm": 0.725938618183136,
+      "learning_rate": 4.921017826612597e-05,
+      "loss": 0.0235,
+      "num_input_tokens_seen": 2508256,
+      "step": 3210
+    },
+    {
+      "epoch": 6.6839916839916835,
+      "grad_norm": 0.05477587506175041,
+      "learning_rate": 4.9207728171238223e-05,
+      "loss": 0.0318,
+      "num_input_tokens_seen": 2512224,
+      "step": 3215
+    },
+    {
+      "epoch": 6.6943866943866945,
+      "grad_norm": 2.5242605209350586,
+      "learning_rate": 4.920527434321458e-05,
+      "loss": 0.0162,
+      "num_input_tokens_seen": 2516192,
+      "step": 3220
+    },
+    {
+      "epoch": 6.704781704781705,
+      "grad_norm": 7.689412593841553,
+      "learning_rate": 4.920281678243345e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 2520128,
+      "step": 3225
+    },
+    {
+      "epoch": 6.715176715176716,
+      "grad_norm": 0.0377560518682003,
+      "learning_rate": 4.920035548927381e-05,
+      "loss": 0.0389,
+      "num_input_tokens_seen": 2524128,
+      "step": 3230
+    },
+    {
+      "epoch": 6.725571725571726,
+      "grad_norm": 0.034012533724308014,
+      "learning_rate": 4.919789046411525e-05,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 2528000,
+      "step": 3235
+    },
+    {
+      "epoch": 6.735966735966736,
+      "grad_norm": 0.21459653973579407,
+      "learning_rate": 4.919542170733787e-05,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 2532032,
+      "step": 3240
+    },
+    {
+      "epoch": 6.746361746361746,
+      "grad_norm": 16.39967918395996,
+      "learning_rate": 4.919294921932242e-05,
+      "loss": 0.0177,
+      "num_input_tokens_seen": 2536128,
+      "step": 3245
+    },
+    {
+      "epoch": 6.756756756756757,
+      "grad_norm": 0.1318790167570114,
+      "learning_rate": 4.919047300045016e-05,
+      "loss": 0.02,
+      "num_input_tokens_seen": 2539904,
+      "step": 3250
+    },
+    {
+      "epoch": 6.767151767151767,
+      "grad_norm": 0.0561412088572979,
+      "learning_rate": 4.918799305110299e-05,
+      "loss": 0.0222,
+      "num_input_tokens_seen": 2543584,
+      "step": 3255
+    },
+    {
+      "epoch": 6.777546777546777,
+      "grad_norm": 10.749137878417969,
+      "learning_rate": 4.918550937166331e-05,
+      "loss": 0.055,
+      "num_input_tokens_seen": 2547456,
+      "step": 3260
+    },
+    {
+      "epoch": 6.787941787941788,
+      "grad_norm": 4.381881237030029,
+      "learning_rate": 4.918302196251415e-05,
+      "loss": 0.0166,
+      "num_input_tokens_seen": 2551232,
+      "step": 3265
+    },
+    {
+      "epoch": 6.798336798336798,
+      "grad_norm": 7.399415493011475,
+      "learning_rate": 4.91805308240391e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 2555072,
+      "step": 3270
+    },
+    {
+      "epoch": 6.8087318087318085,
+      "grad_norm": 2.8986399173736572,
+      "learning_rate": 4.9178035956622326e-05,
+      "loss": 0.0683,
+      "num_input_tokens_seen": 2559008,
+      "step": 3275
+    },
+    {
+      "epoch": 6.8191268191268195,
+      "grad_norm": 1.7759944200515747,
+      "learning_rate": 4.917553736064857e-05,
+      "loss": 0.0238,
+      "num_input_tokens_seen": 2562848,
+      "step": 3280
+    },
+    {
+      "epoch": 6.82952182952183,
+      "grad_norm": 0.3556916415691376,
+      "learning_rate": 4.917303503650314e-05,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 2566816,
+      "step": 3285
+    },
+    {
+      "epoch": 6.83991683991684,
+      "grad_norm": 3.2809507846832275,
+      "learning_rate": 4.917052898457194e-05,
+      "loss": 0.0352,
+      "num_input_tokens_seen": 2570656,
+      "step": 3290
+    },
+    {
+      "epoch": 6.850311850311851,
+      "grad_norm": 9.560991287231445,
+      "learning_rate": 4.916801920524141e-05,
+      "loss": 0.0287,
+      "num_input_tokens_seen": 2574784,
+      "step": 3295
+    },
+    {
+      "epoch": 6.860706860706861,
+      "grad_norm": 0.033985599875450134,
+      "learning_rate": 4.916550569889862e-05,
+      "loss": 0.0045,
+      "num_input_tokens_seen": 2578656,
+      "step": 3300
+    },
+    {
+      "epoch": 6.871101871101871,
+      "grad_norm": 8.830511093139648,
+      "learning_rate": 4.916298846593116e-05,
+      "loss": 0.0299,
+      "num_input_tokens_seen": 2582656,
+      "step": 3305
+    },
+    {
+      "epoch": 6.881496881496881,
+      "grad_norm": 2.2753632068634033,
+      "learning_rate": 4.916046750672722e-05,
+      "loss": 0.058,
+      "num_input_tokens_seen": 2586464,
+      "step": 3310
+    },
+    {
+      "epoch": 6.891891891891892,
+      "grad_norm": 0.06041261553764343,
+      "learning_rate": 4.915794282167559e-05,
+      "loss": 0.0172,
+      "num_input_tokens_seen": 2590336,
+      "step": 3315
+    },
+    {
+      "epoch": 6.902286902286902,
+      "grad_norm": 9.050000190734863,
+      "learning_rate": 4.915541441116558e-05,
+      "loss": 0.03,
+      "num_input_tokens_seen": 2594080,
+      "step": 3320
+    },
+    {
+      "epoch": 6.912681912681912,
+      "grad_norm": 0.46946394443511963,
+      "learning_rate": 4.915288227558711e-05,
+      "loss": 0.0618,
+      "num_input_tokens_seen": 2598080,
+      "step": 3325
+    },
+    {
+      "epoch": 6.923076923076923,
+      "grad_norm": 0.6037935018539429,
+      "learning_rate": 4.915034641533066e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 2602048,
+      "step": 3330
+    },
+    {
+      "epoch": 6.9334719334719335,
+      "grad_norm": 0.9779449701309204,
+      "learning_rate": 4.914780683078731e-05,
+      "loss": 0.0301,
+      "num_input_tokens_seen": 2606016,
+      "step": 3335
+    },
+    {
+      "epoch": 6.943866943866944,
+      "grad_norm": 0.6565176248550415,
+      "learning_rate": 4.9145263522348695e-05,
+      "loss": 0.0275,
+      "num_input_tokens_seen": 2609952,
+      "step": 3340
+    },
+    {
+      "epoch": 6.954261954261955,
+      "grad_norm": 0.99924635887146,
+      "learning_rate": 4.9142716490407e-05,
+      "loss": 0.0324,
+      "num_input_tokens_seen": 2613888,
+      "step": 3345
+    },
+    {
+      "epoch": 6.964656964656965,
+      "grad_norm": 0.9475417733192444,
+      "learning_rate": 4.914016573535504e-05,
+      "loss": 0.0369,
+      "num_input_tokens_seen": 2617728,
+      "step": 3350
+    },
+    {
+      "epoch": 6.975051975051975,
+      "grad_norm": 0.16366787254810333,
+      "learning_rate": 4.9137611257586154e-05,
+      "loss": 0.0202,
+      "num_input_tokens_seen": 2621568,
+      "step": 3355
+    },
+    {
+      "epoch": 6.985446985446986,
+      "grad_norm": 0.24206525087356567,
+      "learning_rate": 4.9135053057494274e-05,
+      "loss": 0.0349,
+      "num_input_tokens_seen": 2625440,
+      "step": 3360
+    },
+    {
+      "epoch": 6.995841995841996,
+      "grad_norm": 0.1584472805261612,
+      "learning_rate": 4.913249113547392e-05,
+      "loss": 0.013,
+      "num_input_tokens_seen": 2629440,
+      "step": 3365
+    },
+    {
+      "epoch": 7.006237006237006,
+      "grad_norm": 3.8973793983459473,
+      "learning_rate": 4.912992549192016e-05,
+      "loss": 0.0067,
+      "num_input_tokens_seen": 2633232,
+      "step": 3370
+    },
+    {
+      "epoch": 7.016632016632016,
+      "grad_norm": 0.1585526019334793,
+      "learning_rate": 4.9127356127228665e-05,
+      "loss": 0.0157,
+      "num_input_tokens_seen": 2637040,
+      "step": 3375
+    },
+    {
+      "epoch": 7.027027027027027,
+      "grad_norm": 0.02379317581653595,
+      "learning_rate": 4.912478304179564e-05,
+      "loss": 0.0584,
+      "num_input_tokens_seen": 2640816,
+      "step": 3380
+    },
+    {
+      "epoch": 7.037422037422037,
+      "grad_norm": 1.0086785554885864,
+      "learning_rate": 4.9122206236017896e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 2644688,
+      "step": 3385
+    },
+    {
+      "epoch": 7.047817047817047,
+      "grad_norm": 0.3364635109901428,
+      "learning_rate": 4.911962571029282e-05,
+      "loss": 0.0231,
+      "num_input_tokens_seen": 2648688,
+      "step": 3390
+    },
+    {
+      "epoch": 7.058212058212058,
+      "grad_norm": 0.027429258450865746,
+      "learning_rate": 4.9117041465018353e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 2652560,
+      "step": 3395
+    },
+    {
+      "epoch": 7.0686070686070686,
+      "grad_norm": 0.20505613088607788,
+      "learning_rate": 4.911445350059302e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 2656400,
+      "step": 3400
+    },
+    {
+      "epoch": 7.0686070686070686,
+      "eval_loss": 0.20138801634311676,
+      "eval_runtime": 11.7152,
+      "eval_samples_per_second": 73.067,
+      "eval_steps_per_second": 18.267,
+      "num_input_tokens_seen": 2656400,
+      "step": 3400
+    },
+    {
+      "epoch": 7.079002079002079,
+      "grad_norm": 0.22735638916492462,
+      "learning_rate": 4.9111861817415905e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 2660240,
+      "step": 3405
+    },
+    {
+      "epoch": 7.08939708939709,
+      "grad_norm": 0.8609936237335205,
+      "learning_rate": 4.91092664158867e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 2664176,
+      "step": 3410
+    },
+    {
+      "epoch": 7.0997920997921,
+      "grad_norm": 0.012282488867640495,
+      "learning_rate": 4.910666729640563e-05,
+      "loss": 0.0202,
+      "num_input_tokens_seen": 2667952,
+      "step": 3415
+    },
+    {
+      "epoch": 7.11018711018711,
+      "grad_norm": 5.507207870483398,
+      "learning_rate": 4.910406445937353e-05,
+      "loss": 0.0158,
+      "num_input_tokens_seen": 2671952,
+      "step": 3420
+    },
+    {
+      "epoch": 7.120582120582121,
+      "grad_norm": 0.04543973505496979,
+      "learning_rate": 4.9101457905191774e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 2675696,
+      "step": 3425
+    },
+    {
+      "epoch": 7.130977130977131,
+      "grad_norm": 3.571209192276001,
+      "learning_rate": 4.909884763426233e-05,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 2679664,
+      "step": 3430
+    },
+    {
+      "epoch": 7.141372141372141,
+      "grad_norm": 0.09967422485351562,
+      "learning_rate": 4.9096233646987736e-05,
+      "loss": 0.0127,
+      "num_input_tokens_seen": 2683568,
+      "step": 3435
+    },
+    {
+      "epoch": 7.151767151767152,
+      "grad_norm": 0.02837456576526165,
+      "learning_rate": 4.9093615943771104e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 2687504,
+      "step": 3440
+    },
+    {
+      "epoch": 7.162162162162162,
+      "grad_norm": 0.025648083537817,
+      "learning_rate": 4.909099452501611e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 2691536,
+      "step": 3445
+    },
+    {
+      "epoch": 7.172557172557172,
+      "grad_norm": 0.026372671127319336,
+      "learning_rate": 4.908836939112702e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 2695248,
+      "step": 3450
+    },
+    {
+      "epoch": 7.182952182952183,
+      "grad_norm": 0.044457968324422836,
+      "learning_rate": 4.908574054250865e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 2699120,
+      "step": 3455
+    },
+    {
+      "epoch": 7.1933471933471935,
+      "grad_norm": 0.011112251318991184,
+      "learning_rate": 4.9083107979566414e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 2702928,
+      "step": 3460
+    },
+    {
+      "epoch": 7.203742203742204,
+      "grad_norm": 0.009859545156359673,
+      "learning_rate": 4.908047170270628e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 2706768,
+      "step": 3465
+    },
+    {
+      "epoch": 7.214137214137214,
+      "grad_norm": 0.008643168956041336,
+      "learning_rate": 4.9077831712334784e-05,
+      "loss": 0.0333,
+      "num_input_tokens_seen": 2710704,
+      "step": 3470
+    },
+    {
+      "epoch": 7.224532224532225,
+      "grad_norm": 0.013558020815253258,
+      "learning_rate": 4.907518800885907e-05,
+      "loss": 0.0252,
+      "num_input_tokens_seen": 2714576,
+      "step": 3475
+    },
+    {
+      "epoch": 7.234927234927235,
+      "grad_norm": 0.05885520949959755,
+      "learning_rate": 4.907254059268681e-05,
+      "loss": 0.0194,
+      "num_input_tokens_seen": 2718448,
+      "step": 3480
+    },
+    {
+      "epoch": 7.245322245322245,
+      "grad_norm": 0.07066506147384644,
+      "learning_rate": 4.906988946422628e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 2722352,
+      "step": 3485
+    },
+    {
+      "epoch": 7.255717255717256,
+      "grad_norm": 0.9059031009674072,
+      "learning_rate": 4.9067234623886315e-05,
+      "loss": 0.0198,
+      "num_input_tokens_seen": 2726160,
+      "step": 3490
+    },
+    {
+      "epoch": 7.266112266112266,
+      "grad_norm": 0.10740066319704056,
+      "learning_rate": 4.9064576072076316e-05,
+      "loss": 0.0304,
+      "num_input_tokens_seen": 2730064,
+      "step": 3495
+    },
+    {
+      "epoch": 7.276507276507276,
+      "grad_norm": 0.2772800028324127,
+      "learning_rate": 4.906191380920628e-05,
+      "loss": 0.0303,
+      "num_input_tokens_seen": 2733936,
+      "step": 3500
+    },
+    {
+      "epoch": 7.286902286902287,
+      "grad_norm": 0.30590516328811646,
+      "learning_rate": 4.905924783568675e-05,
+      "loss": 0.0288,
+      "num_input_tokens_seen": 2737776,
+      "step": 3505
+    },
+    {
+      "epoch": 7.297297297297297,
+      "grad_norm": 1.9509634971618652,
+      "learning_rate": 4.905657815192886e-05,
+      "loss": 0.0188,
+      "num_input_tokens_seen": 2741616,
+      "step": 3510
+    },
+    {
+      "epoch": 7.3076923076923075,
+      "grad_norm": 0.0994449257850647,
+      "learning_rate": 4.90539047583443e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 2745424,
+      "step": 3515
+    },
+    {
+      "epoch": 7.3180873180873185,
+      "grad_norm": 0.1023537740111351,
+      "learning_rate": 4.905122765534534e-05,
+      "loss": 0.0118,
+      "num_input_tokens_seen": 2749392,
+      "step": 3520
+    },
+    {
+      "epoch": 7.328482328482329,
+      "grad_norm": 0.6348842978477478,
+      "learning_rate": 4.9048546843344846e-05,
+      "loss": 0.0145,
+      "num_input_tokens_seen": 2753360,
+      "step": 3525
+    },
+    {
+      "epoch": 7.338877338877339,
+      "grad_norm": 0.03266940265893936,
+      "learning_rate": 4.9045862322756206e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 2757328,
+      "step": 3530
+    },
+    {
+      "epoch": 7.349272349272349,
+      "grad_norm": 5.084386348724365,
+      "learning_rate": 4.904317409399342e-05,
+      "loss": 0.0241,
+      "num_input_tokens_seen": 2761264,
+      "step": 3535
+    },
+    {
+      "epoch": 7.35966735966736,
+      "grad_norm": 0.049261968582868576,
+      "learning_rate": 4.904048215747104e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 2765232,
+      "step": 3540
+    },
+    {
+      "epoch": 7.37006237006237,
+      "grad_norm": 6.875644207000732,
+      "learning_rate": 4.90377865136042e-05,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 2769296,
+      "step": 3545
+    },
+    {
+      "epoch": 7.38045738045738,
+      "grad_norm": 0.017293643206357956,
+      "learning_rate": 4.90350871628086e-05,
+      "loss": 0.0066,
+      "num_input_tokens_seen": 2773264,
+      "step": 3550
+    },
+    {
+      "epoch": 7.390852390852391,
+      "grad_norm": 0.04023009538650513,
+      "learning_rate": 4.903238410550052e-05,
+      "loss": 0.0427,
+      "num_input_tokens_seen": 2777200,
+      "step": 3555
+    },
+    {
+      "epoch": 7.401247401247401,
+      "grad_norm": 11.1991548538208,
+      "learning_rate": 4.90296773420968e-05,
+      "loss": 0.017,
+      "num_input_tokens_seen": 2781008,
+      "step": 3560
+    },
+    {
+      "epoch": 7.411642411642411,
+      "grad_norm": 0.014578435570001602,
+      "learning_rate": 4.902696687301486e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 2785040,
+      "step": 3565
+    },
+    {
+      "epoch": 7.422037422037422,
+      "grad_norm": 0.37953630089759827,
+      "learning_rate": 4.902425269867268e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 2789104,
+      "step": 3570
+    },
+    {
+      "epoch": 7.4324324324324325,
+      "grad_norm": 11.826692581176758,
+      "learning_rate": 4.902153481948883e-05,
+      "loss": 0.0123,
+      "num_input_tokens_seen": 2793104,
+      "step": 3575
+    },
+    {
+      "epoch": 7.442827442827443,
+      "grad_norm": 0.023193754255771637,
+      "learning_rate": 4.901881323588244e-05,
+      "loss": 0.0232,
+      "num_input_tokens_seen": 2797072,
+      "step": 3580
+    },
+    {
+      "epoch": 7.453222453222454,
+      "grad_norm": 1.8994914293289185,
+      "learning_rate": 4.90160879482732e-05,
+      "loss": 0.0147,
+      "num_input_tokens_seen": 2800816,
+      "step": 3585
+    },
+    {
+      "epoch": 7.463617463617464,
+      "grad_norm": 2.010436773300171,
+      "learning_rate": 4.9013358957081405e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 2804752,
+      "step": 3590
+    },
+    {
+      "epoch": 7.474012474012474,
+      "grad_norm": 0.05587797611951828,
+      "learning_rate": 4.901062626272789e-05,
+      "loss": 0.0466,
+      "num_input_tokens_seen": 2808720,
+      "step": 3595
+    },
+    {
+      "epoch": 7.484407484407485,
+      "grad_norm": 0.10555629432201385,
+      "learning_rate": 4.900788986563406e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 2812912,
+      "step": 3600
+    },
+    {
+      "epoch": 7.484407484407485,
+      "eval_loss": 0.23719295859336853,
+      "eval_runtime": 11.7876,
+      "eval_samples_per_second": 72.619,
+      "eval_steps_per_second": 18.155,
+      "num_input_tokens_seen": 2812912,
+      "step": 3600
+    },
+    {
+      "epoch": 7.494802494802495,
+      "grad_norm": 0.043628767132759094,
+      "learning_rate": 4.9005149766221915e-05,
+      "loss": 0.056,
+      "num_input_tokens_seen": 2816912,
+      "step": 3605
+    },
+    {
+      "epoch": 7.505197505197505,
+      "grad_norm": 0.10638926178216934,
+      "learning_rate": 4.9002405964914e-05,
+      "loss": 0.0544,
+      "num_input_tokens_seen": 2820752,
+      "step": 3610
+    },
+    {
+      "epoch": 7.515592515592516,
+      "grad_norm": 0.06740149110555649,
+      "learning_rate": 4.899965846213346e-05,
+      "loss": 0.0105,
+      "num_input_tokens_seen": 2824624,
+      "step": 3615
+    },
+    {
+      "epoch": 7.525987525987526,
+      "grad_norm": 0.19606338441371918,
+      "learning_rate": 4.899690725830399e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 2828688,
+      "step": 3620
+    },
+    {
+      "epoch": 7.536382536382536,
+      "grad_norm": 0.044863954186439514,
+      "learning_rate": 4.899415235384985e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 2832432,
+      "step": 3625
+    },
+    {
+      "epoch": 7.546777546777546,
+      "grad_norm": 0.0860869437456131,
+      "learning_rate": 4.899139374919589e-05,
+      "loss": 0.0573,
+      "num_input_tokens_seen": 2836496,
+      "step": 3630
+    },
+    {
+      "epoch": 7.557172557172557,
+      "grad_norm": 0.4277716875076294,
+      "learning_rate": 4.898863144476752e-05,
+      "loss": 0.0266,
+      "num_input_tokens_seen": 2840496,
+      "step": 3635
+    },
+    {
+      "epoch": 7.5675675675675675,
+      "grad_norm": 1.4758248329162598,
+      "learning_rate": 4.898586544099072e-05,
+      "loss": 0.0328,
+      "num_input_tokens_seen": 2844400,
+      "step": 3640
+    },
+    {
+      "epoch": 7.577962577962578,
+      "grad_norm": 0.22953356802463531,
+      "learning_rate": 4.898309573829204e-05,
+      "loss": 0.0164,
+      "num_input_tokens_seen": 2848336,
+      "step": 3645
+    },
+    {
+      "epoch": 7.588357588357589,
+      "grad_norm": 6.157791614532471,
+      "learning_rate": 4.898032233709862e-05,
+      "loss": 0.0541,
+      "num_input_tokens_seen": 2852272,
+      "step": 3650
+    },
+    {
+      "epoch": 7.598752598752599,
+      "grad_norm": 1.7825745344161987,
+      "learning_rate": 4.8977545237838123e-05,
+      "loss": 0.0175,
+      "num_input_tokens_seen": 2856144,
+      "step": 3655
+    },
+    {
+      "epoch": 7.609147609147609,
+      "grad_norm": 0.185651496052742,
+      "learning_rate": 4.8974764440938836e-05,
+      "loss": 0.04,
+      "num_input_tokens_seen": 2859888,
+      "step": 3660
+    },
+    {
+      "epoch": 7.61954261954262,
+      "grad_norm": 0.21885624527931213,
+      "learning_rate": 4.897197994682959e-05,
+      "loss": 0.0626,
+      "num_input_tokens_seen": 2863632,
+      "step": 3665
+    },
+    {
+      "epoch": 7.62993762993763,
+      "grad_norm": 0.21731393039226532,
+      "learning_rate": 4.8969191755939786e-05,
+      "loss": 0.0284,
+      "num_input_tokens_seen": 2867408,
+      "step": 3670
+    },
+    {
+      "epoch": 7.64033264033264,
+      "grad_norm": 2.1267433166503906,
+      "learning_rate": 4.8966399868699396e-05,
+      "loss": 0.0081,
+      "num_input_tokens_seen": 2871344,
+      "step": 3675
+    },
+    {
+      "epoch": 7.650727650727651,
+      "grad_norm": 0.050839245319366455,
+      "learning_rate": 4.8963604285538965e-05,
+      "loss": 0.0121,
+      "num_input_tokens_seen": 2875312,
+      "step": 3680
+    },
+    {
+      "epoch": 7.661122661122661,
+      "grad_norm": 0.09618256986141205,
+      "learning_rate": 4.8960805006889604e-05,
+      "loss": 0.0225,
+      "num_input_tokens_seen": 2879408,
+      "step": 3685
+    },
+    {
+      "epoch": 7.671517671517671,
+      "grad_norm": 0.04492535442113876,
+      "learning_rate": 4.8958002033183004e-05,
+      "loss": 0.0134,
+      "num_input_tokens_seen": 2883248,
+      "step": 3690
+    },
+    {
+      "epoch": 7.6819126819126815,
+      "grad_norm": 0.06583177298307419,
+      "learning_rate": 4.8955195364851414e-05,
+      "loss": 0.0206,
+      "num_input_tokens_seen": 2887184,
+      "step": 3695
+    },
+    {
+      "epoch": 7.6923076923076925,
+      "grad_norm": 0.2383924424648285,
+      "learning_rate": 4.895238500232766e-05,
+      "loss": 0.0242,
+      "num_input_tokens_seen": 2891216,
+      "step": 3700
+    },
+    {
+      "epoch": 7.702702702702703,
+      "grad_norm": 0.21089376509189606,
+      "learning_rate": 4.8949570946045143e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 2895216,
+      "step": 3705
+    },
+    {
+      "epoch": 7.713097713097713,
+      "grad_norm": 0.14790697395801544,
+      "learning_rate": 4.89467531964378e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 2899088,
+      "step": 3710
+    },
+    {
+      "epoch": 7.723492723492724,
+      "grad_norm": 4.480189323425293,
+      "learning_rate": 4.894393175394019e-05,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 2902864,
+      "step": 3715
+    },
+    {
+      "epoch": 7.733887733887734,
+      "grad_norm": 0.01597491465508938,
+      "learning_rate": 4.8941106618987406e-05,
+      "loss": 0.031,
+      "num_input_tokens_seen": 2906672,
+      "step": 3720
+    },
+    {
+      "epoch": 7.744282744282744,
+      "grad_norm": 8.454855918884277,
+      "learning_rate": 4.893827779201512e-05,
+      "loss": 0.0269,
+      "num_input_tokens_seen": 2910480,
+      "step": 3725
+    },
+    {
+      "epoch": 7.754677754677755,
+      "grad_norm": 0.021925408393144608,
+      "learning_rate": 4.893544527345957e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 2914288,
+      "step": 3730
+    },
+    {
+      "epoch": 7.765072765072765,
+      "grad_norm": 0.32002779841423035,
+      "learning_rate": 4.8932609063757563e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 2918096,
+      "step": 3735
+    },
+    {
+      "epoch": 7.775467775467775,
+      "grad_norm": 0.0655209943652153,
+      "learning_rate": 4.8929769163346484e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 2922032,
+      "step": 3740
+    },
+    {
+      "epoch": 7.785862785862786,
+      "grad_norm": 4.867824554443359,
+      "learning_rate": 4.892692557266429e-05,
+      "loss": 0.053,
+      "num_input_tokens_seen": 2925968,
+      "step": 3745
+    },
+    {
+      "epoch": 7.796257796257796,
+      "grad_norm": 2.9486098289489746,
+      "learning_rate": 4.8924078292149464e-05,
+      "loss": 0.0171,
+      "num_input_tokens_seen": 2929840,
+      "step": 3750
+    },
+    {
+      "epoch": 7.8066528066528065,
+      "grad_norm": 0.06177360564470291,
+      "learning_rate": 4.892122732224114e-05,
+      "loss": 0.0125,
+      "num_input_tokens_seen": 2933680,
+      "step": 3755
+    },
+    {
+      "epoch": 7.817047817047817,
+      "grad_norm": 0.08275920152664185,
+      "learning_rate": 4.8918372663378944e-05,
+      "loss": 0.0211,
+      "num_input_tokens_seen": 2937392,
+      "step": 3760
+    },
+    {
+      "epoch": 7.827442827442828,
+      "grad_norm": 2.2468740940093994,
+      "learning_rate": 4.89155143160031e-05,
+      "loss": 0.0611,
+      "num_input_tokens_seen": 2941424,
+      "step": 3765
+    },
+    {
+      "epoch": 7.837837837837838,
+      "grad_norm": 0.2734047472476959,
+      "learning_rate": 4.891265228055441e-05,
+      "loss": 0.0331,
+      "num_input_tokens_seen": 2945456,
+      "step": 3770
+    },
+    {
+      "epoch": 7.848232848232849,
+      "grad_norm": 7.846740245819092,
+      "learning_rate": 4.890978655747424e-05,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 2949328,
+      "step": 3775
+    },
+    {
+      "epoch": 7.858627858627859,
+      "grad_norm": 0.09487603604793549,
+      "learning_rate": 4.89069171472045e-05,
+      "loss": 0.0184,
+      "num_input_tokens_seen": 2953200,
+      "step": 3780
+    },
+    {
+      "epoch": 7.869022869022869,
+      "grad_norm": 13.041902542114258,
+      "learning_rate": 4.890404405018772e-05,
+      "loss": 0.047,
+      "num_input_tokens_seen": 2957264,
+      "step": 3785
+    },
+    {
+      "epoch": 7.879417879417879,
+      "grad_norm": 0.2936745285987854,
+      "learning_rate": 4.8901167266866934e-05,
+      "loss": 0.0216,
+      "num_input_tokens_seen": 2961168,
+      "step": 3790
+    },
+    {
+      "epoch": 7.88981288981289,
+      "grad_norm": 0.09623939543962479,
+      "learning_rate": 4.88982867976858e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 2965072,
+      "step": 3795
+    },
+    {
+      "epoch": 7.9002079002079,
+      "grad_norm": 0.015466545708477497,
+      "learning_rate": 4.889540264308852e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 2968816,
+      "step": 3800
+    },
+    {
+      "epoch": 7.9002079002079,
+      "eval_loss": 0.20151084661483765,
+      "eval_runtime": 11.7519,
+      "eval_samples_per_second": 72.839,
+      "eval_steps_per_second": 18.21,
+      "num_input_tokens_seen": 2968816,
+      "step": 3800
+    },
+    {
+      "epoch": 7.91060291060291,
+      "grad_norm": 0.01404841523617506,
+      "learning_rate": 4.889251480351986e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 2972720,
+      "step": 3805
+    },
+    {
+      "epoch": 7.920997920997921,
+      "grad_norm": 10.158882141113281,
+      "learning_rate": 4.888962327942517e-05,
+      "loss": 0.0072,
+      "num_input_tokens_seen": 2976624,
+      "step": 3810
+    },
+    {
+      "epoch": 7.9313929313929314,
+      "grad_norm": 1.5284702777862549,
+      "learning_rate": 4.8886728071250356e-05,
+      "loss": 0.0219,
+      "num_input_tokens_seen": 2980464,
+      "step": 3815
+    },
+    {
+      "epoch": 7.941787941787942,
+      "grad_norm": 0.054020170122385025,
+      "learning_rate": 4.8883829179441884e-05,
+      "loss": 0.0181,
+      "num_input_tokens_seen": 2984368,
+      "step": 3820
+    },
+    {
+      "epoch": 7.952182952182953,
+      "grad_norm": 0.026997623965144157,
+      "learning_rate": 4.888092660444682e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 2988112,
+      "step": 3825
+    },
+    {
+      "epoch": 7.962577962577963,
+      "grad_norm": 0.8101741075515747,
+      "learning_rate": 4.887802034671276e-05,
+      "loss": 0.0491,
+      "num_input_tokens_seen": 2992080,
+      "step": 3830
+    },
+    {
+      "epoch": 7.972972972972973,
+      "grad_norm": 0.01841770112514496,
+      "learning_rate": 4.88751104066879e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 2995952,
+      "step": 3835
+    },
+    {
+      "epoch": 7.983367983367984,
+      "grad_norm": 1.8316527605056763,
+      "learning_rate": 4.887219678482098e-05,
+      "loss": 0.0135,
+      "num_input_tokens_seen": 2999920,
+      "step": 3840
+    },
+    {
+      "epoch": 7.993762993762994,
+      "grad_norm": 0.02823369763791561,
+      "learning_rate": 4.8869279481561316e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 3003920,
+      "step": 3845
+    },
+    {
+      "epoch": 8.004158004158004,
+      "grad_norm": 0.06372756510972977,
+      "learning_rate": 4.88663584973588e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 3007648,
+      "step": 3850
+    },
+    {
+      "epoch": 8.014553014553014,
+      "grad_norm": 0.08825688809156418,
+      "learning_rate": 4.8863433832663874e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 3011584,
+      "step": 3855
+    },
+    {
+      "epoch": 8.024948024948024,
+      "grad_norm": 2.7013721466064453,
+      "learning_rate": 4.886050548792757e-05,
+      "loss": 0.0523,
+      "num_input_tokens_seen": 3015456,
+      "step": 3860
+    },
+    {
+      "epoch": 8.035343035343036,
+      "grad_norm": 0.8092334270477295,
+      "learning_rate": 4.8857573463601465e-05,
+      "loss": 0.0184,
+      "num_input_tokens_seen": 3019328,
+      "step": 3865
+    },
+    {
+      "epoch": 8.045738045738046,
+      "grad_norm": 2.392627716064453,
+      "learning_rate": 4.885463776013772e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 3023168,
+      "step": 3870
+    },
+    {
+      "epoch": 8.056133056133056,
+      "grad_norm": 0.03118458017706871,
+      "learning_rate": 4.8851698377989056e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 3027200,
+      "step": 3875
+    },
+    {
+      "epoch": 8.066528066528067,
+      "grad_norm": 0.036838531494140625,
+      "learning_rate": 4.884875531760876e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 3030880,
+      "step": 3880
+    },
+    {
+      "epoch": 8.076923076923077,
+      "grad_norm": 0.03974492847919464,
+      "learning_rate": 4.88458085794507e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 3034752,
+      "step": 3885
+    },
+    {
+      "epoch": 8.087318087318087,
+      "grad_norm": 5.567601203918457,
+      "learning_rate": 4.884285816396929e-05,
+      "loss": 0.0477,
+      "num_input_tokens_seen": 3038528,
+      "step": 3890
+    },
+    {
+      "epoch": 8.097713097713097,
+      "grad_norm": 0.014081713743507862,
+      "learning_rate": 4.8839904071619526e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 3042464,
+      "step": 3895
+    },
+    {
+      "epoch": 8.108108108108109,
+      "grad_norm": 0.012380711734294891,
+      "learning_rate": 4.8836946302856955e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 3046368,
+      "step": 3900
+    },
+    {
+      "epoch": 8.118503118503119,
+      "grad_norm": 0.014421866275370121,
+      "learning_rate": 4.8833984858137715e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 3050336,
+      "step": 3905
+    },
+    {
+      "epoch": 8.128898128898129,
+      "grad_norm": 0.0100218765437603,
+      "learning_rate": 4.8831019737918494e-05,
+      "loss": 0.0089,
+      "num_input_tokens_seen": 3054240,
+      "step": 3910
+    },
+    {
+      "epoch": 8.13929313929314,
+      "grad_norm": 5.467347145080566,
+      "learning_rate": 4.882805094265655e-05,
+      "loss": 0.0088,
+      "num_input_tokens_seen": 3058176,
+      "step": 3915
+    },
+    {
+      "epoch": 8.14968814968815,
+      "grad_norm": 0.024442464113235474,
+      "learning_rate": 4.8825078472809706e-05,
+      "loss": 0.0066,
+      "num_input_tokens_seen": 3062144,
+      "step": 3920
+    },
+    {
+      "epoch": 8.16008316008316,
+      "grad_norm": 0.7753127217292786,
+      "learning_rate": 4.882210232883635e-05,
+      "loss": 0.0283,
+      "num_input_tokens_seen": 3065984,
+      "step": 3925
+    },
+    {
+      "epoch": 8.170478170478171,
+      "grad_norm": 0.4849783480167389,
+      "learning_rate": 4.881912251119546e-05,
+      "loss": 0.051,
+      "num_input_tokens_seen": 3069920,
+      "step": 3930
+    },
+    {
+      "epoch": 8.180873180873181,
+      "grad_norm": 2.7070422172546387,
+      "learning_rate": 4.881613902034654e-05,
+      "loss": 0.0176,
+      "num_input_tokens_seen": 3073856,
+      "step": 3935
+    },
+    {
+      "epoch": 8.191268191268192,
+      "grad_norm": 0.8864293098449707,
+      "learning_rate": 4.88131518567497e-05,
+      "loss": 0.0167,
+      "num_input_tokens_seen": 3077760,
+      "step": 3940
+    },
+    {
+      "epoch": 8.201663201663202,
+      "grad_norm": 0.07755299657583237,
+      "learning_rate": 4.881016102086558e-05,
+      "loss": 0.0174,
+      "num_input_tokens_seen": 3081632,
+      "step": 3945
+    },
+    {
+      "epoch": 8.212058212058212,
+      "grad_norm": 0.08531589806079865,
+      "learning_rate": 4.8807166513155425e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 3085472,
+      "step": 3950
+    },
+    {
+      "epoch": 8.222453222453222,
+      "grad_norm": 0.10235360264778137,
+      "learning_rate": 4.8804168334081004e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 3089376,
+      "step": 3955
+    },
+    {
+      "epoch": 8.232848232848234,
+      "grad_norm": 0.030183851718902588,
+      "learning_rate": 4.880116648410468e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 3093408,
+      "step": 3960
+    },
+    {
+      "epoch": 8.243243243243244,
+      "grad_norm": 0.02350478619337082,
+      "learning_rate": 4.879816096368939e-05,
+      "loss": 0.0216,
+      "num_input_tokens_seen": 3097280,
+      "step": 3965
+    },
+    {
+      "epoch": 8.253638253638254,
+      "grad_norm": 0.02739868313074112,
+      "learning_rate": 4.879515177329861e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 3101152,
+      "step": 3970
+    },
+    {
+      "epoch": 8.264033264033264,
+      "grad_norm": 0.03188429772853851,
+      "learning_rate": 4.8792138913396394e-05,
+      "loss": 0.0089,
+      "num_input_tokens_seen": 3104896,
+      "step": 3975
+    },
+    {
+      "epoch": 8.274428274428274,
+      "grad_norm": 0.06726208329200745,
+      "learning_rate": 4.8789122384447374e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 3108704,
+      "step": 3980
+    },
+    {
+      "epoch": 8.284823284823284,
+      "grad_norm": 0.11809299886226654,
+      "learning_rate": 4.878610218691673e-05,
+      "loss": 0.028,
+      "num_input_tokens_seen": 3112512,
+      "step": 3985
+    },
+    {
+      "epoch": 8.295218295218294,
+      "grad_norm": 0.19482602179050446,
+      "learning_rate": 4.87830783212702e-05,
+      "loss": 0.0127,
+      "num_input_tokens_seen": 3116544,
+      "step": 3990
+    },
+    {
+      "epoch": 8.305613305613306,
+      "grad_norm": 0.10719515383243561,
+      "learning_rate": 4.878005078797413e-05,
+      "loss": 0.0134,
+      "num_input_tokens_seen": 3120384,
+      "step": 3995
+    },
+    {
+      "epoch": 8.316008316008316,
+      "grad_norm": 21.43325424194336,
+      "learning_rate": 4.877701958749539e-05,
+      "loss": 0.0336,
+      "num_input_tokens_seen": 3124448,
+      "step": 4000
+    },
+    {
+      "epoch": 8.316008316008316,
+      "eval_loss": 0.20573952794075012,
+      "eval_runtime": 11.8054,
+      "eval_samples_per_second": 72.509,
+      "eval_steps_per_second": 18.127,
+      "num_input_tokens_seen": 3124448,
+      "step": 4000
+    },
+    {
+      "epoch": 8.326403326403327,
+      "grad_norm": 0.45844659209251404,
+      "learning_rate": 4.877398472030142e-05,
+      "loss": 0.0193,
+      "num_input_tokens_seen": 3128416,
+      "step": 4005
+    },
+    {
+      "epoch": 8.336798336798337,
+      "grad_norm": 0.13104534149169922,
+      "learning_rate": 4.877094618686024e-05,
+      "loss": 0.0307,
+      "num_input_tokens_seen": 3132256,
+      "step": 4010
+    },
+    {
+      "epoch": 8.347193347193347,
+      "grad_norm": 0.0653214231133461,
+      "learning_rate": 4.876790398764045e-05,
+      "loss": 0.0159,
+      "num_input_tokens_seen": 3135968,
+      "step": 4015
+    },
+    {
+      "epoch": 8.357588357588357,
+      "grad_norm": 0.12485301494598389,
+      "learning_rate": 4.8764858123111167e-05,
+      "loss": 0.026,
+      "num_input_tokens_seen": 3139904,
+      "step": 4020
+    },
+    {
+      "epoch": 8.367983367983369,
+      "grad_norm": 0.13592949509620667,
+      "learning_rate": 4.876180859374212e-05,
+      "loss": 0.0133,
+      "num_input_tokens_seen": 3143744,
+      "step": 4025
+    },
+    {
+      "epoch": 8.378378378378379,
+      "grad_norm": 0.4120209515094757,
+      "learning_rate": 4.875875540000357e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 3147680,
+      "step": 4030
+    },
+    {
+      "epoch": 8.388773388773389,
+      "grad_norm": 0.19209295511245728,
+      "learning_rate": 4.8755698542366376e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 3151520,
+      "step": 4035
+    },
+    {
+      "epoch": 8.3991683991684,
+      "grad_norm": 0.02903851307928562,
+      "learning_rate": 4.875263802130193e-05,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 3155488,
+      "step": 4040
+    },
+    {
+      "epoch": 8.40956340956341,
+      "grad_norm": 0.024972299113869667,
+      "learning_rate": 4.8749573837282207e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 3159264,
+      "step": 4045
+    },
+    {
+      "epoch": 8.41995841995842,
+      "grad_norm": 0.03302966430783272,
+      "learning_rate": 4.874650599077974e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 3163200,
+      "step": 4050
+    },
+    {
+      "epoch": 8.43035343035343,
+      "grad_norm": 0.23276501893997192,
+      "learning_rate": 4.874343448226764e-05,
+      "loss": 0.0061,
+      "num_input_tokens_seen": 3167296,
+      "step": 4055
+    },
+    {
+      "epoch": 8.440748440748441,
+      "grad_norm": 0.013871944509446621,
+      "learning_rate": 4.874035931221955e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 3171232,
+      "step": 4060
+    },
+    {
+      "epoch": 8.451143451143452,
+      "grad_norm": 0.011728519573807716,
+      "learning_rate": 4.8737280481109724e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 3175168,
+      "step": 4065
+    },
+    {
+      "epoch": 8.461538461538462,
+      "grad_norm": 0.008907628245651722,
+      "learning_rate": 4.873419798941294e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 3178944,
+      "step": 4070
+    },
+    {
+      "epoch": 8.471933471933472,
+      "grad_norm": 0.026969369500875473,
+      "learning_rate": 4.873111183760458e-05,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 3182848,
+      "step": 4075
+    },
+    {
+      "epoch": 8.482328482328482,
+      "grad_norm": 0.8197475075721741,
+      "learning_rate": 4.8728022026160537e-05,
+      "loss": 0.0223,
+      "num_input_tokens_seen": 3186720,
+      "step": 4080
+    },
+    {
+      "epoch": 8.492723492723492,
+      "grad_norm": 0.04919741302728653,
+      "learning_rate": 4.872492855555732e-05,
+      "loss": 0.041,
+      "num_input_tokens_seen": 3190560,
+      "step": 4085
+    },
+    {
+      "epoch": 8.503118503118504,
+      "grad_norm": 0.06035103648900986,
+      "learning_rate": 4.8721831426271956e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 3194464,
+      "step": 4090
+    },
+    {
+      "epoch": 8.513513513513514,
+      "grad_norm": 9.815949440002441,
+      "learning_rate": 4.87187306387821e-05,
+      "loss": 0.015,
+      "num_input_tokens_seen": 3198464,
+      "step": 4095
+    },
+    {
+      "epoch": 8.523908523908524,
+      "grad_norm": 0.03221115097403526,
+      "learning_rate": 4.87156261935659e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 3202400,
+      "step": 4100
+    },
+    {
+      "epoch": 8.534303534303534,
+      "grad_norm": 0.8433281779289246,
+      "learning_rate": 4.871251809110211e-05,
+      "loss": 0.0566,
+      "num_input_tokens_seen": 3206336,
+      "step": 4105
+    },
+    {
+      "epoch": 8.544698544698544,
+      "grad_norm": 0.051093049347400665,
+      "learning_rate": 4.8709406331870044e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 3210304,
+      "step": 4110
+    },
+    {
+      "epoch": 8.555093555093555,
+      "grad_norm": 0.045183662325143814,
+      "learning_rate": 4.8706290916349574e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 3214208,
+      "step": 4115
+    },
+    {
+      "epoch": 8.565488565488565,
+      "grad_norm": 0.04777225852012634,
+      "learning_rate": 4.8703171845021134e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 3218016,
+      "step": 4120
+    },
+    {
+      "epoch": 8.575883575883577,
+      "grad_norm": 1.2434628009796143,
+      "learning_rate": 4.870004911836572e-05,
+      "loss": 0.0409,
+      "num_input_tokens_seen": 3221760,
+      "step": 4125
+    },
+    {
+      "epoch": 8.586278586278587,
+      "grad_norm": 0.19871462881565094,
+      "learning_rate": 4.869692273686489e-05,
+      "loss": 0.0162,
+      "num_input_tokens_seen": 3225824,
+      "step": 4130
+    },
+    {
+      "epoch": 8.596673596673597,
+      "grad_norm": 0.05036256089806557,
+      "learning_rate": 4.869379270100079e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 3229664,
+      "step": 4135
+    },
+    {
+      "epoch": 8.607068607068607,
+      "grad_norm": 0.07861363142728806,
+      "learning_rate": 4.86906590112561e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 3233632,
+      "step": 4140
+    },
+    {
+      "epoch": 8.617463617463617,
+      "grad_norm": 0.08010400086641312,
+      "learning_rate": 4.8687521668114064e-05,
+      "loss": 0.0082,
+      "num_input_tokens_seen": 3237440,
+      "step": 4145
+    },
+    {
+      "epoch": 8.627858627858627,
+      "grad_norm": 0.04196079447865486,
+      "learning_rate": 4.868438067205853e-05,
+      "loss": 0.0328,
+      "num_input_tokens_seen": 3241280,
+      "step": 4150
+    },
+    {
+      "epoch": 8.638253638253639,
+      "grad_norm": 0.036268796771764755,
+      "learning_rate": 4.8681236023573844e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 3245248,
+      "step": 4155
+    },
+    {
+      "epoch": 8.64864864864865,
+      "grad_norm": 0.038483522832393646,
+      "learning_rate": 4.867808772314497e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 3249120,
+      "step": 4160
+    },
+    {
+      "epoch": 8.65904365904366,
+      "grad_norm": 0.1148298904299736,
+      "learning_rate": 4.867493577125741e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 3253152,
+      "step": 4165
+    },
+    {
+      "epoch": 8.66943866943867,
+      "grad_norm": 0.048769328743219376,
+      "learning_rate": 4.867178016839725e-05,
+      "loss": 0.025,
+      "num_input_tokens_seen": 3257056,
+      "step": 4170
+    },
+    {
+      "epoch": 8.67983367983368,
+      "grad_norm": 0.049257900565862656,
+      "learning_rate": 4.8668620915051094e-05,
+      "loss": 0.06,
+      "num_input_tokens_seen": 3260960,
+      "step": 4175
+    },
+    {
+      "epoch": 8.69022869022869,
+      "grad_norm": 0.03110208921134472,
+      "learning_rate": 4.866545801170616e-05,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 3264832,
+      "step": 4180
+    },
+    {
+      "epoch": 8.700623700623701,
+      "grad_norm": 8.486109733581543,
+      "learning_rate": 4.86622914588502e-05,
+      "loss": 0.0116,
+      "num_input_tokens_seen": 3268736,
+      "step": 4185
+    },
+    {
+      "epoch": 8.711018711018712,
+      "grad_norm": 0.08196869492530823,
+      "learning_rate": 4.865912125697154e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 3272672,
+      "step": 4190
+    },
+    {
+      "epoch": 8.721413721413722,
+      "grad_norm": 3.5260021686553955,
+      "learning_rate": 4.865594740655907e-05,
+      "loss": 0.0393,
+      "num_input_tokens_seen": 3276448,
+      "step": 4195
+    },
+    {
+      "epoch": 8.731808731808732,
+      "grad_norm": 1.599503993988037,
+      "learning_rate": 4.865276990810222e-05,
+      "loss": 0.0461,
+      "num_input_tokens_seen": 3280320,
+      "step": 4200
+    },
+    {
+      "epoch": 8.731808731808732,
+      "eval_loss": 0.2184087485074997,
+      "eval_runtime": 11.7672,
+      "eval_samples_per_second": 72.745,
+      "eval_steps_per_second": 18.186,
+      "num_input_tokens_seen": 3280320,
+      "step": 4200
+    },
+    {
+      "epoch": 8.742203742203742,
+      "grad_norm": 0.08391126245260239,
+      "learning_rate": 4.8649588762091016e-05,
+      "loss": 0.0395,
+      "num_input_tokens_seen": 3284192,
+      "step": 4205
+    },
+    {
+      "epoch": 8.752598752598752,
+      "grad_norm": 4.677314758300781,
+      "learning_rate": 4.8646403969016016e-05,
+      "loss": 0.035,
+      "num_input_tokens_seen": 3288096,
+      "step": 4210
+    },
+    {
+      "epoch": 8.762993762993762,
+      "grad_norm": 0.3902350068092346,
+      "learning_rate": 4.864321552936838e-05,
+      "loss": 0.0196,
+      "num_input_tokens_seen": 3291968,
+      "step": 4215
+    },
+    {
+      "epoch": 8.773388773388774,
+      "grad_norm": 0.21717128157615662,
+      "learning_rate": 4.864002344363978e-05,
+      "loss": 0.0315,
+      "num_input_tokens_seen": 3295936,
+      "step": 4220
+    },
+    {
+      "epoch": 8.783783783783784,
+      "grad_norm": 0.18836237490177155,
+      "learning_rate": 4.863682771232248e-05,
+      "loss": 0.0209,
+      "num_input_tokens_seen": 3299904,
+      "step": 4225
+    },
+    {
+      "epoch": 8.794178794178794,
+      "grad_norm": 0.04174836724996567,
+      "learning_rate": 4.8633628335909324e-05,
+      "loss": 0.0184,
+      "num_input_tokens_seen": 3303776,
+      "step": 4230
+    },
+    {
+      "epoch": 8.804573804573804,
+      "grad_norm": 0.07095040380954742,
+      "learning_rate": 4.8630425314893676e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 3307840,
+      "step": 4235
+    },
+    {
+      "epoch": 8.814968814968815,
+      "grad_norm": 0.031969379633665085,
+      "learning_rate": 4.862721864976948e-05,
+      "loss": 0.05,
+      "num_input_tokens_seen": 3311744,
+      "step": 4240
+    },
+    {
+      "epoch": 8.825363825363825,
+      "grad_norm": 3.7988147735595703,
+      "learning_rate": 4.862400834103125e-05,
+      "loss": 0.028,
+      "num_input_tokens_seen": 3315776,
+      "step": 4245
+    },
+    {
+      "epoch": 8.835758835758837,
+      "grad_norm": 0.058655738830566406,
+      "learning_rate": 4.862079438917406e-05,
+      "loss": 0.0082,
+      "num_input_tokens_seen": 3319872,
+      "step": 4250
+    },
+    {
+      "epoch": 8.846153846153847,
+      "grad_norm": 0.8626923561096191,
+      "learning_rate": 4.8617576794693536e-05,
+      "loss": 0.0327,
+      "num_input_tokens_seen": 3323744,
+      "step": 4255
+    },
+    {
+      "epoch": 8.856548856548857,
+      "grad_norm": 0.0333566889166832,
+      "learning_rate": 4.8614355558085875e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 3327520,
+      "step": 4260
+    },
+    {
+      "epoch": 8.866943866943867,
+      "grad_norm": 0.034962013363838196,
+      "learning_rate": 4.861113067984783e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 3331616,
+      "step": 4265
+    },
+    {
+      "epoch": 8.877338877338877,
+      "grad_norm": 0.019482100382447243,
+      "learning_rate": 4.860790216047671e-05,
+      "loss": 0.0149,
+      "num_input_tokens_seen": 3335520,
+      "step": 4270
+    },
+    {
+      "epoch": 8.887733887733887,
+      "grad_norm": 0.020854836329817772,
+      "learning_rate": 4.860467000047041e-05,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 3339456,
+      "step": 4275
+    },
+    {
+      "epoch": 8.898128898128899,
+      "grad_norm": 0.8835586309432983,
+      "learning_rate": 4.860143420032737e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 3343424,
+      "step": 4280
+    },
+    {
+      "epoch": 8.90852390852391,
+      "grad_norm": 0.013219108805060387,
+      "learning_rate": 4.859819476054657e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 3347392,
+      "step": 4285
+    },
+    {
+      "epoch": 8.91891891891892,
+      "grad_norm": 0.0701918676495552,
+      "learning_rate": 4.859495168162758e-05,
+      "loss": 0.0257,
+      "num_input_tokens_seen": 3351232,
+      "step": 4290
+    },
+    {
+      "epoch": 8.92931392931393,
+      "grad_norm": 0.0749225988984108,
+      "learning_rate": 4.859170496407054e-05,
+      "loss": 0.0318,
+      "num_input_tokens_seen": 3355200,
+      "step": 4295
+    },
+    {
+      "epoch": 8.93970893970894,
+      "grad_norm": 0.28515109419822693,
+      "learning_rate": 4.8588454608376114e-05,
+      "loss": 0.0193,
+      "num_input_tokens_seen": 3359168,
+      "step": 4300
+    },
+    {
+      "epoch": 8.95010395010395,
+      "grad_norm": 0.9071269035339355,
+      "learning_rate": 4.8585200615045555e-05,
+      "loss": 0.0076,
+      "num_input_tokens_seen": 3363104,
+      "step": 4305
+    },
+    {
+      "epoch": 8.96049896049896,
+      "grad_norm": 0.1736389845609665,
+      "learning_rate": 4.8581942984580674e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 3367008,
+      "step": 4310
+    },
+    {
+      "epoch": 8.970893970893972,
+      "grad_norm": 0.023569254204630852,
+      "learning_rate": 4.857868171748384e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 3370720,
+      "step": 4315
+    },
+    {
+      "epoch": 8.981288981288982,
+      "grad_norm": 0.019453419372439384,
+      "learning_rate": 4.8575416814257976e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 3374656,
+      "step": 4320
+    },
+    {
+      "epoch": 8.991683991683992,
+      "grad_norm": 0.008834749460220337,
+      "learning_rate": 4.857214827540657e-05,
+      "loss": 0.0159,
+      "num_input_tokens_seen": 3378592,
+      "step": 4325
+    },
+    {
+      "epoch": 9.002079002079002,
+      "grad_norm": 0.015051688998937607,
+      "learning_rate": 4.856887610143367e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 3382448,
+      "step": 4330
+    },
+    {
+      "epoch": 9.012474012474012,
+      "grad_norm": 3.371316909790039,
+      "learning_rate": 4.8565600292843896e-05,
+      "loss": 0.0573,
+      "num_input_tokens_seen": 3386512,
+      "step": 4335
+    },
+    {
+      "epoch": 9.022869022869022,
+      "grad_norm": 0.07159621268510818,
+      "learning_rate": 4.856232085014241e-05,
+      "loss": 0.0077,
+      "num_input_tokens_seen": 3390416,
+      "step": 4340
+    },
+    {
+      "epoch": 9.033264033264032,
+      "grad_norm": 0.30193865299224854,
+      "learning_rate": 4.855903777383495e-05,
+      "loss": 0.0085,
+      "num_input_tokens_seen": 3394320,
+      "step": 4345
+    },
+    {
+      "epoch": 9.043659043659044,
+      "grad_norm": 0.030544331297278404,
+      "learning_rate": 4.85557510644278e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 3398352,
+      "step": 4350
+    },
+    {
+      "epoch": 9.054054054054054,
+      "grad_norm": 0.022635221481323242,
+      "learning_rate": 4.855246072242782e-05,
+      "loss": 0.0256,
+      "num_input_tokens_seen": 3402224,
+      "step": 4355
+    },
+    {
+      "epoch": 9.064449064449065,
+      "grad_norm": 0.06966003775596619,
+      "learning_rate": 4.8549166748342414e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 3406192,
+      "step": 4360
+    },
+    {
+      "epoch": 9.074844074844075,
+      "grad_norm": 0.020658381283283234,
+      "learning_rate": 4.8545869142679556e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 3409904,
+      "step": 4365
+    },
+    {
+      "epoch": 9.085239085239085,
+      "grad_norm": 0.1102900356054306,
+      "learning_rate": 4.8542567905947776e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 3413744,
+      "step": 4370
+    },
+    {
+      "epoch": 9.095634095634095,
+      "grad_norm": 0.04431593790650368,
+      "learning_rate": 4.853926303865618e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 3417616,
+      "step": 4375
+    },
+    {
+      "epoch": 9.106029106029107,
+      "grad_norm": 0.09042268991470337,
+      "learning_rate": 4.853595454131441e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 3421648,
+      "step": 4380
+    },
+    {
+      "epoch": 9.116424116424117,
+      "grad_norm": 0.011607572436332703,
+      "learning_rate": 4.8532642414432674e-05,
+      "loss": 0.0377,
+      "num_input_tokens_seen": 3425584,
+      "step": 4385
+    },
+    {
+      "epoch": 9.126819126819127,
+      "grad_norm": 0.00593083119019866,
+      "learning_rate": 4.8529326658521754e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 3429488,
+      "step": 4390
+    },
+    {
+      "epoch": 9.137214137214137,
+      "grad_norm": 0.03808402642607689,
+      "learning_rate": 4.8526007274092965e-05,
+      "loss": 0.0286,
+      "num_input_tokens_seen": 3433328,
+      "step": 4395
+    },
+    {
+      "epoch": 9.147609147609147,
+      "grad_norm": 0.018898162990808487,
+      "learning_rate": 4.852268426165822e-05,
+      "loss": 0.0188,
+      "num_input_tokens_seen": 3437072,
+      "step": 4400
+    },
+    {
+      "epoch": 9.147609147609147,
+      "eval_loss": 0.2262478470802307,
+      "eval_runtime": 11.7877,
+      "eval_samples_per_second": 72.618,
+      "eval_steps_per_second": 18.154,
+      "num_input_tokens_seen": 3437072,
+      "step": 4400
+    },
+    {
+      "epoch": 9.158004158004157,
+      "grad_norm": 0.10180994868278503,
+      "learning_rate": 4.851935762172995e-05,
+      "loss": 0.0385,
+      "num_input_tokens_seen": 3440880,
+      "step": 4405
+    },
+    {
+      "epoch": 9.16839916839917,
+      "grad_norm": 0.04300263151526451,
+      "learning_rate": 4.8516027354821175e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 3444688,
+      "step": 4410
+    },
+    {
+      "epoch": 9.17879417879418,
+      "grad_norm": 0.05125404894351959,
+      "learning_rate": 4.851269346144546e-05,
+      "loss": 0.0199,
+      "num_input_tokens_seen": 3448528,
+      "step": 4415
+    },
+    {
+      "epoch": 9.18918918918919,
+      "grad_norm": 0.2032732218503952,
+      "learning_rate": 4.850935594211693e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 3452272,
+      "step": 4420
+    },
+    {
+      "epoch": 9.1995841995842,
+      "grad_norm": 0.831665575504303,
+      "learning_rate": 4.850601479735029e-05,
+      "loss": 0.0188,
+      "num_input_tokens_seen": 3456112,
+      "step": 4425
+    },
+    {
+      "epoch": 9.20997920997921,
+      "grad_norm": 0.08782538771629333,
+      "learning_rate": 4.850267002766076e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 3460048,
+      "step": 4430
+    },
+    {
+      "epoch": 9.22037422037422,
+      "grad_norm": 0.03403269872069359,
+      "learning_rate": 4.849932163356417e-05,
+      "loss": 0.02,
+      "num_input_tokens_seen": 3463952,
+      "step": 4435
+    },
+    {
+      "epoch": 9.23076923076923,
+      "grad_norm": 0.9030600786209106,
+      "learning_rate": 4.8495969615576864e-05,
+      "loss": 0.0398,
+      "num_input_tokens_seen": 3467888,
+      "step": 4440
+    },
+    {
+      "epoch": 9.241164241164242,
+      "grad_norm": 0.09359905123710632,
+      "learning_rate": 4.849261397421577e-05,
+      "loss": 0.0236,
+      "num_input_tokens_seen": 3471920,
+      "step": 4445
+    },
+    {
+      "epoch": 9.251559251559252,
+      "grad_norm": 0.15463358163833618,
+      "learning_rate": 4.848925470999839e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 3475856,
+      "step": 4450
+    },
+    {
+      "epoch": 9.261954261954262,
+      "grad_norm": 0.12877854704856873,
+      "learning_rate": 4.848589182344273e-05,
+      "loss": 0.0366,
+      "num_input_tokens_seen": 3479792,
+      "step": 4455
+    },
+    {
+      "epoch": 9.272349272349272,
+      "grad_norm": 0.12286125868558884,
+      "learning_rate": 4.848252531506742e-05,
+      "loss": 0.0181,
+      "num_input_tokens_seen": 3483824,
+      "step": 4460
+    },
+    {
+      "epoch": 9.282744282744282,
+      "grad_norm": 0.12651143968105316,
+      "learning_rate": 4.847915518539161e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 3487792,
+      "step": 4465
+    },
+    {
+      "epoch": 9.293139293139292,
+      "grad_norm": 0.0358903594315052,
+      "learning_rate": 4.847578143493501e-05,
+      "loss": 0.0167,
+      "num_input_tokens_seen": 3491632,
+      "step": 4470
+    },
+    {
+      "epoch": 9.303534303534304,
+      "grad_norm": 0.076701819896698,
+      "learning_rate": 4.847240406421789e-05,
+      "loss": 0.0149,
+      "num_input_tokens_seen": 3495568,
+      "step": 4475
+    },
+    {
+      "epoch": 9.313929313929314,
+      "grad_norm": 2.334594488143921,
+      "learning_rate": 4.84690230737611e-05,
+      "loss": 0.021,
+      "num_input_tokens_seen": 3499472,
+      "step": 4480
+    },
+    {
+      "epoch": 9.324324324324325,
+      "grad_norm": 2.6680233478546143,
+      "learning_rate": 4.846563846408602e-05,
+      "loss": 0.0274,
+      "num_input_tokens_seen": 3503216,
+      "step": 4485
+    },
+    {
+      "epoch": 9.334719334719335,
+      "grad_norm": 0.15933625400066376,
+      "learning_rate": 4.84622502357146e-05,
+      "loss": 0.0079,
+      "num_input_tokens_seen": 3507120,
+      "step": 4490
+    },
+    {
+      "epoch": 9.345114345114345,
+      "grad_norm": 8.068140029907227,
+      "learning_rate": 4.8458858389169345e-05,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 3511312,
+      "step": 4495
+    },
+    {
+      "epoch": 9.355509355509355,
+      "grad_norm": 0.036514006555080414,
+      "learning_rate": 4.8455462924973334e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 3515088,
+      "step": 4500
+    },
+    {
+      "epoch": 9.365904365904367,
+      "grad_norm": 0.022902660071849823,
+      "learning_rate": 4.845206384365018e-05,
+      "loss": 0.0368,
+      "num_input_tokens_seen": 3519088,
+      "step": 4505
+    },
+    {
+      "epoch": 9.376299376299377,
+      "grad_norm": 0.02011037804186344,
+      "learning_rate": 4.844866114572405e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 3523152,
+      "step": 4510
+    },
+    {
+      "epoch": 9.386694386694387,
+      "grad_norm": 0.029614636674523354,
+      "learning_rate": 4.8445254831719706e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 3526992,
+      "step": 4515
+    },
+    {
+      "epoch": 9.397089397089397,
+      "grad_norm": 0.014102369546890259,
+      "learning_rate": 4.8441844902162434e-05,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 3530896,
+      "step": 4520
+    },
+    {
+      "epoch": 9.407484407484407,
+      "grad_norm": 0.021707022562623024,
+      "learning_rate": 4.843843135757809e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 3534672,
+      "step": 4525
+    },
+    {
+      "epoch": 9.417879417879417,
+      "grad_norm": 0.022668978199362755,
+      "learning_rate": 4.843501419849308e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 3538640,
+      "step": 4530
+    },
+    {
+      "epoch": 9.428274428274428,
+      "grad_norm": 0.02287030592560768,
+      "learning_rate": 4.8431593425434386e-05,
+      "loss": 0.0132,
+      "num_input_tokens_seen": 3542704,
+      "step": 4535
+    },
+    {
+      "epoch": 9.43866943866944,
+      "grad_norm": 0.008507324382662773,
+      "learning_rate": 4.8428169038929526e-05,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 3546608,
+      "step": 4540
+    },
+    {
+      "epoch": 9.44906444906445,
+      "grad_norm": 0.01098943967372179,
+      "learning_rate": 4.8424741039506575e-05,
+      "loss": 0.0104,
+      "num_input_tokens_seen": 3550288,
+      "step": 4545
+    },
+    {
+      "epoch": 9.45945945945946,
+      "grad_norm": 6.795072555541992,
+      "learning_rate": 4.842130942769419e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 3554320,
+      "step": 4550
+    },
+    {
+      "epoch": 9.46985446985447,
+      "grad_norm": 0.019290179014205933,
+      "learning_rate": 4.841787420402156e-05,
+      "loss": 0.0124,
+      "num_input_tokens_seen": 3558384,
+      "step": 4555
+    },
+    {
+      "epoch": 9.48024948024948,
+      "grad_norm": 0.8472408056259155,
+      "learning_rate": 4.841443536901844e-05,
+      "loss": 0.0473,
+      "num_input_tokens_seen": 3562352,
+      "step": 4560
+    },
+    {
+      "epoch": 9.49064449064449,
+      "grad_norm": 0.04898029938340187,
+      "learning_rate": 4.841099292321514e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 3566512,
+      "step": 4565
+    },
+    {
+      "epoch": 9.5010395010395,
+      "grad_norm": 0.06851211935281754,
+      "learning_rate": 4.8407546867142525e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 3570256,
+      "step": 4570
+    },
+    {
+      "epoch": 9.511434511434512,
+      "grad_norm": 0.08732462674379349,
+      "learning_rate": 4.840409720133203e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 3574288,
+      "step": 4575
+    },
+    {
+      "epoch": 9.521829521829522,
+      "grad_norm": 0.03713575750589371,
+      "learning_rate": 4.8400643926315634e-05,
+      "loss": 0.0202,
+      "num_input_tokens_seen": 3578096,
+      "step": 4580
+    },
+    {
+      "epoch": 9.532224532224532,
+      "grad_norm": 0.038661498576402664,
+      "learning_rate": 4.839718704262587e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 3582032,
+      "step": 4585
+    },
+    {
+      "epoch": 9.542619542619542,
+      "grad_norm": 11.609217643737793,
+      "learning_rate": 4.839372655079585e-05,
+      "loss": 0.0272,
+      "num_input_tokens_seen": 3585904,
+      "step": 4590
+    },
+    {
+      "epoch": 9.553014553014552,
+      "grad_norm": 0.22628618776798248,
+      "learning_rate": 4.83902624513592e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 3589808,
+      "step": 4595
+    },
+    {
+      "epoch": 9.563409563409563,
+      "grad_norm": 0.12083947658538818,
+      "learning_rate": 4.838679474485014e-05,
+      "loss": 0.0194,
+      "num_input_tokens_seen": 3593520,
+      "step": 4600
+    },
+    {
+      "epoch": 9.563409563409563,
+      "eval_loss": 0.28447291254997253,
+      "eval_runtime": 11.7282,
+      "eval_samples_per_second": 72.986,
+      "eval_steps_per_second": 18.247,
+      "num_input_tokens_seen": 3593520,
+      "step": 4600
+    },
+    {
+      "epoch": 9.573804573804575,
+      "grad_norm": 0.29484161734580994,
+      "learning_rate": 4.838332343180343e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 3597392,
+      "step": 4605
+    },
+    {
+      "epoch": 9.584199584199585,
+      "grad_norm": 19.186925888061523,
+      "learning_rate": 4.83798485127544e-05,
+      "loss": 0.0155,
+      "num_input_tokens_seen": 3601360,
+      "step": 4610
+    },
+    {
+      "epoch": 9.594594594594595,
+      "grad_norm": 0.01331101544201374,
+      "learning_rate": 4.837636998823892e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 3605232,
+      "step": 4615
+    },
+    {
+      "epoch": 9.604989604989605,
+      "grad_norm": 0.006607442628592253,
+      "learning_rate": 4.8372887858793414e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 3609136,
+      "step": 4620
+    },
+    {
+      "epoch": 9.615384615384615,
+      "grad_norm": 0.681186854839325,
+      "learning_rate": 4.836940212495489e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 3612976,
+      "step": 4625
+    },
+    {
+      "epoch": 9.625779625779625,
+      "grad_norm": 0.33931413292884827,
+      "learning_rate": 4.836591278726087e-05,
+      "loss": 0.0592,
+      "num_input_tokens_seen": 3616944,
+      "step": 4630
+    },
+    {
+      "epoch": 9.636174636174637,
+      "grad_norm": 0.003908243030309677,
+      "learning_rate": 4.836241984624947e-05,
+      "loss": 0.0233,
+      "num_input_tokens_seen": 3620880,
+      "step": 4635
+    },
+    {
+      "epoch": 9.646569646569647,
+      "grad_norm": 12.644734382629395,
+      "learning_rate": 4.8358923302459336e-05,
+      "loss": 0.0346,
+      "num_input_tokens_seen": 3624752,
+      "step": 4640
+    },
+    {
+      "epoch": 9.656964656964657,
+      "grad_norm": 9.816272735595703,
+      "learning_rate": 4.835542315642968e-05,
+      "loss": 0.0469,
+      "num_input_tokens_seen": 3628688,
+      "step": 4645
+    },
+    {
+      "epoch": 9.667359667359667,
+      "grad_norm": 0.0536026656627655,
+      "learning_rate": 4.8351919408700274e-05,
+      "loss": 0.0164,
+      "num_input_tokens_seen": 3632560,
+      "step": 4650
+    },
+    {
+      "epoch": 9.677754677754677,
+      "grad_norm": 0.0873088613152504,
+      "learning_rate": 4.834841205981144e-05,
+      "loss": 0.0242,
+      "num_input_tokens_seen": 3636592,
+      "step": 4655
+    },
+    {
+      "epoch": 9.688149688149688,
+      "grad_norm": 0.10025003552436829,
+      "learning_rate": 4.8344901110304054e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 3640496,
+      "step": 4660
+    },
+    {
+      "epoch": 9.698544698544698,
+      "grad_norm": 0.832689642906189,
+      "learning_rate": 4.8341386560719534e-05,
+      "loss": 0.0179,
+      "num_input_tokens_seen": 3644464,
+      "step": 4665
+    },
+    {
+      "epoch": 9.70893970893971,
+      "grad_norm": 0.08149834722280502,
+      "learning_rate": 4.833786841159989e-05,
+      "loss": 0.0194,
+      "num_input_tokens_seen": 3648368,
+      "step": 4670
+    },
+    {
+      "epoch": 9.71933471933472,
+      "grad_norm": 1.552018642425537,
+      "learning_rate": 4.833434666348765e-05,
+      "loss": 0.015,
+      "num_input_tokens_seen": 3652272,
+      "step": 4675
+    },
+    {
+      "epoch": 9.72972972972973,
+      "grad_norm": 1.3033666610717773,
+      "learning_rate": 4.833082131692592e-05,
+      "loss": 0.038,
+      "num_input_tokens_seen": 3656144,
+      "step": 4680
+    },
+    {
+      "epoch": 9.74012474012474,
+      "grad_norm": 0.42283743619918823,
+      "learning_rate": 4.832729237245835e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 3660112,
+      "step": 4685
+    },
+    {
+      "epoch": 9.75051975051975,
+      "grad_norm": 0.060854945331811905,
+      "learning_rate": 4.8323759830629145e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 3664112,
+      "step": 4690
+    },
+    {
+      "epoch": 9.76091476091476,
+      "grad_norm": 0.024340592324733734,
+      "learning_rate": 4.8320223691983066e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 3667920,
+      "step": 4695
+    },
+    {
+      "epoch": 9.771309771309772,
+      "grad_norm": 2.542741060256958,
+      "learning_rate": 4.831668395706544e-05,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 3671760,
+      "step": 4700
+    },
+    {
+      "epoch": 9.781704781704782,
+      "grad_norm": 0.24724730849266052,
+      "learning_rate": 4.8313140626422125e-05,
+      "loss": 0.0183,
+      "num_input_tokens_seen": 3675728,
+      "step": 4705
+    },
+    {
+      "epoch": 9.792099792099792,
+      "grad_norm": 0.9556503891944885,
+      "learning_rate": 4.830959370059956e-05,
+      "loss": 0.0265,
+      "num_input_tokens_seen": 3679792,
+      "step": 4710
+    },
+    {
+      "epoch": 9.802494802494802,
+      "grad_norm": 0.06612871587276459,
+      "learning_rate": 4.830604318014472e-05,
+      "loss": 0.0045,
+      "num_input_tokens_seen": 3683792,
+      "step": 4715
+    },
+    {
+      "epoch": 9.812889812889813,
+      "grad_norm": 0.0358201339840889,
+      "learning_rate": 4.830248906560514e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 3687728,
+      "step": 4720
+    },
+    {
+      "epoch": 9.823284823284823,
+      "grad_norm": 0.008709483779966831,
+      "learning_rate": 4.829893135752891e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 3691568,
+      "step": 4725
+    },
+    {
+      "epoch": 9.833679833679835,
+      "grad_norm": 0.9549663066864014,
+      "learning_rate": 4.829537005646466e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 3695472,
+      "step": 4730
+    },
+    {
+      "epoch": 9.844074844074845,
+      "grad_norm": 0.1316516101360321,
+      "learning_rate": 4.8291805162961615e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 3699440,
+      "step": 4735
+    },
+    {
+      "epoch": 9.854469854469855,
+      "grad_norm": 0.01093143317848444,
+      "learning_rate": 4.82882366775695e-05,
+      "loss": 0.0277,
+      "num_input_tokens_seen": 3703408,
+      "step": 4740
+    },
+    {
+      "epoch": 9.864864864864865,
+      "grad_norm": 6.285032272338867,
+      "learning_rate": 4.828466460083864e-05,
+      "loss": 0.0258,
+      "num_input_tokens_seen": 3707280,
+      "step": 4745
+    },
+    {
+      "epoch": 9.875259875259875,
+      "grad_norm": 1.3369849920272827,
+      "learning_rate": 4.8281088933319877e-05,
+      "loss": 0.0499,
+      "num_input_tokens_seen": 3711152,
+      "step": 4750
+    },
+    {
+      "epoch": 9.885654885654885,
+      "grad_norm": 5.484418869018555,
+      "learning_rate": 4.827750967556464e-05,
+      "loss": 0.0415,
+      "num_input_tokens_seen": 3715184,
+      "step": 4755
+    },
+    {
+      "epoch": 9.896049896049895,
+      "grad_norm": 0.21396085619926453,
+      "learning_rate": 4.827392682812488e-05,
+      "loss": 0.0548,
+      "num_input_tokens_seen": 3719216,
+      "step": 4760
+    },
+    {
+      "epoch": 9.906444906444907,
+      "grad_norm": 0.227932870388031,
+      "learning_rate": 4.827034039155312e-05,
+      "loss": 0.0045,
+      "num_input_tokens_seen": 3723120,
+      "step": 4765
+    },
+    {
+      "epoch": 9.916839916839917,
+      "grad_norm": 0.19941438734531403,
+      "learning_rate": 4.8266750366402445e-05,
+      "loss": 0.0344,
+      "num_input_tokens_seen": 3727088,
+      "step": 4770
+    },
+    {
+      "epoch": 9.927234927234927,
+      "grad_norm": 1.3860735893249512,
+      "learning_rate": 4.8263156753226476e-05,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 3731024,
+      "step": 4775
+    },
+    {
+      "epoch": 9.937629937629938,
+      "grad_norm": 0.14215907454490662,
+      "learning_rate": 4.8259559552579394e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 3734960,
+      "step": 4780
+    },
+    {
+      "epoch": 9.948024948024948,
+      "grad_norm": 0.1180913969874382,
+      "learning_rate": 4.825595876501593e-05,
+      "loss": 0.0166,
+      "num_input_tokens_seen": 3738832,
+      "step": 4785
+    },
+    {
+      "epoch": 9.958419958419958,
+      "grad_norm": 0.03290320187807083,
+      "learning_rate": 4.825235439109137e-05,
+      "loss": 0.0082,
+      "num_input_tokens_seen": 3742672,
+      "step": 4790
+    },
+    {
+      "epoch": 9.96881496881497,
+      "grad_norm": 5.8379411697387695,
+      "learning_rate": 4.824874643136156e-05,
+      "loss": 0.0237,
+      "num_input_tokens_seen": 3746608,
+      "step": 4795
+    },
+    {
+      "epoch": 9.97920997920998,
+      "grad_norm": 0.0309658944606781,
+      "learning_rate": 4.824513488638288e-05,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 3750544,
+      "step": 4800
+    },
+    {
+      "epoch": 9.97920997920998,
+      "eval_loss": 0.23696501553058624,
+      "eval_runtime": 11.7489,
+      "eval_samples_per_second": 72.858,
+      "eval_steps_per_second": 18.214,
+      "num_input_tokens_seen": 3750544,
+      "step": 4800
+    },
+    {
+      "epoch": 9.98960498960499,
+      "grad_norm": 0.09648259729146957,
+      "learning_rate": 4.8241519756712293e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 3754416,
+      "step": 4805
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.04923223704099655,
+      "learning_rate": 4.8237901042907285e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 3758304,
+      "step": 4810
+    },
+    {
+      "epoch": 10.01039501039501,
+      "grad_norm": 1.7152128219604492,
+      "learning_rate": 4.823427874552591e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 3762176,
+      "step": 4815
+    },
+    {
+      "epoch": 10.02079002079002,
+      "grad_norm": 0.05692080035805702,
+      "learning_rate": 4.823065286512677e-05,
+      "loss": 0.0247,
+      "num_input_tokens_seen": 3766144,
+      "step": 4820
+    },
+    {
+      "epoch": 10.03118503118503,
+      "grad_norm": 0.057389289140701294,
+      "learning_rate": 4.8227023402269025e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 3769920,
+      "step": 4825
+    },
+    {
+      "epoch": 10.041580041580042,
+      "grad_norm": 0.04387771338224411,
+      "learning_rate": 4.822339035751239e-05,
+      "loss": 0.0418,
+      "num_input_tokens_seen": 3773856,
+      "step": 4830
+    },
+    {
+      "epoch": 10.051975051975052,
+      "grad_norm": 0.058387015014886856,
+      "learning_rate": 4.8219753731417104e-05,
+      "loss": 0.0246,
+      "num_input_tokens_seen": 3777792,
+      "step": 4835
+    },
+    {
+      "epoch": 10.062370062370062,
+      "grad_norm": 0.1227274015545845,
+      "learning_rate": 4.821611352454401e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 3781632,
+      "step": 4840
+    },
+    {
+      "epoch": 10.072765072765073,
+      "grad_norm": 1.8203794956207275,
+      "learning_rate": 4.8212469737454444e-05,
+      "loss": 0.01,
+      "num_input_tokens_seen": 3785504,
+      "step": 4845
+    },
+    {
+      "epoch": 10.083160083160083,
+      "grad_norm": 0.7180619239807129,
+      "learning_rate": 4.820882237071035e-05,
+      "loss": 0.0215,
+      "num_input_tokens_seen": 3789280,
+      "step": 4850
+    },
+    {
+      "epoch": 10.093555093555093,
+      "grad_norm": 0.09227979183197021,
+      "learning_rate": 4.820517142487417e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 3793056,
+      "step": 4855
+    },
+    {
+      "epoch": 10.103950103950105,
+      "grad_norm": 1.1498361825942993,
+      "learning_rate": 4.8201516900508956e-05,
+      "loss": 0.0144,
+      "num_input_tokens_seen": 3797024,
+      "step": 4860
+    },
+    {
+      "epoch": 10.114345114345115,
+      "grad_norm": 18.71125602722168,
+      "learning_rate": 4.819785879817827e-05,
+      "loss": 0.008,
+      "num_input_tokens_seen": 3800800,
+      "step": 4865
+    },
+    {
+      "epoch": 10.124740124740125,
+      "grad_norm": 0.0571858286857605,
+      "learning_rate": 4.8194197118446226e-05,
+      "loss": 0.0131,
+      "num_input_tokens_seen": 3804640,
+      "step": 4870
+    },
+    {
+      "epoch": 10.135135135135135,
+      "grad_norm": 0.08828370273113251,
+      "learning_rate": 4.819053186187752e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 3808512,
+      "step": 4875
+    },
+    {
+      "epoch": 10.145530145530145,
+      "grad_norm": 0.03175345063209534,
+      "learning_rate": 4.818686302903736e-05,
+      "loss": 0.0128,
+      "num_input_tokens_seen": 3812416,
+      "step": 4880
+    },
+    {
+      "epoch": 10.155925155925155,
+      "grad_norm": 0.10368512570858002,
+      "learning_rate": 4.818319062049154e-05,
+      "loss": 0.0142,
+      "num_input_tokens_seen": 3816352,
+      "step": 4885
+    },
+    {
+      "epoch": 10.166320166320165,
+      "grad_norm": 0.2868328094482422,
+      "learning_rate": 4.817951463680639e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 3820192,
+      "step": 4890
+    },
+    {
+      "epoch": 10.176715176715177,
+      "grad_norm": 0.09176743030548096,
+      "learning_rate": 4.817583507854879e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 3824192,
+      "step": 4895
+    },
+    {
+      "epoch": 10.187110187110187,
+      "grad_norm": 14.12469482421875,
+      "learning_rate": 4.817215194628617e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 3828096,
+      "step": 4900
+    },
+    {
+      "epoch": 10.197505197505198,
+      "grad_norm": 1.5921155214309692,
+      "learning_rate": 4.816846524058653e-05,
+      "loss": 0.0303,
+      "num_input_tokens_seen": 3832000,
+      "step": 4905
+    },
+    {
+      "epoch": 10.207900207900208,
+      "grad_norm": 0.4759175479412079,
+      "learning_rate": 4.816477496201839e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 3835904,
+      "step": 4910
+    },
+    {
+      "epoch": 10.218295218295218,
+      "grad_norm": 0.049258917570114136,
+      "learning_rate": 4.8161081111150845e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 3839744,
+      "step": 4915
+    },
+    {
+      "epoch": 10.228690228690228,
+      "grad_norm": 0.01703721098601818,
+      "learning_rate": 4.815738368855354e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 3843552,
+      "step": 4920
+    },
+    {
+      "epoch": 10.23908523908524,
+      "grad_norm": 0.13396960496902466,
+      "learning_rate": 4.815368269479664e-05,
+      "loss": 0.0216,
+      "num_input_tokens_seen": 3847424,
+      "step": 4925
+    },
+    {
+      "epoch": 10.24948024948025,
+      "grad_norm": 0.058042798191308975,
+      "learning_rate": 4.814997813045092e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 3851392,
+      "step": 4930
+    },
+    {
+      "epoch": 10.25987525987526,
+      "grad_norm": 1.1204417943954468,
+      "learning_rate": 4.814626999608764e-05,
+      "loss": 0.0228,
+      "num_input_tokens_seen": 3855328,
+      "step": 4935
+    },
+    {
+      "epoch": 10.27027027027027,
+      "grad_norm": 0.03382985666394234,
+      "learning_rate": 4.814255829227865e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 3859232,
+      "step": 4940
+    },
+    {
+      "epoch": 10.28066528066528,
+      "grad_norm": 0.024665044620633125,
+      "learning_rate": 4.813884301959635e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 3863072,
+      "step": 4945
+    },
+    {
+      "epoch": 10.29106029106029,
+      "grad_norm": 1.6747440099716187,
+      "learning_rate": 4.813512417861368e-05,
+      "loss": 0.0236,
+      "num_input_tokens_seen": 3867040,
+      "step": 4950
+    },
+    {
+      "epoch": 10.301455301455302,
+      "grad_norm": 0.07608979940414429,
+      "learning_rate": 4.813140176990411e-05,
+      "loss": 0.0096,
+      "num_input_tokens_seen": 3870976,
+      "step": 4955
+    },
+    {
+      "epoch": 10.311850311850312,
+      "grad_norm": 0.053848907351493835,
+      "learning_rate": 4.8127675794041714e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 3874976,
+      "step": 4960
+    },
+    {
+      "epoch": 10.322245322245323,
+      "grad_norm": 0.016555383801460266,
+      "learning_rate": 4.812394625160107e-05,
+      "loss": 0.0089,
+      "num_input_tokens_seen": 3878816,
+      "step": 4965
+    },
+    {
+      "epoch": 10.332640332640333,
+      "grad_norm": 0.05226912349462509,
+      "learning_rate": 4.812021314315732e-05,
+      "loss": 0.0359,
+      "num_input_tokens_seen": 3882592,
+      "step": 4970
+    },
+    {
+      "epoch": 10.343035343035343,
+      "grad_norm": 0.03985751047730446,
+      "learning_rate": 4.811647646928616e-05,
+      "loss": 0.0148,
+      "num_input_tokens_seen": 3886560,
+      "step": 4975
+    },
+    {
+      "epoch": 10.353430353430353,
+      "grad_norm": 0.05934731289744377,
+      "learning_rate": 4.8112736230563814e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 3890368,
+      "step": 4980
+    },
+    {
+      "epoch": 10.363825363825363,
+      "grad_norm": 0.039196718484163284,
+      "learning_rate": 4.81089924275671e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 3894144,
+      "step": 4985
+    },
+    {
+      "epoch": 10.374220374220375,
+      "grad_norm": 6.2303786277771,
+      "learning_rate": 4.810524506087335e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 3898112,
+      "step": 4990
+    },
+    {
+      "epoch": 10.384615384615385,
+      "grad_norm": 0.15308767557144165,
+      "learning_rate": 4.810149413106044e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 3902016,
+      "step": 4995
+    },
+    {
+      "epoch": 10.395010395010395,
+      "grad_norm": 0.007705794181674719,
+      "learning_rate": 4.809773963870684e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 3905920,
+      "step": 5000
+    },
+    {
+      "epoch": 10.395010395010395,
+      "eval_loss": 0.2759002149105072,
+      "eval_runtime": 11.7335,
+      "eval_samples_per_second": 72.953,
+      "eval_steps_per_second": 18.238,
+      "num_input_tokens_seen": 3905920,
+      "step": 5000
+    },
+    {
+      "epoch": 10.405405405405405,
+      "grad_norm": 1.2631350755691528,
+      "learning_rate": 4.809398158439151e-05,
+      "loss": 0.0475,
+      "num_input_tokens_seen": 3909696,
+      "step": 5005
+    },
+    {
+      "epoch": 10.415800415800415,
+      "grad_norm": 0.0419263057410717,
+      "learning_rate": 4.8090219968694005e-05,
+      "loss": 0.0248,
+      "num_input_tokens_seen": 3913568,
+      "step": 5010
+    },
+    {
+      "epoch": 10.426195426195425,
+      "grad_norm": 0.26723411679267883,
+      "learning_rate": 4.808645479219442e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 3917632,
+      "step": 5015
+    },
+    {
+      "epoch": 10.436590436590437,
+      "grad_norm": 0.050326522439718246,
+      "learning_rate": 4.8082686055473375e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 3921408,
+      "step": 5020
+    },
+    {
+      "epoch": 10.446985446985448,
+      "grad_norm": 3.321608066558838,
+      "learning_rate": 4.8078913759112066e-05,
+      "loss": 0.0362,
+      "num_input_tokens_seen": 3925376,
+      "step": 5025
+    },
+    {
+      "epoch": 10.457380457380458,
+      "grad_norm": 0.012260369025170803,
+      "learning_rate": 4.807513790369223e-05,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 3929248,
+      "step": 5030
+    },
+    {
+      "epoch": 10.467775467775468,
+      "grad_norm": 0.10189782083034515,
+      "learning_rate": 4.8071358489796145e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 3933056,
+      "step": 5035
+    },
+    {
+      "epoch": 10.478170478170478,
+      "grad_norm": 1.129288673400879,
+      "learning_rate": 4.806757551800665e-05,
+      "loss": 0.026,
+      "num_input_tokens_seen": 3937056,
+      "step": 5040
+    },
+    {
+      "epoch": 10.488565488565488,
+      "grad_norm": 8.380205154418945,
+      "learning_rate": 4.806378898890713e-05,
+      "loss": 0.0127,
+      "num_input_tokens_seen": 3940928,
+      "step": 5045
+    },
+    {
+      "epoch": 10.4989604989605,
+      "grad_norm": 0.035134073346853256,
+      "learning_rate": 4.80599989030815e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 3944832,
+      "step": 5050
+    },
+    {
+      "epoch": 10.50935550935551,
+      "grad_norm": 2.0698719024658203,
+      "learning_rate": 4.805620526111426e-05,
+      "loss": 0.0183,
+      "num_input_tokens_seen": 3948736,
+      "step": 5055
+    },
+    {
+      "epoch": 10.51975051975052,
+      "grad_norm": 3.013629913330078,
+      "learning_rate": 4.805240806359042e-05,
+      "loss": 0.0241,
+      "num_input_tokens_seen": 3952704,
+      "step": 5060
+    },
+    {
+      "epoch": 10.53014553014553,
+      "grad_norm": 0.10605139285326004,
+      "learning_rate": 4.804860731109557e-05,
+      "loss": 0.0159,
+      "num_input_tokens_seen": 3956704,
+      "step": 5065
+    },
+    {
+      "epoch": 10.54054054054054,
+      "grad_norm": 0.07914886623620987,
+      "learning_rate": 4.804480300421581e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 3960704,
+      "step": 5070
+    },
+    {
+      "epoch": 10.55093555093555,
+      "grad_norm": 12.41367244720459,
+      "learning_rate": 4.804099514353784e-05,
+      "loss": 0.0259,
+      "num_input_tokens_seen": 3964608,
+      "step": 5075
+    },
+    {
+      "epoch": 10.56133056133056,
+      "grad_norm": 9.098644256591797,
+      "learning_rate": 4.8037183729648867e-05,
+      "loss": 0.0224,
+      "num_input_tokens_seen": 3968512,
+      "step": 5080
+    },
+    {
+      "epoch": 10.571725571725572,
+      "grad_norm": 0.06350910663604736,
+      "learning_rate": 4.803336876313666e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 3972288,
+      "step": 5085
+    },
+    {
+      "epoch": 10.582120582120583,
+      "grad_norm": 0.18019025027751923,
+      "learning_rate": 4.802955024458953e-05,
+      "loss": 0.0453,
+      "num_input_tokens_seen": 3976160,
+      "step": 5090
+    },
+    {
+      "epoch": 10.592515592515593,
+      "grad_norm": 0.18829886615276337,
+      "learning_rate": 4.802572817459634e-05,
+      "loss": 0.0159,
+      "num_input_tokens_seen": 3980192,
+      "step": 5095
+    },
+    {
+      "epoch": 10.602910602910603,
+      "grad_norm": 0.06333538889884949,
+      "learning_rate": 4.802190255374651e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 3984192,
+      "step": 5100
+    },
+    {
+      "epoch": 10.613305613305613,
+      "grad_norm": 0.21939562261104584,
+      "learning_rate": 4.801807338263e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 3988096,
+      "step": 5105
+    },
+    {
+      "epoch": 10.623700623700623,
+      "grad_norm": 0.07072778791189194,
+      "learning_rate": 4.8014240661837306e-05,
+      "loss": 0.027,
+      "num_input_tokens_seen": 3991968,
+      "step": 5110
+    },
+    {
+      "epoch": 10.634095634095633,
+      "grad_norm": 0.09584195166826248,
+      "learning_rate": 4.80104043919595e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 3995904,
+      "step": 5115
+    },
+    {
+      "epoch": 10.644490644490645,
+      "grad_norm": 0.03379153460264206,
+      "learning_rate": 4.800656457358815e-05,
+      "loss": 0.0194,
+      "num_input_tokens_seen": 3999808,
+      "step": 5120
+    },
+    {
+      "epoch": 10.654885654885655,
+      "grad_norm": 0.0455503910779953,
+      "learning_rate": 4.800272120731544e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 4003776,
+      "step": 5125
+    },
+    {
+      "epoch": 10.665280665280665,
+      "grad_norm": 0.40649598836898804,
+      "learning_rate": 4.799887429373404e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 4007712,
+      "step": 5130
+    },
+    {
+      "epoch": 10.675675675675675,
+      "grad_norm": 0.01196593139320612,
+      "learning_rate": 4.79950238334372e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 4011680,
+      "step": 5135
+    },
+    {
+      "epoch": 10.686070686070686,
+      "grad_norm": 0.01183714997023344,
+      "learning_rate": 4.799116982701872e-05,
+      "loss": 0.0246,
+      "num_input_tokens_seen": 4015712,
+      "step": 5140
+    },
+    {
+      "epoch": 10.696465696465696,
+      "grad_norm": 0.012303495779633522,
+      "learning_rate": 4.7987312275072926e-05,
+      "loss": 0.0349,
+      "num_input_tokens_seen": 4019616,
+      "step": 5145
+    },
+    {
+      "epoch": 10.706860706860708,
+      "grad_norm": 0.02791743166744709,
+      "learning_rate": 4.79834511781947e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 4023552,
+      "step": 5150
+    },
+    {
+      "epoch": 10.717255717255718,
+      "grad_norm": 0.034830883145332336,
+      "learning_rate": 4.797958653697947e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 4027712,
+      "step": 5155
+    },
+    {
+      "epoch": 10.727650727650728,
+      "grad_norm": 0.5767345428466797,
+      "learning_rate": 4.7975718352023225e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4031648,
+      "step": 5160
+    },
+    {
+      "epoch": 10.738045738045738,
+      "grad_norm": 0.06368085741996765,
+      "learning_rate": 4.7971846623922476e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 4035712,
+      "step": 5165
+    },
+    {
+      "epoch": 10.748440748440748,
+      "grad_norm": 3.2244069576263428,
+      "learning_rate": 4.7967971353274294e-05,
+      "loss": 0.0688,
+      "num_input_tokens_seen": 4039584,
+      "step": 5170
+    },
+    {
+      "epoch": 10.758835758835758,
+      "grad_norm": 0.06239760294556618,
+      "learning_rate": 4.79640925406763e-05,
+      "loss": 0.0107,
+      "num_input_tokens_seen": 4043424,
+      "step": 5175
+    },
+    {
+      "epoch": 10.76923076923077,
+      "grad_norm": 1.069339394569397,
+      "learning_rate": 4.796021018672664e-05,
+      "loss": 0.0462,
+      "num_input_tokens_seen": 4047296,
+      "step": 5180
+    },
+    {
+      "epoch": 10.77962577962578,
+      "grad_norm": 0.07372652739286423,
+      "learning_rate": 4.795632429202405e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 4051200,
+      "step": 5185
+    },
+    {
+      "epoch": 10.79002079002079,
+      "grad_norm": 0.11431615054607391,
+      "learning_rate": 4.795243485716775e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 4055168,
+      "step": 5190
+    },
+    {
+      "epoch": 10.8004158004158,
+      "grad_norm": 0.10946138203144073,
+      "learning_rate": 4.794854188275757e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 4059136,
+      "step": 5195
+    },
+    {
+      "epoch": 10.81081081081081,
+      "grad_norm": 0.05046825110912323,
+      "learning_rate": 4.794464536939384e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 4063008,
+      "step": 5200
+    },
+    {
+      "epoch": 10.81081081081081,
+      "eval_loss": 0.21500343084335327,
+      "eval_runtime": 11.7629,
+      "eval_samples_per_second": 72.771,
+      "eval_steps_per_second": 18.193,
+      "num_input_tokens_seen": 4063008,
+      "step": 5200
+    },
+    {
+      "epoch": 10.82120582120582,
+      "grad_norm": 0.03688390180468559,
+      "learning_rate": 4.794074531767745e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 4066944,
+      "step": 5205
+    },
+    {
+      "epoch": 10.83160083160083,
+      "grad_norm": 0.016415564343333244,
+      "learning_rate": 4.7936841728209834e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 4070944,
+      "step": 5210
+    },
+    {
+      "epoch": 10.841995841995843,
+      "grad_norm": 0.04815289378166199,
+      "learning_rate": 4.7932934601593e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 4074848,
+      "step": 5215
+    },
+    {
+      "epoch": 10.852390852390853,
+      "grad_norm": 0.018921026960015297,
+      "learning_rate": 4.792902393842943e-05,
+      "loss": 0.0113,
+      "num_input_tokens_seen": 4078656,
+      "step": 5220
+    },
+    {
+      "epoch": 10.862785862785863,
+      "grad_norm": 0.01709054782986641,
+      "learning_rate": 4.792510973932225e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 4082496,
+      "step": 5225
+    },
+    {
+      "epoch": 10.873180873180873,
+      "grad_norm": 0.015574616380035877,
+      "learning_rate": 4.7921192004875036e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4086464,
+      "step": 5230
+    },
+    {
+      "epoch": 10.883575883575883,
+      "grad_norm": 0.02126523293554783,
+      "learning_rate": 4.791727073569198e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 4090304,
+      "step": 5235
+    },
+    {
+      "epoch": 10.893970893970893,
+      "grad_norm": 0.006824342533946037,
+      "learning_rate": 4.7913345932377775e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 4094208,
+      "step": 5240
+    },
+    {
+      "epoch": 10.904365904365905,
+      "grad_norm": 0.004803393501788378,
+      "learning_rate": 4.790941759553769e-05,
+      "loss": 0.0144,
+      "num_input_tokens_seen": 4098144,
+      "step": 5245
+    },
+    {
+      "epoch": 10.914760914760915,
+      "grad_norm": 0.07650803029537201,
+      "learning_rate": 4.79054857257775e-05,
+      "loss": 0.0262,
+      "num_input_tokens_seen": 4101984,
+      "step": 5250
+    },
+    {
+      "epoch": 10.925155925155925,
+      "grad_norm": 0.29701220989227295,
+      "learning_rate": 4.790155032370357e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 4105856,
+      "step": 5255
+    },
+    {
+      "epoch": 10.935550935550935,
+      "grad_norm": 0.016903921961784363,
+      "learning_rate": 4.789761138992278e-05,
+      "loss": 0.0351,
+      "num_input_tokens_seen": 4109760,
+      "step": 5260
+    },
+    {
+      "epoch": 10.945945945945946,
+      "grad_norm": 0.020714566111564636,
+      "learning_rate": 4.7893668925042565e-05,
+      "loss": 0.0127,
+      "num_input_tokens_seen": 4113504,
+      "step": 5265
+    },
+    {
+      "epoch": 10.956340956340956,
+      "grad_norm": 22.26581382751465,
+      "learning_rate": 4.78897229296709e-05,
+      "loss": 0.028,
+      "num_input_tokens_seen": 4117408,
+      "step": 5270
+    },
+    {
+      "epoch": 10.966735966735968,
+      "grad_norm": 0.8819617033004761,
+      "learning_rate": 4.7885773404416315e-05,
+      "loss": 0.0131,
+      "num_input_tokens_seen": 4121248,
+      "step": 5275
+    },
+    {
+      "epoch": 10.977130977130978,
+      "grad_norm": 0.08428268879652023,
+      "learning_rate": 4.788182034988786e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 4125216,
+      "step": 5280
+    },
+    {
+      "epoch": 10.987525987525988,
+      "grad_norm": 0.09387154132127762,
+      "learning_rate": 4.787786376669516e-05,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 4129056,
+      "step": 5285
+    },
+    {
+      "epoch": 10.997920997920998,
+      "grad_norm": 0.04516831040382385,
+      "learning_rate": 4.787390365544837e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 4132928,
+      "step": 5290
+    },
+    {
+      "epoch": 11.008316008316008,
+      "grad_norm": 0.028033094480633736,
+      "learning_rate": 4.786994001675818e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 4136880,
+      "step": 5295
+    },
+    {
+      "epoch": 11.018711018711018,
+      "grad_norm": 0.054609861224889755,
+      "learning_rate": 4.786597285123584e-05,
+      "loss": 0.0099,
+      "num_input_tokens_seen": 4140880,
+      "step": 5300
+    },
+    {
+      "epoch": 11.029106029106028,
+      "grad_norm": 0.0072878627106547356,
+      "learning_rate": 4.7862002159493135e-05,
+      "loss": 0.0106,
+      "num_input_tokens_seen": 4144880,
+      "step": 5305
+    },
+    {
+      "epoch": 11.03950103950104,
+      "grad_norm": 1.0450555086135864,
+      "learning_rate": 4.785802794214239e-05,
+      "loss": 0.0126,
+      "num_input_tokens_seen": 4148784,
+      "step": 5310
+    },
+    {
+      "epoch": 11.04989604989605,
+      "grad_norm": 3.1580328941345215,
+      "learning_rate": 4.7854050199796495e-05,
+      "loss": 0.0084,
+      "num_input_tokens_seen": 4152688,
+      "step": 5315
+    },
+    {
+      "epoch": 11.06029106029106,
+      "grad_norm": 0.06383773684501648,
+      "learning_rate": 4.7850068933068845e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 4156496,
+      "step": 5320
+    },
+    {
+      "epoch": 11.07068607068607,
+      "grad_norm": 0.713450014591217,
+      "learning_rate": 4.7846084142573425e-05,
+      "loss": 0.0129,
+      "num_input_tokens_seen": 4160464,
+      "step": 5325
+    },
+    {
+      "epoch": 11.08108108108108,
+      "grad_norm": 7.949678421020508,
+      "learning_rate": 4.7842095828924725e-05,
+      "loss": 0.0343,
+      "num_input_tokens_seen": 4164368,
+      "step": 5330
+    },
+    {
+      "epoch": 11.09147609147609,
+      "grad_norm": 0.012164674699306488,
+      "learning_rate": 4.783810399273779e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4168432,
+      "step": 5335
+    },
+    {
+      "epoch": 11.101871101871103,
+      "grad_norm": 1.357552409172058,
+      "learning_rate": 4.7834108634628226e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 4172368,
+      "step": 5340
+    },
+    {
+      "epoch": 11.112266112266113,
+      "grad_norm": 0.053929753601551056,
+      "learning_rate": 4.783010975521216e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 4176304,
+      "step": 5345
+    },
+    {
+      "epoch": 11.122661122661123,
+      "grad_norm": 0.014476077631115913,
+      "learning_rate": 4.782610735510626e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 4180272,
+      "step": 5350
+    },
+    {
+      "epoch": 11.133056133056133,
+      "grad_norm": 0.023481080308556557,
+      "learning_rate": 4.782210143492776e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 4184176,
+      "step": 5355
+    },
+    {
+      "epoch": 11.143451143451143,
+      "grad_norm": 0.01540575921535492,
+      "learning_rate": 4.781809199529442e-05,
+      "loss": 0.0513,
+      "num_input_tokens_seen": 4188144,
+      "step": 5360
+    },
+    {
+      "epoch": 11.153846153846153,
+      "grad_norm": 0.047659460455179214,
+      "learning_rate": 4.781407903682454e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 4191984,
+      "step": 5365
+    },
+    {
+      "epoch": 11.164241164241163,
+      "grad_norm": 2.0994277000427246,
+      "learning_rate": 4.781006256013698e-05,
+      "loss": 0.0245,
+      "num_input_tokens_seen": 4195888,
+      "step": 5370
+    },
+    {
+      "epoch": 11.174636174636175,
+      "grad_norm": 0.12447476387023926,
+      "learning_rate": 4.7806042565851115e-05,
+      "loss": 0.0361,
+      "num_input_tokens_seen": 4199856,
+      "step": 5375
+    },
+    {
+      "epoch": 11.185031185031185,
+      "grad_norm": 4.64523983001709,
+      "learning_rate": 4.7802019054586895e-05,
+      "loss": 0.0111,
+      "num_input_tokens_seen": 4203696,
+      "step": 5380
+    },
+    {
+      "epoch": 11.195426195426196,
+      "grad_norm": 0.10893319547176361,
+      "learning_rate": 4.779799202696479e-05,
+      "loss": 0.021,
+      "num_input_tokens_seen": 4207504,
+      "step": 5385
+    },
+    {
+      "epoch": 11.205821205821206,
+      "grad_norm": 0.1436423808336258,
+      "learning_rate": 4.779396148360581e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 4211472,
+      "step": 5390
+    },
+    {
+      "epoch": 11.216216216216216,
+      "grad_norm": 0.1598617434501648,
+      "learning_rate": 4.7789927425131517e-05,
+      "loss": 0.0141,
+      "num_input_tokens_seen": 4215440,
+      "step": 5395
+    },
+    {
+      "epoch": 11.226611226611226,
+      "grad_norm": 0.1840338408946991,
+      "learning_rate": 4.778588985216403e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 4219472,
+      "step": 5400
+    },
+    {
+      "epoch": 11.226611226611226,
+      "eval_loss": 0.1925525814294815,
+      "eval_runtime": 11.7158,
+      "eval_samples_per_second": 73.064,
+      "eval_steps_per_second": 18.266,
+      "num_input_tokens_seen": 4219472,
+      "step": 5400
+    },
+    {
+      "epoch": 11.237006237006238,
+      "grad_norm": 0.019225113093852997,
+      "learning_rate": 4.778184876532598e-05,
+      "loss": 0.0106,
+      "num_input_tokens_seen": 4223248,
+      "step": 5405
+    },
+    {
+      "epoch": 11.247401247401248,
+      "grad_norm": 0.024957939982414246,
+      "learning_rate": 4.7777804165240556e-05,
+      "loss": 0.027,
+      "num_input_tokens_seen": 4227280,
+      "step": 5410
+    },
+    {
+      "epoch": 11.257796257796258,
+      "grad_norm": 0.02111196331679821,
+      "learning_rate": 4.7773756052531485e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 4231120,
+      "step": 5415
+    },
+    {
+      "epoch": 11.268191268191268,
+      "grad_norm": 0.04008803516626358,
+      "learning_rate": 4.7769704427823035e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 4235024,
+      "step": 5420
+    },
+    {
+      "epoch": 11.278586278586278,
+      "grad_norm": 0.026355966925621033,
+      "learning_rate": 4.776564929174003e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 4238960,
+      "step": 5425
+    },
+    {
+      "epoch": 11.288981288981288,
+      "grad_norm": 0.06410873681306839,
+      "learning_rate": 4.7761590644907806e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 4242832,
+      "step": 5430
+    },
+    {
+      "epoch": 11.299376299376299,
+      "grad_norm": 0.3248133361339569,
+      "learning_rate": 4.7757528487952263e-05,
+      "loss": 0.0282,
+      "num_input_tokens_seen": 4246576,
+      "step": 5435
+    },
+    {
+      "epoch": 11.30977130977131,
+      "grad_norm": 0.043612007051706314,
+      "learning_rate": 4.7753462821499836e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 4250416,
+      "step": 5440
+    },
+    {
+      "epoch": 11.32016632016632,
+      "grad_norm": 0.018723925575613976,
+      "learning_rate": 4.774939364617751e-05,
+      "loss": 0.0129,
+      "num_input_tokens_seen": 4254320,
+      "step": 5445
+    },
+    {
+      "epoch": 11.33056133056133,
+      "grad_norm": 0.6330317854881287,
+      "learning_rate": 4.7745320962612795e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 4258288,
+      "step": 5450
+    },
+    {
+      "epoch": 11.34095634095634,
+      "grad_norm": 2.858229160308838,
+      "learning_rate": 4.7741244771433756e-05,
+      "loss": 0.0279,
+      "num_input_tokens_seen": 4262256,
+      "step": 5455
+    },
+    {
+      "epoch": 11.35135135135135,
+      "grad_norm": 0.023076031357049942,
+      "learning_rate": 4.7737165073268985e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4266128,
+      "step": 5460
+    },
+    {
+      "epoch": 11.361746361746361,
+      "grad_norm": 0.02898775413632393,
+      "learning_rate": 4.7733081868747626e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4270096,
+      "step": 5465
+    },
+    {
+      "epoch": 11.372141372141373,
+      "grad_norm": 0.008973234333097935,
+      "learning_rate": 4.772899515849936e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 4274128,
+      "step": 5470
+    },
+    {
+      "epoch": 11.382536382536383,
+      "grad_norm": 0.13322338461875916,
+      "learning_rate": 4.7724904943154414e-05,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 4278032,
+      "step": 5475
+    },
+    {
+      "epoch": 11.392931392931393,
+      "grad_norm": 5.908188819885254,
+      "learning_rate": 4.772081122334354e-05,
+      "loss": 0.0277,
+      "num_input_tokens_seen": 4281904,
+      "step": 5480
+    },
+    {
+      "epoch": 11.403326403326403,
+      "grad_norm": 0.9733322262763977,
+      "learning_rate": 4.771671399969806e-05,
+      "loss": 0.0127,
+      "num_input_tokens_seen": 4285808,
+      "step": 5485
+    },
+    {
+      "epoch": 11.413721413721413,
+      "grad_norm": 0.14214804768562317,
+      "learning_rate": 4.7712613272849794e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 4289744,
+      "step": 5490
+    },
+    {
+      "epoch": 11.424116424116423,
+      "grad_norm": 0.12184374779462814,
+      "learning_rate": 4.770850904343114e-05,
+      "loss": 0.0107,
+      "num_input_tokens_seen": 4293584,
+      "step": 5495
+    },
+    {
+      "epoch": 11.434511434511435,
+      "grad_norm": 0.028937077149748802,
+      "learning_rate": 4.770440131207502e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 4297520,
+      "step": 5500
+    },
+    {
+      "epoch": 11.444906444906445,
+      "grad_norm": 0.02875436842441559,
+      "learning_rate": 4.7700290079414896e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 4301360,
+      "step": 5505
+    },
+    {
+      "epoch": 11.455301455301456,
+      "grad_norm": 0.015668582171201706,
+      "learning_rate": 4.769617534608477e-05,
+      "loss": 0.0243,
+      "num_input_tokens_seen": 4305200,
+      "step": 5510
+    },
+    {
+      "epoch": 11.465696465696466,
+      "grad_norm": 0.016102954745292664,
+      "learning_rate": 4.7692057112719193e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 4309136,
+      "step": 5515
+    },
+    {
+      "epoch": 11.476091476091476,
+      "grad_norm": 0.09143542498350143,
+      "learning_rate": 4.7687935379953234e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 4312976,
+      "step": 5520
+    },
+    {
+      "epoch": 11.486486486486486,
+      "grad_norm": 0.04414770379662514,
+      "learning_rate": 4.7683810148422534e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 4317008,
+      "step": 5525
+    },
+    {
+      "epoch": 11.496881496881496,
+      "grad_norm": 0.033513475209474564,
+      "learning_rate": 4.767968141876324e-05,
+      "loss": 0.0687,
+      "num_input_tokens_seen": 4320912,
+      "step": 5530
+    },
+    {
+      "epoch": 11.507276507276508,
+      "grad_norm": 1.105396032333374,
+      "learning_rate": 4.767554919161207e-05,
+      "loss": 0.0086,
+      "num_input_tokens_seen": 4324816,
+      "step": 5535
+    },
+    {
+      "epoch": 11.517671517671518,
+      "grad_norm": 0.033015619963407516,
+      "learning_rate": 4.767141346760624e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 4328880,
+      "step": 5540
+    },
+    {
+      "epoch": 11.528066528066528,
+      "grad_norm": 0.021686680614948273,
+      "learning_rate": 4.766727424738356e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 4332752,
+      "step": 5545
+    },
+    {
+      "epoch": 11.538461538461538,
+      "grad_norm": 1.1905416250228882,
+      "learning_rate": 4.7663131531582325e-05,
+      "loss": 0.0562,
+      "num_input_tokens_seen": 4336592,
+      "step": 5550
+    },
+    {
+      "epoch": 11.548856548856548,
+      "grad_norm": 0.10105972737073898,
+      "learning_rate": 4.765898532084142e-05,
+      "loss": 0.0159,
+      "num_input_tokens_seen": 4340720,
+      "step": 5555
+    },
+    {
+      "epoch": 11.559251559251559,
+      "grad_norm": 0.10129887610673904,
+      "learning_rate": 4.765483561580022e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 4344528,
+      "step": 5560
+    },
+    {
+      "epoch": 11.56964656964657,
+      "grad_norm": 0.030875446274876595,
+      "learning_rate": 4.7650682417098666e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 4348368,
+      "step": 5565
+    },
+    {
+      "epoch": 11.58004158004158,
+      "grad_norm": 0.04433324933052063,
+      "learning_rate": 4.7646525725377244e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 4352304,
+      "step": 5570
+    },
+    {
+      "epoch": 11.59043659043659,
+      "grad_norm": 0.9995904564857483,
+      "learning_rate": 4.764236554127696e-05,
+      "loss": 0.0297,
+      "num_input_tokens_seen": 4356304,
+      "step": 5575
+    },
+    {
+      "epoch": 11.6008316008316,
+      "grad_norm": 1.6256365776062012,
+      "learning_rate": 4.7638201865439356e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 4360176,
+      "step": 5580
+    },
+    {
+      "epoch": 11.611226611226611,
+      "grad_norm": 1.681180477142334,
+      "learning_rate": 4.7634034698506545e-05,
+      "loss": 0.0282,
+      "num_input_tokens_seen": 4364144,
+      "step": 5585
+    },
+    {
+      "epoch": 11.621621621621621,
+      "grad_norm": 0.35137516260147095,
+      "learning_rate": 4.762986404112115e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 4368048,
+      "step": 5590
+    },
+    {
+      "epoch": 11.632016632016633,
+      "grad_norm": 1.9116151332855225,
+      "learning_rate": 4.762568989392633e-05,
+      "loss": 0.0227,
+      "num_input_tokens_seen": 4372048,
+      "step": 5595
+    },
+    {
+      "epoch": 11.642411642411643,
+      "grad_norm": 6.05643892288208,
+      "learning_rate": 4.76215122575658e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 4376048,
+      "step": 5600
+    },
+    {
+      "epoch": 11.642411642411643,
+      "eval_loss": 0.22330763936042786,
+      "eval_runtime": 11.7611,
+      "eval_samples_per_second": 72.782,
+      "eval_steps_per_second": 18.196,
+      "num_input_tokens_seen": 4376048,
+      "step": 5600
+    },
+    {
+      "epoch": 11.652806652806653,
+      "grad_norm": 0.020298613235354424,
+      "learning_rate": 4.7617331132683795e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 4380016,
+      "step": 5605
+    },
+    {
+      "epoch": 11.663201663201663,
+      "grad_norm": 0.05262778326869011,
+      "learning_rate": 4.7613146519925105e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 4383792,
+      "step": 5610
+    },
+    {
+      "epoch": 11.673596673596673,
+      "grad_norm": 0.10201342403888702,
+      "learning_rate": 4.7608958419935045e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 4387760,
+      "step": 5615
+    },
+    {
+      "epoch": 11.683991683991684,
+      "grad_norm": 0.009988507255911827,
+      "learning_rate": 4.760476683335948e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 4391792,
+      "step": 5620
+    },
+    {
+      "epoch": 11.694386694386694,
+      "grad_norm": 0.05548175796866417,
+      "learning_rate": 4.760057176084479e-05,
+      "loss": 0.0204,
+      "num_input_tokens_seen": 4395664,
+      "step": 5625
+    },
+    {
+      "epoch": 11.704781704781706,
+      "grad_norm": 0.012402224354445934,
+      "learning_rate": 4.759637320303793e-05,
+      "loss": 0.0368,
+      "num_input_tokens_seen": 4399504,
+      "step": 5630
+    },
+    {
+      "epoch": 11.715176715176716,
+      "grad_norm": 0.0378015972673893,
+      "learning_rate": 4.759217116058635e-05,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 4403280,
+      "step": 5635
+    },
+    {
+      "epoch": 11.725571725571726,
+      "grad_norm": 0.17150312662124634,
+      "learning_rate": 4.758796563413807e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 4407280,
+      "step": 5640
+    },
+    {
+      "epoch": 11.735966735966736,
+      "grad_norm": 6.044877052307129,
+      "learning_rate": 4.758375662434163e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 4411152,
+      "step": 5645
+    },
+    {
+      "epoch": 11.746361746361746,
+      "grad_norm": 0.04007073864340782,
+      "learning_rate": 4.7579544131846114e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 4414992,
+      "step": 5650
+    },
+    {
+      "epoch": 11.756756756756756,
+      "grad_norm": 1.9167951345443726,
+      "learning_rate": 4.757532815730114e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 4418800,
+      "step": 5655
+    },
+    {
+      "epoch": 11.767151767151766,
+      "grad_norm": 0.007655569352209568,
+      "learning_rate": 4.7571108701356865e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4422640,
+      "step": 5660
+    },
+    {
+      "epoch": 11.777546777546778,
+      "grad_norm": 0.0205554012209177,
+      "learning_rate": 4.756688576466398e-05,
+      "loss": 0.017,
+      "num_input_tokens_seen": 4426576,
+      "step": 5665
+    },
+    {
+      "epoch": 11.787941787941788,
+      "grad_norm": 0.043982092291116714,
+      "learning_rate": 4.756265934787372e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4430672,
+      "step": 5670
+    },
+    {
+      "epoch": 11.798336798336798,
+      "grad_norm": 0.03664671257138252,
+      "learning_rate": 4.755842945163785e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 4434480,
+      "step": 5675
+    },
+    {
+      "epoch": 11.808731808731808,
+      "grad_norm": 0.04457903280854225,
+      "learning_rate": 4.755419607660867e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 4438448,
+      "step": 5680
+    },
+    {
+      "epoch": 11.819126819126819,
+      "grad_norm": 1.3967039585113525,
+      "learning_rate": 4.7549959223439016e-05,
+      "loss": 0.0261,
+      "num_input_tokens_seen": 4442384,
+      "step": 5685
+    },
+    {
+      "epoch": 11.829521829521829,
+      "grad_norm": 0.01150507852435112,
+      "learning_rate": 4.754571889278228e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4446192,
+      "step": 5690
+    },
+    {
+      "epoch": 11.83991683991684,
+      "grad_norm": 0.037395134568214417,
+      "learning_rate": 4.754147508529235e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 4450032,
+      "step": 5695
+    },
+    {
+      "epoch": 11.85031185031185,
+      "grad_norm": 0.006706413347274065,
+      "learning_rate": 4.75372278016237e-05,
+      "loss": 0.0168,
+      "num_input_tokens_seen": 4453776,
+      "step": 5700
+    },
+    {
+      "epoch": 11.86070686070686,
+      "grad_norm": 3.092329263687134,
+      "learning_rate": 4.753297704243129e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 4457648,
+      "step": 5705
+    },
+    {
+      "epoch": 11.871101871101871,
+      "grad_norm": 0.02166045643389225,
+      "learning_rate": 4.752872280837066e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4461520,
+      "step": 5710
+    },
+    {
+      "epoch": 11.881496881496881,
+      "grad_norm": 0.2925938665866852,
+      "learning_rate": 4.752446510009786e-05,
+      "loss": 0.0234,
+      "num_input_tokens_seen": 4465552,
+      "step": 5715
+    },
+    {
+      "epoch": 11.891891891891891,
+      "grad_norm": 0.04078911617398262,
+      "learning_rate": 4.7520203918269476e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 4469520,
+      "step": 5720
+    },
+    {
+      "epoch": 11.902286902286903,
+      "grad_norm": 0.05736135318875313,
+      "learning_rate": 4.751593926354265e-05,
+      "loss": 0.0234,
+      "num_input_tokens_seen": 4473488,
+      "step": 5725
+    },
+    {
+      "epoch": 11.912681912681913,
+      "grad_norm": 0.034832343459129333,
+      "learning_rate": 4.751167113657503e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 4477296,
+      "step": 5730
+    },
+    {
+      "epoch": 11.923076923076923,
+      "grad_norm": 2.022911548614502,
+      "learning_rate": 4.7507399538024834e-05,
+      "loss": 0.0433,
+      "num_input_tokens_seen": 4481168,
+      "step": 5735
+    },
+    {
+      "epoch": 11.933471933471933,
+      "grad_norm": 0.43225133419036865,
+      "learning_rate": 4.750312446855077e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 4485232,
+      "step": 5740
+    },
+    {
+      "epoch": 11.943866943866944,
+      "grad_norm": 0.08120215684175491,
+      "learning_rate": 4.749884592881212e-05,
+      "loss": 0.0155,
+      "num_input_tokens_seen": 4489040,
+      "step": 5745
+    },
+    {
+      "epoch": 11.954261954261954,
+      "grad_norm": 21.994937896728516,
+      "learning_rate": 4.74945639194687e-05,
+      "loss": 0.0091,
+      "num_input_tokens_seen": 4493136,
+      "step": 5750
+    },
+    {
+      "epoch": 11.964656964656964,
+      "grad_norm": 0.031408678740262985,
+      "learning_rate": 4.749027844118083e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 4496976,
+      "step": 5755
+    },
+    {
+      "epoch": 11.975051975051976,
+      "grad_norm": 0.013192703947424889,
+      "learning_rate": 4.7485989494609395e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4500784,
+      "step": 5760
+    },
+    {
+      "epoch": 11.985446985446986,
+      "grad_norm": 0.004290902987122536,
+      "learning_rate": 4.748169708041581e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 4504720,
+      "step": 5765
+    },
+    {
+      "epoch": 11.995841995841996,
+      "grad_norm": 0.005577556788921356,
+      "learning_rate": 4.7477401199262004e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 4508592,
+      "step": 5770
+    },
+    {
+      "epoch": 12.006237006237006,
+      "grad_norm": 0.005534137599170208,
+      "learning_rate": 4.747310185181048e-05,
+      "loss": 0.0068,
+      "num_input_tokens_seen": 4512392,
+      "step": 5775
+    },
+    {
+      "epoch": 12.016632016632016,
+      "grad_norm": 0.0018818869721144438,
+      "learning_rate": 4.746879903872422e-05,
+      "loss": 0.0183,
+      "num_input_tokens_seen": 4516392,
+      "step": 5780
+    },
+    {
+      "epoch": 12.027027027027026,
+      "grad_norm": 0.05994001775979996,
+      "learning_rate": 4.746449276066679e-05,
+      "loss": 0.014,
+      "num_input_tokens_seen": 4520232,
+      "step": 5785
+    },
+    {
+      "epoch": 12.037422037422038,
+      "grad_norm": 0.15930356085300446,
+      "learning_rate": 4.746018301830227e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 4524008,
+      "step": 5790
+    },
+    {
+      "epoch": 12.047817047817048,
+      "grad_norm": 0.06134532392024994,
+      "learning_rate": 4.7455869812295275e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 4527816,
+      "step": 5795
+    },
+    {
+      "epoch": 12.058212058212058,
+      "grad_norm": 0.04636925086379051,
+      "learning_rate": 4.7451553143310964e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 4531752,
+      "step": 5800
+    },
+    {
+      "epoch": 12.058212058212058,
+      "eval_loss": 0.30834639072418213,
+      "eval_runtime": 11.7253,
+      "eval_samples_per_second": 73.005,
+      "eval_steps_per_second": 18.251,
+      "num_input_tokens_seen": 4531752,
+      "step": 5800
+    },
+    {
+      "epoch": 12.068607068607069,
+      "grad_norm": 0.016977647319436073,
+      "learning_rate": 4.744723301201501e-05,
+      "loss": 0.0326,
+      "num_input_tokens_seen": 4535688,
+      "step": 5805
+    },
+    {
+      "epoch": 12.079002079002079,
+      "grad_norm": 0.07128752022981644,
+      "learning_rate": 4.744290941907364e-05,
+      "loss": 0.0135,
+      "num_input_tokens_seen": 4539720,
+      "step": 5810
+    },
+    {
+      "epoch": 12.089397089397089,
+      "grad_norm": 0.02939283475279808,
+      "learning_rate": 4.7438582365153594e-05,
+      "loss": 0.0154,
+      "num_input_tokens_seen": 4543720,
+      "step": 5815
+    },
+    {
+      "epoch": 12.0997920997921,
+      "grad_norm": 0.024673325940966606,
+      "learning_rate": 4.743425185092217e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 4547496,
+      "step": 5820
+    },
+    {
+      "epoch": 12.11018711018711,
+      "grad_norm": 0.026365164667367935,
+      "learning_rate": 4.742991787704719e-05,
+      "loss": 0.0128,
+      "num_input_tokens_seen": 4551272,
+      "step": 5825
+    },
+    {
+      "epoch": 12.120582120582121,
+      "grad_norm": 0.0711318626999855,
+      "learning_rate": 4.7425580444196994e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 4555176,
+      "step": 5830
+    },
+    {
+      "epoch": 12.130977130977131,
+      "grad_norm": 0.03738310933113098,
+      "learning_rate": 4.742123955304048e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 4558984,
+      "step": 5835
+    },
+    {
+      "epoch": 12.141372141372141,
+      "grad_norm": 2.1672463417053223,
+      "learning_rate": 4.741689520424706e-05,
+      "loss": 0.0224,
+      "num_input_tokens_seen": 4562792,
+      "step": 5840
+    },
+    {
+      "epoch": 12.151767151767151,
+      "grad_norm": 0.19997909665107727,
+      "learning_rate": 4.741254739848669e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 4566632,
+      "step": 5845
+    },
+    {
+      "epoch": 12.162162162162161,
+      "grad_norm": 0.11898019909858704,
+      "learning_rate": 4.740819613642987e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 4570504,
+      "step": 5850
+    },
+    {
+      "epoch": 12.172557172557173,
+      "grad_norm": 1.1623338460922241,
+      "learning_rate": 4.74038414187476e-05,
+      "loss": 0.01,
+      "num_input_tokens_seen": 4574472,
+      "step": 5855
+    },
+    {
+      "epoch": 12.182952182952183,
+      "grad_norm": 0.056794967502355576,
+      "learning_rate": 4.739948324611144e-05,
+      "loss": 0.0082,
+      "num_input_tokens_seen": 4578472,
+      "step": 5860
+    },
+    {
+      "epoch": 12.193347193347194,
+      "grad_norm": 6.193287372589111,
+      "learning_rate": 4.7395121619193465e-05,
+      "loss": 0.0166,
+      "num_input_tokens_seen": 4582376,
+      "step": 5865
+    },
+    {
+      "epoch": 12.203742203742204,
+      "grad_norm": 0.3576969802379608,
+      "learning_rate": 4.7390756538666313e-05,
+      "loss": 0.0339,
+      "num_input_tokens_seen": 4586248,
+      "step": 5870
+    },
+    {
+      "epoch": 12.214137214137214,
+      "grad_norm": 0.23271678388118744,
+      "learning_rate": 4.738638800520311e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 4590152,
+      "step": 5875
+    },
+    {
+      "epoch": 12.224532224532224,
+      "grad_norm": 0.03428751230239868,
+      "learning_rate": 4.738201601947757e-05,
+      "loss": 0.0152,
+      "num_input_tokens_seen": 4594120,
+      "step": 5880
+    },
+    {
+      "epoch": 12.234927234927236,
+      "grad_norm": 0.07020007073879242,
+      "learning_rate": 4.7377640582163876e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 4597992,
+      "step": 5885
+    },
+    {
+      "epoch": 12.245322245322246,
+      "grad_norm": 0.01610688678920269,
+      "learning_rate": 4.7373261693936786e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 4601992,
+      "step": 5890
+    },
+    {
+      "epoch": 12.255717255717256,
+      "grad_norm": 1.8543815612792969,
+      "learning_rate": 4.7368879355471595e-05,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 4605672,
+      "step": 5895
+    },
+    {
+      "epoch": 12.266112266112266,
+      "grad_norm": 1.3990095853805542,
+      "learning_rate": 4.736449356744409e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 4609512,
+      "step": 5900
+    },
+    {
+      "epoch": 12.276507276507276,
+      "grad_norm": 0.0080321179702878,
+      "learning_rate": 4.736010433053064e-05,
+      "loss": 0.0474,
+      "num_input_tokens_seen": 4613256,
+      "step": 5905
+    },
+    {
+      "epoch": 12.286902286902286,
+      "grad_norm": 1.0895771980285645,
+      "learning_rate": 4.73557116454081e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 4617064,
+      "step": 5910
+    },
+    {
+      "epoch": 12.297297297297296,
+      "grad_norm": 0.012539055198431015,
+      "learning_rate": 4.735131551275389e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4621000,
+      "step": 5915
+    },
+    {
+      "epoch": 12.307692307692308,
+      "grad_norm": 0.00727649312466383,
+      "learning_rate": 4.734691593324594e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 4624776,
+      "step": 5920
+    },
+    {
+      "epoch": 12.318087318087318,
+      "grad_norm": 1.8456015586853027,
+      "learning_rate": 4.734251290756272e-05,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 4628584,
+      "step": 5925
+    },
+    {
+      "epoch": 12.328482328482329,
+      "grad_norm": 0.007722186855971813,
+      "learning_rate": 4.7338106436383246e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 4632520,
+      "step": 5930
+    },
+    {
+      "epoch": 12.338877338877339,
+      "grad_norm": 0.025089988484978676,
+      "learning_rate": 4.733369652038703e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 4636424,
+      "step": 5935
+    },
+    {
+      "epoch": 12.349272349272349,
+      "grad_norm": 1.0568286180496216,
+      "learning_rate": 4.7329283160254156e-05,
+      "loss": 0.0297,
+      "num_input_tokens_seen": 4640264,
+      "step": 5940
+    },
+    {
+      "epoch": 12.359667359667359,
+      "grad_norm": 1.0728076696395874,
+      "learning_rate": 4.732486635666521e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 4644200,
+      "step": 5945
+    },
+    {
+      "epoch": 12.37006237006237,
+      "grad_norm": 9.51717758178711,
+      "learning_rate": 4.732044611030132e-05,
+      "loss": 0.0216,
+      "num_input_tokens_seen": 4648104,
+      "step": 5950
+    },
+    {
+      "epoch": 12.380457380457381,
+      "grad_norm": 0.019188659265637398,
+      "learning_rate": 4.731602242184414e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 4652072,
+      "step": 5955
+    },
+    {
+      "epoch": 12.390852390852391,
+      "grad_norm": 1.4538097381591797,
+      "learning_rate": 4.7311595291975864e-05,
+      "loss": 0.0154,
+      "num_input_tokens_seen": 4656136,
+      "step": 5960
+    },
+    {
+      "epoch": 12.401247401247401,
+      "grad_norm": 0.03708979859948158,
+      "learning_rate": 4.7307164721379216e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 4660072,
+      "step": 5965
+    },
+    {
+      "epoch": 12.411642411642411,
+      "grad_norm": 0.05756617337465286,
+      "learning_rate": 4.730273071073743e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 4663912,
+      "step": 5970
+    },
+    {
+      "epoch": 12.422037422037421,
+      "grad_norm": 0.09049296379089355,
+      "learning_rate": 4.729829326073429e-05,
+      "loss": 0.0121,
+      "num_input_tokens_seen": 4667784,
+      "step": 5975
+    },
+    {
+      "epoch": 12.432432432432432,
+      "grad_norm": 0.04940245300531387,
+      "learning_rate": 4.7293852372054126e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 4671656,
+      "step": 5980
+    },
+    {
+      "epoch": 12.442827442827443,
+      "grad_norm": 1.4520052671432495,
+      "learning_rate": 4.728940804538176e-05,
+      "loss": 0.0351,
+      "num_input_tokens_seen": 4675528,
+      "step": 5985
+    },
+    {
+      "epoch": 12.453222453222454,
+      "grad_norm": 0.01196135301142931,
+      "learning_rate": 4.7284960281402556e-05,
+      "loss": 0.0396,
+      "num_input_tokens_seen": 4679336,
+      "step": 5990
+    },
+    {
+      "epoch": 12.463617463617464,
+      "grad_norm": 0.06594218313694,
+      "learning_rate": 4.728050908080244e-05,
+      "loss": 0.0108,
+      "num_input_tokens_seen": 4683144,
+      "step": 5995
+    },
+    {
+      "epoch": 12.474012474012474,
+      "grad_norm": 0.04574023559689522,
+      "learning_rate": 4.727605444426782e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 4687112,
+      "step": 6000
+    },
+    {
+      "epoch": 12.474012474012474,
+      "eval_loss": 0.2677686810493469,
+      "eval_runtime": 11.7606,
+      "eval_samples_per_second": 72.786,
+      "eval_steps_per_second": 18.196,
+      "num_input_tokens_seen": 4687112,
+      "step": 6000
+    },
+    {
+      "epoch": 12.484407484407484,
+      "grad_norm": 0.044395044445991516,
+      "learning_rate": 4.727159637248567e-05,
+      "loss": 0.0126,
+      "num_input_tokens_seen": 4690984,
+      "step": 6005
+    },
+    {
+      "epoch": 12.494802494802494,
+      "grad_norm": 0.22760798037052155,
+      "learning_rate": 4.7267134866143474e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 4695048,
+      "step": 6010
+    },
+    {
+      "epoch": 12.505197505197506,
+      "grad_norm": 12.76021671295166,
+      "learning_rate": 4.726266992592926e-05,
+      "loss": 0.022,
+      "num_input_tokens_seen": 4698856,
+      "step": 6015
+    },
+    {
+      "epoch": 12.515592515592516,
+      "grad_norm": 0.1677766889333725,
+      "learning_rate": 4.725820155253157e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 4702760,
+      "step": 6020
+    },
+    {
+      "epoch": 12.525987525987526,
+      "grad_norm": 0.08048887550830841,
+      "learning_rate": 4.725372974663948e-05,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 4706632,
+      "step": 6025
+    },
+    {
+      "epoch": 12.536382536382536,
+      "grad_norm": 1.9121975898742676,
+      "learning_rate": 4.724925450894262e-05,
+      "loss": 0.0116,
+      "num_input_tokens_seen": 4710440,
+      "step": 6030
+    },
+    {
+      "epoch": 12.546777546777546,
+      "grad_norm": 0.09510556608438492,
+      "learning_rate": 4.72447758401311e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 4714280,
+      "step": 6035
+    },
+    {
+      "epoch": 12.557172557172557,
+      "grad_norm": 0.16276058554649353,
+      "learning_rate": 4.7240293740895616e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4718152,
+      "step": 6040
+    },
+    {
+      "epoch": 12.567567567567568,
+      "grad_norm": 0.027902526780962944,
+      "learning_rate": 4.723580821192733e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4722024,
+      "step": 6045
+    },
+    {
+      "epoch": 12.577962577962579,
+      "grad_norm": 0.028932638466358185,
+      "learning_rate": 4.7231319253917996e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 4725960,
+      "step": 6050
+    },
+    {
+      "epoch": 12.588357588357589,
+      "grad_norm": 0.02022162266075611,
+      "learning_rate": 4.722682686755986e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 4729800,
+      "step": 6055
+    },
+    {
+      "epoch": 12.598752598752599,
+      "grad_norm": 0.016622351482510567,
+      "learning_rate": 4.722233105354569e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 4733768,
+      "step": 6060
+    },
+    {
+      "epoch": 12.609147609147609,
+      "grad_norm": 8.536815643310547,
+      "learning_rate": 4.7217831812568815e-05,
+      "loss": 0.0148,
+      "num_input_tokens_seen": 4737608,
+      "step": 6065
+    },
+    {
+      "epoch": 12.619542619542619,
+      "grad_norm": 0.0035151138436049223,
+      "learning_rate": 4.721332914532307e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 4741704,
+      "step": 6070
+    },
+    {
+      "epoch": 12.62993762993763,
+      "grad_norm": 0.0004089566064067185,
+      "learning_rate": 4.720882305250281e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 4745448,
+      "step": 6075
+    },
+    {
+      "epoch": 12.640332640332641,
+      "grad_norm": 0.5120530128479004,
+      "learning_rate": 4.720431353480295e-05,
+      "loss": 0.0069,
+      "num_input_tokens_seen": 4749320,
+      "step": 6080
+    },
+    {
+      "epoch": 12.650727650727651,
+      "grad_norm": 0.019099649041891098,
+      "learning_rate": 4.719980059291891e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 4753128,
+      "step": 6085
+    },
+    {
+      "epoch": 12.661122661122661,
+      "grad_norm": 0.004581266548484564,
+      "learning_rate": 4.7195284227546634e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 4757160,
+      "step": 6090
+    },
+    {
+      "epoch": 12.671517671517671,
+      "grad_norm": 0.0021090894006192684,
+      "learning_rate": 4.7190764439382604e-05,
+      "loss": 0.0658,
+      "num_input_tokens_seen": 4761096,
+      "step": 6095
+    },
+    {
+      "epoch": 12.681912681912682,
+      "grad_norm": 0.05193767324090004,
+      "learning_rate": 4.7186241229123826e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4765032,
+      "step": 6100
+    },
+    {
+      "epoch": 12.692307692307692,
+      "grad_norm": 0.5144153237342834,
+      "learning_rate": 4.718171459746785e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4768744,
+      "step": 6105
+    },
+    {
+      "epoch": 12.702702702702704,
+      "grad_norm": 0.010315638966858387,
+      "learning_rate": 4.717718454511273e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4772744,
+      "step": 6110
+    },
+    {
+      "epoch": 12.713097713097714,
+      "grad_norm": 0.009729957208037376,
+      "learning_rate": 4.7172651072757056e-05,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 4776648,
+      "step": 6115
+    },
+    {
+      "epoch": 12.723492723492724,
+      "grad_norm": 0.4957438111305237,
+      "learning_rate": 4.7168114181099945e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 4780520,
+      "step": 6120
+    },
+    {
+      "epoch": 12.733887733887734,
+      "grad_norm": 0.027314627543091774,
+      "learning_rate": 4.716357387084105e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 4784328,
+      "step": 6125
+    },
+    {
+      "epoch": 12.744282744282744,
+      "grad_norm": 18.05086326599121,
+      "learning_rate": 4.715903014268054e-05,
+      "loss": 0.0271,
+      "num_input_tokens_seen": 4788232,
+      "step": 6130
+    },
+    {
+      "epoch": 12.754677754677754,
+      "grad_norm": 0.002224838826805353,
+      "learning_rate": 4.715448299731911e-05,
+      "loss": 0.03,
+      "num_input_tokens_seen": 4792168,
+      "step": 6135
+    },
+    {
+      "epoch": 12.765072765072766,
+      "grad_norm": 0.006576491054147482,
+      "learning_rate": 4.7149932435457986e-05,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 4795944,
+      "step": 6140
+    },
+    {
+      "epoch": 12.775467775467776,
+      "grad_norm": 2.61725115776062,
+      "learning_rate": 4.714537845779894e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 4799944,
+      "step": 6145
+    },
+    {
+      "epoch": 12.785862785862786,
+      "grad_norm": 0.005532352719455957,
+      "learning_rate": 4.714082106504423e-05,
+      "loss": 0.0156,
+      "num_input_tokens_seen": 4803912,
+      "step": 6150
+    },
+    {
+      "epoch": 12.796257796257796,
+      "grad_norm": 2.989269495010376,
+      "learning_rate": 4.713626025789667e-05,
+      "loss": 0.0137,
+      "num_input_tokens_seen": 4807784,
+      "step": 6155
+    },
+    {
+      "epoch": 12.806652806652806,
+      "grad_norm": 0.023813817650079727,
+      "learning_rate": 4.7131696037059606e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 4811624,
+      "step": 6160
+    },
+    {
+      "epoch": 12.817047817047817,
+      "grad_norm": 0.008196731097996235,
+      "learning_rate": 4.712712840323689e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4815592,
+      "step": 6165
+    },
+    {
+      "epoch": 12.827442827442827,
+      "grad_norm": 1.2622429132461548,
+      "learning_rate": 4.71225573571329e-05,
+      "loss": 0.0297,
+      "num_input_tokens_seen": 4819624,
+      "step": 6170
+    },
+    {
+      "epoch": 12.837837837837839,
+      "grad_norm": 0.03181881085038185,
+      "learning_rate": 4.711798289945256e-05,
+      "loss": 0.0222,
+      "num_input_tokens_seen": 4823592,
+      "step": 6175
+    },
+    {
+      "epoch": 12.848232848232849,
+      "grad_norm": 0.05753985047340393,
+      "learning_rate": 4.71134050309013e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 4827528,
+      "step": 6180
+    },
+    {
+      "epoch": 12.858627858627859,
+      "grad_norm": 1.033466100692749,
+      "learning_rate": 4.710882375218509e-05,
+      "loss": 0.0194,
+      "num_input_tokens_seen": 4831656,
+      "step": 6185
+    },
+    {
+      "epoch": 12.869022869022869,
+      "grad_norm": 0.01851142756640911,
+      "learning_rate": 4.7104239064010424e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 4835656,
+      "step": 6190
+    },
+    {
+      "epoch": 12.879417879417879,
+      "grad_norm": 0.01913604699075222,
+      "learning_rate": 4.709965096708432e-05,
+      "loss": 0.009,
+      "num_input_tokens_seen": 4839432,
+      "step": 6195
+    },
+    {
+      "epoch": 12.88981288981289,
+      "grad_norm": 0.0591636523604393,
+      "learning_rate": 4.709505946211431e-05,
+      "loss": 0.0072,
+      "num_input_tokens_seen": 4843464,
+      "step": 6200
+    },
+    {
+      "epoch": 12.88981288981289,
+      "eval_loss": 0.28426283597946167,
+      "eval_runtime": 11.7599,
+      "eval_samples_per_second": 72.79,
+      "eval_steps_per_second": 18.197,
+      "num_input_tokens_seen": 4843464,
+      "step": 6200
+    },
+    {
+      "epoch": 12.9002079002079,
+      "grad_norm": 0.006554158870130777,
+      "learning_rate": 4.709046454980846e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 4847336,
+      "step": 6205
+    },
+    {
+      "epoch": 12.910602910602911,
+      "grad_norm": 0.03891647979617119,
+      "learning_rate": 4.708586623087538e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 4851304,
+      "step": 6210
+    },
+    {
+      "epoch": 12.920997920997921,
+      "grad_norm": 0.06868544220924377,
+      "learning_rate": 4.708126450602418e-05,
+      "loss": 0.0121,
+      "num_input_tokens_seen": 4855240,
+      "step": 6215
+    },
+    {
+      "epoch": 12.931392931392931,
+      "grad_norm": 0.008104972541332245,
+      "learning_rate": 4.7076659375964495e-05,
+      "loss": 0.0179,
+      "num_input_tokens_seen": 4859240,
+      "step": 6220
+    },
+    {
+      "epoch": 12.941787941787942,
+      "grad_norm": 4.218451499938965,
+      "learning_rate": 4.707205084140651e-05,
+      "loss": 0.0242,
+      "num_input_tokens_seen": 4862984,
+      "step": 6225
+    },
+    {
+      "epoch": 12.952182952182952,
+      "grad_norm": 0.03341161459684372,
+      "learning_rate": 4.7067438903060904e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 4866856,
+      "step": 6230
+    },
+    {
+      "epoch": 12.962577962577962,
+      "grad_norm": 10.400771141052246,
+      "learning_rate": 4.70628235616389e-05,
+      "loss": 0.0124,
+      "num_input_tokens_seen": 4870792,
+      "step": 6235
+    },
+    {
+      "epoch": 12.972972972972974,
+      "grad_norm": 0.05861322581768036,
+      "learning_rate": 4.7058204817852256e-05,
+      "loss": 0.0194,
+      "num_input_tokens_seen": 4874600,
+      "step": 6240
+    },
+    {
+      "epoch": 12.983367983367984,
+      "grad_norm": 0.010822607204318047,
+      "learning_rate": 4.705358267241322e-05,
+      "loss": 0.0261,
+      "num_input_tokens_seen": 4878760,
+      "step": 6245
+    },
+    {
+      "epoch": 12.993762993762994,
+      "grad_norm": 0.021621907129883766,
+      "learning_rate": 4.704895712603459e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 4882792,
+      "step": 6250
+    },
+    {
+      "epoch": 13.004158004158004,
+      "grad_norm": 1.1583045721054077,
+      "learning_rate": 4.704432817942969e-05,
+      "loss": 0.0369,
+      "num_input_tokens_seen": 4886688,
+      "step": 6255
+    },
+    {
+      "epoch": 13.014553014553014,
+      "grad_norm": 0.032789867371320724,
+      "learning_rate": 4.703969583331236e-05,
+      "loss": 0.0161,
+      "num_input_tokens_seen": 4890560,
+      "step": 6260
+    },
+    {
+      "epoch": 13.024948024948024,
+      "grad_norm": 0.09883461147546768,
+      "learning_rate": 4.7035060088396965e-05,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 4894400,
+      "step": 6265
+    },
+    {
+      "epoch": 13.035343035343036,
+      "grad_norm": 0.10294045507907867,
+      "learning_rate": 4.703042094539839e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 4898304,
+      "step": 6270
+    },
+    {
+      "epoch": 13.045738045738046,
+      "grad_norm": 0.09711601585149765,
+      "learning_rate": 4.702577840503206e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 4902144,
+      "step": 6275
+    },
+    {
+      "epoch": 13.056133056133056,
+      "grad_norm": 0.12752005457878113,
+      "learning_rate": 4.70211324680139e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 4906048,
+      "step": 6280
+    },
+    {
+      "epoch": 13.066528066528067,
+      "grad_norm": 0.02415969781577587,
+      "learning_rate": 4.7016483135060386e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 4909984,
+      "step": 6285
+    },
+    {
+      "epoch": 13.076923076923077,
+      "grad_norm": 0.038762643933296204,
+      "learning_rate": 4.701183040688849e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 4913984,
+      "step": 6290
+    },
+    {
+      "epoch": 13.087318087318087,
+      "grad_norm": 0.023439249023795128,
+      "learning_rate": 4.700717428421573e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 4917888,
+      "step": 6295
+    },
+    {
+      "epoch": 13.097713097713097,
+      "grad_norm": 0.006205381825566292,
+      "learning_rate": 4.700251476776014e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4921856,
+      "step": 6300
+    },
+    {
+      "epoch": 13.108108108108109,
+      "grad_norm": 0.010002137161791325,
+      "learning_rate": 4.699785185824026e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 4925664,
+      "step": 6305
+    },
+    {
+      "epoch": 13.118503118503119,
+      "grad_norm": 0.029517097398638725,
+      "learning_rate": 4.699318555637519e-05,
+      "loss": 0.0166,
+      "num_input_tokens_seen": 4929568,
+      "step": 6310
+    },
+    {
+      "epoch": 13.128898128898129,
+      "grad_norm": 0.08562258630990982,
+      "learning_rate": 4.6988515862884525e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 4933536,
+      "step": 6315
+    },
+    {
+      "epoch": 13.13929313929314,
+      "grad_norm": 0.010957196354866028,
+      "learning_rate": 4.698384277848838e-05,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 4937312,
+      "step": 6320
+    },
+    {
+      "epoch": 13.14968814968815,
+      "grad_norm": 0.2961273193359375,
+      "learning_rate": 4.6979166303907425e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 4941344,
+      "step": 6325
+    },
+    {
+      "epoch": 13.16008316008316,
+      "grad_norm": 0.09368538856506348,
+      "learning_rate": 4.697448643986281e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 4945248,
+      "step": 6330
+    },
+    {
+      "epoch": 13.170478170478171,
+      "grad_norm": 0.0048649548552930355,
+      "learning_rate": 4.696980318707624e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4949056,
+      "step": 6335
+    },
+    {
+      "epoch": 13.180873180873181,
+      "grad_norm": 1.672773003578186,
+      "learning_rate": 4.6965116546269924e-05,
+      "loss": 0.0408,
+      "num_input_tokens_seen": 4952800,
+      "step": 6340
+    },
+    {
+      "epoch": 13.191268191268192,
+      "grad_norm": 0.10491890460252762,
+      "learning_rate": 4.6960426518166615e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 4956640,
+      "step": 6345
+    },
+    {
+      "epoch": 13.201663201663202,
+      "grad_norm": 0.01298011839389801,
+      "learning_rate": 4.6955733103489556e-05,
+      "loss": 0.0072,
+      "num_input_tokens_seen": 4960576,
+      "step": 6350
+    },
+    {
+      "epoch": 13.212058212058212,
+      "grad_norm": 0.0796072855591774,
+      "learning_rate": 4.695103630296255e-05,
+      "loss": 0.007,
+      "num_input_tokens_seen": 4964544,
+      "step": 6355
+    },
+    {
+      "epoch": 13.222453222453222,
+      "grad_norm": 0.11109264194965363,
+      "learning_rate": 4.694633611730988e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 4968512,
+      "step": 6360
+    },
+    {
+      "epoch": 13.232848232848234,
+      "grad_norm": 0.01887710578739643,
+      "learning_rate": 4.694163254725639e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4972416,
+      "step": 6365
+    },
+    {
+      "epoch": 13.243243243243244,
+      "grad_norm": 0.007285556755959988,
+      "learning_rate": 4.693692559352743e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 4976288,
+      "step": 6370
+    },
+    {
+      "epoch": 13.253638253638254,
+      "grad_norm": 0.0032066202256828547,
+      "learning_rate": 4.693221525684886e-05,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 4980224,
+      "step": 6375
+    },
+    {
+      "epoch": 13.264033264033264,
+      "grad_norm": 0.022992907091975212,
+      "learning_rate": 4.6927501537947084e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 4984128,
+      "step": 6380
+    },
+    {
+      "epoch": 13.274428274428274,
+      "grad_norm": 0.002715972950682044,
+      "learning_rate": 4.692278443754901e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 4988000,
+      "step": 6385
+    },
+    {
+      "epoch": 13.284823284823284,
+      "grad_norm": 0.0010495564201846719,
+      "learning_rate": 4.691806395638208e-05,
+      "loss": 0.0085,
+      "num_input_tokens_seen": 4992000,
+      "step": 6390
+    },
+    {
+      "epoch": 13.295218295218294,
+      "grad_norm": 0.0009055665577761829,
+      "learning_rate": 4.6913340095174255e-05,
+      "loss": 0.0066,
+      "num_input_tokens_seen": 4995840,
+      "step": 6395
+    },
+    {
+      "epoch": 13.305613305613306,
+      "grad_norm": 0.09054407477378845,
+      "learning_rate": 4.690861285465399e-05,
+      "loss": 0.0117,
+      "num_input_tokens_seen": 4999648,
+      "step": 6400
+    },
+    {
+      "epoch": 13.305613305613306,
+      "eval_loss": 0.3738417625427246,
+      "eval_runtime": 11.7448,
+      "eval_samples_per_second": 72.884,
+      "eval_steps_per_second": 18.221,
+      "num_input_tokens_seen": 4999648,
+      "step": 6400
+    },
+    {
+      "epoch": 13.316008316008316,
+      "grad_norm": 0.006280720233917236,
+      "learning_rate": 4.690388223555031e-05,
+      "loss": 0.0233,
+      "num_input_tokens_seen": 5003392,
+      "step": 6405
+    },
+    {
+      "epoch": 13.326403326403327,
+      "grad_norm": 2.9333152770996094,
+      "learning_rate": 4.689914823859273e-05,
+      "loss": 0.0309,
+      "num_input_tokens_seen": 5007424,
+      "step": 6410
+    },
+    {
+      "epoch": 13.336798336798337,
+      "grad_norm": 0.13100193440914154,
+      "learning_rate": 4.689441086451129e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5011264,
+      "step": 6415
+    },
+    {
+      "epoch": 13.347193347193347,
+      "grad_norm": 0.007129251956939697,
+      "learning_rate": 4.688967011403655e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 5015136,
+      "step": 6420
+    },
+    {
+      "epoch": 13.357588357588357,
+      "grad_norm": 0.05735008046030998,
+      "learning_rate": 4.68849259878996e-05,
+      "loss": 0.0136,
+      "num_input_tokens_seen": 5019104,
+      "step": 6425
+    },
+    {
+      "epoch": 13.367983367983369,
+      "grad_norm": 0.27618882060050964,
+      "learning_rate": 4.6880178486832036e-05,
+      "loss": 0.0139,
+      "num_input_tokens_seen": 5023136,
+      "step": 6430
+    },
+    {
+      "epoch": 13.378378378378379,
+      "grad_norm": 9.256946563720703,
+      "learning_rate": 4.687542761156598e-05,
+      "loss": 0.0249,
+      "num_input_tokens_seen": 5026944,
+      "step": 6435
+    },
+    {
+      "epoch": 13.388773388773389,
+      "grad_norm": 0.2979963421821594,
+      "learning_rate": 4.6870673362834096e-05,
+      "loss": 0.0125,
+      "num_input_tokens_seen": 5030816,
+      "step": 6440
+    },
+    {
+      "epoch": 13.3991683991684,
+      "grad_norm": 0.06246345117688179,
+      "learning_rate": 4.6865915741369526e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5034816,
+      "step": 6445
+    },
+    {
+      "epoch": 13.40956340956341,
+      "grad_norm": 0.015266713686287403,
+      "learning_rate": 4.686115474790597e-05,
+      "loss": 0.011,
+      "num_input_tokens_seen": 5038720,
+      "step": 6450
+    },
+    {
+      "epoch": 13.41995841995842,
+      "grad_norm": 0.08207783848047256,
+      "learning_rate": 4.685639038317762e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 5042528,
+      "step": 6455
+    },
+    {
+      "epoch": 13.43035343035343,
+      "grad_norm": 2.391996383666992,
+      "learning_rate": 4.685162264791921e-05,
+      "loss": 0.015,
+      "num_input_tokens_seen": 5046560,
+      "step": 6460
+    },
+    {
+      "epoch": 13.440748440748441,
+      "grad_norm": 2.09771728515625,
+      "learning_rate": 4.684685154286599e-05,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 5050464,
+      "step": 6465
+    },
+    {
+      "epoch": 13.451143451143452,
+      "grad_norm": 0.03640754893422127,
+      "learning_rate": 4.684207706875371e-05,
+      "loss": 0.0069,
+      "num_input_tokens_seen": 5054304,
+      "step": 6470
+    },
+    {
+      "epoch": 13.461538461538462,
+      "grad_norm": 0.038274042308330536,
+      "learning_rate": 4.683729922631866e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 5058208,
+      "step": 6475
+    },
+    {
+      "epoch": 13.471933471933472,
+      "grad_norm": 0.008901509456336498,
+      "learning_rate": 4.683251801629765e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 5062144,
+      "step": 6480
+    },
+    {
+      "epoch": 13.482328482328482,
+      "grad_norm": 0.7059668302536011,
+      "learning_rate": 4.6827733439428e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 5066176,
+      "step": 6485
+    },
+    {
+      "epoch": 13.492723492723492,
+      "grad_norm": 3.4031314849853516,
+      "learning_rate": 4.682294549644754e-05,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 5069984,
+      "step": 6490
+    },
+    {
+      "epoch": 13.503118503118504,
+      "grad_norm": 0.0025589480064809322,
+      "learning_rate": 4.681815418809464e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 5074208,
+      "step": 6495
+    },
+    {
+      "epoch": 13.513513513513514,
+      "grad_norm": 25.08586311340332,
+      "learning_rate": 4.681335951510819e-05,
+      "loss": 0.0192,
+      "num_input_tokens_seen": 5078112,
+      "step": 6500
+    },
+    {
+      "epoch": 13.523908523908524,
+      "grad_norm": 9.712923049926758,
+      "learning_rate": 4.6808561478227576e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 5082048,
+      "step": 6505
+    },
+    {
+      "epoch": 13.534303534303534,
+      "grad_norm": 0.7734810709953308,
+      "learning_rate": 4.680376007819271e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 5085984,
+      "step": 6510
+    },
+    {
+      "epoch": 13.544698544698544,
+      "grad_norm": 0.15039609372615814,
+      "learning_rate": 4.679895531574405e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5089920,
+      "step": 6515
+    },
+    {
+      "epoch": 13.555093555093555,
+      "grad_norm": 0.00615449994802475,
+      "learning_rate": 4.679414719162253e-05,
+      "loss": 0.0125,
+      "num_input_tokens_seen": 5093792,
+      "step": 6520
+    },
+    {
+      "epoch": 13.565488565488565,
+      "grad_norm": 20.813987731933594,
+      "learning_rate": 4.6789335706569635e-05,
+      "loss": 0.0717,
+      "num_input_tokens_seen": 5097600,
+      "step": 6525
+    },
+    {
+      "epoch": 13.575883575883577,
+      "grad_norm": 14.720486640930176,
+      "learning_rate": 4.678452086132734e-05,
+      "loss": 0.0292,
+      "num_input_tokens_seen": 5101440,
+      "step": 6530
+    },
+    {
+      "epoch": 13.586278586278587,
+      "grad_norm": 0.0569884330034256,
+      "learning_rate": 4.677970265663818e-05,
+      "loss": 0.0427,
+      "num_input_tokens_seen": 5105408,
+      "step": 6535
+    },
+    {
+      "epoch": 13.596673596673597,
+      "grad_norm": 0.9164248108863831,
+      "learning_rate": 4.677488109324517e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 5109472,
+      "step": 6540
+    },
+    {
+      "epoch": 13.607068607068607,
+      "grad_norm": 0.2685730457305908,
+      "learning_rate": 4.6770056171891846e-05,
+      "loss": 0.0346,
+      "num_input_tokens_seen": 5113504,
+      "step": 6545
+    },
+    {
+      "epoch": 13.617463617463617,
+      "grad_norm": 0.2764210104942322,
+      "learning_rate": 4.6765227893322286e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 5117472,
+      "step": 6550
+    },
+    {
+      "epoch": 13.627858627858627,
+      "grad_norm": 0.010970736853778362,
+      "learning_rate": 4.676039625828107e-05,
+      "loss": 0.021,
+      "num_input_tokens_seen": 5121440,
+      "step": 6555
+    },
+    {
+      "epoch": 13.638253638253639,
+      "grad_norm": 0.035065677016973495,
+      "learning_rate": 4.675556126751328e-05,
+      "loss": 0.0118,
+      "num_input_tokens_seen": 5125440,
+      "step": 6560
+    },
+    {
+      "epoch": 13.64864864864865,
+      "grad_norm": 4.2225775718688965,
+      "learning_rate": 4.6750722921764556e-05,
+      "loss": 0.0396,
+      "num_input_tokens_seen": 5129472,
+      "step": 6565
+    },
+    {
+      "epoch": 13.65904365904366,
+      "grad_norm": 0.036204490810632706,
+      "learning_rate": 4.674588122178102e-05,
+      "loss": 0.0086,
+      "num_input_tokens_seen": 5133312,
+      "step": 6570
+    },
+    {
+      "epoch": 13.66943866943867,
+      "grad_norm": 0.08017393946647644,
+      "learning_rate": 4.674103616830931e-05,
+      "loss": 0.009,
+      "num_input_tokens_seen": 5137312,
+      "step": 6575
+    },
+    {
+      "epoch": 13.67983367983368,
+      "grad_norm": 0.069004587829113,
+      "learning_rate": 4.673618776209663e-05,
+      "loss": 0.0098,
+      "num_input_tokens_seen": 5141280,
+      "step": 6580
+    },
+    {
+      "epoch": 13.69022869022869,
+      "grad_norm": 0.018386686220765114,
+      "learning_rate": 4.673133600389063e-05,
+      "loss": 0.0106,
+      "num_input_tokens_seen": 5145280,
+      "step": 6585
+    },
+    {
+      "epoch": 13.700623700623701,
+      "grad_norm": 0.051297374069690704,
+      "learning_rate": 4.672648089443953e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 5149280,
+      "step": 6590
+    },
+    {
+      "epoch": 13.711018711018712,
+      "grad_norm": 0.16853292286396027,
+      "learning_rate": 4.672162243449204e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 5153312,
+      "step": 6595
+    },
+    {
+      "epoch": 13.721413721413722,
+      "grad_norm": 3.2588884830474854,
+      "learning_rate": 4.67167606247974e-05,
+      "loss": 0.0599,
+      "num_input_tokens_seen": 5157152,
+      "step": 6600
+    },
+    {
+      "epoch": 13.721413721413722,
+      "eval_loss": 0.3713972270488739,
+      "eval_runtime": 11.7473,
+      "eval_samples_per_second": 72.868,
+      "eval_steps_per_second": 18.217,
+      "num_input_tokens_seen": 5157152,
+      "step": 6600
+    },
+    {
+      "epoch": 13.731808731808732,
+      "grad_norm": 0.13597096502780914,
+      "learning_rate": 4.671189546610536e-05,
+      "loss": 0.0257,
+      "num_input_tokens_seen": 5161120,
+      "step": 6605
+    },
+    {
+      "epoch": 13.742203742203742,
+      "grad_norm": 0.239722341299057,
+      "learning_rate": 4.67070269591662e-05,
+      "loss": 0.0141,
+      "num_input_tokens_seen": 5164864,
+      "step": 6610
+    },
+    {
+      "epoch": 13.752598752598752,
+      "grad_norm": 0.13080227375030518,
+      "learning_rate": 4.670215510473068e-05,
+      "loss": 0.0108,
+      "num_input_tokens_seen": 5168768,
+      "step": 6615
+    },
+    {
+      "epoch": 13.762993762993762,
+      "grad_norm": 1.232017993927002,
+      "learning_rate": 4.669727990355013e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 5172544,
+      "step": 6620
+    },
+    {
+      "epoch": 13.773388773388774,
+      "grad_norm": 0.0744221955537796,
+      "learning_rate": 4.669240135637635e-05,
+      "loss": 0.0047,
+      "num_input_tokens_seen": 5176384,
+      "step": 6625
+    },
+    {
+      "epoch": 13.783783783783784,
+      "grad_norm": 0.03181573748588562,
+      "learning_rate": 4.6687519463961675e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 5180192,
+      "step": 6630
+    },
+    {
+      "epoch": 13.794178794178794,
+      "grad_norm": 1.634863257408142,
+      "learning_rate": 4.668263422705896e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 5183936,
+      "step": 6635
+    },
+    {
+      "epoch": 13.804573804573804,
+      "grad_norm": 0.025549843907356262,
+      "learning_rate": 4.667774564642156e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5187808,
+      "step": 6640
+    },
+    {
+      "epoch": 13.814968814968815,
+      "grad_norm": 0.03881119564175606,
+      "learning_rate": 4.6672853722803365e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5191552,
+      "step": 6645
+    },
+    {
+      "epoch": 13.825363825363825,
+      "grad_norm": 0.03980828821659088,
+      "learning_rate": 4.666795845695877e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5195424,
+      "step": 6650
+    },
+    {
+      "epoch": 13.835758835758837,
+      "grad_norm": 0.02431022934615612,
+      "learning_rate": 4.666305984964269e-05,
+      "loss": 0.0549,
+      "num_input_tokens_seen": 5199296,
+      "step": 6655
+    },
+    {
+      "epoch": 13.846153846153847,
+      "grad_norm": 0.01673826016485691,
+      "learning_rate": 4.6658157901610535e-05,
+      "loss": 0.0117,
+      "num_input_tokens_seen": 5203296,
+      "step": 6660
+    },
+    {
+      "epoch": 13.856548856548857,
+      "grad_norm": 0.11518001556396484,
+      "learning_rate": 4.665325261361826e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5207136,
+      "step": 6665
+    },
+    {
+      "epoch": 13.866943866943867,
+      "grad_norm": 0.006322914734482765,
+      "learning_rate": 4.664834398642232e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5210976,
+      "step": 6670
+    },
+    {
+      "epoch": 13.877338877338877,
+      "grad_norm": 0.002222065581008792,
+      "learning_rate": 4.6643432020779686e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5215008,
+      "step": 6675
+    },
+    {
+      "epoch": 13.887733887733887,
+      "grad_norm": 0.0037550954148173332,
+      "learning_rate": 4.663851671744786e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5218976,
+      "step": 6680
+    },
+    {
+      "epoch": 13.898128898128899,
+      "grad_norm": 0.002217038767412305,
+      "learning_rate": 4.6633598077184815e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5222976,
+      "step": 6685
+    },
+    {
+      "epoch": 13.90852390852391,
+      "grad_norm": 0.0027837231755256653,
+      "learning_rate": 4.662867610074908e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5226944,
+      "step": 6690
+    },
+    {
+      "epoch": 13.91891891891892,
+      "grad_norm": 0.0013817045837640762,
+      "learning_rate": 4.6623750788899696e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5230912,
+      "step": 6695
+    },
+    {
+      "epoch": 13.92931392931393,
+      "grad_norm": 1.1704003810882568,
+      "learning_rate": 4.6618822142396195e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5234784,
+      "step": 6700
+    },
+    {
+      "epoch": 13.93970893970894,
+      "grad_norm": 0.008950392715632915,
+      "learning_rate": 4.661389016199864e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5238752,
+      "step": 6705
+    },
+    {
+      "epoch": 13.95010395010395,
+      "grad_norm": 0.043816354125738144,
+      "learning_rate": 4.660895484846761e-05,
+      "loss": 0.0497,
+      "num_input_tokens_seen": 5242592,
+      "step": 6710
+    },
+    {
+      "epoch": 13.96049896049896,
+      "grad_norm": 0.06151091307401657,
+      "learning_rate": 4.660401620256418e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 5246496,
+      "step": 6715
+    },
+    {
+      "epoch": 13.970893970893972,
+      "grad_norm": 4.49084997177124,
+      "learning_rate": 4.659907422504997e-05,
+      "loss": 0.0068,
+      "num_input_tokens_seen": 5250304,
+      "step": 6720
+    },
+    {
+      "epoch": 13.981288981288982,
+      "grad_norm": 1.0494707822799683,
+      "learning_rate": 4.6594128916687074e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 5254176,
+      "step": 6725
+    },
+    {
+      "epoch": 13.991683991683992,
+      "grad_norm": 0.0032667324412614107,
+      "learning_rate": 4.658918027823813e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5258016,
+      "step": 6730
+    },
+    {
+      "epoch": 14.002079002079002,
+      "grad_norm": 0.2955338656902313,
+      "learning_rate": 4.658422831046628e-05,
+      "loss": 0.0556,
+      "num_input_tokens_seen": 5261864,
+      "step": 6735
+    },
+    {
+      "epoch": 14.012474012474012,
+      "grad_norm": 0.0022966559045016766,
+      "learning_rate": 4.657927301413518e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5265672,
+      "step": 6740
+    },
+    {
+      "epoch": 14.022869022869022,
+      "grad_norm": 0.0012907760683447123,
+      "learning_rate": 4.657431439000901e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5269576,
+      "step": 6745
+    },
+    {
+      "epoch": 14.033264033264032,
+      "grad_norm": 0.030306093394756317,
+      "learning_rate": 4.656935243885243e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 5273512,
+      "step": 6750
+    },
+    {
+      "epoch": 14.043659043659044,
+      "grad_norm": 0.0015927221393212676,
+      "learning_rate": 4.656438716143066e-05,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 5277480,
+      "step": 6755
+    },
+    {
+      "epoch": 14.054054054054054,
+      "grad_norm": 0.004528914112597704,
+      "learning_rate": 4.6559418558509384e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5281224,
+      "step": 6760
+    },
+    {
+      "epoch": 14.064449064449065,
+      "grad_norm": 0.0019111852161586285,
+      "learning_rate": 4.6554446630854833e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5285160,
+      "step": 6765
+    },
+    {
+      "epoch": 14.074844074844075,
+      "grad_norm": 0.515924334526062,
+      "learning_rate": 4.654947137923374e-05,
+      "loss": 0.0254,
+      "num_input_tokens_seen": 5289032,
+      "step": 6770
+    },
+    {
+      "epoch": 14.085239085239085,
+      "grad_norm": 0.00843868125230074,
+      "learning_rate": 4.654449280441335e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5292808,
+      "step": 6775
+    },
+    {
+      "epoch": 14.095634095634095,
+      "grad_norm": 0.5000017285346985,
+      "learning_rate": 4.653951090716143e-05,
+      "loss": 0.0117,
+      "num_input_tokens_seen": 5296616,
+      "step": 6780
+    },
+    {
+      "epoch": 14.106029106029107,
+      "grad_norm": 0.03468581289052963,
+      "learning_rate": 4.653452568824625e-05,
+      "loss": 0.0068,
+      "num_input_tokens_seen": 5300520,
+      "step": 6785
+    },
+    {
+      "epoch": 14.116424116424117,
+      "grad_norm": 1.341816782951355,
+      "learning_rate": 4.6529537148436585e-05,
+      "loss": 0.0081,
+      "num_input_tokens_seen": 5304552,
+      "step": 6790
+    },
+    {
+      "epoch": 14.126819126819127,
+      "grad_norm": 0.030397871509194374,
+      "learning_rate": 4.6524545288501734e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 5308488,
+      "step": 6795
+    },
+    {
+      "epoch": 14.137214137214137,
+      "grad_norm": 0.013919285498559475,
+      "learning_rate": 4.6519550109211506e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 5312328,
+      "step": 6800
+    },
+    {
+      "epoch": 14.137214137214137,
+      "eval_loss": 0.2459891140460968,
+      "eval_runtime": 11.8348,
+      "eval_samples_per_second": 72.329,
+      "eval_steps_per_second": 18.082,
+      "num_input_tokens_seen": 5312328,
+      "step": 6800
+    },
+    {
+      "epoch": 14.147609147609147,
+      "grad_norm": 0.14261461794376373,
+      "learning_rate": 4.651455161133622e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 5316232,
+      "step": 6805
+    },
+    {
+      "epoch": 14.158004158004157,
+      "grad_norm": 0.0058163865469396114,
+      "learning_rate": 4.6509549795646704e-05,
+      "loss": 0.0117,
+      "num_input_tokens_seen": 5320200,
+      "step": 6810
+    },
+    {
+      "epoch": 14.16839916839917,
+      "grad_norm": 0.00735585018992424,
+      "learning_rate": 4.6504544662914306e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5323976,
+      "step": 6815
+    },
+    {
+      "epoch": 14.17879417879418,
+      "grad_norm": 0.05532779172062874,
+      "learning_rate": 4.6499536213910876e-05,
+      "loss": 0.0245,
+      "num_input_tokens_seen": 5327752,
+      "step": 6820
+    },
+    {
+      "epoch": 14.18918918918919,
+      "grad_norm": 0.03135530278086662,
+      "learning_rate": 4.6494524449408786e-05,
+      "loss": 0.0059,
+      "num_input_tokens_seen": 5331720,
+      "step": 6825
+    },
+    {
+      "epoch": 14.1995841995842,
+      "grad_norm": 0.03382992744445801,
+      "learning_rate": 4.6489509370180903e-05,
+      "loss": 0.0219,
+      "num_input_tokens_seen": 5335560,
+      "step": 6830
+    },
+    {
+      "epoch": 14.20997920997921,
+      "grad_norm": 0.051179494708776474,
+      "learning_rate": 4.648449097700063e-05,
+      "loss": 0.011,
+      "num_input_tokens_seen": 5339528,
+      "step": 6835
+    },
+    {
+      "epoch": 14.22037422037422,
+      "grad_norm": 0.0022999378852546215,
+      "learning_rate": 4.647946927064185e-05,
+      "loss": 0.0099,
+      "num_input_tokens_seen": 5343496,
+      "step": 6840
+    },
+    {
+      "epoch": 14.23076923076923,
+      "grad_norm": 0.17613603174686432,
+      "learning_rate": 4.647444425187898e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5347464,
+      "step": 6845
+    },
+    {
+      "epoch": 14.241164241164242,
+      "grad_norm": 0.044802144169807434,
+      "learning_rate": 4.646941592148695e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 5351464,
+      "step": 6850
+    },
+    {
+      "epoch": 14.251559251559252,
+      "grad_norm": 7.770108222961426,
+      "learning_rate": 4.646438428024117e-05,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 5355592,
+      "step": 6855
+    },
+    {
+      "epoch": 14.261954261954262,
+      "grad_norm": 0.007711804937571287,
+      "learning_rate": 4.64593493289176e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5359432,
+      "step": 6860
+    },
+    {
+      "epoch": 14.272349272349272,
+      "grad_norm": 0.001642890041694045,
+      "learning_rate": 4.64543110682927e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5363368,
+      "step": 6865
+    },
+    {
+      "epoch": 14.282744282744282,
+      "grad_norm": 0.003832791931927204,
+      "learning_rate": 4.644926949914341e-05,
+      "loss": 0.0067,
+      "num_input_tokens_seen": 5367336,
+      "step": 6870
+    },
+    {
+      "epoch": 14.293139293139292,
+      "grad_norm": 0.0020148465409874916,
+      "learning_rate": 4.644422462224722e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5371112,
+      "step": 6875
+    },
+    {
+      "epoch": 14.303534303534304,
+      "grad_norm": 0.015755318105220795,
+      "learning_rate": 4.643917643838211e-05,
+      "loss": 0.0213,
+      "num_input_tokens_seen": 5374952,
+      "step": 6880
+    },
+    {
+      "epoch": 14.313929313929314,
+      "grad_norm": 0.027270758524537086,
+      "learning_rate": 4.6434124948326564e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5378760,
+      "step": 6885
+    },
+    {
+      "epoch": 14.324324324324325,
+      "grad_norm": 0.016694605350494385,
+      "learning_rate": 4.6429070152859594e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 5382600,
+      "step": 6890
+    },
+    {
+      "epoch": 14.334719334719335,
+      "grad_norm": 0.048291273415088654,
+      "learning_rate": 4.6424012052760714e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5386408,
+      "step": 6895
+    },
+    {
+      "epoch": 14.345114345114345,
+      "grad_norm": 16.030696868896484,
+      "learning_rate": 4.6418950648809945e-05,
+      "loss": 0.0119,
+      "num_input_tokens_seen": 5390248,
+      "step": 6900
+    },
+    {
+      "epoch": 14.355509355509355,
+      "grad_norm": 6.376165866851807,
+      "learning_rate": 4.641388594178782e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 5394024,
+      "step": 6905
+    },
+    {
+      "epoch": 14.365904365904367,
+      "grad_norm": 2.5552306175231934,
+      "learning_rate": 4.640881793247538e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 5397992,
+      "step": 6910
+    },
+    {
+      "epoch": 14.376299376299377,
+      "grad_norm": 0.017878534272313118,
+      "learning_rate": 4.6403746621654173e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 5401800,
+      "step": 6915
+    },
+    {
+      "epoch": 14.386694386694387,
+      "grad_norm": 2.354126453399658,
+      "learning_rate": 4.639867201010626e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 5405864,
+      "step": 6920
+    },
+    {
+      "epoch": 14.397089397089397,
+      "grad_norm": 0.2535584270954132,
+      "learning_rate": 4.6393594098614204e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5409640,
+      "step": 6925
+    },
+    {
+      "epoch": 14.407484407484407,
+      "grad_norm": 0.00797057244926691,
+      "learning_rate": 4.63885128879611e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5413544,
+      "step": 6930
+    },
+    {
+      "epoch": 14.417879417879417,
+      "grad_norm": 0.005766143091022968,
+      "learning_rate": 4.638342837893052e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 5417640,
+      "step": 6935
+    },
+    {
+      "epoch": 14.428274428274428,
+      "grad_norm": 0.0041153086349368095,
+      "learning_rate": 4.6378340572306565e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5421640,
+      "step": 6940
+    },
+    {
+      "epoch": 14.43866943866944,
+      "grad_norm": 0.003298897761851549,
+      "learning_rate": 4.6373249468873833e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 5425544,
+      "step": 6945
+    },
+    {
+      "epoch": 14.44906444906445,
+      "grad_norm": 1.105170488357544,
+      "learning_rate": 4.636815506941744e-05,
+      "loss": 0.0214,
+      "num_input_tokens_seen": 5429352,
+      "step": 6950
+    },
+    {
+      "epoch": 14.45945945945946,
+      "grad_norm": 0.0012863079318776727,
+      "learning_rate": 4.6363057374723004e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 5433256,
+      "step": 6955
+    },
+    {
+      "epoch": 14.46985446985447,
+      "grad_norm": 0.19801227748394012,
+      "learning_rate": 4.635795638557666e-05,
+      "loss": 0.0091,
+      "num_input_tokens_seen": 5437128,
+      "step": 6960
+    },
+    {
+      "epoch": 14.48024948024948,
+      "grad_norm": 0.0006170738488435745,
+      "learning_rate": 4.635285210276504e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5441000,
+      "step": 6965
+    },
+    {
+      "epoch": 14.49064449064449,
+      "grad_norm": 1.7611429691314697,
+      "learning_rate": 4.6347744527075295e-05,
+      "loss": 0.0153,
+      "num_input_tokens_seen": 5444968,
+      "step": 6970
+    },
+    {
+      "epoch": 14.5010395010395,
+      "grad_norm": 0.030574681237339973,
+      "learning_rate": 4.634263365929506e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5448808,
+      "step": 6975
+    },
+    {
+      "epoch": 14.511434511434512,
+      "grad_norm": 0.09469109773635864,
+      "learning_rate": 4.6337519500212515e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5452648,
+      "step": 6980
+    },
+    {
+      "epoch": 14.521829521829522,
+      "grad_norm": 1.1469786167144775,
+      "learning_rate": 4.633240205061632e-05,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 5456680,
+      "step": 6985
+    },
+    {
+      "epoch": 14.532224532224532,
+      "grad_norm": 0.730315625667572,
+      "learning_rate": 4.632728131129565e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 5460648,
+      "step": 6990
+    },
+    {
+      "epoch": 14.542619542619542,
+      "grad_norm": 15.928923606872559,
+      "learning_rate": 4.632215728304018e-05,
+      "loss": 0.0255,
+      "num_input_tokens_seen": 5464712,
+      "step": 6995
+    },
+    {
+      "epoch": 14.553014553014552,
+      "grad_norm": 0.012449370697140694,
+      "learning_rate": 4.63170299666401e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5468680,
+      "step": 7000
+    },
+    {
+      "epoch": 14.553014553014552,
+      "eval_loss": 0.3300459086894989,
+      "eval_runtime": 11.7722,
+      "eval_samples_per_second": 72.714,
+      "eval_steps_per_second": 18.178,
+      "num_input_tokens_seen": 5468680,
+      "step": 7000
+    },
+    {
+      "epoch": 14.563409563409563,
+      "grad_norm": 0.017549606040120125,
+      "learning_rate": 4.631189936288612e-05,
+      "loss": 0.0102,
+      "num_input_tokens_seen": 5472808,
+      "step": 7005
+    },
+    {
+      "epoch": 14.573804573804575,
+      "grad_norm": 0.03761540725827217,
+      "learning_rate": 4.630676547256944e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 5476808,
+      "step": 7010
+    },
+    {
+      "epoch": 14.584199584199585,
+      "grad_norm": 0.0119699751958251,
+      "learning_rate": 4.630162829648176e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 5480616,
+      "step": 7015
+    },
+    {
+      "epoch": 14.594594594594595,
+      "grad_norm": 0.06385606527328491,
+      "learning_rate": 4.629648783541531e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 5484552,
+      "step": 7020
+    },
+    {
+      "epoch": 14.604989604989605,
+      "grad_norm": 0.024332210421562195,
+      "learning_rate": 4.6291344090162804e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5488392,
+      "step": 7025
+    },
+    {
+      "epoch": 14.615384615384615,
+      "grad_norm": 0.0003160382038913667,
+      "learning_rate": 4.628619706151748e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5492392,
+      "step": 7030
+    },
+    {
+      "epoch": 14.625779625779625,
+      "grad_norm": 0.014940068125724792,
+      "learning_rate": 4.628104675027306e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5496232,
+      "step": 7035
+    },
+    {
+      "epoch": 14.636174636174637,
+      "grad_norm": 0.0005278954631648958,
+      "learning_rate": 4.6275893157223805e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5500232,
+      "step": 7040
+    },
+    {
+      "epoch": 14.646569646569647,
+      "grad_norm": 2.1316137313842773,
+      "learning_rate": 4.627073628316445e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 5504232,
+      "step": 7045
+    },
+    {
+      "epoch": 14.656964656964657,
+      "grad_norm": 0.00032588490284979343,
+      "learning_rate": 4.626557612889026e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5508168,
+      "step": 7050
+    },
+    {
+      "epoch": 14.667359667359667,
+      "grad_norm": 0.0011168497148901224,
+      "learning_rate": 4.626041269519699e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5512136,
+      "step": 7055
+    },
+    {
+      "epoch": 14.677754677754677,
+      "grad_norm": 0.0006210098508745432,
+      "learning_rate": 4.6255245982880905e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 5515912,
+      "step": 7060
+    },
+    {
+      "epoch": 14.688149688149688,
+      "grad_norm": 0.002803626237437129,
+      "learning_rate": 4.625007599273879e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5519848,
+      "step": 7065
+    },
+    {
+      "epoch": 14.698544698544698,
+      "grad_norm": 0.0457867756485939,
+      "learning_rate": 4.6244902725567895e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 5523752,
+      "step": 7070
+    },
+    {
+      "epoch": 14.70893970893971,
+      "grad_norm": 0.00037687915028072894,
+      "learning_rate": 4.6239726182166024e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5527688,
+      "step": 7075
+    },
+    {
+      "epoch": 14.71933471933472,
+      "grad_norm": 2.917299270629883,
+      "learning_rate": 4.623454636333147e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5531656,
+      "step": 7080
+    },
+    {
+      "epoch": 14.72972972972973,
+      "grad_norm": 47.633644104003906,
+      "learning_rate": 4.622936326986301e-05,
+      "loss": 0.0997,
+      "num_input_tokens_seen": 5535592,
+      "step": 7085
+    },
+    {
+      "epoch": 14.74012474012474,
+      "grad_norm": 0.31589916348457336,
+      "learning_rate": 4.6224176902559946e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 5539432,
+      "step": 7090
+    },
+    {
+      "epoch": 14.75051975051975,
+      "grad_norm": 0.06226380541920662,
+      "learning_rate": 4.621898726222209e-05,
+      "loss": 0.014,
+      "num_input_tokens_seen": 5543368,
+      "step": 7095
+    },
+    {
+      "epoch": 14.76091476091476,
+      "grad_norm": 18.08425521850586,
+      "learning_rate": 4.6213794349649744e-05,
+      "loss": 0.0091,
+      "num_input_tokens_seen": 5547144,
+      "step": 7100
+    },
+    {
+      "epoch": 14.771309771309772,
+      "grad_norm": 0.08277711272239685,
+      "learning_rate": 4.6208598165643715e-05,
+      "loss": 0.0596,
+      "num_input_tokens_seen": 5550984,
+      "step": 7105
+    },
+    {
+      "epoch": 14.781704781704782,
+      "grad_norm": 0.03559502586722374,
+      "learning_rate": 4.620339871100533e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 5554952,
+      "step": 7110
+    },
+    {
+      "epoch": 14.792099792099792,
+      "grad_norm": 0.7523064613342285,
+      "learning_rate": 4.6198195986536394e-05,
+      "loss": 0.0055,
+      "num_input_tokens_seen": 5558728,
+      "step": 7115
+    },
+    {
+      "epoch": 14.802494802494802,
+      "grad_norm": 0.11872179806232452,
+      "learning_rate": 4.619298999303926e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5562632,
+      "step": 7120
+    },
+    {
+      "epoch": 14.812889812889813,
+      "grad_norm": 0.011281021870672703,
+      "learning_rate": 4.618778073131673e-05,
+      "loss": 0.0182,
+      "num_input_tokens_seen": 5566472,
+      "step": 7125
+    },
+    {
+      "epoch": 14.823284823284823,
+      "grad_norm": 0.03502533957362175,
+      "learning_rate": 4.618256820217215e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5570376,
+      "step": 7130
+    },
+    {
+      "epoch": 14.833679833679835,
+      "grad_norm": 0.03718564659357071,
+      "learning_rate": 4.617735240640936e-05,
+      "loss": 0.0293,
+      "num_input_tokens_seen": 5574248,
+      "step": 7135
+    },
+    {
+      "epoch": 14.844074844074845,
+      "grad_norm": 0.03180214390158653,
+      "learning_rate": 4.6172133344832705e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 5578216,
+      "step": 7140
+    },
+    {
+      "epoch": 14.854469854469855,
+      "grad_norm": 0.04238855838775635,
+      "learning_rate": 4.6166911018247004e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 5582120,
+      "step": 7145
+    },
+    {
+      "epoch": 14.864864864864865,
+      "grad_norm": 0.08221118897199631,
+      "learning_rate": 4.616168542745764e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 5585864,
+      "step": 7150
+    },
+    {
+      "epoch": 14.875259875259875,
+      "grad_norm": 0.03431606665253639,
+      "learning_rate": 4.6156456573270446e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 5589864,
+      "step": 7155
+    },
+    {
+      "epoch": 14.885654885654885,
+      "grad_norm": 2.3205227851867676,
+      "learning_rate": 4.615122445649177e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 5593928,
+      "step": 7160
+    },
+    {
+      "epoch": 14.896049896049895,
+      "grad_norm": 30.62255096435547,
+      "learning_rate": 4.6145989077928486e-05,
+      "loss": 0.0322,
+      "num_input_tokens_seen": 5597864,
+      "step": 7165
+    },
+    {
+      "epoch": 14.906444906444907,
+      "grad_norm": 0.00943776685744524,
+      "learning_rate": 4.6140750438387953e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5601640,
+      "step": 7170
+    },
+    {
+      "epoch": 14.916839916839917,
+      "grad_norm": 0.002834032056853175,
+      "learning_rate": 4.613550853867803e-05,
+      "loss": 0.0093,
+      "num_input_tokens_seen": 5605480,
+      "step": 7175
+    },
+    {
+      "epoch": 14.927234927234927,
+      "grad_norm": 0.007750964257866144,
+      "learning_rate": 4.613026337960708e-05,
+      "loss": 0.0153,
+      "num_input_tokens_seen": 5609320,
+      "step": 7180
+    },
+    {
+      "epoch": 14.937629937629938,
+      "grad_norm": 2.5469861030578613,
+      "learning_rate": 4.612501496198398e-05,
+      "loss": 0.006,
+      "num_input_tokens_seen": 5613128,
+      "step": 7185
+    },
+    {
+      "epoch": 14.948024948024948,
+      "grad_norm": 0.061841659247875214,
+      "learning_rate": 4.61197632866181e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 5617032,
+      "step": 7190
+    },
+    {
+      "epoch": 14.958419958419958,
+      "grad_norm": 2.669663429260254,
+      "learning_rate": 4.611450835431931e-05,
+      "loss": 0.0286,
+      "num_input_tokens_seen": 5620904,
+      "step": 7195
+    },
+    {
+      "epoch": 14.96881496881497,
+      "grad_norm": 0.057192977517843246,
+      "learning_rate": 4.6109250165898e-05,
+      "loss": 0.0206,
+      "num_input_tokens_seen": 5624776,
+      "step": 7200
+    },
+    {
+      "epoch": 14.96881496881497,
+      "eval_loss": 0.3043736219406128,
+      "eval_runtime": 11.8011,
+      "eval_samples_per_second": 72.535,
+      "eval_steps_per_second": 18.134,
+      "num_input_tokens_seen": 5624776,
+      "step": 7200
+    },
+    {
+      "epoch": 14.97920997920998,
+      "grad_norm": 0.004903925582766533,
+      "learning_rate": 4.610398872216503e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 5628776,
+      "step": 7205
+    },
+    {
+      "epoch": 14.98960498960499,
+      "grad_norm": 0.007795864716172218,
+      "learning_rate": 4.6098724023931796e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 5632808,
+      "step": 7210
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 2.751840114593506,
+      "learning_rate": 4.609345607201017e-05,
+      "loss": 0.0298,
+      "num_input_tokens_seen": 5636752,
+      "step": 7215
+    },
+    {
+      "epoch": 15.01039501039501,
+      "grad_norm": 0.055486928671598434,
+      "learning_rate": 4.608818486721254e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 5640848,
+      "step": 7220
+    },
+    {
+      "epoch": 15.02079002079002,
+      "grad_norm": 0.012933751568198204,
+      "learning_rate": 4.608291041035179e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 5644848,
+      "step": 7225
+    },
+    {
+      "epoch": 15.03118503118503,
+      "grad_norm": 0.08094523102045059,
+      "learning_rate": 4.607763270224132e-05,
+      "loss": 0.0072,
+      "num_input_tokens_seen": 5648720,
+      "step": 7230
+    },
+    {
+      "epoch": 15.041580041580042,
+      "grad_norm": 0.0006411690264940262,
+      "learning_rate": 4.6072351743695e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 5652688,
+      "step": 7235
+    },
+    {
+      "epoch": 15.051975051975052,
+      "grad_norm": 0.04732662811875343,
+      "learning_rate": 4.606706753552723e-05,
+      "loss": 0.0179,
+      "num_input_tokens_seen": 5656592,
+      "step": 7240
+    },
+    {
+      "epoch": 15.062370062370062,
+      "grad_norm": 1.6166480779647827,
+      "learning_rate": 4.6061780078552906e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5660624,
+      "step": 7245
+    },
+    {
+      "epoch": 15.072765072765073,
+      "grad_norm": 0.01828949525952339,
+      "learning_rate": 4.605648937358742e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5664464,
+      "step": 7250
+    },
+    {
+      "epoch": 15.083160083160083,
+      "grad_norm": 0.005632069893181324,
+      "learning_rate": 4.605119542144665e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5668368,
+      "step": 7255
+    },
+    {
+      "epoch": 15.093555093555093,
+      "grad_norm": 0.010634763166308403,
+      "learning_rate": 4.604589822294701e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5672432,
+      "step": 7260
+    },
+    {
+      "epoch": 15.103950103950105,
+      "grad_norm": 0.19665808975696564,
+      "learning_rate": 4.604059777890537e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5676272,
+      "step": 7265
+    },
+    {
+      "epoch": 15.114345114345115,
+      "grad_norm": 0.06486106663942337,
+      "learning_rate": 4.6035294090139145e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5680208,
+      "step": 7270
+    },
+    {
+      "epoch": 15.124740124740125,
+      "grad_norm": 0.0008466948638670146,
+      "learning_rate": 4.6029987157466226e-05,
+      "loss": 0.0414,
+      "num_input_tokens_seen": 5684208,
+      "step": 7275
+    },
+    {
+      "epoch": 15.135135135135135,
+      "grad_norm": 0.0026520679239183664,
+      "learning_rate": 4.602467698170502e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5688144,
+      "step": 7280
+    },
+    {
+      "epoch": 15.145530145530145,
+      "grad_norm": 0.012995190918445587,
+      "learning_rate": 4.601936356367439e-05,
+      "loss": 0.0191,
+      "num_input_tokens_seen": 5692016,
+      "step": 7285
+    },
+    {
+      "epoch": 15.155925155925155,
+      "grad_norm": 0.015167951583862305,
+      "learning_rate": 4.601404690419377e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5695856,
+      "step": 7290
+    },
+    {
+      "epoch": 15.166320166320165,
+      "grad_norm": 0.6330662369728088,
+      "learning_rate": 4.600872700408303e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 5699728,
+      "step": 7295
+    },
+    {
+      "epoch": 15.176715176715177,
+      "grad_norm": 0.004072941839694977,
+      "learning_rate": 4.600340386416258e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 5703792,
+      "step": 7300
+    },
+    {
+      "epoch": 15.187110187110187,
+      "grad_norm": 0.003215422388166189,
+      "learning_rate": 4.5998077485253296e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5707568,
+      "step": 7305
+    },
+    {
+      "epoch": 15.197505197505198,
+      "grad_norm": 0.023494796827435493,
+      "learning_rate": 4.59927478681766e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 5711568,
+      "step": 7310
+    },
+    {
+      "epoch": 15.207900207900208,
+      "grad_norm": 0.03548942133784294,
+      "learning_rate": 4.5987415013754366e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5715376,
+      "step": 7315
+    },
+    {
+      "epoch": 15.218295218295218,
+      "grad_norm": 0.15131616592407227,
+      "learning_rate": 4.598207892280899e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5719280,
+      "step": 7320
+    },
+    {
+      "epoch": 15.228690228690228,
+      "grad_norm": 22.958171844482422,
+      "learning_rate": 4.597673959616337e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 5723216,
+      "step": 7325
+    },
+    {
+      "epoch": 15.23908523908524,
+      "grad_norm": 0.002984103513881564,
+      "learning_rate": 4.597139703464089e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5727184,
+      "step": 7330
+    },
+    {
+      "epoch": 15.24948024948025,
+      "grad_norm": 0.01022561639547348,
+      "learning_rate": 4.596605123906545e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5731024,
+      "step": 7335
+    },
+    {
+      "epoch": 15.25987525987526,
+      "grad_norm": 0.09409703314304352,
+      "learning_rate": 4.596070221026143e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5734896,
+      "step": 7340
+    },
+    {
+      "epoch": 15.27027027027027,
+      "grad_norm": 0.0005038998206146061,
+      "learning_rate": 4.595534994905372e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5738640,
+      "step": 7345
+    },
+    {
+      "epoch": 15.28066528066528,
+      "grad_norm": 6.245558261871338,
+      "learning_rate": 4.594999445626771e-05,
+      "loss": 0.0109,
+      "num_input_tokens_seen": 5742800,
+      "step": 7350
+    },
+    {
+      "epoch": 15.29106029106029,
+      "grad_norm": 0.0005823435494676232,
+      "learning_rate": 4.5944635732729276e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5746704,
+      "step": 7355
+    },
+    {
+      "epoch": 15.301455301455302,
+      "grad_norm": 0.0025945166125893593,
+      "learning_rate": 4.5939273779264804e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5750736,
+      "step": 7360
+    },
+    {
+      "epoch": 15.311850311850312,
+      "grad_norm": 0.005077396985143423,
+      "learning_rate": 4.593390859670118e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 5754704,
+      "step": 7365
+    },
+    {
+      "epoch": 15.322245322245323,
+      "grad_norm": 0.03294922411441803,
+      "learning_rate": 4.5928540185865776e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5758736,
+      "step": 7370
+    },
+    {
+      "epoch": 15.332640332640333,
+      "grad_norm": 0.07002729177474976,
+      "learning_rate": 4.592316854758648e-05,
+      "loss": 0.0494,
+      "num_input_tokens_seen": 5762672,
+      "step": 7375
+    },
+    {
+      "epoch": 15.343035343035343,
+      "grad_norm": 1.7875813245773315,
+      "learning_rate": 4.5917793682691646e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 5766640,
+      "step": 7380
+    },
+    {
+      "epoch": 15.353430353430353,
+      "grad_norm": 0.1553148478269577,
+      "learning_rate": 4.5912415592010164e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5770512,
+      "step": 7385
+    },
+    {
+      "epoch": 15.363825363825363,
+      "grad_norm": 0.02190096490085125,
+      "learning_rate": 4.5907034276371386e-05,
+      "loss": 0.0124,
+      "num_input_tokens_seen": 5774352,
+      "step": 7390
+    },
+    {
+      "epoch": 15.374220374220375,
+      "grad_norm": 0.12948505580425262,
+      "learning_rate": 4.5901649736605196e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 5778128,
+      "step": 7395
+    },
+    {
+      "epoch": 15.384615384615385,
+      "grad_norm": 0.21873773634433746,
+      "learning_rate": 4.589626197354195e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 5782032,
+      "step": 7400
+    },
+    {
+      "epoch": 15.384615384615385,
+      "eval_loss": 0.3595716655254364,
+      "eval_runtime": 11.7702,
+      "eval_samples_per_second": 72.726,
+      "eval_steps_per_second": 18.182,
+      "num_input_tokens_seen": 5782032,
+      "step": 7400
+    },
+    {
+      "epoch": 15.395010395010395,
+      "grad_norm": 0.021236218512058258,
+      "learning_rate": 4.5890870988012504e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 5785872,
+      "step": 7405
+    },
+    {
+      "epoch": 15.405405405405405,
+      "grad_norm": 0.0029206043109297752,
+      "learning_rate": 4.5885476780848226e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 5789744,
+      "step": 7410
+    },
+    {
+      "epoch": 15.415800415800415,
+      "grad_norm": 0.03966590762138367,
+      "learning_rate": 4.5880079352880964e-05,
+      "loss": 0.0099,
+      "num_input_tokens_seen": 5793680,
+      "step": 7415
+    },
+    {
+      "epoch": 15.426195426195425,
+      "grad_norm": 0.03188794106245041,
+      "learning_rate": 4.5874678704943065e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 5797616,
+      "step": 7420
+    },
+    {
+      "epoch": 15.436590436590437,
+      "grad_norm": 0.05273277312517166,
+      "learning_rate": 4.5869274837867394e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 5801456,
+      "step": 7425
+    },
+    {
+      "epoch": 15.446985446985448,
+      "grad_norm": 0.022159524261951447,
+      "learning_rate": 4.5863867752487275e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5805296,
+      "step": 7430
+    },
+    {
+      "epoch": 15.457380457380458,
+      "grad_norm": 0.007698631379753351,
+      "learning_rate": 4.5858457449636554e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5809296,
+      "step": 7435
+    },
+    {
+      "epoch": 15.467775467775468,
+      "grad_norm": 0.011302124708890915,
+      "learning_rate": 4.5853043930149574e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 5813168,
+      "step": 7440
+    },
+    {
+      "epoch": 15.478170478170478,
+      "grad_norm": 0.0012454447569325566,
+      "learning_rate": 4.584762719486117e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 5817168,
+      "step": 7445
+    },
+    {
+      "epoch": 15.488565488565488,
+      "grad_norm": 0.006228930316865444,
+      "learning_rate": 4.584220724460665e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5821072,
+      "step": 7450
+    },
+    {
+      "epoch": 15.4989604989605,
+      "grad_norm": 0.0024062900338321924,
+      "learning_rate": 4.5836784080221865e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5825072,
+      "step": 7455
+    },
+    {
+      "epoch": 15.50935550935551,
+      "grad_norm": 0.001959446584805846,
+      "learning_rate": 4.583135770254312e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 5829008,
+      "step": 7460
+    },
+    {
+      "epoch": 15.51975051975052,
+      "grad_norm": 0.0002505350857973099,
+      "learning_rate": 4.5825928112407236e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 5832848,
+      "step": 7465
+    },
+    {
+      "epoch": 15.53014553014553,
+      "grad_norm": 0.0002905271539930254,
+      "learning_rate": 4.582049531065152e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5836784,
+      "step": 7470
+    },
+    {
+      "epoch": 15.54054054054054,
+      "grad_norm": 0.0013178534572944045,
+      "learning_rate": 4.5815059298113783e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5840656,
+      "step": 7475
+    },
+    {
+      "epoch": 15.55093555093555,
+      "grad_norm": 0.00099087820854038,
+      "learning_rate": 4.580962007563232e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5844560,
+      "step": 7480
+    },
+    {
+      "epoch": 15.56133056133056,
+      "grad_norm": 0.00021940964506939054,
+      "learning_rate": 4.5804177644045935e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5848400,
+      "step": 7485
+    },
+    {
+      "epoch": 15.571725571725572,
+      "grad_norm": 0.0018596386071294546,
+      "learning_rate": 4.579873200419391e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5852304,
+      "step": 7490
+    },
+    {
+      "epoch": 15.582120582120583,
+      "grad_norm": 0.00022652081679552794,
+      "learning_rate": 4.5793283156916046e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5856208,
+      "step": 7495
+    },
+    {
+      "epoch": 15.592515592515593,
+      "grad_norm": 0.0022898537572473288,
+      "learning_rate": 4.578783110305261e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5860048,
+      "step": 7500
+    },
+    {
+      "epoch": 15.602910602910603,
+      "grad_norm": 8.80543957464397e-05,
+      "learning_rate": 4.578237584344438e-05,
+      "loss": 0.076,
+      "num_input_tokens_seen": 5863984,
+      "step": 7505
+    },
+    {
+      "epoch": 15.613305613305613,
+      "grad_norm": 0.00256293173879385,
+      "learning_rate": 4.577691737893263e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 5868016,
+      "step": 7510
+    },
+    {
+      "epoch": 15.623700623700623,
+      "grad_norm": 0.014912741258740425,
+      "learning_rate": 4.577145571035912e-05,
+      "loss": 0.0068,
+      "num_input_tokens_seen": 5871920,
+      "step": 7515
+    },
+    {
+      "epoch": 15.634095634095633,
+      "grad_norm": 0.10057365149259567,
+      "learning_rate": 4.576599083856611e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 5875664,
+      "step": 7520
+    },
+    {
+      "epoch": 15.644490644490645,
+      "grad_norm": 0.014837207272648811,
+      "learning_rate": 4.576052276439635e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5879600,
+      "step": 7525
+    },
+    {
+      "epoch": 15.654885654885655,
+      "grad_norm": 0.06529439240694046,
+      "learning_rate": 4.575505148869308e-05,
+      "loss": 0.0162,
+      "num_input_tokens_seen": 5883344,
+      "step": 7530
+    },
+    {
+      "epoch": 15.665280665280665,
+      "grad_norm": 0.012321981601417065,
+      "learning_rate": 4.574957701230006e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5887344,
+      "step": 7535
+    },
+    {
+      "epoch": 15.675675675675675,
+      "grad_norm": 0.004208738449960947,
+      "learning_rate": 4.57440993360615e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 5891312,
+      "step": 7540
+    },
+    {
+      "epoch": 15.686070686070686,
+      "grad_norm": 0.004973672796040773,
+      "learning_rate": 4.5738618460822134e-05,
+      "loss": 0.0238,
+      "num_input_tokens_seen": 5895216,
+      "step": 7545
+    },
+    {
+      "epoch": 15.696465696465696,
+      "grad_norm": 0.002069343812763691,
+      "learning_rate": 4.573313438742719e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5899216,
+      "step": 7550
+    },
+    {
+      "epoch": 15.706860706860708,
+      "grad_norm": 0.46913856267929077,
+      "learning_rate": 4.5727647116722374e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 5903152,
+      "step": 7555
+    },
+    {
+      "epoch": 15.717255717255718,
+      "grad_norm": 77.39089965820312,
+      "learning_rate": 4.5722156649553884e-05,
+      "loss": 0.0567,
+      "num_input_tokens_seen": 5907056,
+      "step": 7560
+    },
+    {
+      "epoch": 15.727650727650728,
+      "grad_norm": 0.26517659425735474,
+      "learning_rate": 4.571666298676843e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 5910832,
+      "step": 7565
+    },
+    {
+      "epoch": 15.738045738045738,
+      "grad_norm": 0.004737752489745617,
+      "learning_rate": 4.571116612921321e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 5914864,
+      "step": 7570
+    },
+    {
+      "epoch": 15.748440748440748,
+      "grad_norm": 2.010793447494507,
+      "learning_rate": 4.57056660777359e-05,
+      "loss": 0.0124,
+      "num_input_tokens_seen": 5918704,
+      "step": 7575
+    },
+    {
+      "epoch": 15.758835758835758,
+      "grad_norm": 2.8011481761932373,
+      "learning_rate": 4.5700162833184666e-05,
+      "loss": 0.0236,
+      "num_input_tokens_seen": 5922640,
+      "step": 7580
+    },
+    {
+      "epoch": 15.76923076923077,
+      "grad_norm": 0.0033244644291698933,
+      "learning_rate": 4.5694656396408195e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 5926416,
+      "step": 7585
+    },
+    {
+      "epoch": 15.77962577962578,
+      "grad_norm": 0.10715188086032867,
+      "learning_rate": 4.5689146768255646e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5930320,
+      "step": 7590
+    },
+    {
+      "epoch": 15.79002079002079,
+      "grad_norm": 0.0040803453885018826,
+      "learning_rate": 4.568363394957667e-05,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 5934160,
+      "step": 7595
+    },
+    {
+      "epoch": 15.8004158004158,
+      "grad_norm": 0.0015644116792827845,
+      "learning_rate": 4.567811794122141e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5938000,
+      "step": 7600
+    },
+    {
+      "epoch": 15.8004158004158,
+      "eval_loss": 0.4537171423435211,
+      "eval_runtime": 11.7201,
+      "eval_samples_per_second": 73.037,
+      "eval_steps_per_second": 18.259,
+      "num_input_tokens_seen": 5938000,
+      "step": 7600
+    },
+    {
+      "epoch": 15.81081081081081,
+      "grad_norm": 0.003845552448183298,
+      "learning_rate": 4.56725987440405e-05,
+      "loss": 0.0098,
+      "num_input_tokens_seen": 5941936,
+      "step": 7605
+    },
+    {
+      "epoch": 15.82120582120582,
+      "grad_norm": 0.0036989073269069195,
+      "learning_rate": 4.566707635888508e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5945808,
+      "step": 7610
+    },
+    {
+      "epoch": 15.83160083160083,
+      "grad_norm": 0.003745192429050803,
+      "learning_rate": 4.566155078660677e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 5949872,
+      "step": 7615
+    },
+    {
+      "epoch": 15.841995841995843,
+      "grad_norm": 0.036712221801280975,
+      "learning_rate": 4.565602202805768e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 5953776,
+      "step": 7620
+    },
+    {
+      "epoch": 15.852390852390853,
+      "grad_norm": 0.0012895691907033324,
+      "learning_rate": 4.56504900840904e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 5957616,
+      "step": 7625
+    },
+    {
+      "epoch": 15.862785862785863,
+      "grad_norm": 0.0023723712656646967,
+      "learning_rate": 4.564495495555805e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 5961488,
+      "step": 7630
+    },
+    {
+      "epoch": 15.873180873180873,
+      "grad_norm": 0.0027660259511321783,
+      "learning_rate": 4.5639416643314204e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 5965360,
+      "step": 7635
+    },
+    {
+      "epoch": 15.883575883575883,
+      "grad_norm": 0.0012920337030664086,
+      "learning_rate": 4.5633875148212946e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 5969328,
+      "step": 7640
+    },
+    {
+      "epoch": 15.893970893970893,
+      "grad_norm": 0.0018102313624694943,
+      "learning_rate": 4.562833047110883e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 5973168,
+      "step": 7645
+    },
+    {
+      "epoch": 15.904365904365905,
+      "grad_norm": 0.006578972563147545,
+      "learning_rate": 4.5622782612856923e-05,
+      "loss": 0.0327,
+      "num_input_tokens_seen": 5976912,
+      "step": 7650
+    },
+    {
+      "epoch": 15.914760914760915,
+      "grad_norm": 1.6755412817001343,
+      "learning_rate": 4.561723157431278e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 5980688,
+      "step": 7655
+    },
+    {
+      "epoch": 15.925155925155925,
+      "grad_norm": 0.09045970439910889,
+      "learning_rate": 4.5611677356332435e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 5984624,
+      "step": 7660
+    },
+    {
+      "epoch": 15.935550935550935,
+      "grad_norm": 0.0038761969190090895,
+      "learning_rate": 4.560611995977242e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 5988464,
+      "step": 7665
+    },
+    {
+      "epoch": 15.945945945945946,
+      "grad_norm": 0.005257665179669857,
+      "learning_rate": 4.560055938548975e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 5992176,
+      "step": 7670
+    },
+    {
+      "epoch": 15.956340956340956,
+      "grad_norm": 0.003505527041852474,
+      "learning_rate": 4.5594995634341944e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 5996080,
+      "step": 7675
+    },
+    {
+      "epoch": 15.966735966735968,
+      "grad_norm": 0.006819556467235088,
+      "learning_rate": 4.5589428707187e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6000176,
+      "step": 7680
+    },
+    {
+      "epoch": 15.977130977130978,
+      "grad_norm": 1.9728459119796753,
+      "learning_rate": 4.55838586048834e-05,
+      "loss": 0.0396,
+      "num_input_tokens_seen": 6004176,
+      "step": 7685
+    },
+    {
+      "epoch": 15.987525987525988,
+      "grad_norm": 0.013420788571238518,
+      "learning_rate": 4.557828532829013e-05,
+      "loss": 0.0237,
+      "num_input_tokens_seen": 6008080,
+      "step": 7690
+    },
+    {
+      "epoch": 15.997920997920998,
+      "grad_norm": 4.7596235275268555,
+      "learning_rate": 4.557270887826667e-05,
+      "loss": 0.0303,
+      "num_input_tokens_seen": 6012016,
+      "step": 7695
+    },
+    {
+      "epoch": 16.008316008316008,
+      "grad_norm": 0.04280000552535057,
+      "learning_rate": 4.556712925567296e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6016008,
+      "step": 7700
+    },
+    {
+      "epoch": 16.01871101871102,
+      "grad_norm": 1.853678822517395,
+      "learning_rate": 4.5561546461369454e-05,
+      "loss": 0.0143,
+      "num_input_tokens_seen": 6020136,
+      "step": 7705
+    },
+    {
+      "epoch": 16.02910602910603,
+      "grad_norm": 0.09575894474983215,
+      "learning_rate": 4.55559604962171e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 6024200,
+      "step": 7710
+    },
+    {
+      "epoch": 16.03950103950104,
+      "grad_norm": 0.012401796877384186,
+      "learning_rate": 4.55503713610773e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6028040,
+      "step": 7715
+    },
+    {
+      "epoch": 16.04989604989605,
+      "grad_norm": 0.005576007533818483,
+      "learning_rate": 4.5544779056812e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6031976,
+      "step": 7720
+    },
+    {
+      "epoch": 16.06029106029106,
+      "grad_norm": 1.0480237007141113,
+      "learning_rate": 4.553918358428358e-05,
+      "loss": 0.0099,
+      "num_input_tokens_seen": 6035816,
+      "step": 7725
+    },
+    {
+      "epoch": 16.070686070686072,
+      "grad_norm": 0.020325014367699623,
+      "learning_rate": 4.553358494435494e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 6039752,
+      "step": 7730
+    },
+    {
+      "epoch": 16.08108108108108,
+      "grad_norm": 0.007197919301688671,
+      "learning_rate": 4.5527983137889464e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6043752,
+      "step": 7735
+    },
+    {
+      "epoch": 16.091476091476093,
+      "grad_norm": 0.002279489068314433,
+      "learning_rate": 4.5522378165751015e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 6047528,
+      "step": 7740
+    },
+    {
+      "epoch": 16.1018711018711,
+      "grad_norm": 0.0015998255694285035,
+      "learning_rate": 4.5516770028803954e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6051368,
+      "step": 7745
+    },
+    {
+      "epoch": 16.112266112266113,
+      "grad_norm": 0.1031498983502388,
+      "learning_rate": 4.5511158727913116e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6055144,
+      "step": 7750
+    },
+    {
+      "epoch": 16.12266112266112,
+      "grad_norm": 0.0017241965979337692,
+      "learning_rate": 4.5505544263943856e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6059048,
+      "step": 7755
+    },
+    {
+      "epoch": 16.133056133056133,
+      "grad_norm": 0.0004968007560819387,
+      "learning_rate": 4.549992663776197e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6062952,
+      "step": 7760
+    },
+    {
+      "epoch": 16.143451143451145,
+      "grad_norm": 0.0028085263911634684,
+      "learning_rate": 4.5494305850233786e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6066888,
+      "step": 7765
+    },
+    {
+      "epoch": 16.153846153846153,
+      "grad_norm": 0.0005490938783623278,
+      "learning_rate": 4.5488681902226094e-05,
+      "loss": 0.017,
+      "num_input_tokens_seen": 6070856,
+      "step": 7770
+    },
+    {
+      "epoch": 16.164241164241165,
+      "grad_norm": 45.05674743652344,
+      "learning_rate": 4.5483054794606174e-05,
+      "loss": 0.0095,
+      "num_input_tokens_seen": 6074728,
+      "step": 7775
+    },
+    {
+      "epoch": 16.174636174636174,
+      "grad_norm": 1.2453941106796265,
+      "learning_rate": 4.547742452824179e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 6078696,
+      "step": 7780
+    },
+    {
+      "epoch": 16.185031185031185,
+      "grad_norm": 0.0003694011829793453,
+      "learning_rate": 4.5471791104001215e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6082760,
+      "step": 7785
+    },
+    {
+      "epoch": 16.195426195426194,
+      "grad_norm": 0.0023320212494581938,
+      "learning_rate": 4.546615452275319e-05,
+      "loss": 0.0203,
+      "num_input_tokens_seen": 6086600,
+      "step": 7790
+    },
+    {
+      "epoch": 16.205821205821206,
+      "grad_norm": 0.024880988523364067,
+      "learning_rate": 4.5460514785366944e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 6090472,
+      "step": 7795
+    },
+    {
+      "epoch": 16.216216216216218,
+      "grad_norm": 0.9836779236793518,
+      "learning_rate": 4.545487189271219e-05,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 6094536,
+      "step": 7800
+    },
+    {
+      "epoch": 16.216216216216218,
+      "eval_loss": 0.40850090980529785,
+      "eval_runtime": 11.858,
+      "eval_samples_per_second": 72.188,
+      "eval_steps_per_second": 18.047,
+      "num_input_tokens_seen": 6094536,
+      "step": 7800
+    },
+    {
+      "epoch": 16.226611226611226,
+      "grad_norm": 0.20902898907661438,
+      "learning_rate": 4.544922584565914e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 6098504,
+      "step": 7805
+    },
+    {
+      "epoch": 16.237006237006238,
+      "grad_norm": 6.749849319458008,
+      "learning_rate": 4.544357664507848e-05,
+      "loss": 0.0505,
+      "num_input_tokens_seen": 6102344,
+      "step": 7810
+    },
+    {
+      "epoch": 16.247401247401246,
+      "grad_norm": 0.06901159137487411,
+      "learning_rate": 4.54379242918414e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 6106248,
+      "step": 7815
+    },
+    {
+      "epoch": 16.257796257796258,
+      "grad_norm": 0.05513492971658707,
+      "learning_rate": 4.543226878681955e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 6110120,
+      "step": 7820
+    },
+    {
+      "epoch": 16.26819126819127,
+      "grad_norm": 19.927392959594727,
+      "learning_rate": 4.5426610130885087e-05,
+      "loss": 0.0045,
+      "num_input_tokens_seen": 6114056,
+      "step": 7825
+    },
+    {
+      "epoch": 16.27858627858628,
+      "grad_norm": 0.0011525977170094848,
+      "learning_rate": 4.542094832491064e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 6118120,
+      "step": 7830
+    },
+    {
+      "epoch": 16.28898128898129,
+      "grad_norm": 0.4577887952327728,
+      "learning_rate": 4.541528336976934e-05,
+      "loss": 0.0106,
+      "num_input_tokens_seen": 6122056,
+      "step": 7835
+    },
+    {
+      "epoch": 16.2993762993763,
+      "grad_norm": 0.012642495334148407,
+      "learning_rate": 4.540961526633479e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6126024,
+      "step": 7840
+    },
+    {
+      "epoch": 16.30977130977131,
+      "grad_norm": 0.0011409998405724764,
+      "learning_rate": 4.540394401548108e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6129896,
+      "step": 7845
+    },
+    {
+      "epoch": 16.32016632016632,
+      "grad_norm": 0.022218482568860054,
+      "learning_rate": 4.539826961808279e-05,
+      "loss": 0.0079,
+      "num_input_tokens_seen": 6133928,
+      "step": 7850
+    },
+    {
+      "epoch": 16.33056133056133,
+      "grad_norm": 0.006972114089876413,
+      "learning_rate": 4.5392592075014994e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6137736,
+      "step": 7855
+    },
+    {
+      "epoch": 16.340956340956343,
+      "grad_norm": 0.4928295314311981,
+      "learning_rate": 4.538691138715322e-05,
+      "loss": 0.0151,
+      "num_input_tokens_seen": 6141640,
+      "step": 7860
+    },
+    {
+      "epoch": 16.35135135135135,
+      "grad_norm": 0.018248209729790688,
+      "learning_rate": 4.5381227555373516e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 6145640,
+      "step": 7865
+    },
+    {
+      "epoch": 16.361746361746363,
+      "grad_norm": 0.003096154658123851,
+      "learning_rate": 4.537554058055239e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6149480,
+      "step": 7870
+    },
+    {
+      "epoch": 16.37214137214137,
+      "grad_norm": 10.395303726196289,
+      "learning_rate": 4.5369850463566865e-05,
+      "loss": 0.0181,
+      "num_input_tokens_seen": 6153384,
+      "step": 7875
+    },
+    {
+      "epoch": 16.382536382536383,
+      "grad_norm": 0.0029550495091825724,
+      "learning_rate": 4.5364157205294404e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6157192,
+      "step": 7880
+    },
+    {
+      "epoch": 16.39293139293139,
+      "grad_norm": 0.002382299629971385,
+      "learning_rate": 4.5358460806612996e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6161096,
+      "step": 7885
+    },
+    {
+      "epoch": 16.403326403326403,
+      "grad_norm": 0.0014511863701045513,
+      "learning_rate": 4.535276126840109e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 6164936,
+      "step": 7890
+    },
+    {
+      "epoch": 16.413721413721415,
+      "grad_norm": 0.01925472542643547,
+      "learning_rate": 4.5347058591537626e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6168872,
+      "step": 7895
+    },
+    {
+      "epoch": 16.424116424116423,
+      "grad_norm": 0.0026381060015410185,
+      "learning_rate": 4.534135277690203e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6172744,
+      "step": 7900
+    },
+    {
+      "epoch": 16.434511434511435,
+      "grad_norm": 0.0007052918081171811,
+      "learning_rate": 4.533564382537421e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6176872,
+      "step": 7905
+    },
+    {
+      "epoch": 16.444906444906444,
+      "grad_norm": 0.004420078359544277,
+      "learning_rate": 4.532993173783456e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6180776,
+      "step": 7910
+    },
+    {
+      "epoch": 16.455301455301456,
+      "grad_norm": 0.22011351585388184,
+      "learning_rate": 4.5324216515163954e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 6184776,
+      "step": 7915
+    },
+    {
+      "epoch": 16.465696465696467,
+      "grad_norm": 0.043546367436647415,
+      "learning_rate": 4.531849815824375e-05,
+      "loss": 0.051,
+      "num_input_tokens_seen": 6188744,
+      "step": 7920
+    },
+    {
+      "epoch": 16.476091476091476,
+      "grad_norm": 0.03377815708518028,
+      "learning_rate": 4.5312776667955795e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6192584,
+      "step": 7925
+    },
+    {
+      "epoch": 16.486486486486488,
+      "grad_norm": 0.08407463878393173,
+      "learning_rate": 4.5307052045182405e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 6196456,
+      "step": 7930
+    },
+    {
+      "epoch": 16.496881496881496,
+      "grad_norm": 3.020683765411377,
+      "learning_rate": 4.53013242908064e-05,
+      "loss": 0.0378,
+      "num_input_tokens_seen": 6200488,
+      "step": 7935
+    },
+    {
+      "epoch": 16.507276507276508,
+      "grad_norm": 1.19234037399292,
+      "learning_rate": 4.529559340571107e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 6204360,
+      "step": 7940
+    },
+    {
+      "epoch": 16.517671517671516,
+      "grad_norm": 0.0012648142874240875,
+      "learning_rate": 4.528985939078018e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6208232,
+      "step": 7945
+    },
+    {
+      "epoch": 16.528066528066528,
+      "grad_norm": 0.007225461769849062,
+      "learning_rate": 4.5284122246898e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6212008,
+      "step": 7950
+    },
+    {
+      "epoch": 16.53846153846154,
+      "grad_norm": 0.00797426700592041,
+      "learning_rate": 4.527838197494926e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6215816,
+      "step": 7955
+    },
+    {
+      "epoch": 16.54885654885655,
+      "grad_norm": 0.04489576444029808,
+      "learning_rate": 4.527263857581918e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6219720,
+      "step": 7960
+    },
+    {
+      "epoch": 16.55925155925156,
+      "grad_norm": 0.03976420313119888,
+      "learning_rate": 4.526689205039347e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 6223528,
+      "step": 7965
+    },
+    {
+      "epoch": 16.56964656964657,
+      "grad_norm": 0.0007705093012191355,
+      "learning_rate": 4.5261142399558324e-05,
+      "loss": 0.0557,
+      "num_input_tokens_seen": 6227400,
+      "step": 7970
+    },
+    {
+      "epoch": 16.58004158004158,
+      "grad_norm": 0.01329643651843071,
+      "learning_rate": 4.525538962420041e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6231304,
+      "step": 7975
+    },
+    {
+      "epoch": 16.59043659043659,
+      "grad_norm": 0.004066470079123974,
+      "learning_rate": 4.524963372520685e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6235240,
+      "step": 7980
+    },
+    {
+      "epoch": 16.6008316008316,
+      "grad_norm": 0.0031231907196342945,
+      "learning_rate": 4.524387470346531e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 6239080,
+      "step": 7985
+    },
+    {
+      "epoch": 16.611226611226613,
+      "grad_norm": 0.003938264213502407,
+      "learning_rate": 4.5238112559863885e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6242984,
+      "step": 7990
+    },
+    {
+      "epoch": 16.62162162162162,
+      "grad_norm": 0.00072537298547104,
+      "learning_rate": 4.5232347295291175e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 6246824,
+      "step": 7995
+    },
+    {
+      "epoch": 16.632016632016633,
+      "grad_norm": 0.007357693277299404,
+      "learning_rate": 4.522657891063626e-05,
+      "loss": 0.0204,
+      "num_input_tokens_seen": 6250760,
+      "step": 8000
+    },
+    {
+      "epoch": 16.632016632016633,
+      "eval_loss": 0.36195802688598633,
+      "eval_runtime": 11.7312,
+      "eval_samples_per_second": 72.968,
+      "eval_steps_per_second": 18.242,
+      "num_input_tokens_seen": 6250760,
+      "step": 8000
+    },
+    {
+      "epoch": 16.64241164241164,
+      "grad_norm": 0.010383481159806252,
+      "learning_rate": 4.52208074067887e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 6254760,
+      "step": 8005
+    },
+    {
+      "epoch": 16.652806652806653,
+      "grad_norm": 0.028110913932323456,
+      "learning_rate": 4.5215032784638516e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6258728,
+      "step": 8010
+    },
+    {
+      "epoch": 16.66320166320166,
+      "grad_norm": 0.14461137354373932,
+      "learning_rate": 4.5209255045076245e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 6262600,
+      "step": 8015
+    },
+    {
+      "epoch": 16.673596673596673,
+      "grad_norm": 0.002005411311984062,
+      "learning_rate": 4.5203474188992875e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6266440,
+      "step": 8020
+    },
+    {
+      "epoch": 16.683991683991685,
+      "grad_norm": 0.014980043284595013,
+      "learning_rate": 4.51976902172799e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6270280,
+      "step": 8025
+    },
+    {
+      "epoch": 16.694386694386694,
+      "grad_norm": 10.187496185302734,
+      "learning_rate": 4.519190313082927e-05,
+      "loss": 0.0117,
+      "num_input_tokens_seen": 6274088,
+      "step": 8030
+    },
+    {
+      "epoch": 16.704781704781706,
+      "grad_norm": 0.008852017112076283,
+      "learning_rate": 4.518611293053343e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6277928,
+      "step": 8035
+    },
+    {
+      "epoch": 16.715176715176714,
+      "grad_norm": 0.001192402676679194,
+      "learning_rate": 4.51803196172853e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 6281896,
+      "step": 8040
+    },
+    {
+      "epoch": 16.725571725571726,
+      "grad_norm": 0.00029703218024224043,
+      "learning_rate": 4.517452319197828e-05,
+      "loss": 0.0107,
+      "num_input_tokens_seen": 6285832,
+      "step": 8045
+    },
+    {
+      "epoch": 16.735966735966738,
+      "grad_norm": 0.4248342514038086,
+      "learning_rate": 4.5168723655506265e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 6289832,
+      "step": 8050
+    },
+    {
+      "epoch": 16.746361746361746,
+      "grad_norm": 0.0021027063485234976,
+      "learning_rate": 4.51629210087636e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6293640,
+      "step": 8055
+    },
+    {
+      "epoch": 16.756756756756758,
+      "grad_norm": 0.024694669991731644,
+      "learning_rate": 4.515711525264513e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 6297544,
+      "step": 8060
+    },
+    {
+      "epoch": 16.767151767151766,
+      "grad_norm": 0.0002857391955330968,
+      "learning_rate": 4.5151306388046175e-05,
+      "loss": 0.0231,
+      "num_input_tokens_seen": 6301480,
+      "step": 8065
+    },
+    {
+      "epoch": 16.777546777546778,
+      "grad_norm": 0.0011725698132067919,
+      "learning_rate": 4.514549441586255e-05,
+      "loss": 0.0129,
+      "num_input_tokens_seen": 6305128,
+      "step": 8070
+    },
+    {
+      "epoch": 16.787941787941786,
+      "grad_norm": 0.0054862042888998985,
+      "learning_rate": 4.513967933699051e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 6309096,
+      "step": 8075
+    },
+    {
+      "epoch": 16.7983367983368,
+      "grad_norm": 0.03842340037226677,
+      "learning_rate": 4.513386115232684e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 6313064,
+      "step": 8080
+    },
+    {
+      "epoch": 16.80873180873181,
+      "grad_norm": 0.14747729897499084,
+      "learning_rate": 4.5128039862768745e-05,
+      "loss": 0.0091,
+      "num_input_tokens_seen": 6317032,
+      "step": 8085
+    },
+    {
+      "epoch": 16.81912681912682,
+      "grad_norm": 0.9815083146095276,
+      "learning_rate": 4.512221546921397e-05,
+      "loss": 0.0211,
+      "num_input_tokens_seen": 6320904,
+      "step": 8090
+    },
+    {
+      "epoch": 16.82952182952183,
+      "grad_norm": 0.0013478652108460665,
+      "learning_rate": 4.5116387972560694e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6324872,
+      "step": 8095
+    },
+    {
+      "epoch": 16.83991683991684,
+      "grad_norm": 0.27748551964759827,
+      "learning_rate": 4.511055737370759e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6328776,
+      "step": 8100
+    },
+    {
+      "epoch": 16.85031185031185,
+      "grad_norm": 0.0018963367911055684,
+      "learning_rate": 4.510472367355383e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6332520,
+      "step": 8105
+    },
+    {
+      "epoch": 16.86070686070686,
+      "grad_norm": 0.7088038325309753,
+      "learning_rate": 4.509888687299901e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6336552,
+      "step": 8110
+    },
+    {
+      "epoch": 16.87110187110187,
+      "grad_norm": 0.00019241106929257512,
+      "learning_rate": 4.5093046972943266e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 6340424,
+      "step": 8115
+    },
+    {
+      "epoch": 16.881496881496883,
+      "grad_norm": 4.72757625579834,
+      "learning_rate": 4.508720397428717e-05,
+      "loss": 0.077,
+      "num_input_tokens_seen": 6344232,
+      "step": 8120
+    },
+    {
+      "epoch": 16.89189189189189,
+      "grad_norm": 0.0004391357651911676,
+      "learning_rate": 4.508135787793178e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6348040,
+      "step": 8125
+    },
+    {
+      "epoch": 16.902286902286903,
+      "grad_norm": 0.0016441376646980643,
+      "learning_rate": 4.5075508684778664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6352136,
+      "step": 8130
+    },
+    {
+      "epoch": 16.91268191268191,
+      "grad_norm": 0.0003971205442212522,
+      "learning_rate": 4.506965639572982e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6356072,
+      "step": 8135
+    },
+    {
+      "epoch": 16.923076923076923,
+      "grad_norm": 0.0026301753241568804,
+      "learning_rate": 4.506380101168774e-05,
+      "loss": 0.0168,
+      "num_input_tokens_seen": 6359944,
+      "step": 8140
+    },
+    {
+      "epoch": 16.933471933471935,
+      "grad_norm": 23.869932174682617,
+      "learning_rate": 4.505794253355542e-05,
+      "loss": 0.0289,
+      "num_input_tokens_seen": 6363880,
+      "step": 8145
+    },
+    {
+      "epoch": 16.943866943866944,
+      "grad_norm": 38.64217758178711,
+      "learning_rate": 4.5052080962236286e-05,
+      "loss": 0.0112,
+      "num_input_tokens_seen": 6367816,
+      "step": 8150
+    },
+    {
+      "epoch": 16.954261954261955,
+      "grad_norm": 1.2362966537475586,
+      "learning_rate": 4.504621629863428e-05,
+      "loss": 0.0328,
+      "num_input_tokens_seen": 6371624,
+      "step": 8155
+    },
+    {
+      "epoch": 16.964656964656964,
+      "grad_norm": 0.006759606767445803,
+      "learning_rate": 4.504034854365381e-05,
+      "loss": 0.0076,
+      "num_input_tokens_seen": 6375400,
+      "step": 8160
+    },
+    {
+      "epoch": 16.975051975051976,
+      "grad_norm": 0.01113686989992857,
+      "learning_rate": 4.503447769819974e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6379336,
+      "step": 8165
+    },
+    {
+      "epoch": 16.985446985446984,
+      "grad_norm": 0.004490755498409271,
+      "learning_rate": 4.502860376317745e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6383272,
+      "step": 8170
+    },
+    {
+      "epoch": 16.995841995841996,
+      "grad_norm": 0.09532777220010757,
+      "learning_rate": 4.502272673949276e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6387144,
+      "step": 8175
+    },
+    {
+      "epoch": 17.006237006237008,
+      "grad_norm": 0.002229373436421156,
+      "learning_rate": 4.501684662805199e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 6390936,
+      "step": 8180
+    },
+    {
+      "epoch": 17.016632016632016,
+      "grad_norm": 0.09374048560857773,
+      "learning_rate": 4.5010963429761924e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6394936,
+      "step": 8185
+    },
+    {
+      "epoch": 17.027027027027028,
+      "grad_norm": 0.0036959426943212748,
+      "learning_rate": 4.500507714552982e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 6398776,
+      "step": 8190
+    },
+    {
+      "epoch": 17.037422037422036,
+      "grad_norm": 0.005745233967900276,
+      "learning_rate": 4.499918777626342e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6402744,
+      "step": 8195
+    },
+    {
+      "epoch": 17.04781704781705,
+      "grad_norm": 0.002384748077020049,
+      "learning_rate": 4.499329532287093e-05,
+      "loss": 0.0135,
+      "num_input_tokens_seen": 6406616,
+      "step": 8200
+    },
+    {
+      "epoch": 17.04781704781705,
+      "eval_loss": 0.5138377547264099,
+      "eval_runtime": 11.7186,
+      "eval_samples_per_second": 73.046,
+      "eval_steps_per_second": 18.262,
+      "num_input_tokens_seen": 6406616,
+      "step": 8200
+    },
+    {
+      "epoch": 17.058212058212057,
+      "grad_norm": 0.003423857968300581,
+      "learning_rate": 4.4987399786261064e-05,
+      "loss": 0.0174,
+      "num_input_tokens_seen": 6410584,
+      "step": 8205
+    },
+    {
+      "epoch": 17.06860706860707,
+      "grad_norm": 0.0015596495941281319,
+      "learning_rate": 4.498150116734297e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 6414552,
+      "step": 8210
+    },
+    {
+      "epoch": 17.07900207900208,
+      "grad_norm": 0.006223027594387531,
+      "learning_rate": 4.4975599467026294e-05,
+      "loss": 0.0268,
+      "num_input_tokens_seen": 6418520,
+      "step": 8215
+    },
+    {
+      "epoch": 17.08939708939709,
+      "grad_norm": 0.0017311853589490056,
+      "learning_rate": 4.496969468622114e-05,
+      "loss": 0.0055,
+      "num_input_tokens_seen": 6422520,
+      "step": 8220
+    },
+    {
+      "epoch": 17.0997920997921,
+      "grad_norm": 0.5798501372337341,
+      "learning_rate": 4.496378682583813e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 6426456,
+      "step": 8225
+    },
+    {
+      "epoch": 17.11018711018711,
+      "grad_norm": 0.018150396645069122,
+      "learning_rate": 4.495787588678829e-05,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 6430264,
+      "step": 8230
+    },
+    {
+      "epoch": 17.12058212058212,
+      "grad_norm": 0.014628610573709011,
+      "learning_rate": 4.4951961869983196e-05,
+      "loss": 0.0109,
+      "num_input_tokens_seen": 6434264,
+      "step": 8235
+    },
+    {
+      "epoch": 17.13097713097713,
+      "grad_norm": 0.016852010041475296,
+      "learning_rate": 4.494604477633485e-05,
+      "loss": 0.0097,
+      "num_input_tokens_seen": 6438232,
+      "step": 8240
+    },
+    {
+      "epoch": 17.14137214137214,
+      "grad_norm": 0.14317458868026733,
+      "learning_rate": 4.4940124606755734e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 6442232,
+      "step": 8245
+    },
+    {
+      "epoch": 17.151767151767153,
+      "grad_norm": 0.005841641221195459,
+      "learning_rate": 4.493420136215882e-05,
+      "loss": 0.0124,
+      "num_input_tokens_seen": 6446168,
+      "step": 8250
+    },
+    {
+      "epoch": 17.16216216216216,
+      "grad_norm": 0.03964950516819954,
+      "learning_rate": 4.492827504345756e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 6450008,
+      "step": 8255
+    },
+    {
+      "epoch": 17.172557172557173,
+      "grad_norm": 0.033274222165346146,
+      "learning_rate": 4.492234565156584e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6454008,
+      "step": 8260
+    },
+    {
+      "epoch": 17.18295218295218,
+      "grad_norm": 0.0023523292038589716,
+      "learning_rate": 4.491641318739807e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 6458008,
+      "step": 8265
+    },
+    {
+      "epoch": 17.193347193347194,
+      "grad_norm": 0.015039775520563126,
+      "learning_rate": 4.4910477651869096e-05,
+      "loss": 0.0119,
+      "num_input_tokens_seen": 6461944,
+      "step": 8270
+    },
+    {
+      "epoch": 17.203742203742205,
+      "grad_norm": 0.02762039378285408,
+      "learning_rate": 4.4904539045894254e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 6465816,
+      "step": 8275
+    },
+    {
+      "epoch": 17.214137214137214,
+      "grad_norm": 0.006152668967843056,
+      "learning_rate": 4.4898597370389364e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6469816,
+      "step": 8280
+    },
+    {
+      "epoch": 17.224532224532226,
+      "grad_norm": 0.0014009707374498248,
+      "learning_rate": 4.489265262627069e-05,
+      "loss": 0.027,
+      "num_input_tokens_seen": 6473496,
+      "step": 8285
+    },
+    {
+      "epoch": 17.234927234927234,
+      "grad_norm": 0.004118416458368301,
+      "learning_rate": 4.488670481445499e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6477496,
+      "step": 8290
+    },
+    {
+      "epoch": 17.245322245322246,
+      "grad_norm": 0.02178024686872959,
+      "learning_rate": 4.488075393585951e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6481528,
+      "step": 8295
+    },
+    {
+      "epoch": 17.255717255717254,
+      "grad_norm": 0.04819168522953987,
+      "learning_rate": 4.487479999140193e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6485336,
+      "step": 8300
+    },
+    {
+      "epoch": 17.266112266112266,
+      "grad_norm": 0.8021442890167236,
+      "learning_rate": 4.4868842982000425e-05,
+      "loss": 0.0205,
+      "num_input_tokens_seen": 6489112,
+      "step": 8305
+    },
+    {
+      "epoch": 17.276507276507278,
+      "grad_norm": 0.04864349961280823,
+      "learning_rate": 4.486288290857365e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 6492952,
+      "step": 8310
+    },
+    {
+      "epoch": 17.286902286902286,
+      "grad_norm": 0.011212851852178574,
+      "learning_rate": 4.4856919772040715e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 6496952,
+      "step": 8315
+    },
+    {
+      "epoch": 17.2972972972973,
+      "grad_norm": 0.004983293358236551,
+      "learning_rate": 4.485095357332122e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6500984,
+      "step": 8320
+    },
+    {
+      "epoch": 17.307692307692307,
+      "grad_norm": 0.02356879413127899,
+      "learning_rate": 4.484498431333521e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6504952,
+      "step": 8325
+    },
+    {
+      "epoch": 17.31808731808732,
+      "grad_norm": 0.0122224111109972,
+      "learning_rate": 4.4839011993003245e-05,
+      "loss": 0.0179,
+      "num_input_tokens_seen": 6508792,
+      "step": 8330
+    },
+    {
+      "epoch": 17.328482328482327,
+      "grad_norm": 0.0200260728597641,
+      "learning_rate": 4.4833036613246305e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6512664,
+      "step": 8335
+    },
+    {
+      "epoch": 17.33887733887734,
+      "grad_norm": 0.11308394372463226,
+      "learning_rate": 4.482705817498589e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 6516504,
+      "step": 8340
+    },
+    {
+      "epoch": 17.34927234927235,
+      "grad_norm": 0.09201828390359879,
+      "learning_rate": 4.4821076679143934e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 6520504,
+      "step": 8345
+    },
+    {
+      "epoch": 17.35966735966736,
+      "grad_norm": 2.8752779960632324,
+      "learning_rate": 4.481509212664288e-05,
+      "loss": 0.0464,
+      "num_input_tokens_seen": 6524344,
+      "step": 8350
+    },
+    {
+      "epoch": 17.37006237006237,
+      "grad_norm": 0.03789002448320389,
+      "learning_rate": 4.480910451840559e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6528344,
+      "step": 8355
+    },
+    {
+      "epoch": 17.38045738045738,
+      "grad_norm": 0.01753596030175686,
+      "learning_rate": 4.480311385535546e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6532184,
+      "step": 8360
+    },
+    {
+      "epoch": 17.39085239085239,
+      "grad_norm": 0.08195172250270844,
+      "learning_rate": 4.47971201384163e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 6536088,
+      "step": 8365
+    },
+    {
+      "epoch": 17.401247401247403,
+      "grad_norm": 0.007267999928444624,
+      "learning_rate": 4.4791123368512446e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6539992,
+      "step": 8370
+    },
+    {
+      "epoch": 17.41164241164241,
+      "grad_norm": 0.015784507617354393,
+      "learning_rate": 4.478512354656864e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6543992,
+      "step": 8375
+    },
+    {
+      "epoch": 17.422037422037423,
+      "grad_norm": 0.0009457683190703392,
+      "learning_rate": 4.477912067351016e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6547992,
+      "step": 8380
+    },
+    {
+      "epoch": 17.43243243243243,
+      "grad_norm": 0.0023589646443724632,
+      "learning_rate": 4.477311475026271e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6551768,
+      "step": 8385
+    },
+    {
+      "epoch": 17.442827442827443,
+      "grad_norm": 0.0008453708724118769,
+      "learning_rate": 4.476710577775248e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6555704,
+      "step": 8390
+    },
+    {
+      "epoch": 17.453222453222452,
+      "grad_norm": 0.00922363344579935,
+      "learning_rate": 4.476109375690612e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 6559544,
+      "step": 8395
+    },
+    {
+      "epoch": 17.463617463617464,
+      "grad_norm": 0.005655642598867416,
+      "learning_rate": 4.4755078688650784e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 6563416,
+      "step": 8400
+    },
+    {
+      "epoch": 17.463617463617464,
+      "eval_loss": 0.4106968641281128,
+      "eval_runtime": 11.76,
+      "eval_samples_per_second": 72.789,
+      "eval_steps_per_second": 18.197,
+      "num_input_tokens_seen": 6563416,
+      "step": 8400
+    },
+    {
+      "epoch": 17.474012474012476,
+      "grad_norm": 0.0008729042019695044,
+      "learning_rate": 4.474906057391406e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6567416,
+      "step": 8405
+    },
+    {
+      "epoch": 17.484407484407484,
+      "grad_norm": 0.0012758030788972974,
+      "learning_rate": 4.4743039413624e-05,
+      "loss": 0.0233,
+      "num_input_tokens_seen": 6571224,
+      "step": 8410
+    },
+    {
+      "epoch": 17.494802494802496,
+      "grad_norm": 0.02750765159726143,
+      "learning_rate": 4.473701520870916e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6575032,
+      "step": 8415
+    },
+    {
+      "epoch": 17.505197505197504,
+      "grad_norm": 0.022524213418364525,
+      "learning_rate": 4.4730987960098544e-05,
+      "loss": 0.0233,
+      "num_input_tokens_seen": 6579096,
+      "step": 8420
+    },
+    {
+      "epoch": 17.515592515592516,
+      "grad_norm": 0.0074014621786773205,
+      "learning_rate": 4.4724957668721635e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6583000,
+      "step": 8425
+    },
+    {
+      "epoch": 17.525987525987524,
+      "grad_norm": 0.019565295428037643,
+      "learning_rate": 4.471892433550836e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6586904,
+      "step": 8430
+    },
+    {
+      "epoch": 17.536382536382536,
+      "grad_norm": 0.5956645607948303,
+      "learning_rate": 4.471288796138916e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 6590840,
+      "step": 8435
+    },
+    {
+      "epoch": 17.546777546777548,
+      "grad_norm": 0.0026397742331027985,
+      "learning_rate": 4.470684854729491e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6594616,
+      "step": 8440
+    },
+    {
+      "epoch": 17.557172557172557,
+      "grad_norm": 0.5044841766357422,
+      "learning_rate": 4.4700806094156955e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 6598520,
+      "step": 8445
+    },
+    {
+      "epoch": 17.56756756756757,
+      "grad_norm": 0.000441287673311308,
+      "learning_rate": 4.469476060290713e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6602520,
+      "step": 8450
+    },
+    {
+      "epoch": 17.577962577962577,
+      "grad_norm": 0.0018722789827734232,
+      "learning_rate": 4.468871207447772e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6606328,
+      "step": 8455
+    },
+    {
+      "epoch": 17.58835758835759,
+      "grad_norm": 0.0006285469280555844,
+      "learning_rate": 4.4682660509801486e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6610136,
+      "step": 8460
+    },
+    {
+      "epoch": 17.598752598752597,
+      "grad_norm": 0.005743082612752914,
+      "learning_rate": 4.467660590981165e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6614008,
+      "step": 8465
+    },
+    {
+      "epoch": 17.60914760914761,
+      "grad_norm": 0.001562782097607851,
+      "learning_rate": 4.467054827544191e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6617816,
+      "step": 8470
+    },
+    {
+      "epoch": 17.61954261954262,
+      "grad_norm": 0.005089858081191778,
+      "learning_rate": 4.4664487607626434e-05,
+      "loss": 0.0406,
+      "num_input_tokens_seen": 6621720,
+      "step": 8475
+    },
+    {
+      "epoch": 17.62993762993763,
+      "grad_norm": 0.03914234787225723,
+      "learning_rate": 4.4658423907299845e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 6625656,
+      "step": 8480
+    },
+    {
+      "epoch": 17.64033264033264,
+      "grad_norm": 0.010528391227126122,
+      "learning_rate": 4.465235717539725e-05,
+      "loss": 0.0142,
+      "num_input_tokens_seen": 6629528,
+      "step": 8485
+    },
+    {
+      "epoch": 17.65072765072765,
+      "grad_norm": 0.013506411574780941,
+      "learning_rate": 4.464628741285421e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 6633592,
+      "step": 8490
+    },
+    {
+      "epoch": 17.66112266112266,
+      "grad_norm": 0.03788122907280922,
+      "learning_rate": 4.4640214620606754e-05,
+      "loss": 0.011,
+      "num_input_tokens_seen": 6637432,
+      "step": 8495
+    },
+    {
+      "epoch": 17.671517671517673,
+      "grad_norm": 22.47655487060547,
+      "learning_rate": 4.46341387995914e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 6641336,
+      "step": 8500
+    },
+    {
+      "epoch": 17.68191268191268,
+      "grad_norm": 0.06966245174407959,
+      "learning_rate": 4.4628059950745106e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 6645144,
+      "step": 8505
+    },
+    {
+      "epoch": 17.692307692307693,
+      "grad_norm": 0.020864883437752724,
+      "learning_rate": 4.4621978075005297e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 6648888,
+      "step": 8510
+    },
+    {
+      "epoch": 17.7027027027027,
+      "grad_norm": 0.6676743030548096,
+      "learning_rate": 4.461589317330989e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 6652856,
+      "step": 8515
+    },
+    {
+      "epoch": 17.713097713097714,
+      "grad_norm": 0.006599752232432365,
+      "learning_rate": 4.460980524659724e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 6656920,
+      "step": 8520
+    },
+    {
+      "epoch": 17.723492723492722,
+      "grad_norm": 0.02141130343079567,
+      "learning_rate": 4.46037142958062e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 6660728,
+      "step": 8525
+    },
+    {
+      "epoch": 17.733887733887734,
+      "grad_norm": 0.8411881327629089,
+      "learning_rate": 4.4597620321876046e-05,
+      "loss": 0.0086,
+      "num_input_tokens_seen": 6664664,
+      "step": 8530
+    },
+    {
+      "epoch": 17.744282744282746,
+      "grad_norm": 0.014439110644161701,
+      "learning_rate": 4.459152332574656e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6668568,
+      "step": 8535
+    },
+    {
+      "epoch": 17.754677754677754,
+      "grad_norm": 0.004729472100734711,
+      "learning_rate": 4.4585423308357985e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6672440,
+      "step": 8540
+    },
+    {
+      "epoch": 17.765072765072766,
+      "grad_norm": 0.01057102158665657,
+      "learning_rate": 4.457932027065102e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6676312,
+      "step": 8545
+    },
+    {
+      "epoch": 17.775467775467774,
+      "grad_norm": 0.00198289449326694,
+      "learning_rate": 4.45732142135668e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6680184,
+      "step": 8550
+    },
+    {
+      "epoch": 17.785862785862786,
+      "grad_norm": 0.01267933938652277,
+      "learning_rate": 4.4567105138046986e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6684120,
+      "step": 8555
+    },
+    {
+      "epoch": 17.796257796257795,
+      "grad_norm": 0.0018415853846818209,
+      "learning_rate": 4.456099304503365e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6688184,
+      "step": 8560
+    },
+    {
+      "epoch": 17.806652806652806,
+      "grad_norm": 0.0009031853987835348,
+      "learning_rate": 4.455487793546939e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 6692120,
+      "step": 8565
+    },
+    {
+      "epoch": 17.81704781704782,
+      "grad_norm": 2.2261483669281006,
+      "learning_rate": 4.454875981029719e-05,
+      "loss": 0.0157,
+      "num_input_tokens_seen": 6696024,
+      "step": 8570
+    },
+    {
+      "epoch": 17.827442827442827,
+      "grad_norm": 0.001802282058633864,
+      "learning_rate": 4.454263867046057e-05,
+      "loss": 0.0301,
+      "num_input_tokens_seen": 6699928,
+      "step": 8575
+    },
+    {
+      "epoch": 17.83783783783784,
+      "grad_norm": 0.03232668712735176,
+      "learning_rate": 4.4536514516903484e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6703736,
+      "step": 8580
+    },
+    {
+      "epoch": 17.848232848232847,
+      "grad_norm": 1.195551872253418,
+      "learning_rate": 4.453038735057034e-05,
+      "loss": 0.0066,
+      "num_input_tokens_seen": 6707736,
+      "step": 8585
+    },
+    {
+      "epoch": 17.85862785862786,
+      "grad_norm": 4.085968971252441,
+      "learning_rate": 4.4524257172406034e-05,
+      "loss": 0.014,
+      "num_input_tokens_seen": 6711608,
+      "step": 8590
+    },
+    {
+      "epoch": 17.86902286902287,
+      "grad_norm": 0.009458834305405617,
+      "learning_rate": 4.451812398335592e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 6715448,
+      "step": 8595
+    },
+    {
+      "epoch": 17.87941787941788,
+      "grad_norm": 0.15226314961910248,
+      "learning_rate": 4.4511987784365805e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 6719288,
+      "step": 8600
+    },
+    {
+      "epoch": 17.87941787941788,
+      "eval_loss": 0.30361926555633545,
+      "eval_runtime": 11.7354,
+      "eval_samples_per_second": 72.942,
+      "eval_steps_per_second": 18.235,
+      "num_input_tokens_seen": 6719288,
+      "step": 8600
+    },
+    {
+      "epoch": 17.88981288981289,
+      "grad_norm": 0.0385698638856411,
+      "learning_rate": 4.450584857638197e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6723192,
+      "step": 8605
+    },
+    {
+      "epoch": 17.9002079002079,
+      "grad_norm": 0.10000857710838318,
+      "learning_rate": 4.449970636035116e-05,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 6727096,
+      "step": 8610
+    },
+    {
+      "epoch": 17.91060291060291,
+      "grad_norm": 0.014321527443826199,
+      "learning_rate": 4.4493561137220574e-05,
+      "loss": 0.0274,
+      "num_input_tokens_seen": 6731032,
+      "step": 8615
+    },
+    {
+      "epoch": 17.92099792099792,
+      "grad_norm": 0.05307680368423462,
+      "learning_rate": 4.44874129079379e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 6734872,
+      "step": 8620
+    },
+    {
+      "epoch": 17.93139293139293,
+      "grad_norm": 0.5392767786979675,
+      "learning_rate": 4.4481261673451255e-05,
+      "loss": 0.0061,
+      "num_input_tokens_seen": 6738776,
+      "step": 8625
+    },
+    {
+      "epoch": 17.941787941787943,
+      "grad_norm": 0.018508339300751686,
+      "learning_rate": 4.4475107434709245e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 6742840,
+      "step": 8630
+    },
+    {
+      "epoch": 17.95218295218295,
+      "grad_norm": 0.008080949075520039,
+      "learning_rate": 4.446895019266093e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6746744,
+      "step": 8635
+    },
+    {
+      "epoch": 17.962577962577964,
+      "grad_norm": 0.002380843972787261,
+      "learning_rate": 4.446278994825583e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6750744,
+      "step": 8640
+    },
+    {
+      "epoch": 17.972972972972972,
+      "grad_norm": 0.019223669543862343,
+      "learning_rate": 4.445662670244394e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 6754616,
+      "step": 8645
+    },
+    {
+      "epoch": 17.983367983367984,
+      "grad_norm": 0.0017484142445027828,
+      "learning_rate": 4.44504604561757e-05,
+      "loss": 0.0091,
+      "num_input_tokens_seen": 6758520,
+      "step": 8650
+    },
+    {
+      "epoch": 17.993762993762992,
+      "grad_norm": 0.0013211109908297658,
+      "learning_rate": 4.4444291210402035e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6762488,
+      "step": 8655
+    },
+    {
+      "epoch": 18.004158004158004,
+      "grad_norm": 0.00235916068777442,
+      "learning_rate": 4.443811896607431e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6766312,
+      "step": 8660
+    },
+    {
+      "epoch": 18.014553014553016,
+      "grad_norm": 0.0028010013047605753,
+      "learning_rate": 4.443194372414436e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6770280,
+      "step": 8665
+    },
+    {
+      "epoch": 18.024948024948024,
+      "grad_norm": 0.006842336617410183,
+      "learning_rate": 4.442576548556449e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6774184,
+      "step": 8670
+    },
+    {
+      "epoch": 18.035343035343036,
+      "grad_norm": 0.0014726038789376616,
+      "learning_rate": 4.441958425128747e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6777992,
+      "step": 8675
+    },
+    {
+      "epoch": 18.045738045738045,
+      "grad_norm": 0.002342812018468976,
+      "learning_rate": 4.4413400022266515e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 6781960,
+      "step": 8680
+    },
+    {
+      "epoch": 18.056133056133056,
+      "grad_norm": 0.001647822093218565,
+      "learning_rate": 4.4407212799455313e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6785832,
+      "step": 8685
+    },
+    {
+      "epoch": 18.066528066528065,
+      "grad_norm": 0.001254828181117773,
+      "learning_rate": 4.4401022583808003e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 6789864,
+      "step": 8690
+    },
+    {
+      "epoch": 18.076923076923077,
+      "grad_norm": 0.1833944022655487,
+      "learning_rate": 4.439482937627921e-05,
+      "loss": 0.0422,
+      "num_input_tokens_seen": 6793800,
+      "step": 8695
+    },
+    {
+      "epoch": 18.08731808731809,
+      "grad_norm": 0.001657387474551797,
+      "learning_rate": 4.4388633177824004e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6797640,
+      "step": 8700
+    },
+    {
+      "epoch": 18.097713097713097,
+      "grad_norm": 4.811999797821045,
+      "learning_rate": 4.4382433989397895e-05,
+      "loss": 0.0101,
+      "num_input_tokens_seen": 6801512,
+      "step": 8705
+    },
+    {
+      "epoch": 18.10810810810811,
+      "grad_norm": 0.047937117516994476,
+      "learning_rate": 4.4376231811956895e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6805288,
+      "step": 8710
+    },
+    {
+      "epoch": 18.118503118503117,
+      "grad_norm": 0.0014649588847532868,
+      "learning_rate": 4.437002664645745e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6809256,
+      "step": 8715
+    },
+    {
+      "epoch": 18.12889812889813,
+      "grad_norm": 0.008437583222985268,
+      "learning_rate": 4.436381849385649e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 6813064,
+      "step": 8720
+    },
+    {
+      "epoch": 18.13929313929314,
+      "grad_norm": 0.0017435267800465226,
+      "learning_rate": 4.435760735511136e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6817096,
+      "step": 8725
+    },
+    {
+      "epoch": 18.14968814968815,
+      "grad_norm": 0.0010306776966899633,
+      "learning_rate": 4.435139323117992e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6821096,
+      "step": 8730
+    },
+    {
+      "epoch": 18.16008316008316,
+      "grad_norm": 0.008486341685056686,
+      "learning_rate": 4.434517612302046e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6824872,
+      "step": 8735
+    },
+    {
+      "epoch": 18.17047817047817,
+      "grad_norm": 0.0004062569933012128,
+      "learning_rate": 4.433895603159174e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6828744,
+      "step": 8740
+    },
+    {
+      "epoch": 18.18087318087318,
+      "grad_norm": 0.008802400901913643,
+      "learning_rate": 4.433273295785296e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6832776,
+      "step": 8745
+    },
+    {
+      "epoch": 18.19126819126819,
+      "grad_norm": 0.008892741985619068,
+      "learning_rate": 4.432650690276382e-05,
+      "loss": 0.0114,
+      "num_input_tokens_seen": 6836680,
+      "step": 8750
+    },
+    {
+      "epoch": 18.2016632016632,
+      "grad_norm": 0.0049385447055101395,
+      "learning_rate": 4.4320277867284435e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6840648,
+      "step": 8755
+    },
+    {
+      "epoch": 18.212058212058214,
+      "grad_norm": 0.9296993017196655,
+      "learning_rate": 4.431404585237541e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 6844520,
+      "step": 8760
+    },
+    {
+      "epoch": 18.222453222453222,
+      "grad_norm": 0.0014056918444111943,
+      "learning_rate": 4.43078108589978e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6848392,
+      "step": 8765
+    },
+    {
+      "epoch": 18.232848232848234,
+      "grad_norm": 0.0024011239875108004,
+      "learning_rate": 4.4301572888113116e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6852264,
+      "step": 8770
+    },
+    {
+      "epoch": 18.243243243243242,
+      "grad_norm": 0.015302462503314018,
+      "learning_rate": 4.4295331940683337e-05,
+      "loss": 0.0075,
+      "num_input_tokens_seen": 6856200,
+      "step": 8775
+    },
+    {
+      "epoch": 18.253638253638254,
+      "grad_norm": 0.002090963302180171,
+      "learning_rate": 4.428908801767089e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6860072,
+      "step": 8780
+    },
+    {
+      "epoch": 18.264033264033262,
+      "grad_norm": 0.07867579907178879,
+      "learning_rate": 4.428284112003868e-05,
+      "loss": 0.0379,
+      "num_input_tokens_seen": 6863944,
+      "step": 8785
+    },
+    {
+      "epoch": 18.274428274428274,
+      "grad_norm": 0.033874984830617905,
+      "learning_rate": 4.4276591248750033e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6867752,
+      "step": 8790
+    },
+    {
+      "epoch": 18.284823284823286,
+      "grad_norm": 0.08844104409217834,
+      "learning_rate": 4.4270338404768774e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6871592,
+      "step": 8795
+    },
+    {
+      "epoch": 18.295218295218294,
+      "grad_norm": 0.012733840383589268,
+      "learning_rate": 4.426408258905917e-05,
+      "loss": 0.0086,
+      "num_input_tokens_seen": 6875592,
+      "step": 8800
+    },
+    {
+      "epoch": 18.295218295218294,
+      "eval_loss": 0.42426058650016785,
+      "eval_runtime": 11.7924,
+      "eval_samples_per_second": 72.589,
+      "eval_steps_per_second": 18.147,
+      "num_input_tokens_seen": 6875592,
+      "step": 8800
+    },
+    {
+      "epoch": 18.305613305613306,
+      "grad_norm": 0.004350714851170778,
+      "learning_rate": 4.425782380258594e-05,
+      "loss": 0.0184,
+      "num_input_tokens_seen": 6879496,
+      "step": 8805
+    },
+    {
+      "epoch": 18.316008316008315,
+      "grad_norm": 0.0014879924710839987,
+      "learning_rate": 4.425156204631427e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6883432,
+      "step": 8810
+    },
+    {
+      "epoch": 18.326403326403327,
+      "grad_norm": 3.1495790481567383,
+      "learning_rate": 4.424529732120981e-05,
+      "loss": 0.0065,
+      "num_input_tokens_seen": 6887176,
+      "step": 8815
+    },
+    {
+      "epoch": 18.33679833679834,
+      "grad_norm": 0.011795809492468834,
+      "learning_rate": 4.423902962823864e-05,
+      "loss": 0.0147,
+      "num_input_tokens_seen": 6891048,
+      "step": 8820
+    },
+    {
+      "epoch": 18.347193347193347,
+      "grad_norm": 0.008360178209841251,
+      "learning_rate": 4.423275896836733e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 6894952,
+      "step": 8825
+    },
+    {
+      "epoch": 18.35758835758836,
+      "grad_norm": 1.0073368549346924,
+      "learning_rate": 4.42264853425629e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 6899016,
+      "step": 8830
+    },
+    {
+      "epoch": 18.367983367983367,
+      "grad_norm": 0.034642551094293594,
+      "learning_rate": 4.4220208751792816e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6902888,
+      "step": 8835
+    },
+    {
+      "epoch": 18.37837837837838,
+      "grad_norm": 0.005602675024420023,
+      "learning_rate": 4.421392919702499e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 6906856,
+      "step": 8840
+    },
+    {
+      "epoch": 18.388773388773387,
+      "grad_norm": 0.461182564496994,
+      "learning_rate": 4.4207646679227846e-05,
+      "loss": 0.012,
+      "num_input_tokens_seen": 6910824,
+      "step": 8845
+    },
+    {
+      "epoch": 18.3991683991684,
+      "grad_norm": 0.2156793773174286,
+      "learning_rate": 4.42013611993702e-05,
+      "loss": 0.023,
+      "num_input_tokens_seen": 6914728,
+      "step": 8850
+    },
+    {
+      "epoch": 18.40956340956341,
+      "grad_norm": 0.014667106792330742,
+      "learning_rate": 4.419507275842135e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 6918568,
+      "step": 8855
+    },
+    {
+      "epoch": 18.41995841995842,
+      "grad_norm": 0.007777804508805275,
+      "learning_rate": 4.418878135735106e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 6922472,
+      "step": 8860
+    },
+    {
+      "epoch": 18.43035343035343,
+      "grad_norm": 0.7429842352867126,
+      "learning_rate": 4.418248699712955e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 6926344,
+      "step": 8865
+    },
+    {
+      "epoch": 18.44074844074844,
+      "grad_norm": 0.006064104847609997,
+      "learning_rate": 4.417618967872748e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 6930280,
+      "step": 8870
+    },
+    {
+      "epoch": 18.45114345114345,
+      "grad_norm": 0.010251617059111595,
+      "learning_rate": 4.4169889403115985e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 6934184,
+      "step": 8875
+    },
+    {
+      "epoch": 18.46153846153846,
+      "grad_norm": 0.0037590546999126673,
+      "learning_rate": 4.4163586171266627e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 6938120,
+      "step": 8880
+    },
+    {
+      "epoch": 18.471933471933472,
+      "grad_norm": 21.746665954589844,
+      "learning_rate": 4.415727998415147e-05,
+      "loss": 0.0079,
+      "num_input_tokens_seen": 6942024,
+      "step": 8885
+    },
+    {
+      "epoch": 18.482328482328484,
+      "grad_norm": 0.020424241200089455,
+      "learning_rate": 4.4150970842742985e-05,
+      "loss": 0.0105,
+      "num_input_tokens_seen": 6945960,
+      "step": 8890
+    },
+    {
+      "epoch": 18.492723492723492,
+      "grad_norm": 0.0014556065434589982,
+      "learning_rate": 4.4144658748014134e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 6949864,
+      "step": 8895
+    },
+    {
+      "epoch": 18.503118503118504,
+      "grad_norm": 14.067486763000488,
+      "learning_rate": 4.413834370093831e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 6953736,
+      "step": 8900
+    },
+    {
+      "epoch": 18.513513513513512,
+      "grad_norm": 0.045841120183467865,
+      "learning_rate": 4.413202570248939e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 6957640,
+      "step": 8905
+    },
+    {
+      "epoch": 18.523908523908524,
+      "grad_norm": 0.0016749849310144782,
+      "learning_rate": 4.412570475364167e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 6961672,
+      "step": 8910
+    },
+    {
+      "epoch": 18.534303534303533,
+      "grad_norm": 0.005154506769031286,
+      "learning_rate": 4.411938085536994e-05,
+      "loss": 0.0305,
+      "num_input_tokens_seen": 6965512,
+      "step": 8915
+    },
+    {
+      "epoch": 18.544698544698544,
+      "grad_norm": 0.0015094536356627941,
+      "learning_rate": 4.41130540086494e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 6969320,
+      "step": 8920
+    },
+    {
+      "epoch": 18.555093555093556,
+      "grad_norm": 1.951533555984497,
+      "learning_rate": 4.4106724214455754e-05,
+      "loss": 0.0084,
+      "num_input_tokens_seen": 6973416,
+      "step": 8925
+    },
+    {
+      "epoch": 18.565488565488565,
+      "grad_norm": 0.0009440477006137371,
+      "learning_rate": 4.4100391473765115e-05,
+      "loss": 0.0155,
+      "num_input_tokens_seen": 6977352,
+      "step": 8930
+    },
+    {
+      "epoch": 18.575883575883577,
+      "grad_norm": 0.0005057752714492381,
+      "learning_rate": 4.409405578755408e-05,
+      "loss": 0.0728,
+      "num_input_tokens_seen": 6981192,
+      "step": 8935
+    },
+    {
+      "epoch": 18.586278586278585,
+      "grad_norm": 0.0005936166271567345,
+      "learning_rate": 4.4087717156799705e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 6985128,
+      "step": 8940
+    },
+    {
+      "epoch": 18.596673596673597,
+      "grad_norm": 0.00036830283352173865,
+      "learning_rate": 4.408137558247946e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 6989032,
+      "step": 8945
+    },
+    {
+      "epoch": 18.60706860706861,
+      "grad_norm": 3.057783365249634,
+      "learning_rate": 4.4075031065571306e-05,
+      "loss": 0.0173,
+      "num_input_tokens_seen": 6993160,
+      "step": 8950
+    },
+    {
+      "epoch": 18.617463617463617,
+      "grad_norm": 0.03817493095993996,
+      "learning_rate": 4.406868360705366e-05,
+      "loss": 0.0349,
+      "num_input_tokens_seen": 6997160,
+      "step": 8955
+    },
+    {
+      "epoch": 18.62785862785863,
+      "grad_norm": 0.035579029470682144,
+      "learning_rate": 4.406233320790536e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 7001000,
+      "step": 8960
+    },
+    {
+      "epoch": 18.638253638253637,
+      "grad_norm": 0.0674932450056076,
+      "learning_rate": 4.4055979869105734e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 7004872,
+      "step": 8965
+    },
+    {
+      "epoch": 18.64864864864865,
+      "grad_norm": 0.059263862669467926,
+      "learning_rate": 4.404962359163454e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7008808,
+      "step": 8970
+    },
+    {
+      "epoch": 18.659043659043657,
+      "grad_norm": 0.028947738930583,
+      "learning_rate": 4.404326437647199e-05,
+      "loss": 0.0467,
+      "num_input_tokens_seen": 7012744,
+      "step": 8975
+    },
+    {
+      "epoch": 18.66943866943867,
+      "grad_norm": 0.03783753141760826,
+      "learning_rate": 4.403690222459877e-05,
+      "loss": 0.0085,
+      "num_input_tokens_seen": 7016712,
+      "step": 8980
+    },
+    {
+      "epoch": 18.67983367983368,
+      "grad_norm": 0.11083272099494934,
+      "learning_rate": 4.4030537136995984e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 7020616,
+      "step": 8985
+    },
+    {
+      "epoch": 18.69022869022869,
+      "grad_norm": 0.024758264422416687,
+      "learning_rate": 4.402416911464523e-05,
+      "loss": 0.0075,
+      "num_input_tokens_seen": 7024520,
+      "step": 8990
+    },
+    {
+      "epoch": 18.7006237006237,
+      "grad_norm": 0.07515788078308105,
+      "learning_rate": 4.4017798158528516e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 7028488,
+      "step": 8995
+    },
+    {
+      "epoch": 18.71101871101871,
+      "grad_norm": 0.011992399580776691,
+      "learning_rate": 4.401142426962834e-05,
+      "loss": 0.0238,
+      "num_input_tokens_seen": 7032392,
+      "step": 9000
+    },
+    {
+      "epoch": 18.71101871101871,
+      "eval_loss": 0.3896433115005493,
+      "eval_runtime": 11.7505,
+      "eval_samples_per_second": 72.848,
+      "eval_steps_per_second": 18.212,
+      "num_input_tokens_seen": 7032392,
+      "step": 9000
+    },
+    {
+      "epoch": 18.72141372141372,
+      "grad_norm": 0.0014524637954309583,
+      "learning_rate": 4.400504744892763e-05,
+      "loss": 0.014,
+      "num_input_tokens_seen": 7036200,
+      "step": 9005
+    },
+    {
+      "epoch": 18.731808731808734,
+      "grad_norm": 0.005674745887517929,
+      "learning_rate": 4.399866769740975e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7040072,
+      "step": 9010
+    },
+    {
+      "epoch": 18.742203742203742,
+      "grad_norm": 2.698385000228882,
+      "learning_rate": 4.399228501605859e-05,
+      "loss": 0.023,
+      "num_input_tokens_seen": 7043912,
+      "step": 9015
+    },
+    {
+      "epoch": 18.752598752598754,
+      "grad_norm": 0.03240244463086128,
+      "learning_rate": 4.398589940585839e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7047784,
+      "step": 9020
+    },
+    {
+      "epoch": 18.762993762993762,
+      "grad_norm": 0.3848036527633667,
+      "learning_rate": 4.3979510867793917e-05,
+      "loss": 0.0147,
+      "num_input_tokens_seen": 7051656,
+      "step": 9025
+    },
+    {
+      "epoch": 18.773388773388774,
+      "grad_norm": 0.02791333571076393,
+      "learning_rate": 4.3973119402850346e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7055560,
+      "step": 9030
+    },
+    {
+      "epoch": 18.783783783783782,
+      "grad_norm": 0.0033901971764862537,
+      "learning_rate": 4.396672501201334e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 7059432,
+      "step": 9035
+    },
+    {
+      "epoch": 18.794178794178794,
+      "grad_norm": 0.01879810355603695,
+      "learning_rate": 4.396032769626899e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7063400,
+      "step": 9040
+    },
+    {
+      "epoch": 18.804573804573806,
+      "grad_norm": 0.0014846312114968896,
+      "learning_rate": 4.395392745660384e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7067432,
+      "step": 9045
+    },
+    {
+      "epoch": 18.814968814968815,
+      "grad_norm": 0.0944225937128067,
+      "learning_rate": 4.394752429400488e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7071400,
+      "step": 9050
+    },
+    {
+      "epoch": 18.825363825363826,
+      "grad_norm": 1.7997299432754517,
+      "learning_rate": 4.394111820945957e-05,
+      "loss": 0.0271,
+      "num_input_tokens_seen": 7075208,
+      "step": 9055
+    },
+    {
+      "epoch": 18.835758835758835,
+      "grad_norm": 0.0301273874938488,
+      "learning_rate": 4.393470920395579e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 7079208,
+      "step": 9060
+    },
+    {
+      "epoch": 18.846153846153847,
+      "grad_norm": 0.24553526937961578,
+      "learning_rate": 4.392829727848192e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 7083176,
+      "step": 9065
+    },
+    {
+      "epoch": 18.856548856548855,
+      "grad_norm": 0.07568057626485825,
+      "learning_rate": 4.392188243402673e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 7087112,
+      "step": 9070
+    },
+    {
+      "epoch": 18.866943866943867,
+      "grad_norm": 0.01672988012433052,
+      "learning_rate": 4.391546467157949e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 7090984,
+      "step": 9075
+    },
+    {
+      "epoch": 18.87733887733888,
+      "grad_norm": 0.07133127003908157,
+      "learning_rate": 4.390904399212988e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 7094856,
+      "step": 9080
+    },
+    {
+      "epoch": 18.887733887733887,
+      "grad_norm": 0.03471151739358902,
+      "learning_rate": 4.390262039666807e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 7098856,
+      "step": 9085
+    },
+    {
+      "epoch": 18.8981288981289,
+      "grad_norm": 0.009242451749742031,
+      "learning_rate": 4.389619388618464e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7102696,
+      "step": 9090
+    },
+    {
+      "epoch": 18.908523908523907,
+      "grad_norm": 0.0017885027918964624,
+      "learning_rate": 4.3889764461670655e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7106600,
+      "step": 9095
+    },
+    {
+      "epoch": 18.91891891891892,
+      "grad_norm": 0.02879387140274048,
+      "learning_rate": 4.38833321241176e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7110504,
+      "step": 9100
+    },
+    {
+      "epoch": 18.929313929313928,
+      "grad_norm": 0.005251833703368902,
+      "learning_rate": 4.3876896874517434e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7114376,
+      "step": 9105
+    },
+    {
+      "epoch": 18.93970893970894,
+      "grad_norm": 0.02801823802292347,
+      "learning_rate": 4.3870458713862554e-05,
+      "loss": 0.0374,
+      "num_input_tokens_seen": 7118248,
+      "step": 9110
+    },
+    {
+      "epoch": 18.95010395010395,
+      "grad_norm": 0.8053313493728638,
+      "learning_rate": 4.386401764314579e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7122248,
+      "step": 9115
+    },
+    {
+      "epoch": 18.96049896049896,
+      "grad_norm": 0.0009069358347915113,
+      "learning_rate": 4.385757366336045e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7126184,
+      "step": 9120
+    },
+    {
+      "epoch": 18.97089397089397,
+      "grad_norm": 0.019884303212165833,
+      "learning_rate": 4.385112677550027e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 7130184,
+      "step": 9125
+    },
+    {
+      "epoch": 18.98128898128898,
+      "grad_norm": 1.6082353591918945,
+      "learning_rate": 4.384467698055945e-05,
+      "loss": 0.0147,
+      "num_input_tokens_seen": 7133992,
+      "step": 9130
+    },
+    {
+      "epoch": 18.991683991683992,
+      "grad_norm": 0.0011145351454615593,
+      "learning_rate": 4.383822427953261e-05,
+      "loss": 0.0131,
+      "num_input_tokens_seen": 7137736,
+      "step": 9135
+    },
+    {
+      "epoch": 19.002079002079004,
+      "grad_norm": 0.0007585136918351054,
+      "learning_rate": 4.3831768673414864e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7141624,
+      "step": 9140
+    },
+    {
+      "epoch": 19.012474012474012,
+      "grad_norm": 0.0013565808767452836,
+      "learning_rate": 4.382531016320173e-05,
+      "loss": 0.0285,
+      "num_input_tokens_seen": 7145528,
+      "step": 9145
+    },
+    {
+      "epoch": 19.022869022869024,
+      "grad_norm": 0.0011094907531514764,
+      "learning_rate": 4.3818848749889184e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7149336,
+      "step": 9150
+    },
+    {
+      "epoch": 19.033264033264032,
+      "grad_norm": 0.0017787219258025289,
+      "learning_rate": 4.381238443447368e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 7153272,
+      "step": 9155
+    },
+    {
+      "epoch": 19.043659043659044,
+      "grad_norm": 0.09489791095256805,
+      "learning_rate": 4.380591721795208e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7157176,
+      "step": 9160
+    },
+    {
+      "epoch": 19.054054054054053,
+      "grad_norm": 0.002361648017540574,
+      "learning_rate": 4.3799447101321723e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 7160888,
+      "step": 9165
+    },
+    {
+      "epoch": 19.064449064449065,
+      "grad_norm": 0.010296989232301712,
+      "learning_rate": 4.379297408558036e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7164760,
+      "step": 9170
+    },
+    {
+      "epoch": 19.074844074844076,
+      "grad_norm": 2.982184410095215,
+      "learning_rate": 4.378649817172624e-05,
+      "loss": 0.0434,
+      "num_input_tokens_seen": 7168600,
+      "step": 9175
+    },
+    {
+      "epoch": 19.085239085239085,
+      "grad_norm": 0.0028615272603929043,
+      "learning_rate": 4.378001936075801e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7172504,
+      "step": 9180
+    },
+    {
+      "epoch": 19.095634095634097,
+      "grad_norm": 3.9629344940185547,
+      "learning_rate": 4.377353765367479e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 7176536,
+      "step": 9185
+    },
+    {
+      "epoch": 19.106029106029105,
+      "grad_norm": 0.0015214100712910295,
+      "learning_rate": 4.376705305147614e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7180472,
+      "step": 9190
+    },
+    {
+      "epoch": 19.116424116424117,
+      "grad_norm": 0.01849008910357952,
+      "learning_rate": 4.376056555516206e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7184312,
+      "step": 9195
+    },
+    {
+      "epoch": 19.126819126819125,
+      "grad_norm": 1.0551828145980835,
+      "learning_rate": 4.375407516573302e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7188120,
+      "step": 9200
+    },
+    {
+      "epoch": 19.126819126819125,
+      "eval_loss": 0.32237207889556885,
+      "eval_runtime": 11.7123,
+      "eval_samples_per_second": 73.085,
+      "eval_steps_per_second": 18.271,
+      "num_input_tokens_seen": 7188120,
+      "step": 9200
+    },
+    {
+      "epoch": 19.137214137214137,
+      "grad_norm": 0.0013765387702733278,
+      "learning_rate": 4.3747581884189913e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7192152,
+      "step": 9205
+    },
+    {
+      "epoch": 19.14760914760915,
+      "grad_norm": 0.007411059457808733,
+      "learning_rate": 4.374108571153408e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7195928,
+      "step": 9210
+    },
+    {
+      "epoch": 19.158004158004157,
+      "grad_norm": 0.0025674644857645035,
+      "learning_rate": 4.3734586648767316e-05,
+      "loss": 0.0216,
+      "num_input_tokens_seen": 7199896,
+      "step": 9215
+    },
+    {
+      "epoch": 19.16839916839917,
+      "grad_norm": 0.021920178085565567,
+      "learning_rate": 4.372808469689186e-05,
+      "loss": 0.0601,
+      "num_input_tokens_seen": 7203896,
+      "step": 9220
+    },
+    {
+      "epoch": 19.178794178794178,
+      "grad_norm": 0.1206963062286377,
+      "learning_rate": 4.372157985691039e-05,
+      "loss": 0.0314,
+      "num_input_tokens_seen": 7207768,
+      "step": 9225
+    },
+    {
+      "epoch": 19.18918918918919,
+      "grad_norm": 0.07288124412298203,
+      "learning_rate": 4.371507212982603e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 7211576,
+      "step": 9230
+    },
+    {
+      "epoch": 19.1995841995842,
+      "grad_norm": 0.014435326680541039,
+      "learning_rate": 4.370856151664236e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7215416,
+      "step": 9235
+    },
+    {
+      "epoch": 19.20997920997921,
+      "grad_norm": 0.005922358483076096,
+      "learning_rate": 4.3702048018363404e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7219256,
+      "step": 9240
+    },
+    {
+      "epoch": 19.22037422037422,
+      "grad_norm": 0.0345078706741333,
+      "learning_rate": 4.369553163599362e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7223160,
+      "step": 9245
+    },
+    {
+      "epoch": 19.23076923076923,
+      "grad_norm": 0.004253263119608164,
+      "learning_rate": 4.3689012370537904e-05,
+      "loss": 0.0339,
+      "num_input_tokens_seen": 7227064,
+      "step": 9250
+    },
+    {
+      "epoch": 19.241164241164242,
+      "grad_norm": 0.006758962757885456,
+      "learning_rate": 4.368249022300164e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7230872,
+      "step": 9255
+    },
+    {
+      "epoch": 19.25155925155925,
+      "grad_norm": 0.006864566821604967,
+      "learning_rate": 4.367596519439059e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7234904,
+      "step": 9260
+    },
+    {
+      "epoch": 19.261954261954262,
+      "grad_norm": 0.001782537205144763,
+      "learning_rate": 4.366943728571101e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7238584,
+      "step": 9265
+    },
+    {
+      "epoch": 19.272349272349274,
+      "grad_norm": 0.007178550120443106,
+      "learning_rate": 4.366290649796959e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7242456,
+      "step": 9270
+    },
+    {
+      "epoch": 19.282744282744282,
+      "grad_norm": 0.02892821468412876,
+      "learning_rate": 4.3656372832173456e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7246488,
+      "step": 9275
+    },
+    {
+      "epoch": 19.293139293139294,
+      "grad_norm": 0.013718009926378727,
+      "learning_rate": 4.364983628933017e-05,
+      "loss": 0.0284,
+      "num_input_tokens_seen": 7250456,
+      "step": 9280
+    },
+    {
+      "epoch": 19.303534303534303,
+      "grad_norm": 0.021066170185804367,
+      "learning_rate": 4.364329687044777e-05,
+      "loss": 0.0093,
+      "num_input_tokens_seen": 7254264,
+      "step": 9285
+    },
+    {
+      "epoch": 19.313929313929314,
+      "grad_norm": 0.031627386808395386,
+      "learning_rate": 4.36367545765347e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 7258200,
+      "step": 9290
+    },
+    {
+      "epoch": 19.324324324324323,
+      "grad_norm": 0.007449266966432333,
+      "learning_rate": 4.363020940859988e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 7262200,
+      "step": 9295
+    },
+    {
+      "epoch": 19.334719334719335,
+      "grad_norm": 0.039484117180109024,
+      "learning_rate": 4.362366136765263e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7266264,
+      "step": 9300
+    },
+    {
+      "epoch": 19.345114345114347,
+      "grad_norm": 0.0008576915715821087,
+      "learning_rate": 4.361711045470278e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 7270200,
+      "step": 9305
+    },
+    {
+      "epoch": 19.355509355509355,
+      "grad_norm": 0.016613058745861053,
+      "learning_rate": 4.3610556670760524e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7273976,
+      "step": 9310
+    },
+    {
+      "epoch": 19.365904365904367,
+      "grad_norm": 0.0413687527179718,
+      "learning_rate": 4.360400001683657e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7277816,
+      "step": 9315
+    },
+    {
+      "epoch": 19.376299376299375,
+      "grad_norm": 0.023947447538375854,
+      "learning_rate": 4.3597440493942e-05,
+      "loss": 0.0113,
+      "num_input_tokens_seen": 7281720,
+      "step": 9320
+    },
+    {
+      "epoch": 19.386694386694387,
+      "grad_norm": 0.013270124793052673,
+      "learning_rate": 4.3590878103088405e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 7285816,
+      "step": 9325
+    },
+    {
+      "epoch": 19.397089397089395,
+      "grad_norm": 0.001659382600337267,
+      "learning_rate": 4.358431284528779e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7289720,
+      "step": 9330
+    },
+    {
+      "epoch": 19.407484407484407,
+      "grad_norm": 0.01401410810649395,
+      "learning_rate": 4.357774472155257e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7293752,
+      "step": 9335
+    },
+    {
+      "epoch": 19.41787941787942,
+      "grad_norm": 0.0006796792149543762,
+      "learning_rate": 4.3571173732895664e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 7297720,
+      "step": 9340
+    },
+    {
+      "epoch": 19.428274428274428,
+      "grad_norm": 0.005369341000914574,
+      "learning_rate": 4.356459988033039e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7301624,
+      "step": 9345
+    },
+    {
+      "epoch": 19.43866943866944,
+      "grad_norm": 0.004104485269635916,
+      "learning_rate": 4.355802316487051e-05,
+      "loss": 0.0217,
+      "num_input_tokens_seen": 7305624,
+      "step": 9350
+    },
+    {
+      "epoch": 19.449064449064448,
+      "grad_norm": 0.0017564091831445694,
+      "learning_rate": 4.355144358753025e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7309656,
+      "step": 9355
+    },
+    {
+      "epoch": 19.45945945945946,
+      "grad_norm": 0.15754827857017517,
+      "learning_rate": 4.354486114932425e-05,
+      "loss": 0.0117,
+      "num_input_tokens_seen": 7313464,
+      "step": 9360
+    },
+    {
+      "epoch": 19.46985446985447,
+      "grad_norm": 0.010223554447293282,
+      "learning_rate": 4.353827585126762e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7317432,
+      "step": 9365
+    },
+    {
+      "epoch": 19.48024948024948,
+      "grad_norm": 1.203948736190796,
+      "learning_rate": 4.353168769437588e-05,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 7321176,
+      "step": 9370
+    },
+    {
+      "epoch": 19.490644490644492,
+      "grad_norm": 0.037610556930303574,
+      "learning_rate": 4.3525096679665014e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7325080,
+      "step": 9375
+    },
+    {
+      "epoch": 19.5010395010395,
+      "grad_norm": 0.008778461255133152,
+      "learning_rate": 4.351850280815144e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7329080,
+      "step": 9380
+    },
+    {
+      "epoch": 19.511434511434512,
+      "grad_norm": 0.05225607380270958,
+      "learning_rate": 4.3511906080852014e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 7333016,
+      "step": 9385
+    },
+    {
+      "epoch": 19.52182952182952,
+      "grad_norm": 0.05260981619358063,
+      "learning_rate": 4.350530649878404e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7337016,
+      "step": 9390
+    },
+    {
+      "epoch": 19.532224532224532,
+      "grad_norm": 0.04012509807944298,
+      "learning_rate": 4.3498704062965246e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7340952,
+      "step": 9395
+    },
+    {
+      "epoch": 19.542619542619544,
+      "grad_norm": 0.009017516858875751,
+      "learning_rate": 4.3492098774413815e-05,
+      "loss": 0.0335,
+      "num_input_tokens_seen": 7344760,
+      "step": 9400
+    },
+    {
+      "epoch": 19.542619542619544,
+      "eval_loss": 0.3270372152328491,
+      "eval_runtime": 11.7548,
+      "eval_samples_per_second": 72.821,
+      "eval_steps_per_second": 18.205,
+      "num_input_tokens_seen": 7344760,
+      "step": 9400
+    },
+    {
+      "epoch": 19.553014553014552,
+      "grad_norm": 0.5346773862838745,
+      "learning_rate": 4.3485490634148375e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 7348600,
+      "step": 9405
+    },
+    {
+      "epoch": 19.563409563409564,
+      "grad_norm": 0.11308003216981888,
+      "learning_rate": 4.347887964318797e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 7352536,
+      "step": 9410
+    },
+    {
+      "epoch": 19.573804573804573,
+      "grad_norm": 6.700404167175293,
+      "learning_rate": 4.34722658025521e-05,
+      "loss": 0.055,
+      "num_input_tokens_seen": 7356376,
+      "step": 9415
+    },
+    {
+      "epoch": 19.584199584199585,
+      "grad_norm": 0.022852640599012375,
+      "learning_rate": 4.346564911326071e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 7360376,
+      "step": 9420
+    },
+    {
+      "epoch": 19.594594594594593,
+      "grad_norm": 0.0008559370180591941,
+      "learning_rate": 4.345902957633418e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7364440,
+      "step": 9425
+    },
+    {
+      "epoch": 19.604989604989605,
+      "grad_norm": 0.008153757080435753,
+      "learning_rate": 4.345240719279331e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7368536,
+      "step": 9430
+    },
+    {
+      "epoch": 19.615384615384617,
+      "grad_norm": 0.0009798218961805105,
+      "learning_rate": 4.3445781963659374e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7372408,
+      "step": 9435
+    },
+    {
+      "epoch": 19.625779625779625,
+      "grad_norm": 0.0005604480975307524,
+      "learning_rate": 4.3439153889954045e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7376376,
+      "step": 9440
+    },
+    {
+      "epoch": 19.636174636174637,
+      "grad_norm": 0.0005390907172113657,
+      "learning_rate": 4.343252297269946e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 7380248,
+      "step": 9445
+    },
+    {
+      "epoch": 19.646569646569645,
+      "grad_norm": 0.0011029522866010666,
+      "learning_rate": 4.342588921291821e-05,
+      "loss": 0.009,
+      "num_input_tokens_seen": 7384152,
+      "step": 9450
+    },
+    {
+      "epoch": 19.656964656964657,
+      "grad_norm": 0.004384808707982302,
+      "learning_rate": 4.341925261163328e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 7388024,
+      "step": 9455
+    },
+    {
+      "epoch": 19.66735966735967,
+      "grad_norm": 3.778301239013672,
+      "learning_rate": 4.341261316986813e-05,
+      "loss": 0.0197,
+      "num_input_tokens_seen": 7391992,
+      "step": 9460
+    },
+    {
+      "epoch": 19.677754677754677,
+      "grad_norm": 0.02075347863137722,
+      "learning_rate": 4.340597088864664e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 7395896,
+      "step": 9465
+    },
+    {
+      "epoch": 19.68814968814969,
+      "grad_norm": 0.0073738787323236465,
+      "learning_rate": 4.339932576899313e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 7399704,
+      "step": 9470
+    },
+    {
+      "epoch": 19.698544698544698,
+      "grad_norm": 0.022166702896356583,
+      "learning_rate": 4.3392677811932375e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7403544,
+      "step": 9475
+    },
+    {
+      "epoch": 19.70893970893971,
+      "grad_norm": 0.0029769104439765215,
+      "learning_rate": 4.338602701848956e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7407320,
+      "step": 9480
+    },
+    {
+      "epoch": 19.719334719334718,
+      "grad_norm": 0.0029098275117576122,
+      "learning_rate": 4.337937338969033e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7411256,
+      "step": 9485
+    },
+    {
+      "epoch": 19.72972972972973,
+      "grad_norm": 0.020923037081956863,
+      "learning_rate": 4.337271692656075e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7415064,
+      "step": 9490
+    },
+    {
+      "epoch": 19.74012474012474,
+      "grad_norm": 0.029448986053466797,
+      "learning_rate": 4.336605763012733e-05,
+      "loss": 0.0137,
+      "num_input_tokens_seen": 7419032,
+      "step": 9495
+    },
+    {
+      "epoch": 19.75051975051975,
+      "grad_norm": 0.6414199471473694,
+      "learning_rate": 4.3359395501417026e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 7422840,
+      "step": 9500
+    },
+    {
+      "epoch": 19.760914760914762,
+      "grad_norm": 0.1569661647081375,
+      "learning_rate": 4.335273054145722e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 7426808,
+      "step": 9505
+    },
+    {
+      "epoch": 19.77130977130977,
+      "grad_norm": 6.036850452423096,
+      "learning_rate": 4.334606275127572e-05,
+      "loss": 0.0504,
+      "num_input_tokens_seen": 7430680,
+      "step": 9510
+    },
+    {
+      "epoch": 19.781704781704782,
+      "grad_norm": 0.000966002931818366,
+      "learning_rate": 4.33393921319008e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7434456,
+      "step": 9515
+    },
+    {
+      "epoch": 19.79209979209979,
+      "grad_norm": 0.016039792448282242,
+      "learning_rate": 4.3332718684361146e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7438328,
+      "step": 9520
+    },
+    {
+      "epoch": 19.802494802494802,
+      "grad_norm": 0.004024949390441179,
+      "learning_rate": 4.332604240968588e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7442200,
+      "step": 9525
+    },
+    {
+      "epoch": 19.812889812889814,
+      "grad_norm": 0.005892149638384581,
+      "learning_rate": 4.331936330890459e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 7446008,
+      "step": 9530
+    },
+    {
+      "epoch": 19.823284823284823,
+      "grad_norm": 0.0015447679907083511,
+      "learning_rate": 4.331268138304725e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7449976,
+      "step": 9535
+    },
+    {
+      "epoch": 19.833679833679835,
+      "grad_norm": 0.005089654121547937,
+      "learning_rate": 4.330599663314431e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7453784,
+      "step": 9540
+    },
+    {
+      "epoch": 19.844074844074843,
+      "grad_norm": 0.0011815736070275307,
+      "learning_rate": 4.329930906022665e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7457656,
+      "step": 9545
+    },
+    {
+      "epoch": 19.854469854469855,
+      "grad_norm": 0.9251693487167358,
+      "learning_rate": 4.3292618665325564e-05,
+      "loss": 0.0047,
+      "num_input_tokens_seen": 7461464,
+      "step": 9550
+    },
+    {
+      "epoch": 19.864864864864863,
+      "grad_norm": 0.0022294684313237667,
+      "learning_rate": 4.3285925449472796e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7465400,
+      "step": 9555
+    },
+    {
+      "epoch": 19.875259875259875,
+      "grad_norm": 0.0004831206169910729,
+      "learning_rate": 4.327922941370054e-05,
+      "loss": 0.0678,
+      "num_input_tokens_seen": 7469624,
+      "step": 9560
+    },
+    {
+      "epoch": 19.885654885654887,
+      "grad_norm": 0.0129434484988451,
+      "learning_rate": 4.3272530559041384e-05,
+      "loss": 0.034,
+      "num_input_tokens_seen": 7473464,
+      "step": 9565
+    },
+    {
+      "epoch": 19.896049896049895,
+      "grad_norm": 0.0026881112717092037,
+      "learning_rate": 4.32658288865284e-05,
+      "loss": 0.0156,
+      "num_input_tokens_seen": 7477432,
+      "step": 9570
+    },
+    {
+      "epoch": 19.906444906444907,
+      "grad_norm": 0.2697642743587494,
+      "learning_rate": 4.325912439719505e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 7481464,
+      "step": 9575
+    },
+    {
+      "epoch": 19.916839916839916,
+      "grad_norm": 0.1304720640182495,
+      "learning_rate": 4.3252417092075266e-05,
+      "loss": 0.0275,
+      "num_input_tokens_seen": 7485336,
+      "step": 9580
+    },
+    {
+      "epoch": 19.927234927234927,
+      "grad_norm": 0.013042273931205273,
+      "learning_rate": 4.3245706972203385e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 7489304,
+      "step": 9585
+    },
+    {
+      "epoch": 19.93762993762994,
+      "grad_norm": 0.04536634683609009,
+      "learning_rate": 4.323899403861421e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 7493432,
+      "step": 9590
+    },
+    {
+      "epoch": 19.948024948024948,
+      "grad_norm": 0.008753183297812939,
+      "learning_rate": 4.3232278292342935e-05,
+      "loss": 0.015,
+      "num_input_tokens_seen": 7497336,
+      "step": 9595
+    },
+    {
+      "epoch": 19.95841995841996,
+      "grad_norm": 0.018148479983210564,
+      "learning_rate": 4.322555973442524e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 7501144,
+      "step": 9600
+    },
+    {
+      "epoch": 19.95841995841996,
+      "eval_loss": 0.29398223757743835,
+      "eval_runtime": 11.7051,
+      "eval_samples_per_second": 73.13,
+      "eval_steps_per_second": 18.283,
+      "num_input_tokens_seen": 7501144,
+      "step": 9600
+    },
+    {
+      "epoch": 19.968814968814968,
+      "grad_norm": 0.25274112820625305,
+      "learning_rate": 4.3218838365897184e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 7504984,
+      "step": 9605
+    },
+    {
+      "epoch": 19.97920997920998,
+      "grad_norm": 0.02003517933189869,
+      "learning_rate": 4.3212114187795306e-05,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 7508984,
+      "step": 9610
+    },
+    {
+      "epoch": 19.989604989604988,
+      "grad_norm": 0.05211924761533737,
+      "learning_rate": 4.320538720115656e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7512888,
+      "step": 9615
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 0.006104409694671631,
+      "learning_rate": 4.319865740701831e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7516712,
+      "step": 9620
+    },
+    {
+      "epoch": 20.010395010395012,
+      "grad_norm": 0.35356977581977844,
+      "learning_rate": 4.3191924806418396e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 7520616,
+      "step": 9625
+    },
+    {
+      "epoch": 20.02079002079002,
+      "grad_norm": 0.0016144376713782549,
+      "learning_rate": 4.318518940039507e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7524552,
+      "step": 9630
+    },
+    {
+      "epoch": 20.031185031185032,
+      "grad_norm": 0.7551239728927612,
+      "learning_rate": 4.3178451189987e-05,
+      "loss": 0.0092,
+      "num_input_tokens_seen": 7528328,
+      "step": 9635
+    },
+    {
+      "epoch": 20.04158004158004,
+      "grad_norm": 0.0019132076995447278,
+      "learning_rate": 4.3171710176233315e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 7532136,
+      "step": 9640
+    },
+    {
+      "epoch": 20.051975051975052,
+      "grad_norm": 0.011274375952780247,
+      "learning_rate": 4.316496636017355e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 7535912,
+      "step": 9645
+    },
+    {
+      "epoch": 20.06237006237006,
+      "grad_norm": 2.416999578475952,
+      "learning_rate": 4.315821974284771e-05,
+      "loss": 0.0076,
+      "num_input_tokens_seen": 7539912,
+      "step": 9650
+    },
+    {
+      "epoch": 20.072765072765073,
+      "grad_norm": 0.004423514008522034,
+      "learning_rate": 4.315147032529619e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7543944,
+      "step": 9655
+    },
+    {
+      "epoch": 20.083160083160084,
+      "grad_norm": 0.015618329867720604,
+      "learning_rate": 4.3144718108559845e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7547656,
+      "step": 9660
+    },
+    {
+      "epoch": 20.093555093555093,
+      "grad_norm": 0.0008989518973976374,
+      "learning_rate": 4.3137963093679945e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7551528,
+      "step": 9665
+    },
+    {
+      "epoch": 20.103950103950105,
+      "grad_norm": 0.004108930006623268,
+      "learning_rate": 4.31312052816982e-05,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 7555528,
+      "step": 9670
+    },
+    {
+      "epoch": 20.114345114345113,
+      "grad_norm": 0.0017329229740425944,
+      "learning_rate": 4.312444467365675e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7559432,
+      "step": 9675
+    },
+    {
+      "epoch": 20.124740124740125,
+      "grad_norm": 0.016278591006994247,
+      "learning_rate": 4.311768127059816e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7563272,
+      "step": 9680
+    },
+    {
+      "epoch": 20.135135135135137,
+      "grad_norm": 0.0005227483925409615,
+      "learning_rate": 4.3110915073565444e-05,
+      "loss": 0.0529,
+      "num_input_tokens_seen": 7567144,
+      "step": 9685
+    },
+    {
+      "epoch": 20.145530145530145,
+      "grad_norm": 0.005774090997874737,
+      "learning_rate": 4.310414608360203e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7571080,
+      "step": 9690
+    },
+    {
+      "epoch": 20.155925155925157,
+      "grad_norm": 0.002052737632766366,
+      "learning_rate": 4.309737430175177e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 7574984,
+      "step": 9695
+    },
+    {
+      "epoch": 20.166320166320165,
+      "grad_norm": 0.00956485141068697,
+      "learning_rate": 4.309059972905897e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7578760,
+      "step": 9700
+    },
+    {
+      "epoch": 20.176715176715177,
+      "grad_norm": 0.0013072338188067079,
+      "learning_rate": 4.308382236656836e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 7582600,
+      "step": 9705
+    },
+    {
+      "epoch": 20.187110187110186,
+      "grad_norm": 0.003211019327864051,
+      "learning_rate": 4.307704221532507e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 7586504,
+      "step": 9710
+    },
+    {
+      "epoch": 20.197505197505198,
+      "grad_norm": 0.006205777171999216,
+      "learning_rate": 4.307025927637471e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7590344,
+      "step": 9715
+    },
+    {
+      "epoch": 20.20790020790021,
+      "grad_norm": 0.7139180898666382,
+      "learning_rate": 4.306347355076328e-05,
+      "loss": 0.0103,
+      "num_input_tokens_seen": 7594344,
+      "step": 9720
+    },
+    {
+      "epoch": 20.218295218295218,
+      "grad_norm": 0.0015376611845567822,
+      "learning_rate": 4.305668503953724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7598312,
+      "step": 9725
+    },
+    {
+      "epoch": 20.22869022869023,
+      "grad_norm": 0.000662686419673264,
+      "learning_rate": 4.3049893743743436e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7602152,
+      "step": 9730
+    },
+    {
+      "epoch": 20.239085239085238,
+      "grad_norm": 0.06495756655931473,
+      "learning_rate": 4.304309966442919e-05,
+      "loss": 0.0241,
+      "num_input_tokens_seen": 7606088,
+      "step": 9735
+    },
+    {
+      "epoch": 20.24948024948025,
+      "grad_norm": 0.0008402467938140035,
+      "learning_rate": 4.303630280264224e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7610056,
+      "step": 9740
+    },
+    {
+      "epoch": 20.25987525987526,
+      "grad_norm": 0.0012562820920720696,
+      "learning_rate": 4.302950315943074e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7613960,
+      "step": 9745
+    },
+    {
+      "epoch": 20.27027027027027,
+      "grad_norm": 0.042528197169303894,
+      "learning_rate": 4.3022700735843275e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7617896,
+      "step": 9750
+    },
+    {
+      "epoch": 20.280665280665282,
+      "grad_norm": 0.004081972874701023,
+      "learning_rate": 4.301589553292887e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7621768,
+      "step": 9755
+    },
+    {
+      "epoch": 20.29106029106029,
+      "grad_norm": 0.0010060222120955586,
+      "learning_rate": 4.300908755173697e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7625960,
+      "step": 9760
+    },
+    {
+      "epoch": 20.301455301455302,
+      "grad_norm": 0.000550690689124167,
+      "learning_rate": 4.300227679331745e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 7629896,
+      "step": 9765
+    },
+    {
+      "epoch": 20.31185031185031,
+      "grad_norm": 5.239345550537109,
+      "learning_rate": 4.299546325872063e-05,
+      "loss": 0.0281,
+      "num_input_tokens_seen": 7633896,
+      "step": 9770
+    },
+    {
+      "epoch": 20.322245322245323,
+      "grad_norm": 3.378838300704956,
+      "learning_rate": 4.2988646948997225e-05,
+      "loss": 0.0093,
+      "num_input_tokens_seen": 7637864,
+      "step": 9775
+    },
+    {
+      "epoch": 20.33264033264033,
+      "grad_norm": 0.00798058696091175,
+      "learning_rate": 4.29818278651984e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7641800,
+      "step": 9780
+    },
+    {
+      "epoch": 20.343035343035343,
+      "grad_norm": 0.0008979014819487929,
+      "learning_rate": 4.297500600837574e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7645672,
+      "step": 9785
+    },
+    {
+      "epoch": 20.353430353430355,
+      "grad_norm": 0.7875456809997559,
+      "learning_rate": 4.2968181379581276e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 7649512,
+      "step": 9790
+    },
+    {
+      "epoch": 20.363825363825363,
+      "grad_norm": 0.0674014613032341,
+      "learning_rate": 4.296135397986743e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7653320,
+      "step": 9795
+    },
+    {
+      "epoch": 20.374220374220375,
+      "grad_norm": 0.1976960152387619,
+      "learning_rate": 4.295452381028709e-05,
+      "loss": 0.029,
+      "num_input_tokens_seen": 7657160,
+      "step": 9800
+    },
+    {
+      "epoch": 20.374220374220375,
+      "eval_loss": 0.2817394733428955,
+      "eval_runtime": 11.7137,
+      "eval_samples_per_second": 73.077,
+      "eval_steps_per_second": 18.269,
+      "num_input_tokens_seen": 7657160,
+      "step": 9800
+    },
+    {
+      "epoch": 20.384615384615383,
+      "grad_norm": 0.003965442534536123,
+      "learning_rate": 4.294769087189354e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7661032,
+      "step": 9805
+    },
+    {
+      "epoch": 20.395010395010395,
+      "grad_norm": 0.11588193476200104,
+      "learning_rate": 4.294085516574052e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7664904,
+      "step": 9810
+    },
+    {
+      "epoch": 20.405405405405407,
+      "grad_norm": 0.015173057094216347,
+      "learning_rate": 4.2934016692882176e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7668872,
+      "step": 9815
+    },
+    {
+      "epoch": 20.415800415800415,
+      "grad_norm": 0.01588415540754795,
+      "learning_rate": 4.292717545437308e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7672776,
+      "step": 9820
+    },
+    {
+      "epoch": 20.426195426195427,
+      "grad_norm": 0.02777923084795475,
+      "learning_rate": 4.292033145126825e-05,
+      "loss": 0.0257,
+      "num_input_tokens_seen": 7676584,
+      "step": 9825
+    },
+    {
+      "epoch": 20.436590436590436,
+      "grad_norm": 0.04639751464128494,
+      "learning_rate": 4.29134846846231e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 7680456,
+      "step": 9830
+    },
+    {
+      "epoch": 20.446985446985448,
+      "grad_norm": 0.018501723185181618,
+      "learning_rate": 4.29066351554935e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7684360,
+      "step": 9835
+    },
+    {
+      "epoch": 20.457380457380456,
+      "grad_norm": 0.020827632397413254,
+      "learning_rate": 4.289978286493574e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7688360,
+      "step": 9840
+    },
+    {
+      "epoch": 20.467775467775468,
+      "grad_norm": 0.0032526308204978704,
+      "learning_rate": 4.28929278140065e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7692392,
+      "step": 9845
+    },
+    {
+      "epoch": 20.47817047817048,
+      "grad_norm": 0.008946762420237064,
+      "learning_rate": 4.288607000376295e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7696328,
+      "step": 9850
+    },
+    {
+      "epoch": 20.488565488565488,
+      "grad_norm": 1.0407580137252808,
+      "learning_rate": 4.2879209435262624e-05,
+      "loss": 0.0113,
+      "num_input_tokens_seen": 7700328,
+      "step": 9855
+    },
+    {
+      "epoch": 20.4989604989605,
+      "grad_norm": 0.08180486410856247,
+      "learning_rate": 4.287234610956353e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7704232,
+      "step": 9860
+    },
+    {
+      "epoch": 20.509355509355508,
+      "grad_norm": 0.0032193181104958057,
+      "learning_rate": 4.2865480027724056e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7708200,
+      "step": 9865
+    },
+    {
+      "epoch": 20.51975051975052,
+      "grad_norm": 0.006638789549469948,
+      "learning_rate": 4.285861119080306e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7712040,
+      "step": 9870
+    },
+    {
+      "epoch": 20.53014553014553,
+      "grad_norm": 0.0022654649801552296,
+      "learning_rate": 4.2851739599859784e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7716040,
+      "step": 9875
+    },
+    {
+      "epoch": 20.54054054054054,
+      "grad_norm": 0.006353930104523897,
+      "learning_rate": 4.2844865255953934e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7719944,
+      "step": 9880
+    },
+    {
+      "epoch": 20.550935550935552,
+      "grad_norm": 0.007507271133363247,
+      "learning_rate": 4.2837988160145605e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7723656,
+      "step": 9885
+    },
+    {
+      "epoch": 20.56133056133056,
+      "grad_norm": 0.0014374034944921732,
+      "learning_rate": 4.2831108313495336e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7727528,
+      "step": 9890
+    },
+    {
+      "epoch": 20.571725571725572,
+      "grad_norm": 0.003429212374612689,
+      "learning_rate": 4.282422571706408e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7731400,
+      "step": 9895
+    },
+    {
+      "epoch": 20.58212058212058,
+      "grad_norm": 0.0021545749623328447,
+      "learning_rate": 4.281734037191323e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7735208,
+      "step": 9900
+    },
+    {
+      "epoch": 20.592515592515593,
+      "grad_norm": 10.110888481140137,
+      "learning_rate": 4.281045227910459e-05,
+      "loss": 0.0661,
+      "num_input_tokens_seen": 7739080,
+      "step": 9905
+    },
+    {
+      "epoch": 20.602910602910605,
+      "grad_norm": 1.5142184495925903,
+      "learning_rate": 4.280356143970038e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7743080,
+      "step": 9910
+    },
+    {
+      "epoch": 20.613305613305613,
+      "grad_norm": 0.03431055694818497,
+      "learning_rate": 4.279666785476327e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7747080,
+      "step": 9915
+    },
+    {
+      "epoch": 20.623700623700625,
+      "grad_norm": 0.02889537811279297,
+      "learning_rate": 4.2789771525356325e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 7751016,
+      "step": 9920
+    },
+    {
+      "epoch": 20.634095634095633,
+      "grad_norm": 0.3675740659236908,
+      "learning_rate": 4.2782872452543056e-05,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 7754920,
+      "step": 9925
+    },
+    {
+      "epoch": 20.644490644490645,
+      "grad_norm": 0.004927989095449448,
+      "learning_rate": 4.2775970637387376e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7758984,
+      "step": 9930
+    },
+    {
+      "epoch": 20.654885654885653,
+      "grad_norm": 0.007411003112792969,
+      "learning_rate": 4.276906608095363e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7762856,
+      "step": 9935
+    },
+    {
+      "epoch": 20.665280665280665,
+      "grad_norm": 0.0014744573272764683,
+      "learning_rate": 4.276215878430661e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7766568,
+      "step": 9940
+    },
+    {
+      "epoch": 20.675675675675677,
+      "grad_norm": 0.005334658548235893,
+      "learning_rate": 4.275524874851149e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7770440,
+      "step": 9945
+    },
+    {
+      "epoch": 20.686070686070686,
+      "grad_norm": 0.29518118500709534,
+      "learning_rate": 4.274833597463388e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 7774440,
+      "step": 9950
+    },
+    {
+      "epoch": 20.696465696465697,
+      "grad_norm": 0.007380020339041948,
+      "learning_rate": 4.2741420463739824e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 7778472,
+      "step": 9955
+    },
+    {
+      "epoch": 20.706860706860706,
+      "grad_norm": 0.003582435194402933,
+      "learning_rate": 4.273450221689578e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 7782472,
+      "step": 9960
+    },
+    {
+      "epoch": 20.717255717255718,
+      "grad_norm": 0.0014012668980285525,
+      "learning_rate": 4.272758123516863e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7786376,
+      "step": 9965
+    },
+    {
+      "epoch": 20.727650727650726,
+      "grad_norm": 0.0005345281097106636,
+      "learning_rate": 4.272065751962567e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7790184,
+      "step": 9970
+    },
+    {
+      "epoch": 20.738045738045738,
+      "grad_norm": 0.0006469987565651536,
+      "learning_rate": 4.271373107133464e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7794056,
+      "step": 9975
+    },
+    {
+      "epoch": 20.74844074844075,
+      "grad_norm": 0.008332164026796818,
+      "learning_rate": 4.270680189136366e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7797864,
+      "step": 9980
+    },
+    {
+      "epoch": 20.758835758835758,
+      "grad_norm": 0.0003797918907366693,
+      "learning_rate": 4.269986998078132e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 7801672,
+      "step": 9985
+    },
+    {
+      "epoch": 20.76923076923077,
+      "grad_norm": 0.0007952759624458849,
+      "learning_rate": 4.2692935340656595e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7805512,
+      "step": 9990
+    },
+    {
+      "epoch": 20.77962577962578,
+      "grad_norm": 0.011573133058845997,
+      "learning_rate": 4.26859979720589e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7809288,
+      "step": 9995
+    },
+    {
+      "epoch": 20.79002079002079,
+      "grad_norm": 0.000853057368658483,
+      "learning_rate": 4.267905787605806e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7813128,
+      "step": 10000
+    },
+    {
+      "epoch": 20.79002079002079,
+      "eval_loss": 0.45024314522743225,
+      "eval_runtime": 11.6927,
+      "eval_samples_per_second": 73.208,
+      "eval_steps_per_second": 18.302,
+      "num_input_tokens_seen": 7813128,
+      "step": 10000
+    },
+    {
+      "epoch": 20.8004158004158,
+      "grad_norm": 0.00010054479935206473,
+      "learning_rate": 4.267211505372433e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7817064,
+      "step": 10005
+    },
+    {
+      "epoch": 20.81081081081081,
+      "grad_norm": 0.001086555770598352,
+      "learning_rate": 4.266516950612837e-05,
+      "loss": 0.0074,
+      "num_input_tokens_seen": 7821064,
+      "step": 10010
+    },
+    {
+      "epoch": 20.821205821205822,
+      "grad_norm": 0.332743376493454,
+      "learning_rate": 4.265822123434128e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 7824968,
+      "step": 10015
+    },
+    {
+      "epoch": 20.83160083160083,
+      "grad_norm": 0.0005716517334803939,
+      "learning_rate": 4.265127023943457e-05,
+      "loss": 0.0263,
+      "num_input_tokens_seen": 7828936,
+      "step": 10020
+    },
+    {
+      "epoch": 20.841995841995843,
+      "grad_norm": 1.2713234424591064,
+      "learning_rate": 4.2644316522480176e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7832872,
+      "step": 10025
+    },
+    {
+      "epoch": 20.85239085239085,
+      "grad_norm": 0.0012621866771951318,
+      "learning_rate": 4.263736008455044e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7836872,
+      "step": 10030
+    },
+    {
+      "epoch": 20.862785862785863,
+      "grad_norm": 0.0020855714101344347,
+      "learning_rate": 4.2630400926718125e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7840840,
+      "step": 10035
+    },
+    {
+      "epoch": 20.873180873180875,
+      "grad_norm": 0.0009193101432174444,
+      "learning_rate": 4.262343905005644e-05,
+      "loss": 0.0145,
+      "num_input_tokens_seen": 7844744,
+      "step": 10040
+    },
+    {
+      "epoch": 20.883575883575883,
+      "grad_norm": 0.0011492939665913582,
+      "learning_rate": 4.261647445563897e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7848776,
+      "step": 10045
+    },
+    {
+      "epoch": 20.893970893970895,
+      "grad_norm": 0.0032093185000121593,
+      "learning_rate": 4.260950714453976e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7852648,
+      "step": 10050
+    },
+    {
+      "epoch": 20.904365904365903,
+      "grad_norm": 0.009996030479669571,
+      "learning_rate": 4.2602537117833266e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 7856648,
+      "step": 10055
+    },
+    {
+      "epoch": 20.914760914760915,
+      "grad_norm": 0.0007567820139229298,
+      "learning_rate": 4.259556437659433e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7860488,
+      "step": 10060
+    },
+    {
+      "epoch": 20.925155925155924,
+      "grad_norm": 0.004969093482941389,
+      "learning_rate": 4.258858892189825e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 7864552,
+      "step": 10065
+    },
+    {
+      "epoch": 20.935550935550935,
+      "grad_norm": 0.005523585714399815,
+      "learning_rate": 4.2581610754820725e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7868296,
+      "step": 10070
+    },
+    {
+      "epoch": 20.945945945945947,
+      "grad_norm": 0.002939720405265689,
+      "learning_rate": 4.2574629876437876e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7872328,
+      "step": 10075
+    },
+    {
+      "epoch": 20.956340956340956,
+      "grad_norm": 0.005195950623601675,
+      "learning_rate": 4.256764628782625e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7876136,
+      "step": 10080
+    },
+    {
+      "epoch": 20.966735966735968,
+      "grad_norm": 0.00015745604468975216,
+      "learning_rate": 4.256065999006279e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7880200,
+      "step": 10085
+    },
+    {
+      "epoch": 20.977130977130976,
+      "grad_norm": 0.02083468623459339,
+      "learning_rate": 4.2553670984224885e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7884040,
+      "step": 10090
+    },
+    {
+      "epoch": 20.987525987525988,
+      "grad_norm": 3.666426658630371,
+      "learning_rate": 4.254667927139032e-05,
+      "loss": 0.0348,
+      "num_input_tokens_seen": 7888008,
+      "step": 10095
+    },
+    {
+      "epoch": 20.997920997921,
+      "grad_norm": 0.002642866922542453,
+      "learning_rate": 4.2539684852637295e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7892008,
+      "step": 10100
+    },
+    {
+      "epoch": 21.008316008316008,
+      "grad_norm": 0.006621617823839188,
+      "learning_rate": 4.253268772904446e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7895800,
+      "step": 10105
+    },
+    {
+      "epoch": 21.01871101871102,
+      "grad_norm": 0.039299752563238144,
+      "learning_rate": 4.252568790169085e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7899608,
+      "step": 10110
+    },
+    {
+      "epoch": 21.02910602910603,
+      "grad_norm": 0.507519006729126,
+      "learning_rate": 4.251868537165592e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 7903800,
+      "step": 10115
+    },
+    {
+      "epoch": 21.03950103950104,
+      "grad_norm": 0.0041491324082016945,
+      "learning_rate": 4.251168014001955e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7907768,
+      "step": 10120
+    },
+    {
+      "epoch": 21.04989604989605,
+      "grad_norm": 0.023151593282818794,
+      "learning_rate": 4.250467220786204e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7911608,
+      "step": 10125
+    },
+    {
+      "epoch": 21.06029106029106,
+      "grad_norm": 0.004172730725258589,
+      "learning_rate": 4.249766157626409e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 7915544,
+      "step": 10130
+    },
+    {
+      "epoch": 21.070686070686072,
+      "grad_norm": 0.0013385401107370853,
+      "learning_rate": 4.249064824630684e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7919416,
+      "step": 10135
+    },
+    {
+      "epoch": 21.08108108108108,
+      "grad_norm": 0.012423397041857243,
+      "learning_rate": 4.248363221907183e-05,
+      "loss": 0.0069,
+      "num_input_tokens_seen": 7923320,
+      "step": 10140
+    },
+    {
+      "epoch": 21.091476091476093,
+      "grad_norm": 0.001316474168561399,
+      "learning_rate": 4.2476613495641026e-05,
+      "loss": 0.0249,
+      "num_input_tokens_seen": 7927160,
+      "step": 10145
+    },
+    {
+      "epoch": 21.1018711018711,
+      "grad_norm": 0.004975050687789917,
+      "learning_rate": 4.246959207709679e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 7931032,
+      "step": 10150
+    },
+    {
+      "epoch": 21.112266112266113,
+      "grad_norm": 0.0034241792745888233,
+      "learning_rate": 4.246256796452192e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 7935096,
+      "step": 10155
+    },
+    {
+      "epoch": 21.12266112266112,
+      "grad_norm": 0.02398431859910488,
+      "learning_rate": 4.245554115899962e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7938936,
+      "step": 10160
+    },
+    {
+      "epoch": 21.133056133056133,
+      "grad_norm": 0.05739543214440346,
+      "learning_rate": 4.2448511661613514e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 7942936,
+      "step": 10165
+    },
+    {
+      "epoch": 21.143451143451145,
+      "grad_norm": 0.4030435383319855,
+      "learning_rate": 4.2441479473447635e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 7946776,
+      "step": 10170
+    },
+    {
+      "epoch": 21.153846153846153,
+      "grad_norm": 0.05861431360244751,
+      "learning_rate": 4.243444459558644e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7950648,
+      "step": 10175
+    },
+    {
+      "epoch": 21.164241164241165,
+      "grad_norm": 0.05785026773810387,
+      "learning_rate": 4.24274070291148e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 7954584,
+      "step": 10180
+    },
+    {
+      "epoch": 21.174636174636174,
+      "grad_norm": 0.02645622193813324,
+      "learning_rate": 4.242036677511798e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 7958456,
+      "step": 10185
+    },
+    {
+      "epoch": 21.185031185031185,
+      "grad_norm": 0.0005858977092429996,
+      "learning_rate": 4.241332383468169e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 7962232,
+      "step": 10190
+    },
+    {
+      "epoch": 21.195426195426194,
+      "grad_norm": 0.04568706825375557,
+      "learning_rate": 4.2406278208892034e-05,
+      "loss": 0.028,
+      "num_input_tokens_seen": 7966104,
+      "step": 10195
+    },
+    {
+      "epoch": 21.205821205821206,
+      "grad_norm": 0.5743926167488098,
+      "learning_rate": 4.2399229898835536e-05,
+      "loss": 0.0196,
+      "num_input_tokens_seen": 7969880,
+      "step": 10200
+    },
+    {
+      "epoch": 21.205821205821206,
+      "eval_loss": 0.39671412110328674,
+      "eval_runtime": 11.7031,
+      "eval_samples_per_second": 73.143,
+      "eval_steps_per_second": 18.286,
+      "num_input_tokens_seen": 7969880,
+      "step": 10200
+    },
+    {
+      "epoch": 21.216216216216218,
+      "grad_norm": 0.0008650927920825779,
+      "learning_rate": 4.239217890559914e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7973880,
+      "step": 10205
+    },
+    {
+      "epoch": 21.226611226611226,
+      "grad_norm": 0.0005381290684454143,
+      "learning_rate": 4.238512523027019e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7977784,
+      "step": 10210
+    },
+    {
+      "epoch": 21.237006237006238,
+      "grad_norm": 0.00048651572433300316,
+      "learning_rate": 4.237806887393645e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 7981752,
+      "step": 10215
+    },
+    {
+      "epoch": 21.247401247401246,
+      "grad_norm": 0.01822696626186371,
+      "learning_rate": 4.237100983768611e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 7985720,
+      "step": 10220
+    },
+    {
+      "epoch": 21.257796257796258,
+      "grad_norm": 0.11147459596395493,
+      "learning_rate": 4.2363948122607756e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 7989592,
+      "step": 10225
+    },
+    {
+      "epoch": 21.26819126819127,
+      "grad_norm": 0.0014612386003136635,
+      "learning_rate": 4.235688372979039e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 7993592,
+      "step": 10230
+    },
+    {
+      "epoch": 21.27858627858628,
+      "grad_norm": 2.857466697692871,
+      "learning_rate": 4.234981666032343e-05,
+      "loss": 0.0381,
+      "num_input_tokens_seen": 7997432,
+      "step": 10235
+    },
+    {
+      "epoch": 21.28898128898129,
+      "grad_norm": 0.0012440710561349988,
+      "learning_rate": 4.2342746915296704e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8001208,
+      "step": 10240
+    },
+    {
+      "epoch": 21.2993762993763,
+      "grad_norm": 0.005547380540519953,
+      "learning_rate": 4.233567449580047e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8004984,
+      "step": 10245
+    },
+    {
+      "epoch": 21.30977130977131,
+      "grad_norm": 0.010459447279572487,
+      "learning_rate": 4.232859940292537e-05,
+      "loss": 0.0246,
+      "num_input_tokens_seen": 8008984,
+      "step": 10250
+    },
+    {
+      "epoch": 21.32016632016632,
+      "grad_norm": 0.009231969714164734,
+      "learning_rate": 4.232152163776248e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8012888,
+      "step": 10255
+    },
+    {
+      "epoch": 21.33056133056133,
+      "grad_norm": 0.6765057444572449,
+      "learning_rate": 4.231444120140328e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 8016696,
+      "step": 10260
+    },
+    {
+      "epoch": 21.340956340956343,
+      "grad_norm": 0.017847290262579918,
+      "learning_rate": 4.230735809493967e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8020536,
+      "step": 10265
+    },
+    {
+      "epoch": 21.35135135135135,
+      "grad_norm": 0.001976384548470378,
+      "learning_rate": 4.2300272319463926e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 8024408,
+      "step": 10270
+    },
+    {
+      "epoch": 21.361746361746363,
+      "grad_norm": 0.004378458019345999,
+      "learning_rate": 4.2293183876068786e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8028344,
+      "step": 10275
+    },
+    {
+      "epoch": 21.37214137214137,
+      "grad_norm": 0.005861243698745966,
+      "learning_rate": 4.228609276584737e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8032184,
+      "step": 10280
+    },
+    {
+      "epoch": 21.382536382536383,
+      "grad_norm": 0.006321617402136326,
+      "learning_rate": 4.227899898989323e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8036024,
+      "step": 10285
+    },
+    {
+      "epoch": 21.39293139293139,
+      "grad_norm": 0.0026212893426418304,
+      "learning_rate": 4.2271902549300293e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8040152,
+      "step": 10290
+    },
+    {
+      "epoch": 21.403326403326403,
+      "grad_norm": 0.31769371032714844,
+      "learning_rate": 4.226480344516294e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 8043896,
+      "step": 10295
+    },
+    {
+      "epoch": 21.413721413721415,
+      "grad_norm": 0.003455405356362462,
+      "learning_rate": 4.2257701678575925e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8047896,
+      "step": 10300
+    },
+    {
+      "epoch": 21.424116424116423,
+      "grad_norm": 2.646965265274048,
+      "learning_rate": 4.225059725063444e-05,
+      "loss": 0.0366,
+      "num_input_tokens_seen": 8051736,
+      "step": 10305
+    },
+    {
+      "epoch": 21.434511434511435,
+      "grad_norm": 0.0008555745589546859,
+      "learning_rate": 4.2243490162434074e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8055608,
+      "step": 10310
+    },
+    {
+      "epoch": 21.444906444906444,
+      "grad_norm": 0.011387009173631668,
+      "learning_rate": 4.223638041507083e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8059480,
+      "step": 10315
+    },
+    {
+      "epoch": 21.455301455301456,
+      "grad_norm": 0.019291790202260017,
+      "learning_rate": 4.2229268009641124e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 8063448,
+      "step": 10320
+    },
+    {
+      "epoch": 21.465696465696467,
+      "grad_norm": 0.008640328422188759,
+      "learning_rate": 4.222215294724177e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 8067608,
+      "step": 10325
+    },
+    {
+      "epoch": 21.476091476091476,
+      "grad_norm": 0.013463953509926796,
+      "learning_rate": 4.2215035228970005e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 8071640,
+      "step": 10330
+    },
+    {
+      "epoch": 21.486486486486488,
+      "grad_norm": 0.004780484363436699,
+      "learning_rate": 4.2207914855923464e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8075576,
+      "step": 10335
+    },
+    {
+      "epoch": 21.496881496881496,
+      "grad_norm": 0.0033248027320951223,
+      "learning_rate": 4.220079182920021e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 8079544,
+      "step": 10340
+    },
+    {
+      "epoch": 21.507276507276508,
+      "grad_norm": 0.003869118634611368,
+      "learning_rate": 4.2193666149898705e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8083640,
+      "step": 10345
+    },
+    {
+      "epoch": 21.517671517671516,
+      "grad_norm": 0.003508581081405282,
+      "learning_rate": 4.21865378191178e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8087544,
+      "step": 10350
+    },
+    {
+      "epoch": 21.528066528066528,
+      "grad_norm": 0.0021643151994794607,
+      "learning_rate": 4.217940683795678e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8091480,
+      "step": 10355
+    },
+    {
+      "epoch": 21.53846153846154,
+      "grad_norm": 0.001493962248787284,
+      "learning_rate": 4.217227320751534e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8095416,
+      "step": 10360
+    },
+    {
+      "epoch": 21.54885654885655,
+      "grad_norm": 0.0016686692833900452,
+      "learning_rate": 4.216513692889358e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 8099416,
+      "step": 10365
+    },
+    {
+      "epoch": 21.55925155925156,
+      "grad_norm": 0.007015164941549301,
+      "learning_rate": 4.215799800319199e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 8103352,
+      "step": 10370
+    },
+    {
+      "epoch": 21.56964656964657,
+      "grad_norm": 0.017890840768814087,
+      "learning_rate": 4.2150856431511485e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8107128,
+      "step": 10375
+    },
+    {
+      "epoch": 21.58004158004158,
+      "grad_norm": 0.00245169666595757,
+      "learning_rate": 4.214371221495339e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 8111000,
+      "step": 10380
+    },
+    {
+      "epoch": 21.59043659043659,
+      "grad_norm": 0.0008147587650455534,
+      "learning_rate": 4.213656535461942e-05,
+      "loss": 0.0222,
+      "num_input_tokens_seen": 8114872,
+      "step": 10385
+    },
+    {
+      "epoch": 21.6008316008316,
+      "grad_norm": 0.004925231449306011,
+      "learning_rate": 4.2129415851611734e-05,
+      "loss": 0.0069,
+      "num_input_tokens_seen": 8118744,
+      "step": 10390
+    },
+    {
+      "epoch": 21.611226611226613,
+      "grad_norm": 0.0034758802503347397,
+      "learning_rate": 4.2122263707032855e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8122552,
+      "step": 10395
+    },
+    {
+      "epoch": 21.62162162162162,
+      "grad_norm": 0.05629570037126541,
+      "learning_rate": 4.211510892198574e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 8126392,
+      "step": 10400
+    },
+    {
+      "epoch": 21.62162162162162,
+      "eval_loss": 0.32306256890296936,
+      "eval_runtime": 11.752,
+      "eval_samples_per_second": 72.839,
+      "eval_steps_per_second": 18.21,
+      "num_input_tokens_seen": 8126392,
+      "step": 10400
+    },
+    {
+      "epoch": 21.632016632016633,
+      "grad_norm": 0.05625100061297417,
+      "learning_rate": 4.210795149757375e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 8130456,
+      "step": 10405
+    },
+    {
+      "epoch": 21.64241164241164,
+      "grad_norm": 0.00840848870575428,
+      "learning_rate": 4.210079143490065e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 8134424,
+      "step": 10410
+    },
+    {
+      "epoch": 21.652806652806653,
+      "grad_norm": 0.03980661928653717,
+      "learning_rate": 4.2093628735070604e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 8138392,
+      "step": 10415
+    },
+    {
+      "epoch": 21.66320166320166,
+      "grad_norm": 0.004391055088490248,
+      "learning_rate": 4.208646339918819e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 8142296,
+      "step": 10420
+    },
+    {
+      "epoch": 21.673596673596673,
+      "grad_norm": 0.023647146299481392,
+      "learning_rate": 4.2079295428358414e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8146072,
+      "step": 10425
+    },
+    {
+      "epoch": 21.683991683991685,
+      "grad_norm": 0.889594316482544,
+      "learning_rate": 4.207212482368664e-05,
+      "loss": 0.0129,
+      "num_input_tokens_seen": 8150008,
+      "step": 10430
+    },
+    {
+      "epoch": 21.694386694386694,
+      "grad_norm": 0.010001474991440773,
+      "learning_rate": 4.206495158627867e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8153912,
+      "step": 10435
+    },
+    {
+      "epoch": 21.704781704781706,
+      "grad_norm": 0.0016635637730360031,
+      "learning_rate": 4.205777571724073e-05,
+      "loss": 0.0114,
+      "num_input_tokens_seen": 8157784,
+      "step": 10440
+    },
+    {
+      "epoch": 21.715176715176714,
+      "grad_norm": 0.012015985324978828,
+      "learning_rate": 4.20505972176794e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8161560,
+      "step": 10445
+    },
+    {
+      "epoch": 21.725571725571726,
+      "grad_norm": 0.0029093080665916204,
+      "learning_rate": 4.204341608870171e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8165624,
+      "step": 10450
+    },
+    {
+      "epoch": 21.735966735966738,
+      "grad_norm": 0.0054401978850364685,
+      "learning_rate": 4.203623233141508e-05,
+      "loss": 0.0191,
+      "num_input_tokens_seen": 8169496,
+      "step": 10455
+    },
+    {
+      "epoch": 21.746361746361746,
+      "grad_norm": 0.02350405417382717,
+      "learning_rate": 4.2029045946927334e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 8173304,
+      "step": 10460
+    },
+    {
+      "epoch": 21.756756756756758,
+      "grad_norm": 0.022608410567045212,
+      "learning_rate": 4.20218569363467e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 8177080,
+      "step": 10465
+    },
+    {
+      "epoch": 21.767151767151766,
+      "grad_norm": 0.004833284765481949,
+      "learning_rate": 4.2014665300781834e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 8181048,
+      "step": 10470
+    },
+    {
+      "epoch": 21.777546777546778,
+      "grad_norm": 0.0007255114032886922,
+      "learning_rate": 4.200747104134174e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8184856,
+      "step": 10475
+    },
+    {
+      "epoch": 21.787941787941786,
+      "grad_norm": 0.016469072550535202,
+      "learning_rate": 4.200027415913588e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8188824,
+      "step": 10480
+    },
+    {
+      "epoch": 21.7983367983368,
+      "grad_norm": 0.0015851228963583708,
+      "learning_rate": 4.1993074655274126e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8192760,
+      "step": 10485
+    },
+    {
+      "epoch": 21.80873180873181,
+      "grad_norm": 0.037705518305301666,
+      "learning_rate": 4.198587253086669e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 8196536,
+      "step": 10490
+    },
+    {
+      "epoch": 21.81912681912682,
+      "grad_norm": 0.44079655408859253,
+      "learning_rate": 4.197866778702426e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 8200504,
+      "step": 10495
+    },
+    {
+      "epoch": 21.82952182952183,
+      "grad_norm": 0.010769295506179333,
+      "learning_rate": 4.197146042485789e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8204312,
+      "step": 10500
+    },
+    {
+      "epoch": 21.83991683991684,
+      "grad_norm": 0.008694740012288094,
+      "learning_rate": 4.1964250445479046e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 8208280,
+      "step": 10505
+    },
+    {
+      "epoch": 21.85031185031185,
+      "grad_norm": 0.008870681747794151,
+      "learning_rate": 4.19570378499996e-05,
+      "loss": 0.0138,
+      "num_input_tokens_seen": 8212216,
+      "step": 10510
+    },
+    {
+      "epoch": 21.86070686070686,
+      "grad_norm": 0.01884889043867588,
+      "learning_rate": 4.194982263953182e-05,
+      "loss": 0.0178,
+      "num_input_tokens_seen": 8216184,
+      "step": 10515
+    },
+    {
+      "epoch": 21.87110187110187,
+      "grad_norm": 0.007196555379778147,
+      "learning_rate": 4.194260481518838e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 8219992,
+      "step": 10520
+    },
+    {
+      "epoch": 21.881496881496883,
+      "grad_norm": 0.010208241641521454,
+      "learning_rate": 4.1935384378082366e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8224056,
+      "step": 10525
+    },
+    {
+      "epoch": 21.89189189189189,
+      "grad_norm": 0.0014961560955271125,
+      "learning_rate": 4.1928161329327267e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8227960,
+      "step": 10530
+    },
+    {
+      "epoch": 21.902286902286903,
+      "grad_norm": 0.07921494543552399,
+      "learning_rate": 4.1920935670036945e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 8231864,
+      "step": 10535
+    },
+    {
+      "epoch": 21.91268191268191,
+      "grad_norm": 0.0004476365866139531,
+      "learning_rate": 4.1913707401325705e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8235800,
+      "step": 10540
+    },
+    {
+      "epoch": 21.923076923076923,
+      "grad_norm": 0.001265322440303862,
+      "learning_rate": 4.1906476524308235e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8239640,
+      "step": 10545
+    },
+    {
+      "epoch": 21.933471933471935,
+      "grad_norm": 0.03338364511728287,
+      "learning_rate": 4.189924304009962e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8243544,
+      "step": 10550
+    },
+    {
+      "epoch": 21.943866943866944,
+      "grad_norm": 0.000386355328373611,
+      "learning_rate": 4.189200694981537e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8247544,
+      "step": 10555
+    },
+    {
+      "epoch": 21.954261954261955,
+      "grad_norm": 0.000376728770788759,
+      "learning_rate": 4.188476825457136e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8251480,
+      "step": 10560
+    },
+    {
+      "epoch": 21.964656964656964,
+      "grad_norm": 0.0003988232638221234,
+      "learning_rate": 4.18775269554839e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8255384,
+      "step": 10565
+    },
+    {
+      "epoch": 21.975051975051976,
+      "grad_norm": 0.0019345612963661551,
+      "learning_rate": 4.187028305366969e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8259192,
+      "step": 10570
+    },
+    {
+      "epoch": 21.985446985446984,
+      "grad_norm": 0.004325455985963345,
+      "learning_rate": 4.1863036550245824e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 8263000,
+      "step": 10575
+    },
+    {
+      "epoch": 21.995841995841996,
+      "grad_norm": 0.001980789937078953,
+      "learning_rate": 4.1855787446329806e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8266808,
+      "step": 10580
+    },
+    {
+      "epoch": 22.006237006237008,
+      "grad_norm": 0.0016170068411156535,
+      "learning_rate": 4.184853574303955e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8270672,
+      "step": 10585
+    },
+    {
+      "epoch": 22.016632016632016,
+      "grad_norm": 0.0004887114628218114,
+      "learning_rate": 4.184128144149334e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8274576,
+      "step": 10590
+    },
+    {
+      "epoch": 22.027027027027028,
+      "grad_norm": 0.001893051783554256,
+      "learning_rate": 4.1834024542809896e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8278416,
+      "step": 10595
+    },
+    {
+      "epoch": 22.037422037422036,
+      "grad_norm": 0.0034373116213828325,
+      "learning_rate": 4.1826765048108315e-05,
+      "loss": 0.0102,
+      "num_input_tokens_seen": 8282480,
+      "step": 10600
+    },
+    {
+      "epoch": 22.037422037422036,
+      "eval_loss": 0.4725196361541748,
+      "eval_runtime": 11.704,
+      "eval_samples_per_second": 73.138,
+      "eval_steps_per_second": 18.284,
+      "num_input_tokens_seen": 8282480,
+      "step": 10600
+    },
+    {
+      "epoch": 22.04781704781705,
+      "grad_norm": 0.0007486446411348879,
+      "learning_rate": 4.181950295850811e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8286352,
+      "step": 10605
+    },
+    {
+      "epoch": 22.058212058212057,
+      "grad_norm": 0.002153553068637848,
+      "learning_rate": 4.181223827512918e-05,
+      "loss": 0.0175,
+      "num_input_tokens_seen": 8290128,
+      "step": 10610
+    },
+    {
+      "epoch": 22.06860706860707,
+      "grad_norm": 0.013942810706794262,
+      "learning_rate": 4.180497099909183e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 8294096,
+      "step": 10615
+    },
+    {
+      "epoch": 22.07900207900208,
+      "grad_norm": 0.0031870054081082344,
+      "learning_rate": 4.179770113151677e-05,
+      "loss": 0.0055,
+      "num_input_tokens_seen": 8298000,
+      "step": 10620
+    },
+    {
+      "epoch": 22.08939708939709,
+      "grad_norm": 0.01308682281523943,
+      "learning_rate": 4.179042867352511e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 8302000,
+      "step": 10625
+    },
+    {
+      "epoch": 22.0997920997921,
+      "grad_norm": 0.003126337891444564,
+      "learning_rate": 4.1783153626238334e-05,
+      "loss": 0.0547,
+      "num_input_tokens_seen": 8305776,
+      "step": 10630
+    },
+    {
+      "epoch": 22.11018711018711,
+      "grad_norm": 0.05091722309589386,
+      "learning_rate": 4.177587599077836e-05,
+      "loss": 0.0202,
+      "num_input_tokens_seen": 8309584,
+      "step": 10635
+    },
+    {
+      "epoch": 22.12058212058212,
+      "grad_norm": 0.0227626021951437,
+      "learning_rate": 4.1768595768267494e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8313552,
+      "step": 10640
+    },
+    {
+      "epoch": 22.13097713097713,
+      "grad_norm": 0.03330610319972038,
+      "learning_rate": 4.176131295982843e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 8317488,
+      "step": 10645
+    },
+    {
+      "epoch": 22.14137214137214,
+      "grad_norm": 0.018187399953603745,
+      "learning_rate": 4.1754027566584276e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8321296,
+      "step": 10650
+    },
+    {
+      "epoch": 22.151767151767153,
+      "grad_norm": 0.039499495178461075,
+      "learning_rate": 4.174673958965852e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 8325104,
+      "step": 10655
+    },
+    {
+      "epoch": 22.16216216216216,
+      "grad_norm": 0.05413970351219177,
+      "learning_rate": 4.173944903017507e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 8329040,
+      "step": 10660
+    },
+    {
+      "epoch": 22.172557172557173,
+      "grad_norm": 0.0049431598745286465,
+      "learning_rate": 4.173215588925822e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8332944,
+      "step": 10665
+    },
+    {
+      "epoch": 22.18295218295218,
+      "grad_norm": 0.02106139250099659,
+      "learning_rate": 4.172486016803266e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8336816,
+      "step": 10670
+    },
+    {
+      "epoch": 22.193347193347194,
+      "grad_norm": 0.0068709272891283035,
+      "learning_rate": 4.171756186762349e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8340784,
+      "step": 10675
+    },
+    {
+      "epoch": 22.203742203742205,
+      "grad_norm": 0.0013318727724254131,
+      "learning_rate": 4.171026098915619e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8344688,
+      "step": 10680
+    },
+    {
+      "epoch": 22.214137214137214,
+      "grad_norm": 1.2471702098846436,
+      "learning_rate": 4.170295753375665e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 8348656,
+      "step": 10685
+    },
+    {
+      "epoch": 22.224532224532226,
+      "grad_norm": 0.0013931597350165248,
+      "learning_rate": 4.169565150255117e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8352592,
+      "step": 10690
+    },
+    {
+      "epoch": 22.234927234927234,
+      "grad_norm": 0.003491421230137348,
+      "learning_rate": 4.16883428966664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8356528,
+      "step": 10695
+    },
+    {
+      "epoch": 22.245322245322246,
+      "grad_norm": 0.0027042448054999113,
+      "learning_rate": 4.168103171722944e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8360272,
+      "step": 10700
+    },
+    {
+      "epoch": 22.255717255717254,
+      "grad_norm": 0.0036551542580127716,
+      "learning_rate": 4.167371796536777e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8364176,
+      "step": 10705
+    },
+    {
+      "epoch": 22.266112266112266,
+      "grad_norm": 0.0014781637582927942,
+      "learning_rate": 4.166640164220924e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8368080,
+      "step": 10710
+    },
+    {
+      "epoch": 22.276507276507278,
+      "grad_norm": 0.01966419443488121,
+      "learning_rate": 4.1659082748882144e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8371984,
+      "step": 10715
+    },
+    {
+      "epoch": 22.286902286902286,
+      "grad_norm": 0.005243126302957535,
+      "learning_rate": 4.1651761286515135e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8375952,
+      "step": 10720
+    },
+    {
+      "epoch": 22.2972972972973,
+      "grad_norm": 0.00026018134667538106,
+      "learning_rate": 4.164443725623728e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8379824,
+      "step": 10725
+    },
+    {
+      "epoch": 22.307692307692307,
+      "grad_norm": 0.00017534194921609014,
+      "learning_rate": 4.163711065917802e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8383824,
+      "step": 10730
+    },
+    {
+      "epoch": 22.31808731808732,
+      "grad_norm": 0.000713278423063457,
+      "learning_rate": 4.1629781496467234e-05,
+      "loss": 0.0095,
+      "num_input_tokens_seen": 8387920,
+      "step": 10735
+    },
+    {
+      "epoch": 22.328482328482327,
+      "grad_norm": 0.0010299714049324393,
+      "learning_rate": 4.1622449769235164e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8391888,
+      "step": 10740
+    },
+    {
+      "epoch": 22.33887733887734,
+      "grad_norm": 0.000314643228193745,
+      "learning_rate": 4.161511547861243e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8395728,
+      "step": 10745
+    },
+    {
+      "epoch": 22.34927234927235,
+      "grad_norm": 0.00028536864556372166,
+      "learning_rate": 4.1607778625730104e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 8399728,
+      "step": 10750
+    },
+    {
+      "epoch": 22.35966735966736,
+      "grad_norm": 0.0002475129731465131,
+      "learning_rate": 4.160043921171961e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 8403632,
+      "step": 10755
+    },
+    {
+      "epoch": 22.37006237006237,
+      "grad_norm": 0.0004474653396755457,
+      "learning_rate": 4.159309723771276e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8407408,
+      "step": 10760
+    },
+    {
+      "epoch": 22.38045738045738,
+      "grad_norm": 0.0017092234920710325,
+      "learning_rate": 4.158575270484181e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8411312,
+      "step": 10765
+    },
+    {
+      "epoch": 22.39085239085239,
+      "grad_norm": 0.0012081365566700697,
+      "learning_rate": 4.157840561423936e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8415280,
+      "step": 10770
+    },
+    {
+      "epoch": 22.401247401247403,
+      "grad_norm": 0.05076507106423378,
+      "learning_rate": 4.1571055967038416e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8419280,
+      "step": 10775
+    },
+    {
+      "epoch": 22.41164241164241,
+      "grad_norm": 0.00034672190668061376,
+      "learning_rate": 4.156370376437241e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8423248,
+      "step": 10780
+    },
+    {
+      "epoch": 22.422037422037423,
+      "grad_norm": 0.0028611384332180023,
+      "learning_rate": 4.155634900737513e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 8427184,
+      "step": 10785
+    },
+    {
+      "epoch": 22.43243243243243,
+      "grad_norm": 0.00014948654279578477,
+      "learning_rate": 4.1548991697180764e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8431152,
+      "step": 10790
+    },
+    {
+      "epoch": 22.442827442827443,
+      "grad_norm": 0.0019250065088272095,
+      "learning_rate": 4.1541631834923914e-05,
+      "loss": 0.025,
+      "num_input_tokens_seen": 8435152,
+      "step": 10795
+    },
+    {
+      "epoch": 22.453222453222452,
+      "grad_norm": 0.003967793192714453,
+      "learning_rate": 4.153426942173956e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8438992,
+      "step": 10800
+    },
+    {
+      "epoch": 22.453222453222452,
+      "eval_loss": 0.44221925735473633,
+      "eval_runtime": 11.773,
+      "eval_samples_per_second": 72.709,
+      "eval_steps_per_second": 18.177,
+      "num_input_tokens_seen": 8438992,
+      "step": 10800
+    },
+    {
+      "epoch": 22.463617463617464,
+      "grad_norm": 0.006438181269913912,
+      "learning_rate": 4.152690445876308e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 8442992,
+      "step": 10805
+    },
+    {
+      "epoch": 22.474012474012476,
+      "grad_norm": 0.00015166369848884642,
+      "learning_rate": 4.1519536947130245e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8446960,
+      "step": 10810
+    },
+    {
+      "epoch": 22.484407484407484,
+      "grad_norm": 0.0038169545587152243,
+      "learning_rate": 4.151216688797722e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8451056,
+      "step": 10815
+    },
+    {
+      "epoch": 22.494802494802496,
+      "grad_norm": 0.001761910505592823,
+      "learning_rate": 4.150479428244054e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8454928,
+      "step": 10820
+    },
+    {
+      "epoch": 22.505197505197504,
+      "grad_norm": 0.0008100150153040886,
+      "learning_rate": 4.1497419131657176e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8458704,
+      "step": 10825
+    },
+    {
+      "epoch": 22.515592515592516,
+      "grad_norm": 0.001854707719758153,
+      "learning_rate": 4.149004143676447e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8462640,
+      "step": 10830
+    },
+    {
+      "epoch": 22.525987525987524,
+      "grad_norm": 0.0045076217502355576,
+      "learning_rate": 4.148266119890015e-05,
+      "loss": 0.0366,
+      "num_input_tokens_seen": 8466608,
+      "step": 10835
+    },
+    {
+      "epoch": 22.536382536382536,
+      "grad_norm": 0.0003019407158717513,
+      "learning_rate": 4.1475278419202324e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 8470544,
+      "step": 10840
+    },
+    {
+      "epoch": 22.546777546777548,
+      "grad_norm": 0.012888406403362751,
+      "learning_rate": 4.146789309880953e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 8474352,
+      "step": 10845
+    },
+    {
+      "epoch": 22.557172557172557,
+      "grad_norm": 0.0007105180993676186,
+      "learning_rate": 4.146050523886068e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8478224,
+      "step": 10850
+    },
+    {
+      "epoch": 22.56756756756757,
+      "grad_norm": 0.003361154580488801,
+      "learning_rate": 4.1453114840495055e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8482096,
+      "step": 10855
+    },
+    {
+      "epoch": 22.577962577962577,
+      "grad_norm": 0.4051663279533386,
+      "learning_rate": 4.1445721904852364e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 8486032,
+      "step": 10860
+    },
+    {
+      "epoch": 22.58835758835759,
+      "grad_norm": 3.5888917446136475,
+      "learning_rate": 4.143832643307269e-05,
+      "loss": 0.0195,
+      "num_input_tokens_seen": 8489808,
+      "step": 10865
+    },
+    {
+      "epoch": 22.598752598752597,
+      "grad_norm": 2.045250415802002,
+      "learning_rate": 4.1430928426296503e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 8493648,
+      "step": 10870
+    },
+    {
+      "epoch": 22.60914760914761,
+      "grad_norm": 0.005587457213550806,
+      "learning_rate": 4.142352788566466e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8497456,
+      "step": 10875
+    },
+    {
+      "epoch": 22.61954261954262,
+      "grad_norm": 0.049359384924173355,
+      "learning_rate": 4.1416124812318424e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8501296,
+      "step": 10880
+    },
+    {
+      "epoch": 22.62993762993763,
+      "grad_norm": 0.002898943377658725,
+      "learning_rate": 4.1408719207399453e-05,
+      "loss": 0.0127,
+      "num_input_tokens_seen": 8505232,
+      "step": 10885
+    },
+    {
+      "epoch": 22.64033264033264,
+      "grad_norm": 0.005498939659446478,
+      "learning_rate": 4.140131107204978e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8508976,
+      "step": 10890
+    },
+    {
+      "epoch": 22.65072765072765,
+      "grad_norm": 0.330873966217041,
+      "learning_rate": 4.139390040741182e-05,
+      "loss": 0.01,
+      "num_input_tokens_seen": 8512912,
+      "step": 10895
+    },
+    {
+      "epoch": 22.66112266112266,
+      "grad_norm": 0.010203393176198006,
+      "learning_rate": 4.1386487214628396e-05,
+      "loss": 0.0126,
+      "num_input_tokens_seen": 8516912,
+      "step": 10900
+    },
+    {
+      "epoch": 22.671517671517673,
+      "grad_norm": 0.08627177774906158,
+      "learning_rate": 4.137907149484272e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 8520720,
+      "step": 10905
+    },
+    {
+      "epoch": 22.68191268191268,
+      "grad_norm": 0.04742557927966118,
+      "learning_rate": 4.137165324919839e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 8524464,
+      "step": 10910
+    },
+    {
+      "epoch": 22.692307692307693,
+      "grad_norm": 0.010312621481716633,
+      "learning_rate": 4.136423247883939e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 8528400,
+      "step": 10915
+    },
+    {
+      "epoch": 22.7027027027027,
+      "grad_norm": 0.0019201621180400252,
+      "learning_rate": 4.135680918491009e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 8532240,
+      "step": 10920
+    },
+    {
+      "epoch": 22.713097713097714,
+      "grad_norm": 0.002710499335080385,
+      "learning_rate": 4.1349383368555265e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8536176,
+      "step": 10925
+    },
+    {
+      "epoch": 22.723492723492722,
+      "grad_norm": 9.052775382995605,
+      "learning_rate": 4.1341955030920065e-05,
+      "loss": 0.0085,
+      "num_input_tokens_seen": 8540080,
+      "step": 10930
+    },
+    {
+      "epoch": 22.733887733887734,
+      "grad_norm": 0.0023876423947513103,
+      "learning_rate": 4.1334524173150036e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8544016,
+      "step": 10935
+    },
+    {
+      "epoch": 22.744282744282746,
+      "grad_norm": 0.017139755189418793,
+      "learning_rate": 4.13270907963911e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8547888,
+      "step": 10940
+    },
+    {
+      "epoch": 22.754677754677754,
+      "grad_norm": 0.0023544630967080593,
+      "learning_rate": 4.131965490178959e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8551856,
+      "step": 10945
+    },
+    {
+      "epoch": 22.765072765072766,
+      "grad_norm": 0.004545012954622507,
+      "learning_rate": 4.131221649049222e-05,
+      "loss": 0.0136,
+      "num_input_tokens_seen": 8555792,
+      "step": 10950
+    },
+    {
+      "epoch": 22.775467775467774,
+      "grad_norm": 24.815576553344727,
+      "learning_rate": 4.130477556364606e-05,
+      "loss": 0.0134,
+      "num_input_tokens_seen": 8559760,
+      "step": 10955
+    },
+    {
+      "epoch": 22.785862785862786,
+      "grad_norm": 0.001013016328215599,
+      "learning_rate": 4.129733212239861e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8563632,
+      "step": 10960
+    },
+    {
+      "epoch": 22.796257796257795,
+      "grad_norm": 0.006652127951383591,
+      "learning_rate": 4.128988616789774e-05,
+      "loss": 0.0103,
+      "num_input_tokens_seen": 8567472,
+      "step": 10965
+    },
+    {
+      "epoch": 22.806652806652806,
+      "grad_norm": 0.002152577508240938,
+      "learning_rate": 4.1282437701291724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8571536,
+      "step": 10970
+    },
+    {
+      "epoch": 22.81704781704782,
+      "grad_norm": 0.011120184324681759,
+      "learning_rate": 4.1274986723729184e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8575472,
+      "step": 10975
+    },
+    {
+      "epoch": 22.827442827442827,
+      "grad_norm": 0.16899710893630981,
+      "learning_rate": 4.126753323635917e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8579440,
+      "step": 10980
+    },
+    {
+      "epoch": 22.83783783783784,
+      "grad_norm": 0.001934931380674243,
+      "learning_rate": 4.12600772403311e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 8583408,
+      "step": 10985
+    },
+    {
+      "epoch": 22.848232848232847,
+      "grad_norm": 0.0024344753473997116,
+      "learning_rate": 4.125261873679479e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8587248,
+      "step": 10990
+    },
+    {
+      "epoch": 22.85862785862786,
+      "grad_norm": 0.31642240285873413,
+      "learning_rate": 4.124515772690042e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 8591408,
+      "step": 10995
+    },
+    {
+      "epoch": 22.86902286902287,
+      "grad_norm": 0.000454961002105847,
+      "learning_rate": 4.123769421179858e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 8595376,
+      "step": 11000
+    },
+    {
+      "epoch": 22.86902286902287,
+      "eval_loss": 0.470651775598526,
+      "eval_runtime": 11.8194,
+      "eval_samples_per_second": 72.423,
+      "eval_steps_per_second": 18.106,
+      "num_input_tokens_seen": 8595376,
+      "step": 11000
+    },
+    {
+      "epoch": 22.87941787941788,
+      "grad_norm": 0.0005633487016893923,
+      "learning_rate": 4.1230228192640236e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 8599312,
+      "step": 11005
+    },
+    {
+      "epoch": 22.88981288981289,
+      "grad_norm": 0.0012471689842641354,
+      "learning_rate": 4.122275967057675e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8603216,
+      "step": 11010
+    },
+    {
+      "epoch": 22.9002079002079,
+      "grad_norm": 0.0005833515897393227,
+      "learning_rate": 4.1215288646759846e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 8607024,
+      "step": 11015
+    },
+    {
+      "epoch": 22.91060291060291,
+      "grad_norm": 0.0005591053050011396,
+      "learning_rate": 4.120781512234166e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8610928,
+      "step": 11020
+    },
+    {
+      "epoch": 22.92099792099792,
+      "grad_norm": 0.0002959732373710722,
+      "learning_rate": 4.120033909847471e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8614704,
+      "step": 11025
+    },
+    {
+      "epoch": 22.93139293139293,
+      "grad_norm": 0.0014309894759207964,
+      "learning_rate": 4.119286057631187e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8618608,
+      "step": 11030
+    },
+    {
+      "epoch": 22.941787941787943,
+      "grad_norm": 0.0005855181370861828,
+      "learning_rate": 4.118537955700646e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 8622576,
+      "step": 11035
+    },
+    {
+      "epoch": 22.95218295218295,
+      "grad_norm": 0.0006009259959682822,
+      "learning_rate": 4.11778960417121e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8626448,
+      "step": 11040
+    },
+    {
+      "epoch": 22.962577962577964,
+      "grad_norm": 0.00017561900313012302,
+      "learning_rate": 4.117041003158288e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8630320,
+      "step": 11045
+    },
+    {
+      "epoch": 22.972972972972972,
+      "grad_norm": 0.00020520217367447913,
+      "learning_rate": 4.1162921527773215e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8634160,
+      "step": 11050
+    },
+    {
+      "epoch": 22.983367983367984,
+      "grad_norm": 0.007315618917346001,
+      "learning_rate": 4.115543053143794e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8638128,
+      "step": 11055
+    },
+    {
+      "epoch": 22.993762993762992,
+      "grad_norm": 0.00034922006307169795,
+      "learning_rate": 4.114793704373226e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8642160,
+      "step": 11060
+    },
+    {
+      "epoch": 23.004158004158004,
+      "grad_norm": 92.76158142089844,
+      "learning_rate": 4.114044106581175e-05,
+      "loss": 0.0145,
+      "num_input_tokens_seen": 8645944,
+      "step": 11065
+    },
+    {
+      "epoch": 23.014553014553016,
+      "grad_norm": 0.0005176915437914431,
+      "learning_rate": 4.11329425988324e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8649880,
+      "step": 11070
+    },
+    {
+      "epoch": 23.024948024948024,
+      "grad_norm": 0.00010142924293177202,
+      "learning_rate": 4.112544164395056e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8653752,
+      "step": 11075
+    },
+    {
+      "epoch": 23.035343035343036,
+      "grad_norm": 0.0002398152428213507,
+      "learning_rate": 4.111793820232297e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8657688,
+      "step": 11080
+    },
+    {
+      "epoch": 23.045738045738045,
+      "grad_norm": 0.0001075253458111547,
+      "learning_rate": 4.1110432275106767e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8661464,
+      "step": 11085
+    },
+    {
+      "epoch": 23.056133056133056,
+      "grad_norm": 0.00011313192953821272,
+      "learning_rate": 4.110292386345944e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8665464,
+      "step": 11090
+    },
+    {
+      "epoch": 23.066528066528065,
+      "grad_norm": 0.00011341348727000877,
+      "learning_rate": 4.109541296853891e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8669432,
+      "step": 11095
+    },
+    {
+      "epoch": 23.076923076923077,
+      "grad_norm": 0.0001111254605348222,
+      "learning_rate": 4.108789959150341e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8673432,
+      "step": 11100
+    },
+    {
+      "epoch": 23.08731808731809,
+      "grad_norm": 0.0003350689075887203,
+      "learning_rate": 4.108038373351163e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8677400,
+      "step": 11105
+    },
+    {
+      "epoch": 23.097713097713097,
+      "grad_norm": 0.00024902267614379525,
+      "learning_rate": 4.10728653957226e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8681272,
+      "step": 11110
+    },
+    {
+      "epoch": 23.10810810810811,
+      "grad_norm": 0.000873545475769788,
+      "learning_rate": 4.106534457929575e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8685272,
+      "step": 11115
+    },
+    {
+      "epoch": 23.118503118503117,
+      "grad_norm": 0.00022555605391971767,
+      "learning_rate": 4.105782128539086e-05,
+      "loss": 0.0108,
+      "num_input_tokens_seen": 8689304,
+      "step": 11120
+    },
+    {
+      "epoch": 23.12889812889813,
+      "grad_norm": 12.371990203857422,
+      "learning_rate": 4.1050295515168144e-05,
+      "loss": 0.0434,
+      "num_input_tokens_seen": 8693208,
+      "step": 11125
+    },
+    {
+      "epoch": 23.13929313929314,
+      "grad_norm": 0.00023144808073993772,
+      "learning_rate": 4.1042767269788155e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8696952,
+      "step": 11130
+    },
+    {
+      "epoch": 23.14968814968815,
+      "grad_norm": 0.00026577169774100184,
+      "learning_rate": 4.103523655041185e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 8700856,
+      "step": 11135
+    },
+    {
+      "epoch": 23.16008316008316,
+      "grad_norm": 1.1934047937393188,
+      "learning_rate": 4.102770335820055e-05,
+      "loss": 0.0104,
+      "num_input_tokens_seen": 8704696,
+      "step": 11140
+    },
+    {
+      "epoch": 23.17047817047817,
+      "grad_norm": 0.001877050963230431,
+      "learning_rate": 4.1020167694315984e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 8708376,
+      "step": 11145
+    },
+    {
+      "epoch": 23.18087318087318,
+      "grad_norm": 0.029695594683289528,
+      "learning_rate": 4.101262955992023e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 8712184,
+      "step": 11150
+    },
+    {
+      "epoch": 23.19126819126819,
+      "grad_norm": 0.011965983547270298,
+      "learning_rate": 4.100508895617578e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 8716056,
+      "step": 11155
+    },
+    {
+      "epoch": 23.2016632016632,
+      "grad_norm": 0.007364668883383274,
+      "learning_rate": 4.099754588424547e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 8720056,
+      "step": 11160
+    },
+    {
+      "epoch": 23.212058212058214,
+      "grad_norm": 0.0012132509145885706,
+      "learning_rate": 4.0990000345292546e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8723928,
+      "step": 11165
+    },
+    {
+      "epoch": 23.222453222453222,
+      "grad_norm": 0.157748281955719,
+      "learning_rate": 4.098245234048064e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 8727832,
+      "step": 11170
+    },
+    {
+      "epoch": 23.232848232848234,
+      "grad_norm": 0.01231328584253788,
+      "learning_rate": 4.0974901870973726e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 8731800,
+      "step": 11175
+    },
+    {
+      "epoch": 23.243243243243242,
+      "grad_norm": 0.06383783370256424,
+      "learning_rate": 4.096734893793619e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8735640,
+      "step": 11180
+    },
+    {
+      "epoch": 23.253638253638254,
+      "grad_norm": 7.957134948810562e-05,
+      "learning_rate": 4.095979354253279e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8739608,
+      "step": 11185
+    },
+    {
+      "epoch": 23.264033264033262,
+      "grad_norm": 0.37657028436660767,
+      "learning_rate": 4.0952235685928656e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 8743448,
+      "step": 11190
+    },
+    {
+      "epoch": 23.274428274428274,
+      "grad_norm": 0.00014692670083604753,
+      "learning_rate": 4.094467536928932e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 8747448,
+      "step": 11195
+    },
+    {
+      "epoch": 23.284823284823286,
+      "grad_norm": 0.0009069291409105062,
+      "learning_rate": 4.093711259378067e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8751352,
+      "step": 11200
+    },
+    {
+      "epoch": 23.284823284823286,
+      "eval_loss": 0.5724512934684753,
+      "eval_runtime": 11.7591,
+      "eval_samples_per_second": 72.795,
+      "eval_steps_per_second": 18.199,
+      "num_input_tokens_seen": 8751352,
+      "step": 11200
+    },
+    {
+      "epoch": 23.295218295218294,
+      "grad_norm": 0.00020998416584916413,
+      "learning_rate": 4.092954736056897e-05,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 8755352,
+      "step": 11205
+    },
+    {
+      "epoch": 23.305613305613306,
+      "grad_norm": 17.45784568786621,
+      "learning_rate": 4.09219796708209e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 8759256,
+      "step": 11210
+    },
+    {
+      "epoch": 23.316008316008315,
+      "grad_norm": 0.00035445287358015776,
+      "learning_rate": 4.0914409525703464e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8763128,
+      "step": 11215
+    },
+    {
+      "epoch": 23.326403326403327,
+      "grad_norm": 0.0005635005072690547,
+      "learning_rate": 4.090683692638408e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 8767128,
+      "step": 11220
+    },
+    {
+      "epoch": 23.33679833679834,
+      "grad_norm": 0.00015881237050052732,
+      "learning_rate": 4.089926187403056e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 8770936,
+      "step": 11225
+    },
+    {
+      "epoch": 23.347193347193347,
+      "grad_norm": 0.0014476149808615446,
+      "learning_rate": 4.0891684369811044e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8774904,
+      "step": 11230
+    },
+    {
+      "epoch": 23.35758835758836,
+      "grad_norm": 0.33889955282211304,
+      "learning_rate": 4.0884104414894107e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 8778776,
+      "step": 11235
+    },
+    {
+      "epoch": 23.367983367983367,
+      "grad_norm": 0.0006126546068117023,
+      "learning_rate": 4.087652201044864e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8782904,
+      "step": 11240
+    },
+    {
+      "epoch": 23.37837837837838,
+      "grad_norm": 0.0005325282691046596,
+      "learning_rate": 4.086893715764397e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8786808,
+      "step": 11245
+    },
+    {
+      "epoch": 23.388773388773387,
+      "grad_norm": 0.00026396181783638895,
+      "learning_rate": 4.086134985764977e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8790680,
+      "step": 11250
+    },
+    {
+      "epoch": 23.3991683991684,
+      "grad_norm": 0.00013210276665631682,
+      "learning_rate": 4.0853760111636085e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8794776,
+      "step": 11255
+    },
+    {
+      "epoch": 23.40956340956341,
+      "grad_norm": 0.002336389385163784,
+      "learning_rate": 4.084616792077337e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8798584,
+      "step": 11260
+    },
+    {
+      "epoch": 23.41995841995842,
+      "grad_norm": 0.0007969197467900813,
+      "learning_rate": 4.083857328623243e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8802552,
+      "step": 11265
+    },
+    {
+      "epoch": 23.43035343035343,
+      "grad_norm": 0.004144653212279081,
+      "learning_rate": 4.083097620918444e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8806584,
+      "step": 11270
+    },
+    {
+      "epoch": 23.44074844074844,
+      "grad_norm": 0.001529625034891069,
+      "learning_rate": 4.082337669080097e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 8810648,
+      "step": 11275
+    },
+    {
+      "epoch": 23.45114345114345,
+      "grad_norm": 0.0012066808994859457,
+      "learning_rate": 4.081577473225398e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8814488,
+      "step": 11280
+    },
+    {
+      "epoch": 23.46153846153846,
+      "grad_norm": 0.00011300267215119675,
+      "learning_rate": 4.080817033471577e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8818264,
+      "step": 11285
+    },
+    {
+      "epoch": 23.471933471933472,
+      "grad_norm": 0.009770617820322514,
+      "learning_rate": 4.080056349935903e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8822040,
+      "step": 11290
+    },
+    {
+      "epoch": 23.482328482328484,
+      "grad_norm": 0.00040875503327697515,
+      "learning_rate": 4.079295422735684e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8825912,
+      "step": 11295
+    },
+    {
+      "epoch": 23.492723492723492,
+      "grad_norm": 0.0003148166579194367,
+      "learning_rate": 4.078534251988264e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8829784,
+      "step": 11300
+    },
+    {
+      "epoch": 23.503118503118504,
+      "grad_norm": 0.00025484192883595824,
+      "learning_rate": 4.077772837811025e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8833624,
+      "step": 11305
+    },
+    {
+      "epoch": 23.513513513513512,
+      "grad_norm": 0.001336139626801014,
+      "learning_rate": 4.0770111803213874e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8837464,
+      "step": 11310
+    },
+    {
+      "epoch": 23.523908523908524,
+      "grad_norm": 0.0035836156457662582,
+      "learning_rate": 4.076249279636807e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8841464,
+      "step": 11315
+    },
+    {
+      "epoch": 23.534303534303533,
+      "grad_norm": 0.05130091682076454,
+      "learning_rate": 4.075487135874781e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 8845496,
+      "step": 11320
+    },
+    {
+      "epoch": 23.544698544698544,
+      "grad_norm": 0.00013018377649132162,
+      "learning_rate": 4.074724749152837e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8849528,
+      "step": 11325
+    },
+    {
+      "epoch": 23.555093555093556,
+      "grad_norm": 0.00013370024680625647,
+      "learning_rate": 4.07396211958855e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8853368,
+      "step": 11330
+    },
+    {
+      "epoch": 23.565488565488565,
+      "grad_norm": 0.0001457570615457371,
+      "learning_rate": 4.073199247299523e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 8857272,
+      "step": 11335
+    },
+    {
+      "epoch": 23.575883575883577,
+      "grad_norm": 9.353934729006141e-05,
+      "learning_rate": 4.072436132403403e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8861208,
+      "step": 11340
+    },
+    {
+      "epoch": 23.586278586278585,
+      "grad_norm": 0.0001280648575630039,
+      "learning_rate": 4.0716727750178704e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8865080,
+      "step": 11345
+    },
+    {
+      "epoch": 23.596673596673597,
+      "grad_norm": 0.0007912616129033267,
+      "learning_rate": 4.0709091752606455e-05,
+      "loss": 0.0082,
+      "num_input_tokens_seen": 8869016,
+      "step": 11350
+    },
+    {
+      "epoch": 23.60706860706861,
+      "grad_norm": 0.33545202016830444,
+      "learning_rate": 4.070145333249484e-05,
+      "loss": 0.0059,
+      "num_input_tokens_seen": 8872920,
+      "step": 11355
+    },
+    {
+      "epoch": 23.617463617463617,
+      "grad_norm": 0.1995973438024521,
+      "learning_rate": 4.069381249102181e-05,
+      "loss": 0.0103,
+      "num_input_tokens_seen": 8876856,
+      "step": 11360
+    },
+    {
+      "epoch": 23.62785862785863,
+      "grad_norm": 0.0009655342437326908,
+      "learning_rate": 4.0686169229365665e-05,
+      "loss": 0.0111,
+      "num_input_tokens_seen": 8880824,
+      "step": 11365
+    },
+    {
+      "epoch": 23.638253638253637,
+      "grad_norm": 0.0024399373214691877,
+      "learning_rate": 4.067852354870511e-05,
+      "loss": 0.0087,
+      "num_input_tokens_seen": 8884728,
+      "step": 11370
+    },
+    {
+      "epoch": 23.64864864864865,
+      "grad_norm": 0.40652695298194885,
+      "learning_rate": 4.067087545021919e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 8888504,
+      "step": 11375
+    },
+    {
+      "epoch": 23.659043659043657,
+      "grad_norm": 0.0006478984141722322,
+      "learning_rate": 4.066322493508734e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8892312,
+      "step": 11380
+    },
+    {
+      "epoch": 23.66943866943867,
+      "grad_norm": 0.002704931888729334,
+      "learning_rate": 4.065557200448937e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8896088,
+      "step": 11385
+    },
+    {
+      "epoch": 23.67983367983368,
+      "grad_norm": 0.0033192376140505075,
+      "learning_rate": 4.064791665960546e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8900056,
+      "step": 11390
+    },
+    {
+      "epoch": 23.69022869022869,
+      "grad_norm": 0.006430488545447588,
+      "learning_rate": 4.064025890161615e-05,
+      "loss": 0.0099,
+      "num_input_tokens_seen": 8904152,
+      "step": 11395
+    },
+    {
+      "epoch": 23.7006237006237,
+      "grad_norm": 0.10335345566272736,
+      "learning_rate": 4.0632598731702373e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 8907960,
+      "step": 11400
+    },
+    {
+      "epoch": 23.7006237006237,
+      "eval_loss": 0.3873137831687927,
+      "eval_runtime": 11.8493,
+      "eval_samples_per_second": 72.241,
+      "eval_steps_per_second": 18.06,
+      "num_input_tokens_seen": 8907960,
+      "step": 11400
+    },
+    {
+      "epoch": 23.71101871101871,
+      "grad_norm": 22.409069061279297,
+      "learning_rate": 4.0624936151045426e-05,
+      "loss": 0.04,
+      "num_input_tokens_seen": 8911832,
+      "step": 11405
+    },
+    {
+      "epoch": 23.72141372141372,
+      "grad_norm": 0.0016517124604433775,
+      "learning_rate": 4.061727116082696e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8915800,
+      "step": 11410
+    },
+    {
+      "epoch": 23.731808731808734,
+      "grad_norm": 0.007716527208685875,
+      "learning_rate": 4.060960376222903e-05,
+      "loss": 0.0112,
+      "num_input_tokens_seen": 8919736,
+      "step": 11415
+    },
+    {
+      "epoch": 23.742203742203742,
+      "grad_norm": 0.04401121288537979,
+      "learning_rate": 4.0601933956434034e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8923608,
+      "step": 11420
+    },
+    {
+      "epoch": 23.752598752598754,
+      "grad_norm": 0.0065982164815068245,
+      "learning_rate": 4.059426174462476e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8927704,
+      "step": 11425
+    },
+    {
+      "epoch": 23.762993762993762,
+      "grad_norm": 0.009032467380166054,
+      "learning_rate": 4.058658712798435e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8931544,
+      "step": 11430
+    },
+    {
+      "epoch": 23.773388773388774,
+      "grad_norm": 10.641914367675781,
+      "learning_rate": 4.0578910107696336e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 8935480,
+      "step": 11435
+    },
+    {
+      "epoch": 23.783783783783782,
+      "grad_norm": 0.016287362203001976,
+      "learning_rate": 4.05712306849446e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8939416,
+      "step": 11440
+    },
+    {
+      "epoch": 23.794178794178794,
+      "grad_norm": 0.008111546747386456,
+      "learning_rate": 4.0563548860913415e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8943384,
+      "step": 11445
+    },
+    {
+      "epoch": 23.804573804573806,
+      "grad_norm": 0.022973019629716873,
+      "learning_rate": 4.0555864636787414e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8947224,
+      "step": 11450
+    },
+    {
+      "epoch": 23.814968814968815,
+      "grad_norm": 0.016543639823794365,
+      "learning_rate": 4.054817801375159e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 8951096,
+      "step": 11455
+    },
+    {
+      "epoch": 23.825363825363826,
+      "grad_norm": 0.0003519717720337212,
+      "learning_rate": 4.054048899299134e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8954936,
+      "step": 11460
+    },
+    {
+      "epoch": 23.835758835758835,
+      "grad_norm": 0.00028478202875703573,
+      "learning_rate": 4.0532797575692385e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8958840,
+      "step": 11465
+    },
+    {
+      "epoch": 23.846153846153847,
+      "grad_norm": 0.0019958640914410353,
+      "learning_rate": 4.052510376304085e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8962680,
+      "step": 11470
+    },
+    {
+      "epoch": 23.856548856548855,
+      "grad_norm": 0.0002532335347495973,
+      "learning_rate": 4.051740755622321e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 8966552,
+      "step": 11475
+    },
+    {
+      "epoch": 23.866943866943867,
+      "grad_norm": 0.0007327420171350241,
+      "learning_rate": 4.050970895642632e-05,
+      "loss": 0.007,
+      "num_input_tokens_seen": 8970296,
+      "step": 11480
+    },
+    {
+      "epoch": 23.87733887733888,
+      "grad_norm": 0.0010916898027062416,
+      "learning_rate": 4.050200796483741e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8974200,
+      "step": 11485
+    },
+    {
+      "epoch": 23.887733887733887,
+      "grad_norm": 0.001244420069269836,
+      "learning_rate": 4.049430458264405e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 8977944,
+      "step": 11490
+    },
+    {
+      "epoch": 23.8981288981289,
+      "grad_norm": 0.010195116512477398,
+      "learning_rate": 4.048659881103422e-05,
+      "loss": 0.0353,
+      "num_input_tokens_seen": 8981752,
+      "step": 11495
+    },
+    {
+      "epoch": 23.908523908523907,
+      "grad_norm": 0.0019312743097543716,
+      "learning_rate": 4.0478890651196235e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8985656,
+      "step": 11500
+    },
+    {
+      "epoch": 23.91891891891892,
+      "grad_norm": 0.05725308507680893,
+      "learning_rate": 4.047118010431879e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8989624,
+      "step": 11505
+    },
+    {
+      "epoch": 23.929313929313928,
+      "grad_norm": 0.009562688879668713,
+      "learning_rate": 4.046346717159094e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 8993528,
+      "step": 11510
+    },
+    {
+      "epoch": 23.93970893970894,
+      "grad_norm": 0.031273581087589264,
+      "learning_rate": 4.045575185420214e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 8997400,
+      "step": 11515
+    },
+    {
+      "epoch": 23.95010395010395,
+      "grad_norm": 0.016927218064665794,
+      "learning_rate": 4.0448034153342165e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9001400,
+      "step": 11520
+    },
+    {
+      "epoch": 23.96049896049896,
+      "grad_norm": 0.0009944562334567308,
+      "learning_rate": 4.0440314070201194e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9005240,
+      "step": 11525
+    },
+    {
+      "epoch": 23.97089397089397,
+      "grad_norm": 0.0012882334413006902,
+      "learning_rate": 4.043259160596976e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9009144,
+      "step": 11530
+    },
+    {
+      "epoch": 23.98128898128898,
+      "grad_norm": 0.008147922344505787,
+      "learning_rate": 4.0424866761838767e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 9013112,
+      "step": 11535
+    },
+    {
+      "epoch": 23.991683991683992,
+      "grad_norm": 0.0013344971230253577,
+      "learning_rate": 4.041713953899948e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 9016920,
+      "step": 11540
+    },
+    {
+      "epoch": 24.002079002079004,
+      "grad_norm": 0.0011157917324453592,
+      "learning_rate": 4.0409409938643515e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9020872,
+      "step": 11545
+    },
+    {
+      "epoch": 24.012474012474012,
+      "grad_norm": 0.004946100525557995,
+      "learning_rate": 4.0401677961962904e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 9024840,
+      "step": 11550
+    },
+    {
+      "epoch": 24.022869022869024,
+      "grad_norm": 0.04083969444036484,
+      "learning_rate": 4.039394361015001e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 9028936,
+      "step": 11555
+    },
+    {
+      "epoch": 24.033264033264032,
+      "grad_norm": 7.409892559051514,
+      "learning_rate": 4.038620688439755e-05,
+      "loss": 0.0294,
+      "num_input_tokens_seen": 9032936,
+      "step": 11560
+    },
+    {
+      "epoch": 24.043659043659044,
+      "grad_norm": 0.00023290231183636934,
+      "learning_rate": 4.037846778589862e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9036840,
+      "step": 11565
+    },
+    {
+      "epoch": 24.054054054054053,
+      "grad_norm": 0.004929251503199339,
+      "learning_rate": 4.0370726315846715e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9040904,
+      "step": 11570
+    },
+    {
+      "epoch": 24.064449064449065,
+      "grad_norm": 0.002363956533372402,
+      "learning_rate": 4.036298247543565e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9044904,
+      "step": 11575
+    },
+    {
+      "epoch": 24.074844074844076,
+      "grad_norm": 0.022853512316942215,
+      "learning_rate": 4.035523626585962e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9048808,
+      "step": 11580
+    },
+    {
+      "epoch": 24.085239085239085,
+      "grad_norm": 0.00453548738732934,
+      "learning_rate": 4.0347487688313194e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 9052808,
+      "step": 11585
+    },
+    {
+      "epoch": 24.095634095634097,
+      "grad_norm": 0.020585373044013977,
+      "learning_rate": 4.0339736743991296e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9056744,
+      "step": 11590
+    },
+    {
+      "epoch": 24.106029106029105,
+      "grad_norm": 0.00022729256306774914,
+      "learning_rate": 4.0331983434089227e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9060584,
+      "step": 11595
+    },
+    {
+      "epoch": 24.116424116424117,
+      "grad_norm": 0.0018072726670652628,
+      "learning_rate": 4.032422775980264e-05,
+      "loss": 0.0116,
+      "num_input_tokens_seen": 9064424,
+      "step": 11600
+    },
+    {
+      "epoch": 24.116424116424117,
+      "eval_loss": 0.4293324053287506,
+      "eval_runtime": 11.7395,
+      "eval_samples_per_second": 72.916,
+      "eval_steps_per_second": 18.229,
+      "num_input_tokens_seen": 9064424,
+      "step": 11600
+    },
+    {
+      "epoch": 24.126819126819125,
+      "grad_norm": 0.0038625309243798256,
+      "learning_rate": 4.031646972232754e-05,
+      "loss": 0.0088,
+      "num_input_tokens_seen": 9068296,
+      "step": 11605
+    },
+    {
+      "epoch": 24.137214137214137,
+      "grad_norm": 0.8182888627052307,
+      "learning_rate": 4.0308709322860344e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 9072072,
+      "step": 11610
+    },
+    {
+      "epoch": 24.14760914760915,
+      "grad_norm": 0.0009588079410605133,
+      "learning_rate": 4.0300946562597784e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 9075944,
+      "step": 11615
+    },
+    {
+      "epoch": 24.158004158004157,
+      "grad_norm": 0.000580871244892478,
+      "learning_rate": 4.029318144273698e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 9079944,
+      "step": 11620
+    },
+    {
+      "epoch": 24.16839916839917,
+      "grad_norm": 0.016922196373343468,
+      "learning_rate": 4.0285413964475415e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9083816,
+      "step": 11625
+    },
+    {
+      "epoch": 24.178794178794178,
+      "grad_norm": 0.0003227711422368884,
+      "learning_rate": 4.0277644129010927e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9087688,
+      "step": 11630
+    },
+    {
+      "epoch": 24.18918918918919,
+      "grad_norm": 0.0004805186181329191,
+      "learning_rate": 4.0269871937541724e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 9091560,
+      "step": 11635
+    },
+    {
+      "epoch": 24.1995841995842,
+      "grad_norm": 11.055603981018066,
+      "learning_rate": 4.026209739126637e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 9095464,
+      "step": 11640
+    },
+    {
+      "epoch": 24.20997920997921,
+      "grad_norm": 0.001977174775674939,
+      "learning_rate": 4.025432049138381e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9099496,
+      "step": 11645
+    },
+    {
+      "epoch": 24.22037422037422,
+      "grad_norm": 0.0081306928768754,
+      "learning_rate": 4.0246541239093325e-05,
+      "loss": 0.0373,
+      "num_input_tokens_seen": 9103624,
+      "step": 11650
+    },
+    {
+      "epoch": 24.23076923076923,
+      "grad_norm": 0.008868390694260597,
+      "learning_rate": 4.023875963559459e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9107592,
+      "step": 11655
+    },
+    {
+      "epoch": 24.241164241164242,
+      "grad_norm": 0.0006513276020996273,
+      "learning_rate": 4.023097568208761e-05,
+      "loss": 0.0288,
+      "num_input_tokens_seen": 9111496,
+      "step": 11660
+    },
+    {
+      "epoch": 24.25155925155925,
+      "grad_norm": 0.0033711139112710953,
+      "learning_rate": 4.022318937977277e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9115304,
+      "step": 11665
+    },
+    {
+      "epoch": 24.261954261954262,
+      "grad_norm": 0.0020683652255684137,
+      "learning_rate": 4.021540072985084e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 9119240,
+      "step": 11670
+    },
+    {
+      "epoch": 24.272349272349274,
+      "grad_norm": 0.023096797987818718,
+      "learning_rate": 4.020760973352289e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9123144,
+      "step": 11675
+    },
+    {
+      "epoch": 24.282744282744282,
+      "grad_norm": 0.036461252719163895,
+      "learning_rate": 4.019981639199042e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 9127048,
+      "step": 11680
+    },
+    {
+      "epoch": 24.293139293139294,
+      "grad_norm": 0.04493029788136482,
+      "learning_rate": 4.0192020706455245e-05,
+      "loss": 0.0232,
+      "num_input_tokens_seen": 9130856,
+      "step": 11685
+    },
+    {
+      "epoch": 24.303534303534303,
+      "grad_norm": 0.017157265916466713,
+      "learning_rate": 4.018422267811956e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9134824,
+      "step": 11690
+    },
+    {
+      "epoch": 24.313929313929314,
+      "grad_norm": 0.048129476606845856,
+      "learning_rate": 4.017642230818592e-05,
+      "loss": 0.017,
+      "num_input_tokens_seen": 9138696,
+      "step": 11695
+    },
+    {
+      "epoch": 24.324324324324323,
+      "grad_norm": 0.4371778666973114,
+      "learning_rate": 4.0168619597857246e-05,
+      "loss": 0.0074,
+      "num_input_tokens_seen": 9142632,
+      "step": 11700
+    },
+    {
+      "epoch": 24.334719334719335,
+      "grad_norm": 0.018233055248856544,
+      "learning_rate": 4.016081454833681e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 9146440,
+      "step": 11705
+    },
+    {
+      "epoch": 24.345114345114347,
+      "grad_norm": 0.006335253361612558,
+      "learning_rate": 4.0153007160828245e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 9150280,
+      "step": 11710
+    },
+    {
+      "epoch": 24.355509355509355,
+      "grad_norm": 0.002298814244568348,
+      "learning_rate": 4.0145197436535555e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 9154120,
+      "step": 11715
+    },
+    {
+      "epoch": 24.365904365904367,
+      "grad_norm": 0.0018283287063241005,
+      "learning_rate": 4.0137385376663095e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9158056,
+      "step": 11720
+    },
+    {
+      "epoch": 24.376299376299375,
+      "grad_norm": 0.002213051775470376,
+      "learning_rate": 4.012957098241558e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9161896,
+      "step": 11725
+    },
+    {
+      "epoch": 24.386694386694387,
+      "grad_norm": 0.0016793679678812623,
+      "learning_rate": 4.0121754254998076e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 9165736,
+      "step": 11730
+    },
+    {
+      "epoch": 24.397089397089395,
+      "grad_norm": 0.0015353064518421888,
+      "learning_rate": 4.011393519561606e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9169640,
+      "step": 11735
+    },
+    {
+      "epoch": 24.407484407484407,
+      "grad_norm": 0.0022984084207564592,
+      "learning_rate": 4.010611380547529e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9173608,
+      "step": 11740
+    },
+    {
+      "epoch": 24.41787941787942,
+      "grad_norm": 0.031232647597789764,
+      "learning_rate": 4.009829008578192e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9177416,
+      "step": 11745
+    },
+    {
+      "epoch": 24.428274428274428,
+      "grad_norm": 0.0010752398520708084,
+      "learning_rate": 4.00904640377425e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9181448,
+      "step": 11750
+    },
+    {
+      "epoch": 24.43866943866944,
+      "grad_norm": 0.008092864416539669,
+      "learning_rate": 4.0082635662563886e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9185224,
+      "step": 11755
+    },
+    {
+      "epoch": 24.449064449064448,
+      "grad_norm": 0.00044008539407514036,
+      "learning_rate": 4.007480496145331e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9189096,
+      "step": 11760
+    },
+    {
+      "epoch": 24.45945945945946,
+      "grad_norm": 0.0005000830860808492,
+      "learning_rate": 4.006697193561837e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9193192,
+      "step": 11765
+    },
+    {
+      "epoch": 24.46985446985447,
+      "grad_norm": 0.0047407569363713264,
+      "learning_rate": 4.005913658626701e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9196968,
+      "step": 11770
+    },
+    {
+      "epoch": 24.48024948024948,
+      "grad_norm": 0.7721876502037048,
+      "learning_rate": 4.005129891460754e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 9200808,
+      "step": 11775
+    },
+    {
+      "epoch": 24.490644490644492,
+      "grad_norm": 0.0017292045522481203,
+      "learning_rate": 4.004345892184864e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9204744,
+      "step": 11780
+    },
+    {
+      "epoch": 24.5010395010395,
+      "grad_norm": 4.110837459564209,
+      "learning_rate": 4.003561660919932e-05,
+      "loss": 0.0596,
+      "num_input_tokens_seen": 9208584,
+      "step": 11785
+    },
+    {
+      "epoch": 24.511434511434512,
+      "grad_norm": 0.0005236705183051527,
+      "learning_rate": 4.002777197786897e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9212392,
+      "step": 11790
+    },
+    {
+      "epoch": 24.52182952182952,
+      "grad_norm": 0.0027354906778782606,
+      "learning_rate": 4.0019925029067326e-05,
+      "loss": 0.0241,
+      "num_input_tokens_seen": 9216456,
+      "step": 11795
+    },
+    {
+      "epoch": 24.532224532224532,
+      "grad_norm": 0.01956312172114849,
+      "learning_rate": 4.0012075764004495e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9220456,
+      "step": 11800
+    },
+    {
+      "epoch": 24.532224532224532,
+      "eval_loss": 0.39914968609809875,
+      "eval_runtime": 11.7388,
+      "eval_samples_per_second": 72.921,
+      "eval_steps_per_second": 18.23,
+      "num_input_tokens_seen": 9220456,
+      "step": 11800
+    },
+    {
+      "epoch": 24.542619542619544,
+      "grad_norm": 0.00835867878049612,
+      "learning_rate": 4.000422418389094e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 9224264,
+      "step": 11805
+    },
+    {
+      "epoch": 24.553014553014552,
+      "grad_norm": 0.012436875142157078,
+      "learning_rate": 3.999637028993744e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 9228296,
+      "step": 11810
+    },
+    {
+      "epoch": 24.563409563409564,
+      "grad_norm": 19.367063522338867,
+      "learning_rate": 3.99885140833552e-05,
+      "loss": 0.0127,
+      "num_input_tokens_seen": 9232136,
+      "step": 11815
+    },
+    {
+      "epoch": 24.573804573804573,
+      "grad_norm": 0.07254007458686829,
+      "learning_rate": 3.998065556535572e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 9236008,
+      "step": 11820
+    },
+    {
+      "epoch": 24.584199584199585,
+      "grad_norm": 0.0053211296908557415,
+      "learning_rate": 3.9972794737150895e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9239784,
+      "step": 11825
+    },
+    {
+      "epoch": 24.594594594594593,
+      "grad_norm": 0.0012193312868475914,
+      "learning_rate": 3.996493159995297e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9243592,
+      "step": 11830
+    },
+    {
+      "epoch": 24.604989604989605,
+      "grad_norm": 0.0028186568524688482,
+      "learning_rate": 3.995706615497453e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9247560,
+      "step": 11835
+    },
+    {
+      "epoch": 24.615384615384617,
+      "grad_norm": 0.012970009818673134,
+      "learning_rate": 3.994919840342852e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9251400,
+      "step": 11840
+    },
+    {
+      "epoch": 24.625779625779625,
+      "grad_norm": 0.0017362326616421342,
+      "learning_rate": 3.994132834652825e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 9255336,
+      "step": 11845
+    },
+    {
+      "epoch": 24.636174636174637,
+      "grad_norm": 0.000593814067542553,
+      "learning_rate": 3.99334559854874e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9259144,
+      "step": 11850
+    },
+    {
+      "epoch": 24.646569646569645,
+      "grad_norm": 1.2461588382720947,
+      "learning_rate": 3.9925581321519955e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 9263080,
+      "step": 11855
+    },
+    {
+      "epoch": 24.656964656964657,
+      "grad_norm": 0.0334169901907444,
+      "learning_rate": 3.991770435584031e-05,
+      "loss": 0.0219,
+      "num_input_tokens_seen": 9267016,
+      "step": 11860
+    },
+    {
+      "epoch": 24.66735966735967,
+      "grad_norm": 0.0037684848066419363,
+      "learning_rate": 3.990982508966319e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9270760,
+      "step": 11865
+    },
+    {
+      "epoch": 24.677754677754677,
+      "grad_norm": 0.459789514541626,
+      "learning_rate": 3.990194352420367e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9274632,
+      "step": 11870
+    },
+    {
+      "epoch": 24.68814968814969,
+      "grad_norm": 0.011655423790216446,
+      "learning_rate": 3.9894059660677184e-05,
+      "loss": 0.0214,
+      "num_input_tokens_seen": 9278504,
+      "step": 11875
+    },
+    {
+      "epoch": 24.698544698544698,
+      "grad_norm": 4.540039539337158,
+      "learning_rate": 3.9886173500299526e-05,
+      "loss": 0.0047,
+      "num_input_tokens_seen": 9282440,
+      "step": 11880
+    },
+    {
+      "epoch": 24.70893970893971,
+      "grad_norm": 0.003093397244811058,
+      "learning_rate": 3.987828504428685e-05,
+      "loss": 0.0118,
+      "num_input_tokens_seen": 9286216,
+      "step": 11885
+    },
+    {
+      "epoch": 24.719334719334718,
+      "grad_norm": 0.0027100832667201757,
+      "learning_rate": 3.987039429385565e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9289992,
+      "step": 11890
+    },
+    {
+      "epoch": 24.72972972972973,
+      "grad_norm": 0.003992415964603424,
+      "learning_rate": 3.986250125022277e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9293928,
+      "step": 11895
+    },
+    {
+      "epoch": 24.74012474012474,
+      "grad_norm": 0.0066984365694224834,
+      "learning_rate": 3.985460591460544e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 9297928,
+      "step": 11900
+    },
+    {
+      "epoch": 24.75051975051975,
+      "grad_norm": 0.001033376669511199,
+      "learning_rate": 3.984670828822118e-05,
+      "loss": 0.0264,
+      "num_input_tokens_seen": 9301896,
+      "step": 11905
+    },
+    {
+      "epoch": 24.760914760914762,
+      "grad_norm": 0.001786477747373283,
+      "learning_rate": 3.983880837228794e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 9305960,
+      "step": 11910
+    },
+    {
+      "epoch": 24.77130977130977,
+      "grad_norm": 0.0006112116971053183,
+      "learning_rate": 3.983090616802396e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9309800,
+      "step": 11915
+    },
+    {
+      "epoch": 24.781704781704782,
+      "grad_norm": 0.0051676505245268345,
+      "learning_rate": 3.982300167664788e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 9313608,
+      "step": 11920
+    },
+    {
+      "epoch": 24.79209979209979,
+      "grad_norm": 0.00045205431524664164,
+      "learning_rate": 3.981509489937868e-05,
+      "loss": 0.0087,
+      "num_input_tokens_seen": 9317384,
+      "step": 11925
+    },
+    {
+      "epoch": 24.802494802494802,
+      "grad_norm": 0.0017313075950369239,
+      "learning_rate": 3.9807185837435643e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9321256,
+      "step": 11930
+    },
+    {
+      "epoch": 24.812889812889814,
+      "grad_norm": 0.0785297378897667,
+      "learning_rate": 3.9799274492038484e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9325256,
+      "step": 11935
+    },
+    {
+      "epoch": 24.823284823284823,
+      "grad_norm": 0.0013968079583719373,
+      "learning_rate": 3.979136086440722e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 9329160,
+      "step": 11940
+    },
+    {
+      "epoch": 24.833679833679835,
+      "grad_norm": 0.0032379806507378817,
+      "learning_rate": 3.9783444955762226e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9333096,
+      "step": 11945
+    },
+    {
+      "epoch": 24.844074844074843,
+      "grad_norm": 37.70167541503906,
+      "learning_rate": 3.977552676732424e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 9337032,
+      "step": 11950
+    },
+    {
+      "epoch": 24.854469854469855,
+      "grad_norm": 0.28557559847831726,
+      "learning_rate": 3.976760630031435e-05,
+      "loss": 0.0096,
+      "num_input_tokens_seen": 9340744,
+      "step": 11955
+    },
+    {
+      "epoch": 24.864864864864863,
+      "grad_norm": 0.03279810771346092,
+      "learning_rate": 3.975968355595398e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9344776,
+      "step": 11960
+    },
+    {
+      "epoch": 24.875259875259875,
+      "grad_norm": 0.0030662526842206717,
+      "learning_rate": 3.9751758535464935e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 9348712,
+      "step": 11965
+    },
+    {
+      "epoch": 24.885654885654887,
+      "grad_norm": 0.33685415983200073,
+      "learning_rate": 3.9743831240069326e-05,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 9352552,
+      "step": 11970
+    },
+    {
+      "epoch": 24.896049896049895,
+      "grad_norm": 0.0045855119824409485,
+      "learning_rate": 3.9735901670989675e-05,
+      "loss": 0.0371,
+      "num_input_tokens_seen": 9356296,
+      "step": 11975
+    },
+    {
+      "epoch": 24.906444906444907,
+      "grad_norm": 0.0053409310057759285,
+      "learning_rate": 3.97279698294488e-05,
+      "loss": 0.0162,
+      "num_input_tokens_seen": 9360168,
+      "step": 11980
+    },
+    {
+      "epoch": 24.916839916839916,
+      "grad_norm": 0.04719991236925125,
+      "learning_rate": 3.9720035716669876e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 9364200,
+      "step": 11985
+    },
+    {
+      "epoch": 24.927234927234927,
+      "grad_norm": 0.010540124028921127,
+      "learning_rate": 3.9712099333876474e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 9368168,
+      "step": 11990
+    },
+    {
+      "epoch": 24.93762993762994,
+      "grad_norm": 0.006145063787698746,
+      "learning_rate": 3.9704160682292475e-05,
+      "loss": 0.0158,
+      "num_input_tokens_seen": 9372360,
+      "step": 11995
+    },
+    {
+      "epoch": 24.948024948024948,
+      "grad_norm": 0.005399067420512438,
+      "learning_rate": 3.9696219763142106e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9376488,
+      "step": 12000
+    },
+    {
+      "epoch": 24.948024948024948,
+      "eval_loss": 0.31668147444725037,
+      "eval_runtime": 11.7982,
+      "eval_samples_per_second": 72.553,
+      "eval_steps_per_second": 18.138,
+      "num_input_tokens_seen": 9376488,
+      "step": 12000
+    },
+    {
+      "epoch": 24.95841995841996,
+      "grad_norm": 0.003340402850881219,
+      "learning_rate": 3.968827657764997e-05,
+      "loss": 0.0094,
+      "num_input_tokens_seen": 9380328,
+      "step": 12005
+    },
+    {
+      "epoch": 24.968814968814968,
+      "grad_norm": 0.008255667984485626,
+      "learning_rate": 3.9680331127041e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 9384264,
+      "step": 12010
+    },
+    {
+      "epoch": 24.97920997920998,
+      "grad_norm": 0.9654374122619629,
+      "learning_rate": 3.9672383412540495e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 9388104,
+      "step": 12015
+    },
+    {
+      "epoch": 24.989604989604988,
+      "grad_norm": 0.029117103666067123,
+      "learning_rate": 3.966443343537407e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 9391912,
+      "step": 12020
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 0.0019423632184043527,
+      "learning_rate": 3.965648119676772e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9395864,
+      "step": 12025
+    },
+    {
+      "epoch": 25.010395010395012,
+      "grad_norm": 0.0013133285101503134,
+      "learning_rate": 3.96485266979478e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9399736,
+      "step": 12030
+    },
+    {
+      "epoch": 25.02079002079002,
+      "grad_norm": 0.0024368315935134888,
+      "learning_rate": 3.9640569940140974e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 9403704,
+      "step": 12035
+    },
+    {
+      "epoch": 25.031185031185032,
+      "grad_norm": 0.026714494451880455,
+      "learning_rate": 3.963261092457428e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 9407672,
+      "step": 12040
+    },
+    {
+      "epoch": 25.04158004158004,
+      "grad_norm": 0.0031088488176465034,
+      "learning_rate": 3.962464965247509e-05,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 9411576,
+      "step": 12045
+    },
+    {
+      "epoch": 25.051975051975052,
+      "grad_norm": 0.000190288046724163,
+      "learning_rate": 3.9616686125071135e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 9415352,
+      "step": 12050
+    },
+    {
+      "epoch": 25.06237006237006,
+      "grad_norm": 0.009025782346725464,
+      "learning_rate": 3.9608720343590506e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9419224,
+      "step": 12055
+    },
+    {
+      "epoch": 25.072765072765073,
+      "grad_norm": 0.0014618682907894254,
+      "learning_rate": 3.960075230926161e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9423032,
+      "step": 12060
+    },
+    {
+      "epoch": 25.083160083160084,
+      "grad_norm": 0.05897584185004234,
+      "learning_rate": 3.959278202331322e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9426936,
+      "step": 12065
+    },
+    {
+      "epoch": 25.093555093555093,
+      "grad_norm": 0.0003839487035293132,
+      "learning_rate": 3.958480948697446e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9430904,
+      "step": 12070
+    },
+    {
+      "epoch": 25.103950103950105,
+      "grad_norm": 0.002946611726656556,
+      "learning_rate": 3.95768347014748e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9434904,
+      "step": 12075
+    },
+    {
+      "epoch": 25.114345114345113,
+      "grad_norm": 0.0033081755973398685,
+      "learning_rate": 3.956885766804404e-05,
+      "loss": 0.0074,
+      "num_input_tokens_seen": 9438872,
+      "step": 12080
+    },
+    {
+      "epoch": 25.124740124740125,
+      "grad_norm": 0.0002962992584798485,
+      "learning_rate": 3.956087838791235e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9442712,
+      "step": 12085
+    },
+    {
+      "epoch": 25.135135135135137,
+      "grad_norm": 0.0005126342293806374,
+      "learning_rate": 3.955289686231022e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9446648,
+      "step": 12090
+    },
+    {
+      "epoch": 25.145530145530145,
+      "grad_norm": 0.0001459710911149159,
+      "learning_rate": 3.9544913092468504e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9450616,
+      "step": 12095
+    },
+    {
+      "epoch": 25.155925155925157,
+      "grad_norm": 0.0006413222290575504,
+      "learning_rate": 3.9536927079618425e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9454616,
+      "step": 12100
+    },
+    {
+      "epoch": 25.166320166320165,
+      "grad_norm": 0.00029552835621871054,
+      "learning_rate": 3.9528938824991494e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 9458296,
+      "step": 12105
+    },
+    {
+      "epoch": 25.176715176715177,
+      "grad_norm": 0.001267689629457891,
+      "learning_rate": 3.952094832981962e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9462264,
+      "step": 12110
+    },
+    {
+      "epoch": 25.187110187110186,
+      "grad_norm": 0.000867947528604418,
+      "learning_rate": 3.951295559533503e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 9466328,
+      "step": 12115
+    },
+    {
+      "epoch": 25.197505197505198,
+      "grad_norm": 0.0001407829549862072,
+      "learning_rate": 3.95049606227703e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9470328,
+      "step": 12120
+    },
+    {
+      "epoch": 25.20790020790021,
+      "grad_norm": 0.2450876235961914,
+      "learning_rate": 3.949696341335838e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 9474168,
+      "step": 12125
+    },
+    {
+      "epoch": 25.218295218295218,
+      "grad_norm": 7.970082515385002e-05,
+      "learning_rate": 3.9488963968332503e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9478136,
+      "step": 12130
+    },
+    {
+      "epoch": 25.22869022869023,
+      "grad_norm": 0.0001044732125592418,
+      "learning_rate": 3.948096228892631e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9481944,
+      "step": 12135
+    },
+    {
+      "epoch": 25.239085239085238,
+      "grad_norm": 0.000809551274869591,
+      "learning_rate": 3.947295837637375e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 9485816,
+      "step": 12140
+    },
+    {
+      "epoch": 25.24948024948025,
+      "grad_norm": 7.274244126165286e-05,
+      "learning_rate": 3.9464952231909135e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9489816,
+      "step": 12145
+    },
+    {
+      "epoch": 25.25987525987526,
+      "grad_norm": 0.00031865513301454484,
+      "learning_rate": 3.945694385676711e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9493912,
+      "step": 12150
+    },
+    {
+      "epoch": 25.27027027027027,
+      "grad_norm": 22.00404167175293,
+      "learning_rate": 3.944893325218265e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 9497784,
+      "step": 12155
+    },
+    {
+      "epoch": 25.280665280665282,
+      "grad_norm": 0.00022738224652130157,
+      "learning_rate": 3.944092041939112e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9501752,
+      "step": 12160
+    },
+    {
+      "epoch": 25.29106029106029,
+      "grad_norm": 0.0009353089844807982,
+      "learning_rate": 3.943290535962818e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9505688,
+      "step": 12165
+    },
+    {
+      "epoch": 25.301455301455302,
+      "grad_norm": 0.0010032287100329995,
+      "learning_rate": 3.942488807412985e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 9509688,
+      "step": 12170
+    },
+    {
+      "epoch": 25.31185031185031,
+      "grad_norm": 5.338428672985174e-05,
+      "learning_rate": 3.941686856413251e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 9513592,
+      "step": 12175
+    },
+    {
+      "epoch": 25.322245322245323,
+      "grad_norm": 0.012060320004820824,
+      "learning_rate": 3.9408846830872874e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9517656,
+      "step": 12180
+    },
+    {
+      "epoch": 25.33264033264033,
+      "grad_norm": 0.05772276967763901,
+      "learning_rate": 3.940082287558798e-05,
+      "loss": 0.0265,
+      "num_input_tokens_seen": 9521528,
+      "step": 12185
+    },
+    {
+      "epoch": 25.343035343035343,
+      "grad_norm": 0.1970878690481186,
+      "learning_rate": 3.939279669951522e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9525368,
+      "step": 12190
+    },
+    {
+      "epoch": 25.353430353430355,
+      "grad_norm": 0.0002769012935459614,
+      "learning_rate": 3.938476830389234e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9529336,
+      "step": 12195
+    },
+    {
+      "epoch": 25.363825363825363,
+      "grad_norm": 0.004143981728702784,
+      "learning_rate": 3.937673768995742e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9533208,
+      "step": 12200
+    },
+    {
+      "epoch": 25.363825363825363,
+      "eval_loss": 0.5012140870094299,
+      "eval_runtime": 11.721,
+      "eval_samples_per_second": 73.031,
+      "eval_steps_per_second": 18.258,
+      "num_input_tokens_seen": 9533208,
+      "step": 12200
+    },
+    {
+      "epoch": 25.374220374220375,
+      "grad_norm": 0.775452733039856,
+      "learning_rate": 3.936870485894888e-05,
+      "loss": 0.0102,
+      "num_input_tokens_seen": 9537080,
+      "step": 12205
+    },
+    {
+      "epoch": 25.384615384615383,
+      "grad_norm": 0.4744768440723419,
+      "learning_rate": 3.9360669812105475e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 9540920,
+      "step": 12210
+    },
+    {
+      "epoch": 25.395010395010395,
+      "grad_norm": 0.0005979675916023552,
+      "learning_rate": 3.9352632550666325e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9544824,
+      "step": 12215
+    },
+    {
+      "epoch": 25.405405405405407,
+      "grad_norm": 0.0002941834391094744,
+      "learning_rate": 3.9344593075870866e-05,
+      "loss": 0.0098,
+      "num_input_tokens_seen": 9548632,
+      "step": 12220
+    },
+    {
+      "epoch": 25.415800415800415,
+      "grad_norm": 0.000617809419054538,
+      "learning_rate": 3.933655138895889e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9552632,
+      "step": 12225
+    },
+    {
+      "epoch": 25.426195426195427,
+      "grad_norm": 0.00041510327719151974,
+      "learning_rate": 3.932850749117053e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9556600,
+      "step": 12230
+    },
+    {
+      "epoch": 25.436590436590436,
+      "grad_norm": 0.00021387585729826242,
+      "learning_rate": 3.932046138374624e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9560472,
+      "step": 12235
+    },
+    {
+      "epoch": 25.446985446985448,
+      "grad_norm": 0.007330484688282013,
+      "learning_rate": 3.9312413067926854e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 9564280,
+      "step": 12240
+    },
+    {
+      "epoch": 25.457380457380456,
+      "grad_norm": 0.02672692947089672,
+      "learning_rate": 3.9304362544953506e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 9568312,
+      "step": 12245
+    },
+    {
+      "epoch": 25.467775467775468,
+      "grad_norm": 0.0005159727297723293,
+      "learning_rate": 3.929630981606769e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9572280,
+      "step": 12250
+    },
+    {
+      "epoch": 25.47817047817048,
+      "grad_norm": 0.23528362810611725,
+      "learning_rate": 3.928825488251124e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 9576216,
+      "step": 12255
+    },
+    {
+      "epoch": 25.488565488565488,
+      "grad_norm": 0.0009234441677108407,
+      "learning_rate": 3.9280197745526344e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 9580152,
+      "step": 12260
+    },
+    {
+      "epoch": 25.4989604989605,
+      "grad_norm": 0.0036656640004366636,
+      "learning_rate": 3.9272138406355495e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9584088,
+      "step": 12265
+    },
+    {
+      "epoch": 25.509355509355508,
+      "grad_norm": 0.0011062429985031486,
+      "learning_rate": 3.926407686624154e-05,
+      "loss": 0.0111,
+      "num_input_tokens_seen": 9588024,
+      "step": 12270
+    },
+    {
+      "epoch": 25.51975051975052,
+      "grad_norm": 0.0028567828703671694,
+      "learning_rate": 3.9256013126427684e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9591832,
+      "step": 12275
+    },
+    {
+      "epoch": 25.53014553014553,
+      "grad_norm": 0.6964496374130249,
+      "learning_rate": 3.9247947188157455e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 9595672,
+      "step": 12280
+    },
+    {
+      "epoch": 25.54054054054054,
+      "grad_norm": 0.0009332301560789347,
+      "learning_rate": 3.9239879052674715e-05,
+      "loss": 0.0073,
+      "num_input_tokens_seen": 9599608,
+      "step": 12285
+    },
+    {
+      "epoch": 25.550935550935552,
+      "grad_norm": 0.004235813394188881,
+      "learning_rate": 3.9231808721223673e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9603448,
+      "step": 12290
+    },
+    {
+      "epoch": 25.56133056133056,
+      "grad_norm": 0.0052426462061703205,
+      "learning_rate": 3.9223736195048886e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9607224,
+      "step": 12295
+    },
+    {
+      "epoch": 25.571725571725572,
+      "grad_norm": 0.0009138259338214993,
+      "learning_rate": 3.921566147539523e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9611128,
+      "step": 12300
+    },
+    {
+      "epoch": 25.58212058212058,
+      "grad_norm": 0.0005009612650610507,
+      "learning_rate": 3.920758456350792e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9615128,
+      "step": 12305
+    },
+    {
+      "epoch": 25.592515592515593,
+      "grad_norm": 0.00021097615535836667,
+      "learning_rate": 3.919950546063253e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9619064,
+      "step": 12310
+    },
+    {
+      "epoch": 25.602910602910605,
+      "grad_norm": 0.000701827579177916,
+      "learning_rate": 3.919142416801496e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9622968,
+      "step": 12315
+    },
+    {
+      "epoch": 25.613305613305613,
+      "grad_norm": 0.0023456786293536425,
+      "learning_rate": 3.918334068690144e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9626872,
+      "step": 12320
+    },
+    {
+      "epoch": 25.623700623700625,
+      "grad_norm": 0.00048007210716605186,
+      "learning_rate": 3.917525501853855e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 9630808,
+      "step": 12325
+    },
+    {
+      "epoch": 25.634095634095633,
+      "grad_norm": 0.00022442259069066495,
+      "learning_rate": 3.916716716417319e-05,
+      "loss": 0.0115,
+      "num_input_tokens_seen": 9634808,
+      "step": 12330
+    },
+    {
+      "epoch": 25.644490644490645,
+      "grad_norm": 0.00015545341011602432,
+      "learning_rate": 3.915907712505263e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9638872,
+      "step": 12335
+    },
+    {
+      "epoch": 25.654885654885653,
+      "grad_norm": 0.014265073463320732,
+      "learning_rate": 3.915098490242444e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9642776,
+      "step": 12340
+    },
+    {
+      "epoch": 25.665280665280665,
+      "grad_norm": 0.046078555285930634,
+      "learning_rate": 3.914289049753654e-05,
+      "loss": 0.0137,
+      "num_input_tokens_seen": 9646584,
+      "step": 12345
+    },
+    {
+      "epoch": 25.675675675675677,
+      "grad_norm": 0.00020056117500644177,
+      "learning_rate": 3.913479391163719e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9650392,
+      "step": 12350
+    },
+    {
+      "epoch": 25.686070686070686,
+      "grad_norm": 0.05163440853357315,
+      "learning_rate": 3.9126695145975e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9654424,
+      "step": 12355
+    },
+    {
+      "epoch": 25.696465696465697,
+      "grad_norm": 0.0079534025862813,
+      "learning_rate": 3.911859420179889e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9658264,
+      "step": 12360
+    },
+    {
+      "epoch": 25.706860706860706,
+      "grad_norm": 0.01539752259850502,
+      "learning_rate": 3.911049108035813e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9662168,
+      "step": 12365
+    },
+    {
+      "epoch": 25.717255717255718,
+      "grad_norm": 0.18074706196784973,
+      "learning_rate": 3.910238578290232e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 9666072,
+      "step": 12370
+    },
+    {
+      "epoch": 25.727650727650726,
+      "grad_norm": 0.036737244576215744,
+      "learning_rate": 3.90942783106814e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 9670072,
+      "step": 12375
+    },
+    {
+      "epoch": 25.738045738045738,
+      "grad_norm": 0.0022400598973035812,
+      "learning_rate": 3.908616866494564e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9673912,
+      "step": 12380
+    },
+    {
+      "epoch": 25.74844074844075,
+      "grad_norm": 0.007298966404050589,
+      "learning_rate": 3.907805684694566e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 9677784,
+      "step": 12385
+    },
+    {
+      "epoch": 25.758835758835758,
+      "grad_norm": 0.006731868721544743,
+      "learning_rate": 3.90699428579324e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9681784,
+      "step": 12390
+    },
+    {
+      "epoch": 25.76923076923077,
+      "grad_norm": 0.03113432228565216,
+      "learning_rate": 3.906182669915713e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 9685560,
+      "step": 12395
+    },
+    {
+      "epoch": 25.77962577962578,
+      "grad_norm": 0.0009280033991672099,
+      "learning_rate": 3.9053708371871476e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9689464,
+      "step": 12400
+    },
+    {
+      "epoch": 25.77962577962578,
+      "eval_loss": 0.4631843864917755,
+      "eval_runtime": 11.7329,
+      "eval_samples_per_second": 72.957,
+      "eval_steps_per_second": 18.239,
+      "num_input_tokens_seen": 9689464,
+      "step": 12400
+    },
+    {
+      "epoch": 25.79002079002079,
+      "grad_norm": 0.06268870830535889,
+      "learning_rate": 3.904558787732738e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 9693368,
+      "step": 12405
+    },
+    {
+      "epoch": 25.8004158004158,
+      "grad_norm": 2.611764648463577e-05,
+      "learning_rate": 3.9037465216777135e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9697240,
+      "step": 12410
+    },
+    {
+      "epoch": 25.81081081081081,
+      "grad_norm": 2.994503483932931e-05,
+      "learning_rate": 3.902934039147334e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9701016,
+      "step": 12415
+    },
+    {
+      "epoch": 25.821205821205822,
+      "grad_norm": 0.0005291198031045496,
+      "learning_rate": 3.902121340266894e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9704856,
+      "step": 12420
+    },
+    {
+      "epoch": 25.83160083160083,
+      "grad_norm": 0.0007556850905530155,
+      "learning_rate": 3.9013084251617246e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9708728,
+      "step": 12425
+    },
+    {
+      "epoch": 25.841995841995843,
+      "grad_norm": 2.6196788894594647e-05,
+      "learning_rate": 3.9004952939571865e-05,
+      "loss": 0.0267,
+      "num_input_tokens_seen": 9712600,
+      "step": 12430
+    },
+    {
+      "epoch": 25.85239085239085,
+      "grad_norm": 4.093574534635991e-05,
+      "learning_rate": 3.899681946778673e-05,
+      "loss": 0.0066,
+      "num_input_tokens_seen": 9716568,
+      "step": 12435
+    },
+    {
+      "epoch": 25.862785862785863,
+      "grad_norm": 0.00017459828814025968,
+      "learning_rate": 3.898868383751615e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9720440,
+      "step": 12440
+    },
+    {
+      "epoch": 25.873180873180875,
+      "grad_norm": 0.00023710515233688056,
+      "learning_rate": 3.8980546050014724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9724312,
+      "step": 12445
+    },
+    {
+      "epoch": 25.883575883575883,
+      "grad_norm": 0.00043990419362671673,
+      "learning_rate": 3.897240610653741e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9728184,
+      "step": 12450
+    },
+    {
+      "epoch": 25.893970893970895,
+      "grad_norm": 0.00014312191342469305,
+      "learning_rate": 3.896426400833948e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9732120,
+      "step": 12455
+    },
+    {
+      "epoch": 25.904365904365903,
+      "grad_norm": 9.264399704989046e-05,
+      "learning_rate": 3.895611975667656e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 9736120,
+      "step": 12460
+    },
+    {
+      "epoch": 25.914760914760915,
+      "grad_norm": 0.0027683735825121403,
+      "learning_rate": 3.8947973352804584e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 9740024,
+      "step": 12465
+    },
+    {
+      "epoch": 25.925155925155924,
+      "grad_norm": 0.0004861181369051337,
+      "learning_rate": 3.893982479797984e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9743928,
+      "step": 12470
+    },
+    {
+      "epoch": 25.935550935550935,
+      "grad_norm": 0.00021190221013966948,
+      "learning_rate": 3.8931674093458926e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9747800,
+      "step": 12475
+    },
+    {
+      "epoch": 25.945945945945947,
+      "grad_norm": 14.994208335876465,
+      "learning_rate": 3.89235212404988e-05,
+      "loss": 0.0274,
+      "num_input_tokens_seen": 9751768,
+      "step": 12480
+    },
+    {
+      "epoch": 25.956340956340956,
+      "grad_norm": 0.0007023972575552762,
+      "learning_rate": 3.891536624035672e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9755640,
+      "step": 12485
+    },
+    {
+      "epoch": 25.966735966735968,
+      "grad_norm": 0.012476206757128239,
+      "learning_rate": 3.8907209094290295e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9759448,
+      "step": 12490
+    },
+    {
+      "epoch": 25.977130977130976,
+      "grad_norm": 0.00048803951358422637,
+      "learning_rate": 3.8899049803557466e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9763288,
+      "step": 12495
+    },
+    {
+      "epoch": 25.987525987525988,
+      "grad_norm": 0.0007652196218259633,
+      "learning_rate": 3.889088836941648e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 9767096,
+      "step": 12500
+    },
+    {
+      "epoch": 25.997920997921,
+      "grad_norm": 0.0012953310506418347,
+      "learning_rate": 3.8882724793125946e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9770872,
+      "step": 12505
+    },
+    {
+      "epoch": 26.008316008316008,
+      "grad_norm": 0.003661304246634245,
+      "learning_rate": 3.8874559075944794e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9774552,
+      "step": 12510
+    },
+    {
+      "epoch": 26.01871101871102,
+      "grad_norm": 0.0044044265523552895,
+      "learning_rate": 3.886639121913227e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9778456,
+      "step": 12515
+    },
+    {
+      "epoch": 26.02910602910603,
+      "grad_norm": 0.00029593994258902967,
+      "learning_rate": 3.885822122394797e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9782488,
+      "step": 12520
+    },
+    {
+      "epoch": 26.03950103950104,
+      "grad_norm": 6.919619083404541,
+      "learning_rate": 3.8850049091651794e-05,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 9786296,
+      "step": 12525
+    },
+    {
+      "epoch": 26.04989604989605,
+      "grad_norm": 0.053698014467954636,
+      "learning_rate": 3.8841874823504e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 9790232,
+      "step": 12530
+    },
+    {
+      "epoch": 26.06029106029106,
+      "grad_norm": 0.00030316950869746506,
+      "learning_rate": 3.8833698420765157e-05,
+      "loss": 0.0355,
+      "num_input_tokens_seen": 9794168,
+      "step": 12535
+    },
+    {
+      "epoch": 26.070686070686072,
+      "grad_norm": 0.0004605670692399144,
+      "learning_rate": 3.882551988469618e-05,
+      "loss": 0.0085,
+      "num_input_tokens_seen": 9798136,
+      "step": 12540
+    },
+    {
+      "epoch": 26.08108108108108,
+      "grad_norm": 0.0012041924055665731,
+      "learning_rate": 3.881733921655829e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9802168,
+      "step": 12545
+    },
+    {
+      "epoch": 26.091476091476093,
+      "grad_norm": 0.00045515556121245027,
+      "learning_rate": 3.8809156417613054e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9806072,
+      "step": 12550
+    },
+    {
+      "epoch": 26.1018711018711,
+      "grad_norm": 0.00044072436867281795,
+      "learning_rate": 3.8800971489122364e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9810008,
+      "step": 12555
+    },
+    {
+      "epoch": 26.112266112266113,
+      "grad_norm": 0.0013903917279094458,
+      "learning_rate": 3.8792784432348434e-05,
+      "loss": 0.0097,
+      "num_input_tokens_seen": 9813848,
+      "step": 12560
+    },
+    {
+      "epoch": 26.12266112266112,
+      "grad_norm": 0.00011648023064481094,
+      "learning_rate": 3.878459524855381e-05,
+      "loss": 0.0109,
+      "num_input_tokens_seen": 9817656,
+      "step": 12565
+    },
+    {
+      "epoch": 26.133056133056133,
+      "grad_norm": 0.0008071967167779803,
+      "learning_rate": 3.8776403939001384e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9821624,
+      "step": 12570
+    },
+    {
+      "epoch": 26.143451143451145,
+      "grad_norm": 0.00013244755973573774,
+      "learning_rate": 3.876821050495433e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9825464,
+      "step": 12575
+    },
+    {
+      "epoch": 26.153846153846153,
+      "grad_norm": 0.00040395426913164556,
+      "learning_rate": 3.87600149476762e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9829432,
+      "step": 12580
+    },
+    {
+      "epoch": 26.164241164241165,
+      "grad_norm": 0.00024344230769202113,
+      "learning_rate": 3.8751817268430843e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9833304,
+      "step": 12585
+    },
+    {
+      "epoch": 26.174636174636174,
+      "grad_norm": 0.001915386994369328,
+      "learning_rate": 3.8743617468482464e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9837208,
+      "step": 12590
+    },
+    {
+      "epoch": 26.185031185031185,
+      "grad_norm": 0.0001497021730756387,
+      "learning_rate": 3.8735415549095535e-05,
+      "loss": 0.0074,
+      "num_input_tokens_seen": 9841112,
+      "step": 12595
+    },
+    {
+      "epoch": 26.195426195426194,
+      "grad_norm": 0.020571861416101456,
+      "learning_rate": 3.8727211511534934e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9845048,
+      "step": 12600
+    },
+    {
+      "epoch": 26.195426195426194,
+      "eval_loss": 0.4233267307281494,
+      "eval_runtime": 11.7043,
+      "eval_samples_per_second": 73.135,
+      "eval_steps_per_second": 18.284,
+      "num_input_tokens_seen": 9845048,
+      "step": 12600
+    },
+    {
+      "epoch": 26.205821205821206,
+      "grad_norm": 0.0009978790767490864,
+      "learning_rate": 3.8719005357065804e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9848824,
+      "step": 12605
+    },
+    {
+      "epoch": 26.216216216216218,
+      "grad_norm": 0.002913514617830515,
+      "learning_rate": 3.8710797086953645e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 9852664,
+      "step": 12610
+    },
+    {
+      "epoch": 26.226611226611226,
+      "grad_norm": 0.00038888363633304834,
+      "learning_rate": 3.870258670246427e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9856632,
+      "step": 12615
+    },
+    {
+      "epoch": 26.237006237006238,
+      "grad_norm": 0.008093463256955147,
+      "learning_rate": 3.869437420486384e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9860632,
+      "step": 12620
+    },
+    {
+      "epoch": 26.247401247401246,
+      "grad_norm": 0.0011530896881595254,
+      "learning_rate": 3.8686159595418805e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 9864568,
+      "step": 12625
+    },
+    {
+      "epoch": 26.257796257796258,
+      "grad_norm": 0.0027560109738260508,
+      "learning_rate": 3.867794287539597e-05,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 9868568,
+      "step": 12630
+    },
+    {
+      "epoch": 26.26819126819127,
+      "grad_norm": 0.0027277451008558273,
+      "learning_rate": 3.866972404606245e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 9872376,
+      "step": 12635
+    },
+    {
+      "epoch": 26.27858627858628,
+      "grad_norm": 6.764470163034275e-05,
+      "learning_rate": 3.866150310868571e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9876440,
+      "step": 12640
+    },
+    {
+      "epoch": 26.28898128898129,
+      "grad_norm": 0.0009982308838516474,
+      "learning_rate": 3.8653280064533506e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9880216,
+      "step": 12645
+    },
+    {
+      "epoch": 26.2993762993763,
+      "grad_norm": 0.0003413147642277181,
+      "learning_rate": 3.864505491487394e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 9884120,
+      "step": 12650
+    },
+    {
+      "epoch": 26.30977130977131,
+      "grad_norm": 0.0003052472311537713,
+      "learning_rate": 3.8636827660975414e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9887992,
+      "step": 12655
+    },
+    {
+      "epoch": 26.32016632016632,
+      "grad_norm": 9.371410124003887e-05,
+      "learning_rate": 3.862859830410671e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 9891960,
+      "step": 12660
+    },
+    {
+      "epoch": 26.33056133056133,
+      "grad_norm": 0.0006032183882780373,
+      "learning_rate": 3.862036684553688e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9895768,
+      "step": 12665
+    },
+    {
+      "epoch": 26.340956340956343,
+      "grad_norm": 0.00017355055024381727,
+      "learning_rate": 3.8612133286535314e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 9899736,
+      "step": 12670
+    },
+    {
+      "epoch": 26.35135135135135,
+      "grad_norm": 0.00040676051867194474,
+      "learning_rate": 3.860389762837173e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9903448,
+      "step": 12675
+    },
+    {
+      "epoch": 26.361746361746363,
+      "grad_norm": 0.005567324813455343,
+      "learning_rate": 3.859565987231618e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9907448,
+      "step": 12680
+    },
+    {
+      "epoch": 26.37214137214137,
+      "grad_norm": 0.00015635798627045006,
+      "learning_rate": 3.858742001963902e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9911608,
+      "step": 12685
+    },
+    {
+      "epoch": 26.382536382536383,
+      "grad_norm": 4.356209754943848,
+      "learning_rate": 3.857917807161094e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 9915608,
+      "step": 12690
+    },
+    {
+      "epoch": 26.39293139293139,
+      "grad_norm": 0.00128124188631773,
+      "learning_rate": 3.857093402950296e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9919416,
+      "step": 12695
+    },
+    {
+      "epoch": 26.403326403326403,
+      "grad_norm": 2.787566336337477e-05,
+      "learning_rate": 3.8562687894586414e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9923320,
+      "step": 12700
+    },
+    {
+      "epoch": 26.413721413721415,
+      "grad_norm": 3.3918131521204486e-05,
+      "learning_rate": 3.8554439668132946e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9927320,
+      "step": 12705
+    },
+    {
+      "epoch": 26.424116424116423,
+      "grad_norm": 0.00022408099903259426,
+      "learning_rate": 3.854618935141455e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9931032,
+      "step": 12710
+    },
+    {
+      "epoch": 26.434511434511435,
+      "grad_norm": 0.1555420160293579,
+      "learning_rate": 3.8537936945703525e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9934872,
+      "step": 12715
+    },
+    {
+      "epoch": 26.444906444906444,
+      "grad_norm": 3.0520361178787425e-05,
+      "learning_rate": 3.852968245227249e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9938840,
+      "step": 12720
+    },
+    {
+      "epoch": 26.455301455301456,
+      "grad_norm": 7.265902968356386e-05,
+      "learning_rate": 3.85214258723944e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9942552,
+      "step": 12725
+    },
+    {
+      "epoch": 26.465696465696467,
+      "grad_norm": 1.4051913022994995,
+      "learning_rate": 3.8513167207342524e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9946648,
+      "step": 12730
+    },
+    {
+      "epoch": 26.476091476091476,
+      "grad_norm": 0.00017810014833230525,
+      "learning_rate": 3.850490645839044e-05,
+      "loss": 0.0086,
+      "num_input_tokens_seen": 9950424,
+      "step": 12735
+    },
+    {
+      "epoch": 26.486486486486488,
+      "grad_norm": 0.002442054683342576,
+      "learning_rate": 3.849664362681207e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9954488,
+      "step": 12740
+    },
+    {
+      "epoch": 26.496881496881496,
+      "grad_norm": 0.0009844610467553139,
+      "learning_rate": 3.848837871388165e-05,
+      "loss": 0.0103,
+      "num_input_tokens_seen": 9958456,
+      "step": 12745
+    },
+    {
+      "epoch": 26.507276507276508,
+      "grad_norm": 0.36058029532432556,
+      "learning_rate": 3.848011172087371e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 9962424,
+      "step": 12750
+    },
+    {
+      "epoch": 26.517671517671516,
+      "grad_norm": 0.0014489000895991921,
+      "learning_rate": 3.847184264906315e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 9966232,
+      "step": 12755
+    },
+    {
+      "epoch": 26.528066528066528,
+      "grad_norm": 0.01071934774518013,
+      "learning_rate": 3.846357149972516e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9970328,
+      "step": 12760
+    },
+    {
+      "epoch": 26.53846153846154,
+      "grad_norm": 0.0005472128977999091,
+      "learning_rate": 3.8455298274135246e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9974264,
+      "step": 12765
+    },
+    {
+      "epoch": 26.54885654885655,
+      "grad_norm": 0.01331627182662487,
+      "learning_rate": 3.8447022973569254e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9978232,
+      "step": 12770
+    },
+    {
+      "epoch": 26.55925155925156,
+      "grad_norm": 0.0017119119875133038,
+      "learning_rate": 3.843874559930332e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 9982104,
+      "step": 12775
+    },
+    {
+      "epoch": 26.56964656964657,
+      "grad_norm": 0.0002902589912991971,
+      "learning_rate": 3.843046615261394e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 9986008,
+      "step": 12780
+    },
+    {
+      "epoch": 26.58004158004158,
+      "grad_norm": 0.004183475859463215,
+      "learning_rate": 3.842218463477791e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 9989912,
+      "step": 12785
+    },
+    {
+      "epoch": 26.59043659043659,
+      "grad_norm": 0.00018399821419734508,
+      "learning_rate": 3.841390104707233e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9993848,
+      "step": 12790
+    },
+    {
+      "epoch": 26.6008316008316,
+      "grad_norm": 0.0011142173316329718,
+      "learning_rate": 3.8405615390774643e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 9997752,
+      "step": 12795
+    },
+    {
+      "epoch": 26.611226611226613,
+      "grad_norm": 0.00012341343972366303,
+      "learning_rate": 3.839732766716259e-05,
+      "loss": 0.009,
+      "num_input_tokens_seen": 10001784,
+      "step": 12800
+    },
+    {
+      "epoch": 26.611226611226613,
+      "eval_loss": 0.4824027717113495,
+      "eval_runtime": 11.7005,
+      "eval_samples_per_second": 73.159,
+      "eval_steps_per_second": 18.29,
+      "num_input_tokens_seen": 10001784,
+      "step": 12800
+    },
+    {
+      "epoch": 26.62162162162162,
+      "grad_norm": 7.819393067620695e-05,
+      "learning_rate": 3.838903787751425e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10005752,
+      "step": 12805
+    },
+    {
+      "epoch": 26.632016632016633,
+      "grad_norm": 0.0002471734187565744,
+      "learning_rate": 3.838074602310802e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 10009592,
+      "step": 12810
+    },
+    {
+      "epoch": 26.64241164241164,
+      "grad_norm": 6.570939876837656e-05,
+      "learning_rate": 3.837245210522258e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10013656,
+      "step": 12815
+    },
+    {
+      "epoch": 26.652806652806653,
+      "grad_norm": 0.0012677264166995883,
+      "learning_rate": 3.8364156125136996e-05,
+      "loss": 0.0302,
+      "num_input_tokens_seen": 10017464,
+      "step": 12820
+    },
+    {
+      "epoch": 26.66320166320166,
+      "grad_norm": 0.00010089840361615643,
+      "learning_rate": 3.835585808413059e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10021432,
+      "step": 12825
+    },
+    {
+      "epoch": 26.673596673596673,
+      "grad_norm": 0.000277720537269488,
+      "learning_rate": 3.8347557983483024e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 10025336,
+      "step": 12830
+    },
+    {
+      "epoch": 26.683991683991685,
+      "grad_norm": 0.0003267799620516598,
+      "learning_rate": 3.833925582447428e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 10029208,
+      "step": 12835
+    },
+    {
+      "epoch": 26.694386694386694,
+      "grad_norm": 0.0010996555211022496,
+      "learning_rate": 3.8330951608384656e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10033080,
+      "step": 12840
+    },
+    {
+      "epoch": 26.704781704781706,
+      "grad_norm": 0.0002579839783720672,
+      "learning_rate": 3.832264533649477e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10036984,
+      "step": 12845
+    },
+    {
+      "epoch": 26.715176715176714,
+      "grad_norm": 0.0002937758981715888,
+      "learning_rate": 3.8314337010085555e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10040952,
+      "step": 12850
+    },
+    {
+      "epoch": 26.725571725571726,
+      "grad_norm": 0.0003313784836791456,
+      "learning_rate": 3.830602663043824e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 10044888,
+      "step": 12855
+    },
+    {
+      "epoch": 26.735966735966738,
+      "grad_norm": 0.0015876221004873514,
+      "learning_rate": 3.8297714198834414e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10048952,
+      "step": 12860
+    },
+    {
+      "epoch": 26.746361746361746,
+      "grad_norm": 0.000208708705031313,
+      "learning_rate": 3.828939971655595e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10052856,
+      "step": 12865
+    },
+    {
+      "epoch": 26.756756756756758,
+      "grad_norm": 0.002927327761426568,
+      "learning_rate": 3.828108318488505e-05,
+      "loss": 0.0156,
+      "num_input_tokens_seen": 10056760,
+      "step": 12870
+    },
+    {
+      "epoch": 26.767151767151766,
+      "grad_norm": 0.001364299445413053,
+      "learning_rate": 3.8272764605104216e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10060632,
+      "step": 12875
+    },
+    {
+      "epoch": 26.777546777546778,
+      "grad_norm": 0.0033943841699510813,
+      "learning_rate": 3.826444397849628e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10064440,
+      "step": 12880
+    },
+    {
+      "epoch": 26.787941787941786,
+      "grad_norm": 0.006886174902319908,
+      "learning_rate": 3.825612130634439e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10068184,
+      "step": 12885
+    },
+    {
+      "epoch": 26.7983367983368,
+      "grad_norm": 0.0013887407258152962,
+      "learning_rate": 3.824779658993202e-05,
+      "loss": 0.011,
+      "num_input_tokens_seen": 10072152,
+      "step": 12890
+    },
+    {
+      "epoch": 26.80873180873181,
+      "grad_norm": 0.0001387308439007029,
+      "learning_rate": 3.823946983054292e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10076088,
+      "step": 12895
+    },
+    {
+      "epoch": 26.81912681912682,
+      "grad_norm": 0.00020518679230008274,
+      "learning_rate": 3.82311410294612e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10079896,
+      "step": 12900
+    },
+    {
+      "epoch": 26.82952182952183,
+      "grad_norm": 0.1266907900571823,
+      "learning_rate": 3.822281018797127e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 10083896,
+      "step": 12905
+    },
+    {
+      "epoch": 26.83991683991684,
+      "grad_norm": 0.0002919606340583414,
+      "learning_rate": 3.821447730735783e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10087832,
+      "step": 12910
+    },
+    {
+      "epoch": 26.85031185031185,
+      "grad_norm": 5.18152883159928e-05,
+      "learning_rate": 3.820614238890592e-05,
+      "loss": 0.012,
+      "num_input_tokens_seen": 10091608,
+      "step": 12915
+    },
+    {
+      "epoch": 26.86070686070686,
+      "grad_norm": 0.004728805739432573,
+      "learning_rate": 3.819780543390091e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 10095608,
+      "step": 12920
+    },
+    {
+      "epoch": 26.87110187110187,
+      "grad_norm": 0.0017111690249294043,
+      "learning_rate": 3.818946644362844e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10099416,
+      "step": 12925
+    },
+    {
+      "epoch": 26.881496881496883,
+      "grad_norm": 0.00021728992578573525,
+      "learning_rate": 3.81811254193745e-05,
+      "loss": 0.0087,
+      "num_input_tokens_seen": 10103512,
+      "step": 12930
+    },
+    {
+      "epoch": 26.89189189189189,
+      "grad_norm": 0.0746416449546814,
+      "learning_rate": 3.8172782362425366e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 10107448,
+      "step": 12935
+    },
+    {
+      "epoch": 26.902286902286903,
+      "grad_norm": 0.0039043880533427,
+      "learning_rate": 3.816443727406765e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 10111256,
+      "step": 12940
+    },
+    {
+      "epoch": 26.91268191268191,
+      "grad_norm": 0.026268869638442993,
+      "learning_rate": 3.815609015558829e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10115128,
+      "step": 12945
+    },
+    {
+      "epoch": 26.923076923076923,
+      "grad_norm": 0.4189640283584595,
+      "learning_rate": 3.814774100827448e-05,
+      "loss": 0.0067,
+      "num_input_tokens_seen": 10118936,
+      "step": 12950
+    },
+    {
+      "epoch": 26.933471933471935,
+      "grad_norm": 0.000595387420617044,
+      "learning_rate": 3.813938983341379e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10122808,
+      "step": 12955
+    },
+    {
+      "epoch": 26.943866943866944,
+      "grad_norm": 0.4984425902366638,
+      "learning_rate": 3.813103663229407e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 10126552,
+      "step": 12960
+    },
+    {
+      "epoch": 26.954261954261955,
+      "grad_norm": 0.03181999921798706,
+      "learning_rate": 3.812268140620349e-05,
+      "loss": 0.0151,
+      "num_input_tokens_seen": 10130488,
+      "step": 12965
+    },
+    {
+      "epoch": 26.964656964656964,
+      "grad_norm": 7.459265179932117e-05,
+      "learning_rate": 3.811432415643051e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10134360,
+      "step": 12970
+    },
+    {
+      "epoch": 26.975051975051976,
+      "grad_norm": 1.0633781130309217e-05,
+      "learning_rate": 3.8105964884263954e-05,
+      "loss": 0.0074,
+      "num_input_tokens_seen": 10138200,
+      "step": 12975
+    },
+    {
+      "epoch": 26.985446985446984,
+      "grad_norm": 0.04983150213956833,
+      "learning_rate": 3.809760359099291e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10142168,
+      "step": 12980
+    },
+    {
+      "epoch": 26.995841995841996,
+      "grad_norm": 2.0360097551019862e-05,
+      "learning_rate": 3.8089240277906804e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 10146008,
+      "step": 12985
+    },
+    {
+      "epoch": 27.006237006237008,
+      "grad_norm": 0.0003351811319589615,
+      "learning_rate": 3.808087494629535e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10150056,
+      "step": 12990
+    },
+    {
+      "epoch": 27.016632016632016,
+      "grad_norm": 0.022291861474514008,
+      "learning_rate": 3.8072507597448595e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10153864,
+      "step": 12995
+    },
+    {
+      "epoch": 27.027027027027028,
+      "grad_norm": 0.50629723072052,
+      "learning_rate": 3.806413823265689e-05,
+      "loss": 0.008,
+      "num_input_tokens_seen": 10157800,
+      "step": 13000
+    },
+    {
+      "epoch": 27.027027027027028,
+      "eval_loss": 0.5093829035758972,
+      "eval_runtime": 11.7416,
+      "eval_samples_per_second": 72.903,
+      "eval_steps_per_second": 18.226,
+      "num_input_tokens_seen": 10157800,
+      "step": 13000
+    },
+    {
+      "epoch": 27.037422037422036,
+      "grad_norm": 5.165270340512507e-05,
+      "learning_rate": 3.805576685321089e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10161736,
+      "step": 13005
+    },
+    {
+      "epoch": 27.04781704781705,
+      "grad_norm": 0.0009045067708939314,
+      "learning_rate": 3.804739346040158e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10165640,
+      "step": 13010
+    },
+    {
+      "epoch": 27.058212058212057,
+      "grad_norm": 0.00016864111239556223,
+      "learning_rate": 3.8039018055520234e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10169544,
+      "step": 13015
+    },
+    {
+      "epoch": 27.06860706860707,
+      "grad_norm": 0.00041307308129034936,
+      "learning_rate": 3.803064063985844e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10173448,
+      "step": 13020
+    },
+    {
+      "epoch": 27.07900207900208,
+      "grad_norm": 5.7139009186357725e-06,
+      "learning_rate": 3.802226121470811e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10177128,
+      "step": 13025
+    },
+    {
+      "epoch": 27.08939708939709,
+      "grad_norm": 1.6268899344140664e-05,
+      "learning_rate": 3.801387978136145e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10180904,
+      "step": 13030
+    },
+    {
+      "epoch": 27.0997920997921,
+      "grad_norm": 14.961146354675293,
+      "learning_rate": 3.800549634111099e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 10184936,
+      "step": 13035
+    },
+    {
+      "epoch": 27.11018711018711,
+      "grad_norm": 0.24317745864391327,
+      "learning_rate": 3.799711089524955e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 10188776,
+      "step": 13040
+    },
+    {
+      "epoch": 27.12058212058212,
+      "grad_norm": 2.3059928935253993e-05,
+      "learning_rate": 3.7988723445070285e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10192616,
+      "step": 13045
+    },
+    {
+      "epoch": 27.13097713097713,
+      "grad_norm": 0.0017215340631082654,
+      "learning_rate": 3.798033399186663e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10196776,
+      "step": 13050
+    },
+    {
+      "epoch": 27.14137214137214,
+      "grad_norm": 0.00024086657504085451,
+      "learning_rate": 3.797194253693237e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 10200616,
+      "step": 13055
+    },
+    {
+      "epoch": 27.151767151767153,
+      "grad_norm": 0.000613218464422971,
+      "learning_rate": 3.796354908156153e-05,
+      "loss": 0.0167,
+      "num_input_tokens_seen": 10204392,
+      "step": 13060
+    },
+    {
+      "epoch": 27.16216216216216,
+      "grad_norm": 2.7620371838565916e-05,
+      "learning_rate": 3.795515362704853e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10208136,
+      "step": 13065
+    },
+    {
+      "epoch": 27.172557172557173,
+      "grad_norm": 1.3078595657134429e-05,
+      "learning_rate": 3.794675617468803e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10212072,
+      "step": 13070
+    },
+    {
+      "epoch": 27.18295218295218,
+      "grad_norm": 0.00012767172302119434,
+      "learning_rate": 3.793835672577503e-05,
+      "loss": 0.029,
+      "num_input_tokens_seen": 10216040,
+      "step": 13075
+    },
+    {
+      "epoch": 27.193347193347194,
+      "grad_norm": 0.2490777224302292,
+      "learning_rate": 3.7929955281604826e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 10219944,
+      "step": 13080
+    },
+    {
+      "epoch": 27.203742203742205,
+      "grad_norm": 0.0005962157156318426,
+      "learning_rate": 3.7921551843473036e-05,
+      "loss": 0.0227,
+      "num_input_tokens_seen": 10224008,
+      "step": 13085
+    },
+    {
+      "epoch": 27.214137214137214,
+      "grad_norm": 0.7550966143608093,
+      "learning_rate": 3.791314641267557e-05,
+      "loss": 0.0545,
+      "num_input_tokens_seen": 10227880,
+      "step": 13090
+    },
+    {
+      "epoch": 27.224532224532226,
+      "grad_norm": 0.01234173309057951,
+      "learning_rate": 3.790473899050864e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10231752,
+      "step": 13095
+    },
+    {
+      "epoch": 27.234927234927234,
+      "grad_norm": 0.006726630963385105,
+      "learning_rate": 3.7896329578268794e-05,
+      "loss": 0.0498,
+      "num_input_tokens_seen": 10235592,
+      "step": 13100
+    },
+    {
+      "epoch": 27.245322245322246,
+      "grad_norm": 0.0017125156009569764,
+      "learning_rate": 3.7887918177252855e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10239496,
+      "step": 13105
+    },
+    {
+      "epoch": 27.255717255717254,
+      "grad_norm": 0.32761499285697937,
+      "learning_rate": 3.787950478875798e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 10243336,
+      "step": 13110
+    },
+    {
+      "epoch": 27.266112266112266,
+      "grad_norm": 0.0017599640414118767,
+      "learning_rate": 3.787108941408162e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 10247080,
+      "step": 13115
+    },
+    {
+      "epoch": 27.276507276507278,
+      "grad_norm": 0.0002435842907289043,
+      "learning_rate": 3.786267205452151e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10251048,
+      "step": 13120
+    },
+    {
+      "epoch": 27.286902286902286,
+      "grad_norm": 0.00570564903318882,
+      "learning_rate": 3.785425271137573e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10255016,
+      "step": 13125
+    },
+    {
+      "epoch": 27.2972972972973,
+      "grad_norm": 0.003407764481380582,
+      "learning_rate": 3.7845831385942655e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10258856,
+      "step": 13130
+    },
+    {
+      "epoch": 27.307692307692307,
+      "grad_norm": 0.00914065632969141,
+      "learning_rate": 3.7837408079520944e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 10262632,
+      "step": 13135
+    },
+    {
+      "epoch": 27.31808731808732,
+      "grad_norm": 0.009358005598187447,
+      "learning_rate": 3.782898279340957e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10266632,
+      "step": 13140
+    },
+    {
+      "epoch": 27.328482328482327,
+      "grad_norm": 0.0024033368099480867,
+      "learning_rate": 3.782055552890784e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 10270536,
+      "step": 13145
+    },
+    {
+      "epoch": 27.33887733887734,
+      "grad_norm": 0.008353152312338352,
+      "learning_rate": 3.781212628731534e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 10274376,
+      "step": 13150
+    },
+    {
+      "epoch": 27.34927234927235,
+      "grad_norm": 0.009285354055464268,
+      "learning_rate": 3.7803695069931946e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10278248,
+      "step": 13155
+    },
+    {
+      "epoch": 27.35966735966736,
+      "grad_norm": 0.0046609933488070965,
+      "learning_rate": 3.779526187805789e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10282216,
+      "step": 13160
+    },
+    {
+      "epoch": 27.37006237006237,
+      "grad_norm": 0.0004199325921945274,
+      "learning_rate": 3.778682671299364e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10286056,
+      "step": 13165
+    },
+    {
+      "epoch": 27.38045738045738,
+      "grad_norm": 0.004431691020727158,
+      "learning_rate": 3.777838957604003e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10289960,
+      "step": 13170
+    },
+    {
+      "epoch": 27.39085239085239,
+      "grad_norm": 0.0034617604687809944,
+      "learning_rate": 3.776995046849816e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10293832,
+      "step": 13175
+    },
+    {
+      "epoch": 27.401247401247403,
+      "grad_norm": 0.009071651846170425,
+      "learning_rate": 3.776150939166945e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10297736,
+      "step": 13180
+    },
+    {
+      "epoch": 27.41164241164241,
+      "grad_norm": 0.00046101637417450547,
+      "learning_rate": 3.775306634685562e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10301608,
+      "step": 13185
+    },
+    {
+      "epoch": 27.422037422037423,
+      "grad_norm": 0.011331824585795403,
+      "learning_rate": 3.7744621335358696e-05,
+      "loss": 0.0245,
+      "num_input_tokens_seen": 10305416,
+      "step": 13190
+    },
+    {
+      "epoch": 27.43243243243243,
+      "grad_norm": 0.00021578306041192263,
+      "learning_rate": 3.7736174358481e-05,
+      "loss": 0.0121,
+      "num_input_tokens_seen": 10309256,
+      "step": 13195
+    },
+    {
+      "epoch": 27.442827442827443,
+      "grad_norm": 0.0037617203779518604,
+      "learning_rate": 3.7727725417525175e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10313128,
+      "step": 13200
+    },
+    {
+      "epoch": 27.442827442827443,
+      "eval_loss": 0.39982056617736816,
+      "eval_runtime": 11.7287,
+      "eval_samples_per_second": 72.984,
+      "eval_steps_per_second": 18.246,
+      "num_input_tokens_seen": 10313128,
+      "step": 13200
+    },
+    {
+      "epoch": 27.453222453222452,
+      "grad_norm": 0.008611735887825489,
+      "learning_rate": 3.771927451379414e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10317032,
+      "step": 13205
+    },
+    {
+      "epoch": 27.463617463617464,
+      "grad_norm": 0.0002831498277373612,
+      "learning_rate": 3.7710821648591135e-05,
+      "loss": 0.0413,
+      "num_input_tokens_seen": 10320936,
+      "step": 13210
+    },
+    {
+      "epoch": 27.474012474012476,
+      "grad_norm": 6.7324748039245605,
+      "learning_rate": 3.7702366823219694e-05,
+      "loss": 0.0289,
+      "num_input_tokens_seen": 10324776,
+      "step": 13215
+    },
+    {
+      "epoch": 27.484407484407484,
+      "grad_norm": 0.08898922801017761,
+      "learning_rate": 3.769391003898366e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10328712,
+      "step": 13220
+    },
+    {
+      "epoch": 27.494802494802496,
+      "grad_norm": 0.016431832686066628,
+      "learning_rate": 3.768545129718718e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10332584,
+      "step": 13225
+    },
+    {
+      "epoch": 27.505197505197504,
+      "grad_norm": 1.3069605827331543,
+      "learning_rate": 3.7676990599134686e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 10336552,
+      "step": 13230
+    },
+    {
+      "epoch": 27.515592515592516,
+      "grad_norm": 0.0043723089620471,
+      "learning_rate": 3.766852794613095e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10340552,
+      "step": 13235
+    },
+    {
+      "epoch": 27.525987525987524,
+      "grad_norm": 0.009219416417181492,
+      "learning_rate": 3.766006333948099e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10344488,
+      "step": 13240
+    },
+    {
+      "epoch": 27.536382536382536,
+      "grad_norm": 0.2589837610721588,
+      "learning_rate": 3.765159678049017e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 10348200,
+      "step": 13245
+    },
+    {
+      "epoch": 27.546777546777548,
+      "grad_norm": 0.2217240035533905,
+      "learning_rate": 3.7643128270464134e-05,
+      "loss": 0.0119,
+      "num_input_tokens_seen": 10352168,
+      "step": 13250
+    },
+    {
+      "epoch": 27.557172557172557,
+      "grad_norm": 0.014035699889063835,
+      "learning_rate": 3.763465781070884e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 10356040,
+      "step": 13255
+    },
+    {
+      "epoch": 27.56756756756757,
+      "grad_norm": 0.000802947033662349,
+      "learning_rate": 3.762618540253052e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10359912,
+      "step": 13260
+    },
+    {
+      "epoch": 27.577962577962577,
+      "grad_norm": 0.19370464980602264,
+      "learning_rate": 3.761771104723576e-05,
+      "loss": 0.0244,
+      "num_input_tokens_seen": 10363784,
+      "step": 13265
+    },
+    {
+      "epoch": 27.58835758835759,
+      "grad_norm": 0.009906965307891369,
+      "learning_rate": 3.7609234746131386e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 10367688,
+      "step": 13270
+    },
+    {
+      "epoch": 27.598752598752597,
+      "grad_norm": 0.009944583289325237,
+      "learning_rate": 3.7600756500524556e-05,
+      "loss": 0.0239,
+      "num_input_tokens_seen": 10371496,
+      "step": 13275
+    },
+    {
+      "epoch": 27.60914760914761,
+      "grad_norm": 0.02455483190715313,
+      "learning_rate": 3.759227631172271e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10375464,
+      "step": 13280
+    },
+    {
+      "epoch": 27.61954261954262,
+      "grad_norm": 0.004260304383933544,
+      "learning_rate": 3.758379418103363e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 10379304,
+      "step": 13285
+    },
+    {
+      "epoch": 27.62993762993763,
+      "grad_norm": 0.0008778144838288426,
+      "learning_rate": 3.757531010976534e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10383176,
+      "step": 13290
+    },
+    {
+      "epoch": 27.64033264033264,
+      "grad_norm": 0.0570853129029274,
+      "learning_rate": 3.75668240992262e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10387208,
+      "step": 13295
+    },
+    {
+      "epoch": 27.65072765072765,
+      "grad_norm": 0.0010933447629213333,
+      "learning_rate": 3.7558336150724865e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10391240,
+      "step": 13300
+    },
+    {
+      "epoch": 27.66112266112266,
+      "grad_norm": 0.009132804349064827,
+      "learning_rate": 3.754984626557028e-05,
+      "loss": 0.0683,
+      "num_input_tokens_seen": 10395144,
+      "step": 13305
+    },
+    {
+      "epoch": 27.671517671517673,
+      "grad_norm": 0.13214372098445892,
+      "learning_rate": 3.754135444507168e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10399208,
+      "step": 13310
+    },
+    {
+      "epoch": 27.68191268191268,
+      "grad_norm": 0.6014215350151062,
+      "learning_rate": 3.753286069053863e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 10403048,
+      "step": 13315
+    },
+    {
+      "epoch": 27.692307692307693,
+      "grad_norm": 0.7916502356529236,
+      "learning_rate": 3.7524365003280945e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 10406920,
+      "step": 13320
+    },
+    {
+      "epoch": 27.7027027027027,
+      "grad_norm": 0.18373142182826996,
+      "learning_rate": 3.75158673846088e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 10410920,
+      "step": 13325
+    },
+    {
+      "epoch": 27.713097713097714,
+      "grad_norm": 6.310108661651611,
+      "learning_rate": 3.750736783583262e-05,
+      "loss": 0.0393,
+      "num_input_tokens_seen": 10414632,
+      "step": 13330
+    },
+    {
+      "epoch": 27.723492723492722,
+      "grad_norm": 0.43808215856552124,
+      "learning_rate": 3.7498866358263144e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10418376,
+      "step": 13335
+    },
+    {
+      "epoch": 27.733887733887734,
+      "grad_norm": 0.0214722640812397,
+      "learning_rate": 3.74903629532114e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 10422216,
+      "step": 13340
+    },
+    {
+      "epoch": 27.744282744282746,
+      "grad_norm": 0.009390564635396004,
+      "learning_rate": 3.748185762198873e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10426184,
+      "step": 13345
+    },
+    {
+      "epoch": 27.754677754677754,
+      "grad_norm": 0.001108654891140759,
+      "learning_rate": 3.747335036590676e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 10430184,
+      "step": 13350
+    },
+    {
+      "epoch": 27.765072765072766,
+      "grad_norm": 0.03246646746993065,
+      "learning_rate": 3.7464841186277405e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 10434152,
+      "step": 13355
+    },
+    {
+      "epoch": 27.775467775467774,
+      "grad_norm": 0.038955315947532654,
+      "learning_rate": 3.7456330084412896e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 10438120,
+      "step": 13360
+    },
+    {
+      "epoch": 27.785862785862786,
+      "grad_norm": 0.0011374898022040725,
+      "learning_rate": 3.744781706162576e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10442024,
+      "step": 13365
+    },
+    {
+      "epoch": 27.796257796257795,
+      "grad_norm": 0.00045425377902574837,
+      "learning_rate": 3.743930211922879e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10445896,
+      "step": 13370
+    },
+    {
+      "epoch": 27.806652806652806,
+      "grad_norm": 0.131572887301445,
+      "learning_rate": 3.743078525853513e-05,
+      "loss": 0.0261,
+      "num_input_tokens_seen": 10449864,
+      "step": 13375
+    },
+    {
+      "epoch": 27.81704781704782,
+      "grad_norm": 0.0025252404157072306,
+      "learning_rate": 3.7422266480858154e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 10453800,
+      "step": 13380
+    },
+    {
+      "epoch": 27.827442827442827,
+      "grad_norm": 0.5964541435241699,
+      "learning_rate": 3.741374578751158e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 10457640,
+      "step": 13385
+    },
+    {
+      "epoch": 27.83783783783784,
+      "grad_norm": 0.001099314889870584,
+      "learning_rate": 3.740522317980941e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10461480,
+      "step": 13390
+    },
+    {
+      "epoch": 27.848232848232847,
+      "grad_norm": 0.0046085575595498085,
+      "learning_rate": 3.739669865906593e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10465416,
+      "step": 13395
+    },
+    {
+      "epoch": 27.85862785862786,
+      "grad_norm": 0.0009839776903390884,
+      "learning_rate": 3.738817222659573e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 10469384,
+      "step": 13400
+    },
+    {
+      "epoch": 27.85862785862786,
+      "eval_loss": 0.3729362189769745,
+      "eval_runtime": 11.7393,
+      "eval_samples_per_second": 72.917,
+      "eval_steps_per_second": 18.229,
+      "num_input_tokens_seen": 10469384,
+      "step": 13400
+    },
+    {
+      "epoch": 27.86902286902287,
+      "grad_norm": 0.0002493943611625582,
+      "learning_rate": 3.73796438837137e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10473256,
+      "step": 13405
+    },
+    {
+      "epoch": 27.87941787941788,
+      "grad_norm": 0.07127159088850021,
+      "learning_rate": 3.7371113631735e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10477064,
+      "step": 13410
+    },
+    {
+      "epoch": 27.88981288981289,
+      "grad_norm": 0.0048734997399151325,
+      "learning_rate": 3.736258147197512e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10480968,
+      "step": 13415
+    },
+    {
+      "epoch": 27.9002079002079,
+      "grad_norm": 0.1491585373878479,
+      "learning_rate": 3.735404740574981e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 10485032,
+      "step": 13420
+    },
+    {
+      "epoch": 27.91060291060291,
+      "grad_norm": 0.024489954113960266,
+      "learning_rate": 3.7345511434375145e-05,
+      "loss": 0.0302,
+      "num_input_tokens_seen": 10488872,
+      "step": 13425
+    },
+    {
+      "epoch": 27.92099792099792,
+      "grad_norm": 0.0003594648151192814,
+      "learning_rate": 3.733697355916748e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10492776,
+      "step": 13430
+    },
+    {
+      "epoch": 27.93139293139293,
+      "grad_norm": 0.000644454441498965,
+      "learning_rate": 3.732843378144345e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 10496616,
+      "step": 13435
+    },
+    {
+      "epoch": 27.941787941787943,
+      "grad_norm": 0.0010607321746647358,
+      "learning_rate": 3.7319892102519995e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10500520,
+      "step": 13440
+    },
+    {
+      "epoch": 27.95218295218295,
+      "grad_norm": 0.0048095001839101315,
+      "learning_rate": 3.731134852371436e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10504424,
+      "step": 13445
+    },
+    {
+      "epoch": 27.962577962577964,
+      "grad_norm": 0.0006881621666252613,
+      "learning_rate": 3.730280304634408e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10508264,
+      "step": 13450
+    },
+    {
+      "epoch": 27.972972972972972,
+      "grad_norm": 18.03462028503418,
+      "learning_rate": 3.729425567172696e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 10512264,
+      "step": 13455
+    },
+    {
+      "epoch": 27.983367983367984,
+      "grad_norm": 0.0009456913685426116,
+      "learning_rate": 3.728570640118111e-05,
+      "loss": 0.0075,
+      "num_input_tokens_seen": 10516168,
+      "step": 13460
+    },
+    {
+      "epoch": 27.993762993762992,
+      "grad_norm": 0.0009338217205367982,
+      "learning_rate": 3.727715523602494e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10520168,
+      "step": 13465
+    },
+    {
+      "epoch": 28.004158004158004,
+      "grad_norm": 0.002189803635701537,
+      "learning_rate": 3.726860217757715e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10524024,
+      "step": 13470
+    },
+    {
+      "epoch": 28.014553014553016,
+      "grad_norm": 0.2433677762746811,
+      "learning_rate": 3.726004722715673e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 10527864,
+      "step": 13475
+    },
+    {
+      "epoch": 28.024948024948024,
+      "grad_norm": 0.0008582946611568332,
+      "learning_rate": 3.725149038608296e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 10531768,
+      "step": 13480
+    },
+    {
+      "epoch": 28.035343035343036,
+      "grad_norm": 0.00047174099017865956,
+      "learning_rate": 3.7242931655675404e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10535640,
+      "step": 13485
+    },
+    {
+      "epoch": 28.045738045738045,
+      "grad_norm": 0.0038608035538345575,
+      "learning_rate": 3.7234371037253937e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 10539448,
+      "step": 13490
+    },
+    {
+      "epoch": 28.056133056133056,
+      "grad_norm": 0.0006925812922418118,
+      "learning_rate": 3.7225808532138705e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10543256,
+      "step": 13495
+    },
+    {
+      "epoch": 28.066528066528065,
+      "grad_norm": 0.001050072954967618,
+      "learning_rate": 3.721724414165016e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10547128,
+      "step": 13500
+    },
+    {
+      "epoch": 28.076923076923077,
+      "grad_norm": 0.03840238228440285,
+      "learning_rate": 3.720867786710904e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10551288,
+      "step": 13505
+    },
+    {
+      "epoch": 28.08731808731809,
+      "grad_norm": 0.0006207757396623492,
+      "learning_rate": 3.7200109709836366e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10555128,
+      "step": 13510
+    },
+    {
+      "epoch": 28.097713097713097,
+      "grad_norm": 0.15154850482940674,
+      "learning_rate": 3.7191539671153465e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10558904,
+      "step": 13515
+    },
+    {
+      "epoch": 28.10810810810811,
+      "grad_norm": 0.00249314634129405,
+      "learning_rate": 3.718296775238193e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10562936,
+      "step": 13520
+    },
+    {
+      "epoch": 28.118503118503117,
+      "grad_norm": 0.3670170307159424,
+      "learning_rate": 3.7174393954843675e-05,
+      "loss": 0.0075,
+      "num_input_tokens_seen": 10566808,
+      "step": 13525
+    },
+    {
+      "epoch": 28.12889812889813,
+      "grad_norm": 0.44398921728134155,
+      "learning_rate": 3.716581827986087e-05,
+      "loss": 0.006,
+      "num_input_tokens_seen": 10570936,
+      "step": 13530
+    },
+    {
+      "epoch": 28.13929313929314,
+      "grad_norm": 0.0015350010944530368,
+      "learning_rate": 3.7157240728756004e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 10574904,
+      "step": 13535
+    },
+    {
+      "epoch": 28.14968814968815,
+      "grad_norm": 0.0013419950846582651,
+      "learning_rate": 3.714866130285184e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10578968,
+      "step": 13540
+    },
+    {
+      "epoch": 28.16008316008316,
+      "grad_norm": 0.03639829158782959,
+      "learning_rate": 3.714008000347143e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10582680,
+      "step": 13545
+    },
+    {
+      "epoch": 28.17047817047817,
+      "grad_norm": 0.0007232759380713105,
+      "learning_rate": 3.7131496831938126e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 10586488,
+      "step": 13550
+    },
+    {
+      "epoch": 28.18087318087318,
+      "grad_norm": 0.0011682086624205112,
+      "learning_rate": 3.7122911789575565e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10590552,
+      "step": 13555
+    },
+    {
+      "epoch": 28.19126819126819,
+      "grad_norm": 0.000714315683580935,
+      "learning_rate": 3.711432487770765e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10594488,
+      "step": 13560
+    },
+    {
+      "epoch": 28.2016632016632,
+      "grad_norm": 0.011370931752026081,
+      "learning_rate": 3.710573609765861e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10598488,
+      "step": 13565
+    },
+    {
+      "epoch": 28.212058212058214,
+      "grad_norm": 0.00013477171887643635,
+      "learning_rate": 3.709714545075292e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10602392,
+      "step": 13570
+    },
+    {
+      "epoch": 28.222453222453222,
+      "grad_norm": 0.0001729174400679767,
+      "learning_rate": 3.708855293831538e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 10606360,
+      "step": 13575
+    },
+    {
+      "epoch": 28.232848232848234,
+      "grad_norm": 0.0019054394215345383,
+      "learning_rate": 3.707995856167107e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10610328,
+      "step": 13580
+    },
+    {
+      "epoch": 28.243243243243242,
+      "grad_norm": 0.001396474428474903,
+      "learning_rate": 3.707136232214534e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10614232,
+      "step": 13585
+    },
+    {
+      "epoch": 28.253638253638254,
+      "grad_norm": 0.3173108696937561,
+      "learning_rate": 3.7062764221063844e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 10618200,
+      "step": 13590
+    },
+    {
+      "epoch": 28.264033264033262,
+      "grad_norm": 0.033917300403118134,
+      "learning_rate": 3.705416425975252e-05,
+      "loss": 0.0045,
+      "num_input_tokens_seen": 10622104,
+      "step": 13595
+    },
+    {
+      "epoch": 28.274428274428274,
+      "grad_norm": 0.0018566518556326628,
+      "learning_rate": 3.704556243953758e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10625944,
+      "step": 13600
+    },
+    {
+      "epoch": 28.274428274428274,
+      "eval_loss": 0.4657649099826813,
+      "eval_runtime": 11.755,
+      "eval_samples_per_second": 72.82,
+      "eval_steps_per_second": 18.205,
+      "num_input_tokens_seen": 10625944,
+      "step": 13600
+    },
+    {
+      "epoch": 28.284823284823286,
+      "grad_norm": 0.1807466298341751,
+      "learning_rate": 3.7036958761745535e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 10629816,
+      "step": 13605
+    },
+    {
+      "epoch": 28.295218295218294,
+      "grad_norm": 0.008468594402074814,
+      "learning_rate": 3.702835322770318e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10633720,
+      "step": 13610
+    },
+    {
+      "epoch": 28.305613305613306,
+      "grad_norm": 0.00876565184444189,
+      "learning_rate": 3.701974583873761e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10637752,
+      "step": 13615
+    },
+    {
+      "epoch": 28.316008316008315,
+      "grad_norm": 0.0011609909124672413,
+      "learning_rate": 3.701113659617618e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10641688,
+      "step": 13620
+    },
+    {
+      "epoch": 28.326403326403327,
+      "grad_norm": 0.018613332882523537,
+      "learning_rate": 3.7002525501346535e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10645560,
+      "step": 13625
+    },
+    {
+      "epoch": 28.33679833679834,
+      "grad_norm": 0.008620056323707104,
+      "learning_rate": 3.699391255557664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10649528,
+      "step": 13630
+    },
+    {
+      "epoch": 28.347193347193347,
+      "grad_norm": 0.0007894114824011922,
+      "learning_rate": 3.69852977601947e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10653432,
+      "step": 13635
+    },
+    {
+      "epoch": 28.35758835758836,
+      "grad_norm": 0.0006244945107027888,
+      "learning_rate": 3.697668111652922e-05,
+      "loss": 0.019,
+      "num_input_tokens_seen": 10657304,
+      "step": 13640
+    },
+    {
+      "epoch": 28.367983367983367,
+      "grad_norm": 0.010529277846217155,
+      "learning_rate": 3.6968062625909005e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10661368,
+      "step": 13645
+    },
+    {
+      "epoch": 28.37837837837838,
+      "grad_norm": 0.004611719865351915,
+      "learning_rate": 3.6959442289663135e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10665304,
+      "step": 13650
+    },
+    {
+      "epoch": 28.388773388773387,
+      "grad_norm": 0.004102529026567936,
+      "learning_rate": 3.695082010912098e-05,
+      "loss": 0.0106,
+      "num_input_tokens_seen": 10669304,
+      "step": 13655
+    },
+    {
+      "epoch": 28.3991683991684,
+      "grad_norm": 0.0013222454581409693,
+      "learning_rate": 3.694219608561217e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 10673208,
+      "step": 13660
+    },
+    {
+      "epoch": 28.40956340956341,
+      "grad_norm": 0.0003221674123778939,
+      "learning_rate": 3.693357022046665e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 10677112,
+      "step": 13665
+    },
+    {
+      "epoch": 28.41995841995842,
+      "grad_norm": 0.023802954703569412,
+      "learning_rate": 3.6924942515014644e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 10681016,
+      "step": 13670
+    },
+    {
+      "epoch": 28.43035343035343,
+      "grad_norm": 0.007106301840394735,
+      "learning_rate": 3.691631297058664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10684888,
+      "step": 13675
+    },
+    {
+      "epoch": 28.44074844074844,
+      "grad_norm": 0.0002778360212687403,
+      "learning_rate": 3.6907681588513424e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10688824,
+      "step": 13680
+    },
+    {
+      "epoch": 28.45114345114345,
+      "grad_norm": 9.470462828176096e-05,
+      "learning_rate": 3.689904837012606e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10692568,
+      "step": 13685
+    },
+    {
+      "epoch": 28.46153846153846,
+      "grad_norm": 0.0012605079682543874,
+      "learning_rate": 3.689041331675591e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10696504,
+      "step": 13690
+    },
+    {
+      "epoch": 28.471933471933472,
+      "grad_norm": 0.0009403428412042558,
+      "learning_rate": 3.688177642973461e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 10700440,
+      "step": 13695
+    },
+    {
+      "epoch": 28.482328482328484,
+      "grad_norm": 0.00047889669076539576,
+      "learning_rate": 3.687313771039406e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10704376,
+      "step": 13700
+    },
+    {
+      "epoch": 28.492723492723492,
+      "grad_norm": 0.000195531829376705,
+      "learning_rate": 3.686449716006647e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10708248,
+      "step": 13705
+    },
+    {
+      "epoch": 28.503118503118504,
+      "grad_norm": 8.166990280151367,
+      "learning_rate": 3.685585478008432e-05,
+      "loss": 0.0418,
+      "num_input_tokens_seen": 10712024,
+      "step": 13710
+    },
+    {
+      "epoch": 28.513513513513512,
+      "grad_norm": 0.00032158478279598057,
+      "learning_rate": 3.6847210571780364e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10716056,
+      "step": 13715
+    },
+    {
+      "epoch": 28.523908523908524,
+      "grad_norm": 0.00911564938724041,
+      "learning_rate": 3.683856453648767e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 10720184,
+      "step": 13720
+    },
+    {
+      "epoch": 28.534303534303533,
+      "grad_norm": 0.19513629376888275,
+      "learning_rate": 3.682991667553954e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 10724024,
+      "step": 13725
+    },
+    {
+      "epoch": 28.544698544698544,
+      "grad_norm": 0.0016311685321852565,
+      "learning_rate": 3.6821266990269606e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10728024,
+      "step": 13730
+    },
+    {
+      "epoch": 28.555093555093556,
+      "grad_norm": 0.26107895374298096,
+      "learning_rate": 3.681261548201174e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 10731864,
+      "step": 13735
+    },
+    {
+      "epoch": 28.565488565488565,
+      "grad_norm": 0.11700304597616196,
+      "learning_rate": 3.6803962152100125e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 10735640,
+      "step": 13740
+    },
+    {
+      "epoch": 28.575883575883577,
+      "grad_norm": 0.00017559542902745306,
+      "learning_rate": 3.67953070018692e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10739544,
+      "step": 13745
+    },
+    {
+      "epoch": 28.586278586278585,
+      "grad_norm": 0.0010464024962857366,
+      "learning_rate": 3.678665003265371e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10743352,
+      "step": 13750
+    },
+    {
+      "epoch": 28.596673596673597,
+      "grad_norm": 0.0003325501747895032,
+      "learning_rate": 3.677799124578867e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10747192,
+      "step": 13755
+    },
+    {
+      "epoch": 28.60706860706861,
+      "grad_norm": 0.04188085347414017,
+      "learning_rate": 3.676933064260937e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10751160,
+      "step": 13760
+    },
+    {
+      "epoch": 28.617463617463617,
+      "grad_norm": 0.1060439944267273,
+      "learning_rate": 3.6760668224451365e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 10755032,
+      "step": 13765
+    },
+    {
+      "epoch": 28.62785862785863,
+      "grad_norm": 0.0002459993411321193,
+      "learning_rate": 3.675200399265054e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10758936,
+      "step": 13770
+    },
+    {
+      "epoch": 28.638253638253637,
+      "grad_norm": 0.0036996749695390463,
+      "learning_rate": 3.6743337948543014e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 10762776,
+      "step": 13775
+    },
+    {
+      "epoch": 28.64864864864865,
+      "grad_norm": 0.01225738599896431,
+      "learning_rate": 3.6734670093465204e-05,
+      "loss": 0.044,
+      "num_input_tokens_seen": 10766840,
+      "step": 13780
+    },
+    {
+      "epoch": 28.659043659043657,
+      "grad_norm": 0.01491709053516388,
+      "learning_rate": 3.672600042875379e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 10770648,
+      "step": 13785
+    },
+    {
+      "epoch": 28.66943866943867,
+      "grad_norm": 0.4516465365886688,
+      "learning_rate": 3.671732895574575e-05,
+      "loss": 0.0103,
+      "num_input_tokens_seen": 10774712,
+      "step": 13790
+    },
+    {
+      "epoch": 28.67983367983368,
+      "grad_norm": 0.02054547518491745,
+      "learning_rate": 3.670865567577834e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 10778552,
+      "step": 13795
+    },
+    {
+      "epoch": 28.69022869022869,
+      "grad_norm": 0.005614214576780796,
+      "learning_rate": 3.669998059018909e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 10782456,
+      "step": 13800
+    },
+    {
+      "epoch": 28.69022869022869,
+      "eval_loss": 0.3126485347747803,
+      "eval_runtime": 11.8064,
+      "eval_samples_per_second": 72.503,
+      "eval_steps_per_second": 18.126,
+      "num_input_tokens_seen": 10782456,
+      "step": 13800
+    },
+    {
+      "epoch": 28.7006237006237,
+      "grad_norm": 0.0025218119844794273,
+      "learning_rate": 3.6691303700315796e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10786328,
+      "step": 13805
+    },
+    {
+      "epoch": 28.71101871101871,
+      "grad_norm": 0.008610566146671772,
+      "learning_rate": 3.668262500749655e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10790328,
+      "step": 13810
+    },
+    {
+      "epoch": 28.72141372141372,
+      "grad_norm": 0.007141931913793087,
+      "learning_rate": 3.667394451306971e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10794232,
+      "step": 13815
+    },
+    {
+      "epoch": 28.731808731808734,
+      "grad_norm": 0.0033440005499869585,
+      "learning_rate": 3.666526221837393e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10798104,
+      "step": 13820
+    },
+    {
+      "epoch": 28.742203742203742,
+      "grad_norm": 0.021961510181427002,
+      "learning_rate": 3.665657812474812e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10802040,
+      "step": 13825
+    },
+    {
+      "epoch": 28.752598752598754,
+      "grad_norm": 0.0014173714444041252,
+      "learning_rate": 3.664789223353147e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10805976,
+      "step": 13830
+    },
+    {
+      "epoch": 28.762993762993762,
+      "grad_norm": 0.0029946775175631046,
+      "learning_rate": 3.663920454606347e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10809880,
+      "step": 13835
+    },
+    {
+      "epoch": 28.773388773388774,
+      "grad_norm": 0.026583010330796242,
+      "learning_rate": 3.6630515063683856e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 10813720,
+      "step": 13840
+    },
+    {
+      "epoch": 28.783783783783782,
+      "grad_norm": 0.0021463632583618164,
+      "learning_rate": 3.662182378773267e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 10817624,
+      "step": 13845
+    },
+    {
+      "epoch": 28.794178794178794,
+      "grad_norm": 0.0006469921790994704,
+      "learning_rate": 3.66131307195502e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 10821528,
+      "step": 13850
+    },
+    {
+      "epoch": 28.804573804573806,
+      "grad_norm": 0.0005891009932383895,
+      "learning_rate": 3.6604435860477034e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 10825400,
+      "step": 13855
+    },
+    {
+      "epoch": 28.814968814968815,
+      "grad_norm": 0.003359222551807761,
+      "learning_rate": 3.6595739211854025e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10829240,
+      "step": 13860
+    },
+    {
+      "epoch": 28.825363825363826,
+      "grad_norm": 0.004693704191595316,
+      "learning_rate": 3.658704077502231e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 10833112,
+      "step": 13865
+    },
+    {
+      "epoch": 28.835758835758835,
+      "grad_norm": 0.020937567576766014,
+      "learning_rate": 3.65783405513233e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10836984,
+      "step": 13870
+    },
+    {
+      "epoch": 28.846153846153847,
+      "grad_norm": 0.00022639281814917922,
+      "learning_rate": 3.656963854209867e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10840664,
+      "step": 13875
+    },
+    {
+      "epoch": 28.856548856548855,
+      "grad_norm": 0.0019807962235063314,
+      "learning_rate": 3.656093474869038e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10844504,
+      "step": 13880
+    },
+    {
+      "epoch": 28.866943866943867,
+      "grad_norm": 0.0020590494386851788,
+      "learning_rate": 3.655222917244068e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10848504,
+      "step": 13885
+    },
+    {
+      "epoch": 28.87733887733888,
+      "grad_norm": 2.8089332580566406,
+      "learning_rate": 3.6543521814692054e-05,
+      "loss": 0.0347,
+      "num_input_tokens_seen": 10852504,
+      "step": 13890
+    },
+    {
+      "epoch": 28.887733887733887,
+      "grad_norm": 11.378747940063477,
+      "learning_rate": 3.653481267678731e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 10856504,
+      "step": 13895
+    },
+    {
+      "epoch": 28.8981288981289,
+      "grad_norm": 0.002609043847769499,
+      "learning_rate": 3.652610176006949e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 10860280,
+      "step": 13900
+    },
+    {
+      "epoch": 28.908523908523907,
+      "grad_norm": 0.03447100520133972,
+      "learning_rate": 3.6517389065881925e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 10864248,
+      "step": 13905
+    },
+    {
+      "epoch": 28.91891891891892,
+      "grad_norm": 0.6630992889404297,
+      "learning_rate": 3.650867459556824e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 10868184,
+      "step": 13910
+    },
+    {
+      "epoch": 28.929313929313928,
+      "grad_norm": 0.0004895153688266873,
+      "learning_rate": 3.64999583504723e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10872056,
+      "step": 13915
+    },
+    {
+      "epoch": 28.93970893970894,
+      "grad_norm": 0.002246311167255044,
+      "learning_rate": 3.649124033193827e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10876024,
+      "step": 13920
+    },
+    {
+      "epoch": 28.95010395010395,
+      "grad_norm": 0.0002838648797478527,
+      "learning_rate": 3.648252054131057e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10879896,
+      "step": 13925
+    },
+    {
+      "epoch": 28.96049896049896,
+      "grad_norm": 0.10585273802280426,
+      "learning_rate": 3.647379897993391e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 10883736,
+      "step": 13930
+    },
+    {
+      "epoch": 28.97089397089397,
+      "grad_norm": 0.0013718486297875643,
+      "learning_rate": 3.646507564915325e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10887576,
+      "step": 13935
+    },
+    {
+      "epoch": 28.98128898128898,
+      "grad_norm": 0.00017019531514961272,
+      "learning_rate": 3.645635055031385e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10891544,
+      "step": 13940
+    },
+    {
+      "epoch": 28.991683991683992,
+      "grad_norm": 0.0012185325613245368,
+      "learning_rate": 3.6447623684761224e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10895384,
+      "step": 13945
+    },
+    {
+      "epoch": 29.002079002079004,
+      "grad_norm": 8.488646562909707e-05,
+      "learning_rate": 3.643889505384117e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10899296,
+      "step": 13950
+    },
+    {
+      "epoch": 29.012474012474012,
+      "grad_norm": 0.0006224823882803321,
+      "learning_rate": 3.6430164658899744e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 10903168,
+      "step": 13955
+    },
+    {
+      "epoch": 29.022869022869024,
+      "grad_norm": 0.00026295779389329255,
+      "learning_rate": 3.642143250128329e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10907072,
+      "step": 13960
+    },
+    {
+      "epoch": 29.033264033264032,
+      "grad_norm": 0.00017225489136762917,
+      "learning_rate": 3.641269858233841e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10911040,
+      "step": 13965
+    },
+    {
+      "epoch": 29.043659043659044,
+      "grad_norm": 0.0014017550274729729,
+      "learning_rate": 3.640396290341199e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10914848,
+      "step": 13970
+    },
+    {
+      "epoch": 29.054054054054053,
+      "grad_norm": 0.0004714478272944689,
+      "learning_rate": 3.639522546585118e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10918784,
+      "step": 13975
+    },
+    {
+      "epoch": 29.064449064449065,
+      "grad_norm": 0.00024439021944999695,
+      "learning_rate": 3.6386486271003404e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10922688,
+      "step": 13980
+    },
+    {
+      "epoch": 29.074844074844076,
+      "grad_norm": 0.00011284026550129056,
+      "learning_rate": 3.6377745320216346e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10926592,
+      "step": 13985
+    },
+    {
+      "epoch": 29.085239085239085,
+      "grad_norm": 0.02229120209813118,
+      "learning_rate": 3.636900261483798e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10930528,
+      "step": 13990
+    },
+    {
+      "epoch": 29.095634095634097,
+      "grad_norm": 0.00016478601901326329,
+      "learning_rate": 3.636025815621654e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10934528,
+      "step": 13995
+    },
+    {
+      "epoch": 29.106029106029105,
+      "grad_norm": 0.000634235271718353,
+      "learning_rate": 3.635151194570054e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10938304,
+      "step": 14000
+    },
+    {
+      "epoch": 29.106029106029105,
+      "eval_loss": 0.4559760093688965,
+      "eval_runtime": 11.7554,
+      "eval_samples_per_second": 72.817,
+      "eval_steps_per_second": 18.204,
+      "num_input_tokens_seen": 10938304,
+      "step": 14000
+    },
+    {
+      "epoch": 29.116424116424117,
+      "grad_norm": 0.0015800597611814737,
+      "learning_rate": 3.634276398463873e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10942112,
+      "step": 14005
+    },
+    {
+      "epoch": 29.126819126819125,
+      "grad_norm": 0.00022832861577626318,
+      "learning_rate": 3.633401427438018e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10946080,
+      "step": 14010
+    },
+    {
+      "epoch": 29.137214137214137,
+      "grad_norm": 0.0010723589221015573,
+      "learning_rate": 3.63252628162742e-05,
+      "loss": 0.007,
+      "num_input_tokens_seen": 10950080,
+      "step": 14015
+    },
+    {
+      "epoch": 29.14760914760915,
+      "grad_norm": 0.12162215262651443,
+      "learning_rate": 3.6316509611670364e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 10953952,
+      "step": 14020
+    },
+    {
+      "epoch": 29.158004158004157,
+      "grad_norm": 0.0001590830652276054,
+      "learning_rate": 3.630775466191854e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10957888,
+      "step": 14025
+    },
+    {
+      "epoch": 29.16839916839917,
+      "grad_norm": 0.000220654605072923,
+      "learning_rate": 3.629899796836884e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 10961664,
+      "step": 14030
+    },
+    {
+      "epoch": 29.178794178794178,
+      "grad_norm": 8.001969337463379,
+      "learning_rate": 3.6290239532371666e-05,
+      "loss": 0.0148,
+      "num_input_tokens_seen": 10965504,
+      "step": 14035
+    },
+    {
+      "epoch": 29.18918918918919,
+      "grad_norm": 0.0006194995949044824,
+      "learning_rate": 3.628147935527767e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10969376,
+      "step": 14040
+    },
+    {
+      "epoch": 29.1995841995842,
+      "grad_norm": 7.819995403289795,
+      "learning_rate": 3.627271743843779e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 10973280,
+      "step": 14045
+    },
+    {
+      "epoch": 29.20997920997921,
+      "grad_norm": 0.0016053746221587062,
+      "learning_rate": 3.626395378320321e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10977184,
+      "step": 14050
+    },
+    {
+      "epoch": 29.22037422037422,
+      "grad_norm": 0.0005530240596272051,
+      "learning_rate": 3.625518839092541e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 10981312,
+      "step": 14055
+    },
+    {
+      "epoch": 29.23076923076923,
+      "grad_norm": 0.0045501054264605045,
+      "learning_rate": 3.624642126295612e-05,
+      "loss": 0.0264,
+      "num_input_tokens_seen": 10985376,
+      "step": 14060
+    },
+    {
+      "epoch": 29.241164241164242,
+      "grad_norm": 0.0005321731441654265,
+      "learning_rate": 3.6237652400647345e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10989088,
+      "step": 14065
+    },
+    {
+      "epoch": 29.25155925155925,
+      "grad_norm": 0.0035862650256603956,
+      "learning_rate": 3.622888180535134e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 10993056,
+      "step": 14070
+    },
+    {
+      "epoch": 29.261954261954262,
+      "grad_norm": 0.2870819568634033,
+      "learning_rate": 3.6220109478420655e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 10996992,
+      "step": 14075
+    },
+    {
+      "epoch": 29.272349272349274,
+      "grad_norm": 0.005229281261563301,
+      "learning_rate": 3.6211335421208084e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11000832,
+      "step": 14080
+    },
+    {
+      "epoch": 29.282744282744282,
+      "grad_norm": 0.008985071443021297,
+      "learning_rate": 3.62025596350667e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 11004768,
+      "step": 14085
+    },
+    {
+      "epoch": 29.293139293139294,
+      "grad_norm": 0.0003884135221596807,
+      "learning_rate": 3.619378212134984e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 11008608,
+      "step": 14090
+    },
+    {
+      "epoch": 29.303534303534303,
+      "grad_norm": 0.0006716604111716151,
+      "learning_rate": 3.618500288141111e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 11012512,
+      "step": 14095
+    },
+    {
+      "epoch": 29.313929313929314,
+      "grad_norm": 0.00041130423778668046,
+      "learning_rate": 3.617622191660438e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11016384,
+      "step": 14100
+    },
+    {
+      "epoch": 29.324324324324323,
+      "grad_norm": 0.0004150904715061188,
+      "learning_rate": 3.616743922828377e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11020480,
+      "step": 14105
+    },
+    {
+      "epoch": 29.334719334719335,
+      "grad_norm": 0.0009282892569899559,
+      "learning_rate": 3.615865481780371e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 11024416,
+      "step": 14110
+    },
+    {
+      "epoch": 29.345114345114347,
+      "grad_norm": 0.001973488600924611,
+      "learning_rate": 3.614986868651883e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11028352,
+      "step": 14115
+    },
+    {
+      "epoch": 29.355509355509355,
+      "grad_norm": 0.0018863034201785922,
+      "learning_rate": 3.614108083578409e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11032384,
+      "step": 14120
+    },
+    {
+      "epoch": 29.365904365904367,
+      "grad_norm": 0.0004136479110457003,
+      "learning_rate": 3.613229126695467e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11036192,
+      "step": 14125
+    },
+    {
+      "epoch": 29.376299376299375,
+      "grad_norm": 0.0012232240987941623,
+      "learning_rate": 3.612349998138605e-05,
+      "loss": 0.0103,
+      "num_input_tokens_seen": 11040128,
+      "step": 14130
+    },
+    {
+      "epoch": 29.386694386694387,
+      "grad_norm": 0.001006035483442247,
+      "learning_rate": 3.6114706980433946e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11044032,
+      "step": 14135
+    },
+    {
+      "epoch": 29.397089397089395,
+      "grad_norm": 0.0002864337293431163,
+      "learning_rate": 3.610591226545435e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11048160,
+      "step": 14140
+    },
+    {
+      "epoch": 29.407484407484407,
+      "grad_norm": 0.0009622670477256179,
+      "learning_rate": 3.6097115837803505e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11052032,
+      "step": 14145
+    },
+    {
+      "epoch": 29.41787941787942,
+      "grad_norm": 0.00017375573224853724,
+      "learning_rate": 3.608831769883795e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11055936,
+      "step": 14150
+    },
+    {
+      "epoch": 29.428274428274428,
+      "grad_norm": 0.00156535638961941,
+      "learning_rate": 3.607951784991446e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 11059680,
+      "step": 14155
+    },
+    {
+      "epoch": 29.43866943866944,
+      "grad_norm": 0.0006230518338270485,
+      "learning_rate": 3.6070716292390085e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11063488,
+      "step": 14160
+    },
+    {
+      "epoch": 29.449064449064448,
+      "grad_norm": 0.0003842904407065362,
+      "learning_rate": 3.606191302762213e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11067392,
+      "step": 14165
+    },
+    {
+      "epoch": 29.45945945945946,
+      "grad_norm": 0.00017944994033314288,
+      "learning_rate": 3.605310805696818e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 11071360,
+      "step": 14170
+    },
+    {
+      "epoch": 29.46985446985447,
+      "grad_norm": 0.0007122172974050045,
+      "learning_rate": 3.6044301381786067e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11075136,
+      "step": 14175
+    },
+    {
+      "epoch": 29.48024948024948,
+      "grad_norm": 0.34698519110679626,
+      "learning_rate": 3.6035493003433883e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 11079072,
+      "step": 14180
+    },
+    {
+      "epoch": 29.490644490644492,
+      "grad_norm": 0.001599398790858686,
+      "learning_rate": 3.6026682923269994e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 11082880,
+      "step": 14185
+    },
+    {
+      "epoch": 29.5010395010395,
+      "grad_norm": 0.0011665400816127658,
+      "learning_rate": 3.6017871142653034e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11086592,
+      "step": 14190
+    },
+    {
+      "epoch": 29.511434511434512,
+      "grad_norm": 0.000571794284041971,
+      "learning_rate": 3.600905766294189e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11090528,
+      "step": 14195
+    },
+    {
+      "epoch": 29.52182952182952,
+      "grad_norm": 0.00012259937648195773,
+      "learning_rate": 3.60002424854957e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 11094528,
+      "step": 14200
+    },
+    {
+      "epoch": 29.52182952182952,
+      "eval_loss": 0.424032986164093,
+      "eval_runtime": 11.7937,
+      "eval_samples_per_second": 72.581,
+      "eval_steps_per_second": 18.145,
+      "num_input_tokens_seen": 11094528,
+      "step": 14200
+    },
+    {
+      "epoch": 29.532224532224532,
+      "grad_norm": 0.00019527811673469841,
+      "learning_rate": 3.5991425611673876e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11098688,
+      "step": 14205
+    },
+    {
+      "epoch": 29.542619542619544,
+      "grad_norm": 0.0005544907180592418,
+      "learning_rate": 3.5982607042836105e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11102560,
+      "step": 14210
+    },
+    {
+      "epoch": 29.553014553014552,
+      "grad_norm": 0.00023749440151732415,
+      "learning_rate": 3.597378678034231e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11106464,
+      "step": 14215
+    },
+    {
+      "epoch": 29.563409563409564,
+      "grad_norm": 0.00012644303205888718,
+      "learning_rate": 3.596496482555269e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11110272,
+      "step": 14220
+    },
+    {
+      "epoch": 29.573804573804573,
+      "grad_norm": 0.0006554077262990177,
+      "learning_rate": 3.595614117982769e-05,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 11114240,
+      "step": 14225
+    },
+    {
+      "epoch": 29.584199584199585,
+      "grad_norm": 0.0080733522772789,
+      "learning_rate": 3.594731584452805e-05,
+      "loss": 0.0059,
+      "num_input_tokens_seen": 11118208,
+      "step": 14230
+    },
+    {
+      "epoch": 29.594594594594593,
+      "grad_norm": 0.00012477456766646355,
+      "learning_rate": 3.593848882101472e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11122240,
+      "step": 14235
+    },
+    {
+      "epoch": 29.604989604989605,
+      "grad_norm": 0.0006544949137605727,
+      "learning_rate": 3.592966011064896e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11126304,
+      "step": 14240
+    },
+    {
+      "epoch": 29.615384615384617,
+      "grad_norm": 0.0005033321212977171,
+      "learning_rate": 3.592082971479226e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11130048,
+      "step": 14245
+    },
+    {
+      "epoch": 29.625779625779625,
+      "grad_norm": 0.00026588927721604705,
+      "learning_rate": 3.5911997634806385e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11133952,
+      "step": 14250
+    },
+    {
+      "epoch": 29.636174636174637,
+      "grad_norm": 0.0010071878787130117,
+      "learning_rate": 3.5903163872053336e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 11137952,
+      "step": 14255
+    },
+    {
+      "epoch": 29.646569646569645,
+      "grad_norm": 0.0004351470561232418,
+      "learning_rate": 3.58943284278954e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11141888,
+      "step": 14260
+    },
+    {
+      "epoch": 29.656964656964657,
+      "grad_norm": 0.0004755902918986976,
+      "learning_rate": 3.588549130369512e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11145824,
+      "step": 14265
+    },
+    {
+      "epoch": 29.66735966735967,
+      "grad_norm": 0.0003224911342840642,
+      "learning_rate": 3.5876652500815274e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11149632,
+      "step": 14270
+    },
+    {
+      "epoch": 29.677754677754677,
+      "grad_norm": 0.13542714715003967,
+      "learning_rate": 3.586781202061894e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 11153440,
+      "step": 14275
+    },
+    {
+      "epoch": 29.68814968814969,
+      "grad_norm": 0.00013101787772029638,
+      "learning_rate": 3.585896986446942e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11157440,
+      "step": 14280
+    },
+    {
+      "epoch": 29.698544698544698,
+      "grad_norm": 0.0003074249252676964,
+      "learning_rate": 3.585012603373028e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11161536,
+      "step": 14285
+    },
+    {
+      "epoch": 29.70893970893971,
+      "grad_norm": 0.00215039006434381,
+      "learning_rate": 3.584128052976535e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11165408,
+      "step": 14290
+    },
+    {
+      "epoch": 29.719334719334718,
+      "grad_norm": 0.00021237270266283303,
+      "learning_rate": 3.5832433353938724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11169312,
+      "step": 14295
+    },
+    {
+      "epoch": 29.72972972972973,
+      "grad_norm": 0.00037614183383993804,
+      "learning_rate": 3.5823584507614746e-05,
+      "loss": 0.007,
+      "num_input_tokens_seen": 11173120,
+      "step": 14300
+    },
+    {
+      "epoch": 29.74012474012474,
+      "grad_norm": 0.0018015889218077064,
+      "learning_rate": 3.581473399215802e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11177024,
+      "step": 14305
+    },
+    {
+      "epoch": 29.75051975051975,
+      "grad_norm": 0.003784690983593464,
+      "learning_rate": 3.580588180893341e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 11180928,
+      "step": 14310
+    },
+    {
+      "epoch": 29.760914760914762,
+      "grad_norm": 0.00017526319425087422,
+      "learning_rate": 3.579702795930602e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11184928,
+      "step": 14315
+    },
+    {
+      "epoch": 29.77130977130977,
+      "grad_norm": 0.0006926210480742157,
+      "learning_rate": 3.578817244464125e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11188800,
+      "step": 14320
+    },
+    {
+      "epoch": 29.781704781704782,
+      "grad_norm": 0.00016094221791718155,
+      "learning_rate": 3.577931526630471e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11192704,
+      "step": 14325
+    },
+    {
+      "epoch": 29.79209979209979,
+      "grad_norm": 0.0008274421561509371,
+      "learning_rate": 3.577045642566229e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11196544,
+      "step": 14330
+    },
+    {
+      "epoch": 29.802494802494802,
+      "grad_norm": 0.0015559374587610364,
+      "learning_rate": 3.576159592408014e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11200416,
+      "step": 14335
+    },
+    {
+      "epoch": 29.812889812889814,
+      "grad_norm": 0.0005259264726191759,
+      "learning_rate": 3.575273376292466e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11204320,
+      "step": 14340
+    },
+    {
+      "epoch": 29.823284823284823,
+      "grad_norm": 7.187703886302188e-05,
+      "learning_rate": 3.574386994356251e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11208128,
+      "step": 14345
+    },
+    {
+      "epoch": 29.833679833679835,
+      "grad_norm": 0.00016724728629924357,
+      "learning_rate": 3.573500446736059e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11211968,
+      "step": 14350
+    },
+    {
+      "epoch": 29.844074844074843,
+      "grad_norm": 0.00010536941408645362,
+      "learning_rate": 3.5726137335686094e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11215648,
+      "step": 14355
+    },
+    {
+      "epoch": 29.854469854469855,
+      "grad_norm": 6.564168870681897e-05,
+      "learning_rate": 3.571726854990642e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11219552,
+      "step": 14360
+    },
+    {
+      "epoch": 29.864864864864863,
+      "grad_norm": 0.00012686268019024283,
+      "learning_rate": 3.570839811138925e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11223520,
+      "step": 14365
+    },
+    {
+      "epoch": 29.875259875259875,
+      "grad_norm": 0.0019695565570145845,
+      "learning_rate": 3.569952602150252e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11227456,
+      "step": 14370
+    },
+    {
+      "epoch": 29.885654885654887,
+      "grad_norm": 0.00474350992590189,
+      "learning_rate": 3.569065228161442e-05,
+      "loss": 0.0069,
+      "num_input_tokens_seen": 11231456,
+      "step": 14375
+    },
+    {
+      "epoch": 29.896049896049895,
+      "grad_norm": 0.0006752132321707904,
+      "learning_rate": 3.5681776893093395e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11235392,
+      "step": 14380
+    },
+    {
+      "epoch": 29.906444906444907,
+      "grad_norm": 0.000556166167370975,
+      "learning_rate": 3.5672899857308134e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 11239360,
+      "step": 14385
+    },
+    {
+      "epoch": 29.916839916839916,
+      "grad_norm": 0.0008138582343235612,
+      "learning_rate": 3.566402117562759e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11243296,
+      "step": 14390
+    },
+    {
+      "epoch": 29.927234927234927,
+      "grad_norm": 0.0003883943718392402,
+      "learning_rate": 3.565514084942097e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11247136,
+      "step": 14395
+    },
+    {
+      "epoch": 29.93762993762994,
+      "grad_norm": 7.066935359034687e-05,
+      "learning_rate": 3.564625888005773e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 11250976,
+      "step": 14400
+    },
+    {
+      "epoch": 29.93762993762994,
+      "eval_loss": 0.45243746042251587,
+      "eval_runtime": 11.7493,
+      "eval_samples_per_second": 72.855,
+      "eval_steps_per_second": 18.214,
+      "num_input_tokens_seen": 11250976,
+      "step": 14400
+    },
+    {
+      "epoch": 29.948024948024948,
+      "grad_norm": 0.001605113735422492,
+      "learning_rate": 3.563737526890759e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11254848,
+      "step": 14405
+    },
+    {
+      "epoch": 29.95841995841996,
+      "grad_norm": 0.00021015891979914159,
+      "learning_rate": 3.562849001734049e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11258752,
+      "step": 14410
+    },
+    {
+      "epoch": 29.968814968814968,
+      "grad_norm": 0.00033790385350584984,
+      "learning_rate": 3.561960312672667e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 11262688,
+      "step": 14415
+    },
+    {
+      "epoch": 29.97920997920998,
+      "grad_norm": 0.0008364073000848293,
+      "learning_rate": 3.5610714598436596e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11266496,
+      "step": 14420
+    },
+    {
+      "epoch": 29.989604989604988,
+      "grad_norm": 0.00136928993742913,
+      "learning_rate": 3.5601824433840986e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11270304,
+      "step": 14425
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 0.0004381730395834893,
+      "learning_rate": 3.559293263431082e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11274032,
+      "step": 14430
+    },
+    {
+      "epoch": 30.010395010395012,
+      "grad_norm": 0.014508431777358055,
+      "learning_rate": 3.558403920121732e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 11277968,
+      "step": 14435
+    },
+    {
+      "epoch": 30.02079002079002,
+      "grad_norm": 0.0008141271537169814,
+      "learning_rate": 3.557514413593197e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11281840,
+      "step": 14440
+    },
+    {
+      "epoch": 30.031185031185032,
+      "grad_norm": 4.6345281589310616e-05,
+      "learning_rate": 3.55662474398265e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11285584,
+      "step": 14445
+    },
+    {
+      "epoch": 30.04158004158004,
+      "grad_norm": 0.0030319197103381157,
+      "learning_rate": 3.555734911427288e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11289424,
+      "step": 14450
+    },
+    {
+      "epoch": 30.051975051975052,
+      "grad_norm": 0.0018423431320115924,
+      "learning_rate": 3.5548449160643363e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11293424,
+      "step": 14455
+    },
+    {
+      "epoch": 30.06237006237006,
+      "grad_norm": 0.0025444994680583477,
+      "learning_rate": 3.553954758031043e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 11297328,
+      "step": 14460
+    },
+    {
+      "epoch": 30.072765072765073,
+      "grad_norm": 0.00016058332403190434,
+      "learning_rate": 3.5530644374646815e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11301328,
+      "step": 14465
+    },
+    {
+      "epoch": 30.083160083160084,
+      "grad_norm": 0.000473006977699697,
+      "learning_rate": 3.552173954502549e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11305136,
+      "step": 14470
+    },
+    {
+      "epoch": 30.093555093555093,
+      "grad_norm": 0.39435258507728577,
+      "learning_rate": 3.55128330928197e-05,
+      "loss": 0.0076,
+      "num_input_tokens_seen": 11309072,
+      "step": 14475
+    },
+    {
+      "epoch": 30.103950103950105,
+      "grad_norm": 0.00014465022832155228,
+      "learning_rate": 3.550392501940294e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11313168,
+      "step": 14480
+    },
+    {
+      "epoch": 30.114345114345113,
+      "grad_norm": 0.00016107432020362467,
+      "learning_rate": 3.5495015326148945e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11317104,
+      "step": 14485
+    },
+    {
+      "epoch": 30.124740124740125,
+      "grad_norm": 0.0002182667376473546,
+      "learning_rate": 3.548610401443169e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11320944,
+      "step": 14490
+    },
+    {
+      "epoch": 30.135135135135137,
+      "grad_norm": 0.12867315113544464,
+      "learning_rate": 3.547719108562543e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 11324816,
+      "step": 14495
+    },
+    {
+      "epoch": 30.145530145530145,
+      "grad_norm": 0.000132853165268898,
+      "learning_rate": 3.546827654110464e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11328688,
+      "step": 14500
+    },
+    {
+      "epoch": 30.155925155925157,
+      "grad_norm": 0.00015117677685339004,
+      "learning_rate": 3.545936038224405e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11332528,
+      "step": 14505
+    },
+    {
+      "epoch": 30.166320166320165,
+      "grad_norm": 0.00029568205354735255,
+      "learning_rate": 3.545044261041864e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11336432,
+      "step": 14510
+    },
+    {
+      "epoch": 30.176715176715177,
+      "grad_norm": 0.00010394130367785692,
+      "learning_rate": 3.5441523227003657e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11340400,
+      "step": 14515
+    },
+    {
+      "epoch": 30.187110187110186,
+      "grad_norm": 0.2796729505062103,
+      "learning_rate": 3.543260223337459e-05,
+      "loss": 0.0114,
+      "num_input_tokens_seen": 11344272,
+      "step": 14520
+    },
+    {
+      "epoch": 30.197505197505198,
+      "grad_norm": 0.00010421641491120681,
+      "learning_rate": 3.542367963090714e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11348304,
+      "step": 14525
+    },
+    {
+      "epoch": 30.20790020790021,
+      "grad_norm": 7.187626761151478e-05,
+      "learning_rate": 3.5414755420977295e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11352112,
+      "step": 14530
+    },
+    {
+      "epoch": 30.218295218295218,
+      "grad_norm": 0.2355155348777771,
+      "learning_rate": 3.54058296049613e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 11356080,
+      "step": 14535
+    },
+    {
+      "epoch": 30.22869022869023,
+      "grad_norm": 0.00016957285697571933,
+      "learning_rate": 3.53969021842356e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11359920,
+      "step": 14540
+    },
+    {
+      "epoch": 30.239085239085238,
+      "grad_norm": 4.610879113897681e-05,
+      "learning_rate": 3.5387973160176926e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11363824,
+      "step": 14545
+    },
+    {
+      "epoch": 30.24948024948025,
+      "grad_norm": 0.0012686606496572495,
+      "learning_rate": 3.537904253416224e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11367728,
+      "step": 14550
+    },
+    {
+      "epoch": 30.25987525987526,
+      "grad_norm": 0.00027616621810011566,
+      "learning_rate": 3.537011030756878e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11371664,
+      "step": 14555
+    },
+    {
+      "epoch": 30.27027027027027,
+      "grad_norm": 0.11804260313510895,
+      "learning_rate": 3.536117648177399e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 11375536,
+      "step": 14560
+    },
+    {
+      "epoch": 30.280665280665282,
+      "grad_norm": 0.0001901845826068893,
+      "learning_rate": 3.535224105815558e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11379440,
+      "step": 14565
+    },
+    {
+      "epoch": 30.29106029106029,
+      "grad_norm": 3.0409226383198984e-05,
+      "learning_rate": 3.5343304038091494e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11383280,
+      "step": 14570
+    },
+    {
+      "epoch": 30.301455301455302,
+      "grad_norm": 0.0009174898150376976,
+      "learning_rate": 3.5334365422959955e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11387184,
+      "step": 14575
+    },
+    {
+      "epoch": 30.31185031185031,
+      "grad_norm": 0.0003063045733142644,
+      "learning_rate": 3.5325425214139396e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11391024,
+      "step": 14580
+    },
+    {
+      "epoch": 30.322245322245323,
+      "grad_norm": 2.370682341279462e-05,
+      "learning_rate": 3.531648341300851e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11395024,
+      "step": 14585
+    },
+    {
+      "epoch": 30.33264033264033,
+      "grad_norm": 0.0004984554834663868,
+      "learning_rate": 3.530754002094623e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11398928,
+      "step": 14590
+    },
+    {
+      "epoch": 30.343035343035343,
+      "grad_norm": 0.0001063265954144299,
+      "learning_rate": 3.529859503933175e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11402704,
+      "step": 14595
+    },
+    {
+      "epoch": 30.353430353430355,
+      "grad_norm": 0.00022724208247382194,
+      "learning_rate": 3.52896484695445e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11406672,
+      "step": 14600
+    },
+    {
+      "epoch": 30.353430353430355,
+      "eval_loss": 0.46896228194236755,
+      "eval_runtime": 11.7524,
+      "eval_samples_per_second": 72.836,
+      "eval_steps_per_second": 18.209,
+      "num_input_tokens_seen": 11406672,
+      "step": 14600
+    },
+    {
+      "epoch": 30.363825363825363,
+      "grad_norm": 5.86502646910958e-05,
+      "learning_rate": 3.528070031296414e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11410544,
+      "step": 14605
+    },
+    {
+      "epoch": 30.374220374220375,
+      "grad_norm": 0.00011793922749347985,
+      "learning_rate": 3.5271750570970605e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 11414352,
+      "step": 14610
+    },
+    {
+      "epoch": 30.384615384615383,
+      "grad_norm": 0.0003722245164681226,
+      "learning_rate": 3.526279924494405e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 11418384,
+      "step": 14615
+    },
+    {
+      "epoch": 30.395010395010395,
+      "grad_norm": 0.00015445346070919186,
+      "learning_rate": 3.5253846336264874e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11422160,
+      "step": 14620
+    },
+    {
+      "epoch": 30.405405405405407,
+      "grad_norm": 0.00012961190077476203,
+      "learning_rate": 3.5244891846313736e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11425968,
+      "step": 14625
+    },
+    {
+      "epoch": 30.415800415800415,
+      "grad_norm": 0.21768450736999512,
+      "learning_rate": 3.5235935776471527e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 11429808,
+      "step": 14630
+    },
+    {
+      "epoch": 30.426195426195427,
+      "grad_norm": 8.320553752128035e-05,
+      "learning_rate": 3.522697812811939e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11433968,
+      "step": 14635
+    },
+    {
+      "epoch": 30.436590436590436,
+      "grad_norm": 0.002030730713158846,
+      "learning_rate": 3.521801890263871e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11437840,
+      "step": 14640
+    },
+    {
+      "epoch": 30.446985446985448,
+      "grad_norm": 0.00024158760788850486,
+      "learning_rate": 3.5209058101411114e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 11441808,
+      "step": 14645
+    },
+    {
+      "epoch": 30.457380457380456,
+      "grad_norm": 6.259041401790455e-05,
+      "learning_rate": 3.520009572581845e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11445648,
+      "step": 14650
+    },
+    {
+      "epoch": 30.467775467775468,
+      "grad_norm": 2.870574280677829e-05,
+      "learning_rate": 3.519113177724285e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11449488,
+      "step": 14655
+    },
+    {
+      "epoch": 30.47817047817048,
+      "grad_norm": 0.0002760708739515394,
+      "learning_rate": 3.5182166257066656e-05,
+      "loss": 0.0526,
+      "num_input_tokens_seen": 11453456,
+      "step": 14660
+    },
+    {
+      "epoch": 30.488565488565488,
+      "grad_norm": 0.01138673722743988,
+      "learning_rate": 3.517319916667247e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11457424,
+      "step": 14665
+    },
+    {
+      "epoch": 30.4989604989605,
+      "grad_norm": 0.020469896495342255,
+      "learning_rate": 3.516423050744313e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11461264,
+      "step": 14670
+    },
+    {
+      "epoch": 30.509355509355508,
+      "grad_norm": 0.001068352023139596,
+      "learning_rate": 3.5155260280761704e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11465104,
+      "step": 14675
+    },
+    {
+      "epoch": 30.51975051975052,
+      "grad_norm": 0.00361792859621346,
+      "learning_rate": 3.514628848801154e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11469072,
+      "step": 14680
+    },
+    {
+      "epoch": 30.53014553014553,
+      "grad_norm": 0.016692833974957466,
+      "learning_rate": 3.5137315130576174e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11473008,
+      "step": 14685
+    },
+    {
+      "epoch": 30.54054054054054,
+      "grad_norm": 0.0041331215761601925,
+      "learning_rate": 3.512834020983942e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11476784,
+      "step": 14690
+    },
+    {
+      "epoch": 30.550935550935552,
+      "grad_norm": 0.0003548905660863966,
+      "learning_rate": 3.5119363727185334e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11480720,
+      "step": 14695
+    },
+    {
+      "epoch": 30.56133056133056,
+      "grad_norm": 0.0007056756876409054,
+      "learning_rate": 3.511038568399819e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11484752,
+      "step": 14700
+    },
+    {
+      "epoch": 30.571725571725572,
+      "grad_norm": 0.0022967832628637552,
+      "learning_rate": 3.510140608166251e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 11488720,
+      "step": 14705
+    },
+    {
+      "epoch": 30.58212058212058,
+      "grad_norm": 0.003185019828379154,
+      "learning_rate": 3.509242492156308e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11492624,
+      "step": 14710
+    },
+    {
+      "epoch": 30.592515592515593,
+      "grad_norm": 0.0011030326131731272,
+      "learning_rate": 3.5083442205084896e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11496400,
+      "step": 14715
+    },
+    {
+      "epoch": 30.602910602910605,
+      "grad_norm": 0.001722104148939252,
+      "learning_rate": 3.507445793361321e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11500304,
+      "step": 14720
+    },
+    {
+      "epoch": 30.613305613305613,
+      "grad_norm": 0.0037831426598131657,
+      "learning_rate": 3.5065472108533505e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 11504464,
+      "step": 14725
+    },
+    {
+      "epoch": 30.623700623700625,
+      "grad_norm": 0.001482873223721981,
+      "learning_rate": 3.5056484731231504e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11508304,
+      "step": 14730
+    },
+    {
+      "epoch": 30.634095634095633,
+      "grad_norm": 0.002775635337457061,
+      "learning_rate": 3.504749580309319e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11512112,
+      "step": 14735
+    },
+    {
+      "epoch": 30.644490644490645,
+      "grad_norm": 0.0011327258544042706,
+      "learning_rate": 3.5038505325504753e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11515952,
+      "step": 14740
+    },
+    {
+      "epoch": 30.654885654885653,
+      "grad_norm": 0.00993194617331028,
+      "learning_rate": 3.502951329985264e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11519920,
+      "step": 14745
+    },
+    {
+      "epoch": 30.665280665280665,
+      "grad_norm": 0.0017168084159493446,
+      "learning_rate": 3.502051972752354e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 11523792,
+      "step": 14750
+    },
+    {
+      "epoch": 30.675675675675677,
+      "grad_norm": 0.016496583819389343,
+      "learning_rate": 3.5011524609904374e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11527664,
+      "step": 14755
+    },
+    {
+      "epoch": 30.686070686070686,
+      "grad_norm": 0.0008128808694891632,
+      "learning_rate": 3.50025279483823e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11531440,
+      "step": 14760
+    },
+    {
+      "epoch": 30.696465696465697,
+      "grad_norm": 0.001657457323744893,
+      "learning_rate": 3.499352974434472e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11535472,
+      "step": 14765
+    },
+    {
+      "epoch": 30.706860706860706,
+      "grad_norm": 0.00025679482496343553,
+      "learning_rate": 3.498452999917926e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11539312,
+      "step": 14770
+    },
+    {
+      "epoch": 30.717255717255718,
+      "grad_norm": 0.001320001669228077,
+      "learning_rate": 3.4975528714273795e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11543312,
+      "step": 14775
+    },
+    {
+      "epoch": 30.727650727650726,
+      "grad_norm": 0.0013582982355728745,
+      "learning_rate": 3.4966525891016454e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 11547216,
+      "step": 14780
+    },
+    {
+      "epoch": 30.738045738045738,
+      "grad_norm": 0.010473170317709446,
+      "learning_rate": 3.495752153079557e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11551120,
+      "step": 14785
+    },
+    {
+      "epoch": 30.74844074844075,
+      "grad_norm": 0.40787678956985474,
+      "learning_rate": 3.494851563499974e-05,
+      "loss": 0.0338,
+      "num_input_tokens_seen": 11554960,
+      "step": 14790
+    },
+    {
+      "epoch": 30.758835758835758,
+      "grad_norm": 0.12933401763439178,
+      "learning_rate": 3.493950820501777e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 11558896,
+      "step": 14795
+    },
+    {
+      "epoch": 30.76923076923077,
+      "grad_norm": 0.00018289896252099425,
+      "learning_rate": 3.493049924223872e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11562768,
+      "step": 14800
+    },
+    {
+      "epoch": 30.76923076923077,
+      "eval_loss": 0.41134947538375854,
+      "eval_runtime": 11.7862,
+      "eval_samples_per_second": 72.628,
+      "eval_steps_per_second": 18.157,
+      "num_input_tokens_seen": 11562768,
+      "step": 14800
+    },
+    {
+      "epoch": 30.77962577962578,
+      "grad_norm": 0.00024908612249419093,
+      "learning_rate": 3.49214887480519e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11566896,
+      "step": 14805
+    },
+    {
+      "epoch": 30.79002079002079,
+      "grad_norm": 0.0006151153356768191,
+      "learning_rate": 3.4912476723846834e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11570672,
+      "step": 14810
+    },
+    {
+      "epoch": 30.8004158004158,
+      "grad_norm": 35.7307243347168,
+      "learning_rate": 3.490346317101328e-05,
+      "loss": 0.0083,
+      "num_input_tokens_seen": 11574512,
+      "step": 14815
+    },
+    {
+      "epoch": 30.81081081081081,
+      "grad_norm": 0.0008498700917698443,
+      "learning_rate": 3.4894448090941266e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11578384,
+      "step": 14820
+    },
+    {
+      "epoch": 30.821205821205822,
+      "grad_norm": 0.0015044535975903273,
+      "learning_rate": 3.488543148502101e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11582224,
+      "step": 14825
+    },
+    {
+      "epoch": 30.83160083160083,
+      "grad_norm": 0.0003192209987901151,
+      "learning_rate": 3.487641335464299e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11586096,
+      "step": 14830
+    },
+    {
+      "epoch": 30.841995841995843,
+      "grad_norm": 20.570926666259766,
+      "learning_rate": 3.4867393701197914e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 11590000,
+      "step": 14835
+    },
+    {
+      "epoch": 30.85239085239085,
+      "grad_norm": 0.00031892317929305136,
+      "learning_rate": 3.485837252607673e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11594032,
+      "step": 14840
+    },
+    {
+      "epoch": 30.862785862785863,
+      "grad_norm": 0.000570613716263324,
+      "learning_rate": 3.4849349830670615e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 11597968,
+      "step": 14845
+    },
+    {
+      "epoch": 30.873180873180875,
+      "grad_norm": 0.0016679925611242652,
+      "learning_rate": 3.4840325616370976e-05,
+      "loss": 0.0341,
+      "num_input_tokens_seen": 11601808,
+      "step": 14850
+    },
+    {
+      "epoch": 30.883575883575883,
+      "grad_norm": 1.4312427043914795,
+      "learning_rate": 3.483129988456947e-05,
+      "loss": 0.0234,
+      "num_input_tokens_seen": 11605808,
+      "step": 14855
+    },
+    {
+      "epoch": 30.893970893970895,
+      "grad_norm": 0.014804188162088394,
+      "learning_rate": 3.482227263665797e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 11609744,
+      "step": 14860
+    },
+    {
+      "epoch": 30.904365904365903,
+      "grad_norm": 0.011677231639623642,
+      "learning_rate": 3.48132438740286e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11613616,
+      "step": 14865
+    },
+    {
+      "epoch": 30.914760914760915,
+      "grad_norm": 0.0631636381149292,
+      "learning_rate": 3.48042135980737e-05,
+      "loss": 0.0047,
+      "num_input_tokens_seen": 11617616,
+      "step": 14870
+    },
+    {
+      "epoch": 30.925155925155924,
+      "grad_norm": 0.02818349376320839,
+      "learning_rate": 3.479518181018586e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 11621328,
+      "step": 14875
+    },
+    {
+      "epoch": 30.935550935550935,
+      "grad_norm": 0.17593304812908173,
+      "learning_rate": 3.4786148511757886e-05,
+      "loss": 0.0165,
+      "num_input_tokens_seen": 11625232,
+      "step": 14880
+    },
+    {
+      "epoch": 30.945945945945947,
+      "grad_norm": 0.016770778223872185,
+      "learning_rate": 3.477711370418284e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11629040,
+      "step": 14885
+    },
+    {
+      "epoch": 30.956340956340956,
+      "grad_norm": 0.055953070521354675,
+      "learning_rate": 3.476807738885399e-05,
+      "loss": 0.0075,
+      "num_input_tokens_seen": 11632848,
+      "step": 14890
+    },
+    {
+      "epoch": 30.966735966735968,
+      "grad_norm": 0.004463621880859137,
+      "learning_rate": 3.475903956716485e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11636752,
+      "step": 14895
+    },
+    {
+      "epoch": 30.977130977130976,
+      "grad_norm": 0.0011694658314809203,
+      "learning_rate": 3.475000024050917e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11640528,
+      "step": 14900
+    },
+    {
+      "epoch": 30.987525987525988,
+      "grad_norm": 0.2576438784599304,
+      "learning_rate": 3.4740959410280926e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 11644432,
+      "step": 14905
+    },
+    {
+      "epoch": 30.997920997921,
+      "grad_norm": 0.219634011387825,
+      "learning_rate": 3.4731917077874324e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 11648368,
+      "step": 14910
+    },
+    {
+      "epoch": 31.008316008316008,
+      "grad_norm": 0.0059320759028196335,
+      "learning_rate": 3.4722873244683816e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11652136,
+      "step": 14915
+    },
+    {
+      "epoch": 31.01871101871102,
+      "grad_norm": 0.1860600709915161,
+      "learning_rate": 3.4713827912104065e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11656008,
+      "step": 14920
+    },
+    {
+      "epoch": 31.02910602910603,
+      "grad_norm": 0.0017055815551429987,
+      "learning_rate": 3.470478108152998e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11659880,
+      "step": 14925
+    },
+    {
+      "epoch": 31.03950103950104,
+      "grad_norm": 0.0035102562978863716,
+      "learning_rate": 3.4695732754356695e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11663784,
+      "step": 14930
+    },
+    {
+      "epoch": 31.04989604989605,
+      "grad_norm": 0.002301650820299983,
+      "learning_rate": 3.4686682931979576e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11667784,
+      "step": 14935
+    },
+    {
+      "epoch": 31.06029106029106,
+      "grad_norm": 0.00046447664499282837,
+      "learning_rate": 3.467763161579422e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11671752,
+      "step": 14940
+    },
+    {
+      "epoch": 31.070686070686072,
+      "grad_norm": 0.0004559301305562258,
+      "learning_rate": 3.466857880719645e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 11675720,
+      "step": 14945
+    },
+    {
+      "epoch": 31.08108108108108,
+      "grad_norm": 0.003737891325727105,
+      "learning_rate": 3.465952450758233e-05,
+      "loss": 0.0213,
+      "num_input_tokens_seen": 11679752,
+      "step": 14950
+    },
+    {
+      "epoch": 31.091476091476093,
+      "grad_norm": 0.00993003323674202,
+      "learning_rate": 3.4650468718348126e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 11683656,
+      "step": 14955
+    },
+    {
+      "epoch": 31.1018711018711,
+      "grad_norm": 0.038923587650060654,
+      "learning_rate": 3.464141144089038e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11687752,
+      "step": 14960
+    },
+    {
+      "epoch": 31.112266112266113,
+      "grad_norm": 10.148765563964844,
+      "learning_rate": 3.463235267660583e-05,
+      "loss": 0.042,
+      "num_input_tokens_seen": 11691528,
+      "step": 14965
+    },
+    {
+      "epoch": 31.12266112266112,
+      "grad_norm": 0.0007664096192456782,
+      "learning_rate": 3.462329242689145e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 11695464,
+      "step": 14970
+    },
+    {
+      "epoch": 31.133056133056133,
+      "grad_norm": 0.011204468086361885,
+      "learning_rate": 3.461423069314444e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 11699336,
+      "step": 14975
+    },
+    {
+      "epoch": 31.143451143451145,
+      "grad_norm": 0.00910489447414875,
+      "learning_rate": 3.460516747676224e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11703400,
+      "step": 14980
+    },
+    {
+      "epoch": 31.153846153846153,
+      "grad_norm": 0.008991115726530552,
+      "learning_rate": 3.459610277914251e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11707304,
+      "step": 14985
+    },
+    {
+      "epoch": 31.164241164241165,
+      "grad_norm": 0.0008429338922724128,
+      "learning_rate": 3.458703660168314e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11711208,
+      "step": 14990
+    },
+    {
+      "epoch": 31.174636174636174,
+      "grad_norm": 0.06838619709014893,
+      "learning_rate": 3.457796894578224e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11715080,
+      "step": 14995
+    },
+    {
+      "epoch": 31.185031185031185,
+      "grad_norm": 0.001317697111517191,
+      "learning_rate": 3.456889981283817e-05,
+      "loss": 0.0072,
+      "num_input_tokens_seen": 11719016,
+      "step": 15000
+    },
+    {
+      "epoch": 31.185031185031185,
+      "eval_loss": 0.36311855912208557,
+      "eval_runtime": 11.7239,
+      "eval_samples_per_second": 73.013,
+      "eval_steps_per_second": 18.253,
+      "num_input_tokens_seen": 11719016,
+      "step": 15000
+    },
+    {
+      "epoch": 31.195426195426194,
+      "grad_norm": 0.001368009834550321,
+      "learning_rate": 3.45598292042495e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11723016,
+      "step": 15005
+    },
+    {
+      "epoch": 31.205821205821206,
+      "grad_norm": 0.0005349669954739511,
+      "learning_rate": 3.4550757121415035e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11726888,
+      "step": 15010
+    },
+    {
+      "epoch": 31.216216216216218,
+      "grad_norm": 0.0016125019174069166,
+      "learning_rate": 3.454168356573378e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11730856,
+      "step": 15015
+    },
+    {
+      "epoch": 31.226611226611226,
+      "grad_norm": 0.0013526803813874722,
+      "learning_rate": 3.453260853860503e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11734760,
+      "step": 15020
+    },
+    {
+      "epoch": 31.237006237006238,
+      "grad_norm": 0.005583057180047035,
+      "learning_rate": 3.452353204142824e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11738632,
+      "step": 15025
+    },
+    {
+      "epoch": 31.247401247401246,
+      "grad_norm": 0.0012445137836039066,
+      "learning_rate": 3.4514454075603136e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11742664,
+      "step": 15030
+    },
+    {
+      "epoch": 31.257796257796258,
+      "grad_norm": 0.0024660369381308556,
+      "learning_rate": 3.450537464252964e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11746472,
+      "step": 15035
+    },
+    {
+      "epoch": 31.26819126819127,
+      "grad_norm": 0.0012339700479060411,
+      "learning_rate": 3.4496293743607925e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 11750344,
+      "step": 15040
+    },
+    {
+      "epoch": 31.27858627858628,
+      "grad_norm": 0.00033606673241592944,
+      "learning_rate": 3.448721138023838e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11754344,
+      "step": 15045
+    },
+    {
+      "epoch": 31.28898128898129,
+      "grad_norm": 0.001675982610322535,
+      "learning_rate": 3.447812755382162e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11758280,
+      "step": 15050
+    },
+    {
+      "epoch": 31.2993762993763,
+      "grad_norm": 0.020477140322327614,
+      "learning_rate": 3.446904226575847e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11762024,
+      "step": 15055
+    },
+    {
+      "epoch": 31.30977130977131,
+      "grad_norm": 0.026173904538154602,
+      "learning_rate": 3.445995551745002e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11765992,
+      "step": 15060
+    },
+    {
+      "epoch": 31.32016632016632,
+      "grad_norm": 0.0001670955534791574,
+      "learning_rate": 3.445086731029753e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11769608,
+      "step": 15065
+    },
+    {
+      "epoch": 31.33056133056133,
+      "grad_norm": 0.0013972127344459295,
+      "learning_rate": 3.444177764570255e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11773608,
+      "step": 15070
+    },
+    {
+      "epoch": 31.340956340956343,
+      "grad_norm": 0.00015790245379321277,
+      "learning_rate": 3.44326865250668e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 11777512,
+      "step": 15075
+    },
+    {
+      "epoch": 31.35135135135135,
+      "grad_norm": 0.04002118855714798,
+      "learning_rate": 3.442359394979225e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11781384,
+      "step": 15080
+    },
+    {
+      "epoch": 31.361746361746363,
+      "grad_norm": 0.00031966075766831636,
+      "learning_rate": 3.441449992128108e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11785256,
+      "step": 15085
+    },
+    {
+      "epoch": 31.37214137214137,
+      "grad_norm": 0.32756292819976807,
+      "learning_rate": 3.440540444093573e-05,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 11789160,
+      "step": 15090
+    },
+    {
+      "epoch": 31.382536382536383,
+      "grad_norm": 0.0006780122639611363,
+      "learning_rate": 3.43963075101588e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 11793064,
+      "step": 15095
+    },
+    {
+      "epoch": 31.39293139293139,
+      "grad_norm": 0.0014072856865823269,
+      "learning_rate": 3.438720913035318e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11796872,
+      "step": 15100
+    },
+    {
+      "epoch": 31.403326403326403,
+      "grad_norm": 0.0002837973879650235,
+      "learning_rate": 3.437810930292195e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11800744,
+      "step": 15105
+    },
+    {
+      "epoch": 31.413721413721415,
+      "grad_norm": 0.0012612058781087399,
+      "learning_rate": 3.43690080292684e-05,
+      "loss": 0.0067,
+      "num_input_tokens_seen": 11804648,
+      "step": 15110
+    },
+    {
+      "epoch": 31.424116424116423,
+      "grad_norm": 0.000335976219503209,
+      "learning_rate": 3.435990531079608e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11808552,
+      "step": 15115
+    },
+    {
+      "epoch": 31.434511434511435,
+      "grad_norm": 0.00042973709059879184,
+      "learning_rate": 3.435080114890874e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11812520,
+      "step": 15120
+    },
+    {
+      "epoch": 31.444906444906444,
+      "grad_norm": 0.00016908359248191118,
+      "learning_rate": 3.434169554501035e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11816456,
+      "step": 15125
+    },
+    {
+      "epoch": 31.455301455301456,
+      "grad_norm": 0.02905069664120674,
+      "learning_rate": 3.433258850050511e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 11820392,
+      "step": 15130
+    },
+    {
+      "epoch": 31.465696465696467,
+      "grad_norm": 0.0002472845371812582,
+      "learning_rate": 3.4323480016797446e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11824328,
+      "step": 15135
+    },
+    {
+      "epoch": 31.476091476091476,
+      "grad_norm": 0.0002938270044978708,
+      "learning_rate": 3.4314370095291995e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 11828264,
+      "step": 15140
+    },
+    {
+      "epoch": 31.486486486486488,
+      "grad_norm": 0.00035041128285229206,
+      "learning_rate": 3.430525873739363e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11832200,
+      "step": 15145
+    },
+    {
+      "epoch": 31.496881496881496,
+      "grad_norm": 0.00019831134704872966,
+      "learning_rate": 3.429614594450743e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11836200,
+      "step": 15150
+    },
+    {
+      "epoch": 31.507276507276508,
+      "grad_norm": 0.0002681480546016246,
+      "learning_rate": 3.428703171803869e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11840168,
+      "step": 15155
+    },
+    {
+      "epoch": 31.517671517671516,
+      "grad_norm": 0.0001256130781257525,
+      "learning_rate": 3.4277916059392964e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11844072,
+      "step": 15160
+    },
+    {
+      "epoch": 31.528066528066528,
+      "grad_norm": 0.0003157881728839129,
+      "learning_rate": 3.426879896997598e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 11848008,
+      "step": 15165
+    },
+    {
+      "epoch": 31.53846153846154,
+      "grad_norm": 0.0023279583547264338,
+      "learning_rate": 3.425968045119372e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11851912,
+      "step": 15170
+    },
+    {
+      "epoch": 31.54885654885655,
+      "grad_norm": 0.000509297417011112,
+      "learning_rate": 3.425056050445237e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 11855816,
+      "step": 15175
+    },
+    {
+      "epoch": 31.55925155925156,
+      "grad_norm": 0.0006594771402888,
+      "learning_rate": 3.4241439131158336e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 11859784,
+      "step": 15180
+    },
+    {
+      "epoch": 31.56964656964657,
+      "grad_norm": 0.00034633270115591586,
+      "learning_rate": 3.423231633271825e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11863688,
+      "step": 15185
+    },
+    {
+      "epoch": 31.58004158004158,
+      "grad_norm": 0.0006508198566734791,
+      "learning_rate": 3.4223192110538985e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11867560,
+      "step": 15190
+    },
+    {
+      "epoch": 31.59043659043659,
+      "grad_norm": 2.8598326025530696e-05,
+      "learning_rate": 3.4214066466027575e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11871400,
+      "step": 15195
+    },
+    {
+      "epoch": 31.6008316008316,
+      "grad_norm": 0.003094115760177374,
+      "learning_rate": 3.4204939400591325e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11875368,
+      "step": 15200
+    },
+    {
+      "epoch": 31.6008316008316,
+      "eval_loss": 0.41432803869247437,
+      "eval_runtime": 11.7892,
+      "eval_samples_per_second": 72.609,
+      "eval_steps_per_second": 18.152,
+      "num_input_tokens_seen": 11875368,
+      "step": 15200
+    },
+    {
+      "epoch": 31.611226611226613,
+      "grad_norm": 0.001765470253303647,
+      "learning_rate": 3.419581091563775e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11879208,
+      "step": 15205
+    },
+    {
+      "epoch": 31.62162162162162,
+      "grad_norm": 0.00030917450203560293,
+      "learning_rate": 3.418668101257456e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 11882952,
+      "step": 15210
+    },
+    {
+      "epoch": 31.632016632016633,
+      "grad_norm": 0.011982299387454987,
+      "learning_rate": 3.417754969280971e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11886920,
+      "step": 15215
+    },
+    {
+      "epoch": 31.64241164241164,
+      "grad_norm": 0.003281707875430584,
+      "learning_rate": 3.416841695775137e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11890824,
+      "step": 15220
+    },
+    {
+      "epoch": 31.652806652806653,
+      "grad_norm": 0.00016572723689023405,
+      "learning_rate": 3.415928280880792e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 11894696,
+      "step": 15225
+    },
+    {
+      "epoch": 31.66320166320166,
+      "grad_norm": 0.00044659909326583147,
+      "learning_rate": 3.4150147247387965e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11898696,
+      "step": 15230
+    },
+    {
+      "epoch": 31.673596673596673,
+      "grad_norm": 0.0027608931995928288,
+      "learning_rate": 3.4141010274900306e-05,
+      "loss": 0.0514,
+      "num_input_tokens_seen": 11902728,
+      "step": 15235
+    },
+    {
+      "epoch": 31.683991683991685,
+      "grad_norm": 0.00398850254714489,
+      "learning_rate": 3.413187189275399e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11906664,
+      "step": 15240
+    },
+    {
+      "epoch": 31.694386694386694,
+      "grad_norm": 0.009197229519486427,
+      "learning_rate": 3.4122732102358265e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 11910568,
+      "step": 15245
+    },
+    {
+      "epoch": 31.704781704781706,
+      "grad_norm": 0.017483161762356758,
+      "learning_rate": 3.411359090512261e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 11914376,
+      "step": 15250
+    },
+    {
+      "epoch": 31.715176715176714,
+      "grad_norm": 0.0073971678502857685,
+      "learning_rate": 3.410444830245672e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 11918248,
+      "step": 15255
+    },
+    {
+      "epoch": 31.725571725571726,
+      "grad_norm": 0.00348267937079072,
+      "learning_rate": 3.409530429577048e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11922088,
+      "step": 15260
+    },
+    {
+      "epoch": 31.735966735966738,
+      "grad_norm": 0.2137698233127594,
+      "learning_rate": 3.408615888647402e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 11925928,
+      "step": 15265
+    },
+    {
+      "epoch": 31.746361746361746,
+      "grad_norm": 0.0025718535762280226,
+      "learning_rate": 3.4077012075977675e-05,
+      "loss": 0.0177,
+      "num_input_tokens_seen": 11929768,
+      "step": 15270
+    },
+    {
+      "epoch": 31.756756756756758,
+      "grad_norm": 0.9185968041419983,
+      "learning_rate": 3.4067863865692e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 11933544,
+      "step": 15275
+    },
+    {
+      "epoch": 31.767151767151766,
+      "grad_norm": 0.0017637010896578431,
+      "learning_rate": 3.4058714257027755e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11937576,
+      "step": 15280
+    },
+    {
+      "epoch": 31.777546777546778,
+      "grad_norm": 0.0025288225151598454,
+      "learning_rate": 3.404956325139594e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 11941480,
+      "step": 15285
+    },
+    {
+      "epoch": 31.787941787941786,
+      "grad_norm": 0.07465585321187973,
+      "learning_rate": 3.404041085020775e-05,
+      "loss": 0.0252,
+      "num_input_tokens_seen": 11945480,
+      "step": 15290
+    },
+    {
+      "epoch": 31.7983367983368,
+      "grad_norm": 0.0007514621829614043,
+      "learning_rate": 3.403125705487459e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 11949288,
+      "step": 15295
+    },
+    {
+      "epoch": 31.80873180873181,
+      "grad_norm": 0.0027914184611290693,
+      "learning_rate": 3.402210186680811e-05,
+      "loss": 0.0137,
+      "num_input_tokens_seen": 11953128,
+      "step": 15300
+    },
+    {
+      "epoch": 31.81912681912682,
+      "grad_norm": 0.018879299983382225,
+      "learning_rate": 3.4012945287420137e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11956968,
+      "step": 15305
+    },
+    {
+      "epoch": 31.82952182952183,
+      "grad_norm": 0.0013888070825487375,
+      "learning_rate": 3.400378731812274e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11960904,
+      "step": 15310
+    },
+    {
+      "epoch": 31.83991683991684,
+      "grad_norm": 6.313884258270264,
+      "learning_rate": 3.399462796032817e-05,
+      "loss": 0.0332,
+      "num_input_tokens_seen": 11964648,
+      "step": 15315
+    },
+    {
+      "epoch": 31.85031185031185,
+      "grad_norm": 0.0005295804585330188,
+      "learning_rate": 3.3985467215448954e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11968488,
+      "step": 15320
+    },
+    {
+      "epoch": 31.86070686070686,
+      "grad_norm": 16.277265548706055,
+      "learning_rate": 3.3976305084897776e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 11972328,
+      "step": 15325
+    },
+    {
+      "epoch": 31.87110187110187,
+      "grad_norm": 0.00171183655038476,
+      "learning_rate": 3.3967141570087544e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 11976296,
+      "step": 15330
+    },
+    {
+      "epoch": 31.881496881496883,
+      "grad_norm": 1.5770390033721924,
+      "learning_rate": 3.39579766724314e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 11980456,
+      "step": 15335
+    },
+    {
+      "epoch": 31.89189189189189,
+      "grad_norm": 0.0018684507813304663,
+      "learning_rate": 3.3948810393342677e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11984328,
+      "step": 15340
+    },
+    {
+      "epoch": 31.902286902286903,
+      "grad_norm": 0.0015248156851157546,
+      "learning_rate": 3.3939642734234936e-05,
+      "loss": 0.0232,
+      "num_input_tokens_seen": 11988200,
+      "step": 15345
+    },
+    {
+      "epoch": 31.91268191268191,
+      "grad_norm": 17.410343170166016,
+      "learning_rate": 3.393047369652194e-05,
+      "loss": 0.0186,
+      "num_input_tokens_seen": 11992264,
+      "step": 15350
+    },
+    {
+      "epoch": 31.923076923076923,
+      "grad_norm": 0.0006420798017643392,
+      "learning_rate": 3.3921303281617664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11996072,
+      "step": 15355
+    },
+    {
+      "epoch": 31.933471933471935,
+      "grad_norm": 0.001431394019164145,
+      "learning_rate": 3.391213149093632e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 11999912,
+      "step": 15360
+    },
+    {
+      "epoch": 31.943866943866944,
+      "grad_norm": 0.0018790598260238767,
+      "learning_rate": 3.3902958325892303e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12003848,
+      "step": 15365
+    },
+    {
+      "epoch": 31.954261954261955,
+      "grad_norm": 0.01927429996430874,
+      "learning_rate": 3.389378378790023e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12007752,
+      "step": 15370
+    },
+    {
+      "epoch": 31.964656964656964,
+      "grad_norm": 0.002277843654155731,
+      "learning_rate": 3.388460787837493e-05,
+      "loss": 0.0185,
+      "num_input_tokens_seen": 12011720,
+      "step": 15375
+    },
+    {
+      "epoch": 31.975051975051976,
+      "grad_norm": 0.0016122279921546578,
+      "learning_rate": 3.387543059873145e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 12015624,
+      "step": 15380
+    },
+    {
+      "epoch": 31.985446985446984,
+      "grad_norm": 0.001535531715489924,
+      "learning_rate": 3.386625195038503e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 12019560,
+      "step": 15385
+    },
+    {
+      "epoch": 31.995841995841996,
+      "grad_norm": 0.006215881090611219,
+      "learning_rate": 3.3857071934751136e-05,
+      "loss": 0.0079,
+      "num_input_tokens_seen": 12023304,
+      "step": 15390
+    },
+    {
+      "epoch": 32.00623700623701,
+      "grad_norm": 0.0007274384261108935,
+      "learning_rate": 3.384789055324544e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12027208,
+      "step": 15395
+    },
+    {
+      "epoch": 32.016632016632016,
+      "grad_norm": 0.0011402657255530357,
+      "learning_rate": 3.3838707807283843e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 12031048,
+      "step": 15400
+    },
+    {
+      "epoch": 32.016632016632016,
+      "eval_loss": 0.3420463502407074,
+      "eval_runtime": 11.7258,
+      "eval_samples_per_second": 73.001,
+      "eval_steps_per_second": 18.25,
+      "num_input_tokens_seen": 12031048,
+      "step": 15400
+    },
+    {
+      "epoch": 32.027027027027025,
+      "grad_norm": 0.001109330216422677,
+      "learning_rate": 3.382952369828243e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12034824,
+      "step": 15405
+    },
+    {
+      "epoch": 32.03742203742204,
+      "grad_norm": 0.0035345708020031452,
+      "learning_rate": 3.38203382276575e-05,
+      "loss": 0.0392,
+      "num_input_tokens_seen": 12038632,
+      "step": 15410
+    },
+    {
+      "epoch": 32.04781704781705,
+      "grad_norm": 0.0059506515972316265,
+      "learning_rate": 3.381115139682557e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 12042472,
+      "step": 15415
+    },
+    {
+      "epoch": 32.05821205821206,
+      "grad_norm": 0.006402590312063694,
+      "learning_rate": 3.3801963207203366e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12046376,
+      "step": 15420
+    },
+    {
+      "epoch": 32.06860706860707,
+      "grad_norm": 0.0032339817844331264,
+      "learning_rate": 3.379277366020782e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 12050376,
+      "step": 15425
+    },
+    {
+      "epoch": 32.07900207900208,
+      "grad_norm": 0.0003090435639023781,
+      "learning_rate": 3.3783582757256085e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12054408,
+      "step": 15430
+    },
+    {
+      "epoch": 32.08939708939709,
+      "grad_norm": 0.13267658650875092,
+      "learning_rate": 3.3774390499765504e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 12058248,
+      "step": 15435
+    },
+    {
+      "epoch": 32.0997920997921,
+      "grad_norm": 0.0030273515731096268,
+      "learning_rate": 3.376519688915364e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12062280,
+      "step": 15440
+    },
+    {
+      "epoch": 32.11018711018711,
+      "grad_norm": 0.010192153975367546,
+      "learning_rate": 3.3756001926838273e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 12066120,
+      "step": 15445
+    },
+    {
+      "epoch": 32.12058212058212,
+      "grad_norm": 0.0003069512313231826,
+      "learning_rate": 3.374680561423737e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12069960,
+      "step": 15450
+    },
+    {
+      "epoch": 32.13097713097713,
+      "grad_norm": 0.0054826620034873486,
+      "learning_rate": 3.373760795276912e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12073832,
+      "step": 15455
+    },
+    {
+      "epoch": 32.141372141372145,
+      "grad_norm": 0.05052020400762558,
+      "learning_rate": 3.372840894385192e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 12077736,
+      "step": 15460
+    },
+    {
+      "epoch": 32.15176715176715,
+      "grad_norm": 0.00023078716185409576,
+      "learning_rate": 3.3719208588904375e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12081576,
+      "step": 15465
+    },
+    {
+      "epoch": 32.16216216216216,
+      "grad_norm": 0.0015169100370258093,
+      "learning_rate": 3.371000688934529e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 12085480,
+      "step": 15470
+    },
+    {
+      "epoch": 32.17255717255717,
+      "grad_norm": 0.0021029312629252672,
+      "learning_rate": 3.370080384659369e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12089416,
+      "step": 15475
+    },
+    {
+      "epoch": 32.182952182952185,
+      "grad_norm": 0.0006517539150081575,
+      "learning_rate": 3.36915994620688e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12093320,
+      "step": 15480
+    },
+    {
+      "epoch": 32.19334719334719,
+      "grad_norm": 0.0008757871692068875,
+      "learning_rate": 3.3682393737190035e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12097352,
+      "step": 15485
+    },
+    {
+      "epoch": 32.2037422037422,
+      "grad_norm": 0.0013623530976474285,
+      "learning_rate": 3.3673186673377054e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12101160,
+      "step": 15490
+    },
+    {
+      "epoch": 32.21413721413722,
+      "grad_norm": 0.0016147700371220708,
+      "learning_rate": 3.366397827204969e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 12105064,
+      "step": 15495
+    },
+    {
+      "epoch": 32.224532224532226,
+      "grad_norm": 0.0005334567977115512,
+      "learning_rate": 3.3654768534628e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12108936,
+      "step": 15500
+    },
+    {
+      "epoch": 32.234927234927234,
+      "grad_norm": 0.0005200728191994131,
+      "learning_rate": 3.3645557462532245e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 12112904,
+      "step": 15505
+    },
+    {
+      "epoch": 32.24532224532224,
+      "grad_norm": 0.0005026936414651573,
+      "learning_rate": 3.363634505718288e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12116776,
+      "step": 15510
+    },
+    {
+      "epoch": 32.25571725571726,
+      "grad_norm": 0.0008229989907704294,
+      "learning_rate": 3.362713132000057e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12120680,
+      "step": 15515
+    },
+    {
+      "epoch": 32.266112266112266,
+      "grad_norm": 0.0008640118758194149,
+      "learning_rate": 3.36179162524062e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12124648,
+      "step": 15520
+    },
+    {
+      "epoch": 32.276507276507274,
+      "grad_norm": 0.0008969518239609897,
+      "learning_rate": 3.3608699855820846e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12128712,
+      "step": 15525
+    },
+    {
+      "epoch": 32.28690228690229,
+      "grad_norm": 0.000548822688870132,
+      "learning_rate": 3.359948213166578e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12132776,
+      "step": 15530
+    },
+    {
+      "epoch": 32.2972972972973,
+      "grad_norm": 0.0002978936245199293,
+      "learning_rate": 3.359026308136252e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12136776,
+      "step": 15535
+    },
+    {
+      "epoch": 32.30769230769231,
+      "grad_norm": 0.09358056634664536,
+      "learning_rate": 3.358104270633272e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12140616,
+      "step": 15540
+    },
+    {
+      "epoch": 32.318087318087315,
+      "grad_norm": 0.0009216090547852218,
+      "learning_rate": 3.357182100799831e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12144744,
+      "step": 15545
+    },
+    {
+      "epoch": 32.32848232848233,
+      "grad_norm": 0.001561944605782628,
+      "learning_rate": 3.3562597987781384e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12148520,
+      "step": 15550
+    },
+    {
+      "epoch": 32.33887733887734,
+      "grad_norm": 0.00034511846024543047,
+      "learning_rate": 3.355337364710424e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12152488,
+      "step": 15555
+    },
+    {
+      "epoch": 32.34927234927235,
+      "grad_norm": 0.0005188288632780313,
+      "learning_rate": 3.354414798738939e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12156520,
+      "step": 15560
+    },
+    {
+      "epoch": 32.35966735966736,
+      "grad_norm": 0.14977431297302246,
+      "learning_rate": 3.353492101005955e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 12160584,
+      "step": 15565
+    },
+    {
+      "epoch": 32.37006237006237,
+      "grad_norm": 0.0011108998442068696,
+      "learning_rate": 3.352569271653763e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12164456,
+      "step": 15570
+    },
+    {
+      "epoch": 32.38045738045738,
+      "grad_norm": 0.007542000617831945,
+      "learning_rate": 3.351646310824675e-05,
+      "loss": 0.0501,
+      "num_input_tokens_seen": 12168456,
+      "step": 15575
+    },
+    {
+      "epoch": 32.39085239085239,
+      "grad_norm": 0.031583186239004135,
+      "learning_rate": 3.350723218661023e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 12172200,
+      "step": 15580
+    },
+    {
+      "epoch": 32.4012474012474,
+      "grad_norm": 6.532576560974121,
+      "learning_rate": 3.349799995305162e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 12176008,
+      "step": 15585
+    },
+    {
+      "epoch": 32.41164241164241,
+      "grad_norm": 0.008115868084132671,
+      "learning_rate": 3.348876640899461e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12179752,
+      "step": 15590
+    },
+    {
+      "epoch": 32.42203742203742,
+      "grad_norm": 0.0011799208587035537,
+      "learning_rate": 3.3479531555863144e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12183656,
+      "step": 15595
+    },
+    {
+      "epoch": 32.432432432432435,
+      "grad_norm": 0.0036848627496510744,
+      "learning_rate": 3.3470295395081344e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12187432,
+      "step": 15600
+    },
+    {
+      "epoch": 32.432432432432435,
+      "eval_loss": 0.34015458822250366,
+      "eval_runtime": 11.8761,
+      "eval_samples_per_second": 72.077,
+      "eval_steps_per_second": 18.019,
+      "num_input_tokens_seen": 12187432,
+      "step": 15600
+    },
+    {
+      "epoch": 32.44282744282744,
+      "grad_norm": 0.0012579775648191571,
+      "learning_rate": 3.3461057928073556e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12191432,
+      "step": 15605
+    },
+    {
+      "epoch": 32.45322245322245,
+      "grad_norm": 0.0007859464967623353,
+      "learning_rate": 3.345181915626431e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 12195336,
+      "step": 15610
+    },
+    {
+      "epoch": 32.46361746361746,
+      "grad_norm": 0.001636948436498642,
+      "learning_rate": 3.344257908107834e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 12199336,
+      "step": 15615
+    },
+    {
+      "epoch": 32.474012474012476,
+      "grad_norm": 0.0016014029970392585,
+      "learning_rate": 3.343333770394058e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12203240,
+      "step": 15620
+    },
+    {
+      "epoch": 32.484407484407484,
+      "grad_norm": 0.0030354438349604607,
+      "learning_rate": 3.342409502627616e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12207208,
+      "step": 15625
+    },
+    {
+      "epoch": 32.49480249480249,
+      "grad_norm": 0.004749445244669914,
+      "learning_rate": 3.341485104951043e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12211144,
+      "step": 15630
+    },
+    {
+      "epoch": 32.50519750519751,
+      "grad_norm": 0.07623323053121567,
+      "learning_rate": 3.340560577506892e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 12214984,
+      "step": 15635
+    },
+    {
+      "epoch": 32.515592515592516,
+      "grad_norm": 0.0032414288725703955,
+      "learning_rate": 3.339635920437735e-05,
+      "loss": 0.0158,
+      "num_input_tokens_seen": 12218888,
+      "step": 15640
+    },
+    {
+      "epoch": 32.525987525987524,
+      "grad_norm": 0.003292031353339553,
+      "learning_rate": 3.338711133886169e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12222600,
+      "step": 15645
+    },
+    {
+      "epoch": 32.53638253638254,
+      "grad_norm": 0.002518410561606288,
+      "learning_rate": 3.3377862179948064e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 12226536,
+      "step": 15650
+    },
+    {
+      "epoch": 32.54677754677755,
+      "grad_norm": 0.0008377740741707385,
+      "learning_rate": 3.336861172906281e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 12230472,
+      "step": 15655
+    },
+    {
+      "epoch": 32.55717255717256,
+      "grad_norm": 0.0017895435448735952,
+      "learning_rate": 3.335935998763245e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 12234312,
+      "step": 15660
+    },
+    {
+      "epoch": 32.567567567567565,
+      "grad_norm": 0.0023522526025772095,
+      "learning_rate": 3.3350106957083744e-05,
+      "loss": 0.076,
+      "num_input_tokens_seen": 12238280,
+      "step": 15665
+    },
+    {
+      "epoch": 32.57796257796258,
+      "grad_norm": 0.0019743209704756737,
+      "learning_rate": 3.33408526388436e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 12242120,
+      "step": 15670
+    },
+    {
+      "epoch": 32.58835758835759,
+      "grad_norm": 0.028455479070544243,
+      "learning_rate": 3.3331597034339166e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12246120,
+      "step": 15675
+    },
+    {
+      "epoch": 32.5987525987526,
+      "grad_norm": 0.012106985785067081,
+      "learning_rate": 3.3322340144997764e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 12249992,
+      "step": 15680
+    },
+    {
+      "epoch": 32.60914760914761,
+      "grad_norm": 0.004259598907083273,
+      "learning_rate": 3.331308197224693e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12253864,
+      "step": 15685
+    },
+    {
+      "epoch": 32.61954261954262,
+      "grad_norm": 0.007172988262027502,
+      "learning_rate": 3.330382251751438e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12257704,
+      "step": 15690
+    },
+    {
+      "epoch": 32.62993762993763,
+      "grad_norm": 0.005221342667937279,
+      "learning_rate": 3.3294561782228054e-05,
+      "loss": 0.0091,
+      "num_input_tokens_seen": 12261608,
+      "step": 15695
+    },
+    {
+      "epoch": 32.64033264033264,
+      "grad_norm": 0.014960790053009987,
+      "learning_rate": 3.328529976781607e-05,
+      "loss": 0.0157,
+      "num_input_tokens_seen": 12265448,
+      "step": 15700
+    },
+    {
+      "epoch": 32.65072765072765,
+      "grad_norm": 0.0075906733982264996,
+      "learning_rate": 3.327603647570673e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 12269448,
+      "step": 15705
+    },
+    {
+      "epoch": 32.66112266112266,
+      "grad_norm": 0.011691397987306118,
+      "learning_rate": 3.326677190732857e-05,
+      "loss": 0.0168,
+      "num_input_tokens_seen": 12273352,
+      "step": 15710
+    },
+    {
+      "epoch": 32.67151767151767,
+      "grad_norm": 0.01173713430762291,
+      "learning_rate": 3.325750606411029e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12277288,
+      "step": 15715
+    },
+    {
+      "epoch": 32.681912681912685,
+      "grad_norm": 0.027007972821593285,
+      "learning_rate": 3.3248238947480804e-05,
+      "loss": 0.0139,
+      "num_input_tokens_seen": 12281288,
+      "step": 15720
+    },
+    {
+      "epoch": 32.69230769230769,
+      "grad_norm": 0.007882378995418549,
+      "learning_rate": 3.323897055886922e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 12285256,
+      "step": 15725
+    },
+    {
+      "epoch": 32.7027027027027,
+      "grad_norm": 0.003484412096440792,
+      "learning_rate": 3.322970089970484e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12289160,
+      "step": 15730
+    },
+    {
+      "epoch": 32.71309771309771,
+      "grad_norm": 0.00818761344999075,
+      "learning_rate": 3.3220429971417165e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12293064,
+      "step": 15735
+    },
+    {
+      "epoch": 32.723492723492726,
+      "grad_norm": 0.0010752736125141382,
+      "learning_rate": 3.321115777543588e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12296968,
+      "step": 15740
+    },
+    {
+      "epoch": 32.733887733887734,
+      "grad_norm": 0.10327494889497757,
+      "learning_rate": 3.320188431319088e-05,
+      "loss": 0.0263,
+      "num_input_tokens_seen": 12300680,
+      "step": 15745
+    },
+    {
+      "epoch": 32.74428274428274,
+      "grad_norm": 0.012577291578054428,
+      "learning_rate": 3.319260958611224e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 12304488,
+      "step": 15750
+    },
+    {
+      "epoch": 32.75467775467776,
+      "grad_norm": 0.0016673600766807795,
+      "learning_rate": 3.3183333595630256e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12308552,
+      "step": 15755
+    },
+    {
+      "epoch": 32.765072765072766,
+      "grad_norm": 0.01802830956876278,
+      "learning_rate": 3.317405634317538e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 12312488,
+      "step": 15760
+    },
+    {
+      "epoch": 32.775467775467774,
+      "grad_norm": 0.0025805856566876173,
+      "learning_rate": 3.3164777830178315e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12316232,
+      "step": 15765
+    },
+    {
+      "epoch": 32.78586278586278,
+      "grad_norm": 0.19327910244464874,
+      "learning_rate": 3.315549805806989e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 12320168,
+      "step": 15770
+    },
+    {
+      "epoch": 32.7962577962578,
+      "grad_norm": 0.007041838951408863,
+      "learning_rate": 3.314621702828118e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12324136,
+      "step": 15775
+    },
+    {
+      "epoch": 32.80665280665281,
+      "grad_norm": 0.002689527813345194,
+      "learning_rate": 3.313693474224342e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 12327912,
+      "step": 15780
+    },
+    {
+      "epoch": 32.817047817047815,
+      "grad_norm": 0.002366786589846015,
+      "learning_rate": 3.312765120138809e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12331752,
+      "step": 15785
+    },
+    {
+      "epoch": 32.82744282744283,
+      "grad_norm": 0.0030533040408045053,
+      "learning_rate": 3.311836640714679e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12335624,
+      "step": 15790
+    },
+    {
+      "epoch": 32.83783783783784,
+      "grad_norm": 0.007672020699828863,
+      "learning_rate": 3.310908036095137e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12339528,
+      "step": 15795
+    },
+    {
+      "epoch": 32.84823284823285,
+      "grad_norm": 0.0018441564170643687,
+      "learning_rate": 3.309979306423386e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12343432,
+      "step": 15800
+    },
+    {
+      "epoch": 32.84823284823285,
+      "eval_loss": 0.3675895035266876,
+      "eval_runtime": 11.7252,
+      "eval_samples_per_second": 73.005,
+      "eval_steps_per_second": 18.251,
+      "num_input_tokens_seen": 12343432,
+      "step": 15800
+    },
+    {
+      "epoch": 32.858627858627855,
+      "grad_norm": 0.0013550991425290704,
+      "learning_rate": 3.309050451842647e-05,
+      "loss": 0.0086,
+      "num_input_tokens_seen": 12347304,
+      "step": 15805
+    },
+    {
+      "epoch": 32.86902286902287,
+      "grad_norm": 0.3189142346382141,
+      "learning_rate": 3.3081214724961604e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 12351176,
+      "step": 15810
+    },
+    {
+      "epoch": 32.87941787941788,
+      "grad_norm": 0.0014360541244968772,
+      "learning_rate": 3.307192368527188e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 12355048,
+      "step": 15815
+    },
+    {
+      "epoch": 32.88981288981289,
+      "grad_norm": 1.5940278768539429,
+      "learning_rate": 3.306263140079008e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 12359080,
+      "step": 15820
+    },
+    {
+      "epoch": 32.9002079002079,
+      "grad_norm": 0.046801093965768814,
+      "learning_rate": 3.30533378729492e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 12362984,
+      "step": 15825
+    },
+    {
+      "epoch": 32.91060291060291,
+      "grad_norm": 0.0032200785353779793,
+      "learning_rate": 3.304404310318242e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12366696,
+      "step": 15830
+    },
+    {
+      "epoch": 32.92099792099792,
+      "grad_norm": 0.003988068085163832,
+      "learning_rate": 3.3034747092923105e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12370696,
+      "step": 15835
+    },
+    {
+      "epoch": 32.931392931392935,
+      "grad_norm": 0.001969020813703537,
+      "learning_rate": 3.3025449843604806e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12374600,
+      "step": 15840
+    },
+    {
+      "epoch": 32.94178794178794,
+      "grad_norm": 0.007740877568721771,
+      "learning_rate": 3.30161513566613e-05,
+      "loss": 0.0108,
+      "num_input_tokens_seen": 12378472,
+      "step": 15845
+    },
+    {
+      "epoch": 32.95218295218295,
+      "grad_norm": 0.009311944246292114,
+      "learning_rate": 3.3006851633526506e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12382280,
+      "step": 15850
+    },
+    {
+      "epoch": 32.96257796257796,
+      "grad_norm": 0.005139147397130728,
+      "learning_rate": 3.2997550675634584e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12386216,
+      "step": 15855
+    },
+    {
+      "epoch": 32.972972972972975,
+      "grad_norm": 0.001358217210508883,
+      "learning_rate": 3.2988248484419825e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12390088,
+      "step": 15860
+    },
+    {
+      "epoch": 32.983367983367984,
+      "grad_norm": 0.009423032402992249,
+      "learning_rate": 3.2978945061316776e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12394120,
+      "step": 15865
+    },
+    {
+      "epoch": 32.99376299376299,
+      "grad_norm": 0.0018393625505268574,
+      "learning_rate": 3.296964040776013e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12397928,
+      "step": 15870
+    },
+    {
+      "epoch": 33.00415800415801,
+      "grad_norm": 0.27112752199172974,
+      "learning_rate": 3.296033452518478e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 12401912,
+      "step": 15875
+    },
+    {
+      "epoch": 33.014553014553016,
+      "grad_norm": 0.0007961143273860216,
+      "learning_rate": 3.2951027415025806e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12405848,
+      "step": 15880
+    },
+    {
+      "epoch": 33.024948024948024,
+      "grad_norm": 0.0009383549913764,
+      "learning_rate": 3.294171907871849e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12409912,
+      "step": 15885
+    },
+    {
+      "epoch": 33.03534303534303,
+      "grad_norm": 0.0021004516165703535,
+      "learning_rate": 3.293240951769828e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12413720,
+      "step": 15890
+    },
+    {
+      "epoch": 33.04573804573805,
+      "grad_norm": 0.004248692188411951,
+      "learning_rate": 3.2923098733400846e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12417528,
+      "step": 15895
+    },
+    {
+      "epoch": 33.056133056133056,
+      "grad_norm": 0.0012758144875988364,
+      "learning_rate": 3.291378672726202e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12421496,
+      "step": 15900
+    },
+    {
+      "epoch": 33.066528066528065,
+      "grad_norm": 0.01426286157220602,
+      "learning_rate": 3.2904473500717824e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12425208,
+      "step": 15905
+    },
+    {
+      "epoch": 33.07692307692308,
+      "grad_norm": 0.0018325691344216466,
+      "learning_rate": 3.289515905520449e-05,
+      "loss": 0.02,
+      "num_input_tokens_seen": 12429240,
+      "step": 15910
+    },
+    {
+      "epoch": 33.08731808731809,
+      "grad_norm": 0.001752551645040512,
+      "learning_rate": 3.288584339215841e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12433208,
+      "step": 15915
+    },
+    {
+      "epoch": 33.0977130977131,
+      "grad_norm": 0.006810096558183432,
+      "learning_rate": 3.287652651301617e-05,
+      "loss": 0.0132,
+      "num_input_tokens_seen": 12437272,
+      "step": 15920
+    },
+    {
+      "epoch": 33.108108108108105,
+      "grad_norm": 0.3453133702278137,
+      "learning_rate": 3.286720841921457e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 12441240,
+      "step": 15925
+    },
+    {
+      "epoch": 33.11850311850312,
+      "grad_norm": 0.0004638487007468939,
+      "learning_rate": 3.285788911219056e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 12445208,
+      "step": 15930
+    },
+    {
+      "epoch": 33.12889812889813,
+      "grad_norm": 0.0003445899928919971,
+      "learning_rate": 3.284856859338131e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12449080,
+      "step": 15935
+    },
+    {
+      "epoch": 33.13929313929314,
+      "grad_norm": 0.0018228143453598022,
+      "learning_rate": 3.283924686422414e-05,
+      "loss": 0.0275,
+      "num_input_tokens_seen": 12453016,
+      "step": 15940
+    },
+    {
+      "epoch": 33.14968814968815,
+      "grad_norm": 0.0009313684422522783,
+      "learning_rate": 3.282992392615659e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12457048,
+      "step": 15945
+    },
+    {
+      "epoch": 33.16008316008316,
+      "grad_norm": 0.00028236949583515525,
+      "learning_rate": 3.282059978061638e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12460856,
+      "step": 15950
+    },
+    {
+      "epoch": 33.17047817047817,
+      "grad_norm": 0.011678874492645264,
+      "learning_rate": 3.28112744290414e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12464952,
+      "step": 15955
+    },
+    {
+      "epoch": 33.18087318087318,
+      "grad_norm": 0.0036447332240641117,
+      "learning_rate": 3.280194787286974e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 12468920,
+      "step": 15960
+    },
+    {
+      "epoch": 33.19126819126819,
+      "grad_norm": 0.005544251296669245,
+      "learning_rate": 3.2792620113539674e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12472824,
+      "step": 15965
+    },
+    {
+      "epoch": 33.2016632016632,
+      "grad_norm": 0.00019013886048924178,
+      "learning_rate": 3.278329115248966e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12476920,
+      "step": 15970
+    },
+    {
+      "epoch": 33.21205821205821,
+      "grad_norm": 0.0003659058129414916,
+      "learning_rate": 3.277396099115834e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12480824,
+      "step": 15975
+    },
+    {
+      "epoch": 33.222453222453225,
+      "grad_norm": 0.0013583321124315262,
+      "learning_rate": 3.276462963098454e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12484728,
+      "step": 15980
+    },
+    {
+      "epoch": 33.232848232848234,
+      "grad_norm": 0.0012789593311026692,
+      "learning_rate": 3.275529707340728e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 12488632,
+      "step": 15985
+    },
+    {
+      "epoch": 33.24324324324324,
+      "grad_norm": 0.0004454974550753832,
+      "learning_rate": 3.274596331986574e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12492600,
+      "step": 15990
+    },
+    {
+      "epoch": 33.25363825363825,
+      "grad_norm": 0.0051170517690479755,
+      "learning_rate": 3.273662837179932e-05,
+      "loss": 0.0212,
+      "num_input_tokens_seen": 12496632,
+      "step": 15995
+    },
+    {
+      "epoch": 33.264033264033266,
+      "grad_norm": 0.0013609976740553975,
+      "learning_rate": 3.272729223064758e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12500472,
+      "step": 16000
+    },
+    {
+      "epoch": 33.264033264033266,
+      "eval_loss": 0.3827119767665863,
+      "eval_runtime": 11.746,
+      "eval_samples_per_second": 72.876,
+      "eval_steps_per_second": 18.219,
+      "num_input_tokens_seen": 12500472,
+      "step": 16000
+    },
+    {
+      "epoch": 33.274428274428274,
+      "grad_norm": 0.0011203483445569873,
+      "learning_rate": 3.2717954897850264e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12504376,
+      "step": 16005
+    },
+    {
+      "epoch": 33.28482328482328,
+      "grad_norm": 0.005053396336734295,
+      "learning_rate": 3.270861637484733e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12508248,
+      "step": 16010
+    },
+    {
+      "epoch": 33.2952182952183,
+      "grad_norm": 0.6805107593536377,
+      "learning_rate": 3.2699276663078867e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12511960,
+      "step": 16015
+    },
+    {
+      "epoch": 33.305613305613306,
+      "grad_norm": 0.0022347650956362486,
+      "learning_rate": 3.268993576398519e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12515832,
+      "step": 16020
+    },
+    {
+      "epoch": 33.316008316008315,
+      "grad_norm": 0.0035279274452477694,
+      "learning_rate": 3.268059367900678e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12519768,
+      "step": 16025
+    },
+    {
+      "epoch": 33.32640332640332,
+      "grad_norm": 0.006889192387461662,
+      "learning_rate": 3.26712504095843e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12523704,
+      "step": 16030
+    },
+    {
+      "epoch": 33.33679833679834,
+      "grad_norm": 0.004268264397978783,
+      "learning_rate": 3.2661905957158615e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12527576,
+      "step": 16035
+    },
+    {
+      "epoch": 33.34719334719335,
+      "grad_norm": 0.0021784950513392687,
+      "learning_rate": 3.2652560323170734e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12531288,
+      "step": 16040
+    },
+    {
+      "epoch": 33.357588357588355,
+      "grad_norm": 0.001953372498974204,
+      "learning_rate": 3.264321350906189e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12535096,
+      "step": 16045
+    },
+    {
+      "epoch": 33.36798336798337,
+      "grad_norm": 0.0030417086090892553,
+      "learning_rate": 3.263386551627346e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 12538936,
+      "step": 16050
+    },
+    {
+      "epoch": 33.37837837837838,
+      "grad_norm": 0.0010388261871412396,
+      "learning_rate": 3.2624516346247055e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12542872,
+      "step": 16055
+    },
+    {
+      "epoch": 33.38877338877339,
+      "grad_norm": 0.0014267786173149943,
+      "learning_rate": 3.2615166000424404e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 12546712,
+      "step": 16060
+    },
+    {
+      "epoch": 33.3991683991684,
+      "grad_norm": 0.00014673150144517422,
+      "learning_rate": 3.260581448024745e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12550584,
+      "step": 16065
+    },
+    {
+      "epoch": 33.40956340956341,
+      "grad_norm": 0.17930103838443756,
+      "learning_rate": 3.2596461787158335e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 12554552,
+      "step": 16070
+    },
+    {
+      "epoch": 33.41995841995842,
+      "grad_norm": 0.0005658988375216722,
+      "learning_rate": 3.258710792259934e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12558424,
+      "step": 16075
+    },
+    {
+      "epoch": 33.43035343035343,
+      "grad_norm": 0.00023736804723739624,
+      "learning_rate": 3.257775288801296e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12562232,
+      "step": 16080
+    },
+    {
+      "epoch": 33.44074844074844,
+      "grad_norm": 0.0006934339762665331,
+      "learning_rate": 3.256839668484186e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12566264,
+      "step": 16085
+    },
+    {
+      "epoch": 33.45114345114345,
+      "grad_norm": 32.45772171020508,
+      "learning_rate": 3.255903931452888e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 12570200,
+      "step": 16090
+    },
+    {
+      "epoch": 33.46153846153846,
+      "grad_norm": 0.001866971724666655,
+      "learning_rate": 3.2549680778517045e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12574104,
+      "step": 16095
+    },
+    {
+      "epoch": 33.471933471933475,
+      "grad_norm": 0.001243234728462994,
+      "learning_rate": 3.2540321078249556e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12578040,
+      "step": 16100
+    },
+    {
+      "epoch": 33.482328482328484,
+      "grad_norm": 20.315349578857422,
+      "learning_rate": 3.2530960215169795e-05,
+      "loss": 0.0177,
+      "num_input_tokens_seen": 12582008,
+      "step": 16105
+    },
+    {
+      "epoch": 33.49272349272349,
+      "grad_norm": 0.0008155488176271319,
+      "learning_rate": 3.2521598190721345e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 12585848,
+      "step": 16110
+    },
+    {
+      "epoch": 33.5031185031185,
+      "grad_norm": 0.008977388963103294,
+      "learning_rate": 3.251223500634792e-05,
+      "loss": 0.0104,
+      "num_input_tokens_seen": 12589624,
+      "step": 16115
+    },
+    {
+      "epoch": 33.513513513513516,
+      "grad_norm": 0.05472652614116669,
+      "learning_rate": 3.2502870663493445e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 12593624,
+      "step": 16120
+    },
+    {
+      "epoch": 33.523908523908524,
+      "grad_norm": 0.010058941319584846,
+      "learning_rate": 3.249350516360203e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 12597560,
+      "step": 16125
+    },
+    {
+      "epoch": 33.53430353430353,
+      "grad_norm": 0.21484583616256714,
+      "learning_rate": 3.248413850811797e-05,
+      "loss": 0.0045,
+      "num_input_tokens_seen": 12601496,
+      "step": 16130
+    },
+    {
+      "epoch": 33.54469854469855,
+      "grad_norm": 0.22884784638881683,
+      "learning_rate": 3.2474770698485677e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 12605432,
+      "step": 16135
+    },
+    {
+      "epoch": 33.555093555093556,
+      "grad_norm": 0.0005409850855357945,
+      "learning_rate": 3.246540173614983e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 12609336,
+      "step": 16140
+    },
+    {
+      "epoch": 33.565488565488565,
+      "grad_norm": 0.0019687360618263483,
+      "learning_rate": 3.2456031622555197e-05,
+      "loss": 0.0396,
+      "num_input_tokens_seen": 12613208,
+      "step": 16145
+    },
+    {
+      "epoch": 33.57588357588357,
+      "grad_norm": 0.1649021953344345,
+      "learning_rate": 3.2446660359146794e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 12617112,
+      "step": 16150
+    },
+    {
+      "epoch": 33.58627858627859,
+      "grad_norm": 0.00943729467689991,
+      "learning_rate": 3.2437287947369786e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12620984,
+      "step": 16155
+    },
+    {
+      "epoch": 33.5966735966736,
+      "grad_norm": 0.0066843111999332905,
+      "learning_rate": 3.2427914388669525e-05,
+      "loss": 0.0086,
+      "num_input_tokens_seen": 12624952,
+      "step": 16160
+    },
+    {
+      "epoch": 33.607068607068605,
+      "grad_norm": 0.0041614412330091,
+      "learning_rate": 3.241853968449151e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12628888,
+      "step": 16165
+    },
+    {
+      "epoch": 33.61746361746362,
+      "grad_norm": 0.002070261398330331,
+      "learning_rate": 3.240916383628144e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12632952,
+      "step": 16170
+    },
+    {
+      "epoch": 33.62785862785863,
+      "grad_norm": 0.00751803582534194,
+      "learning_rate": 3.239978684548521e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12636824,
+      "step": 16175
+    },
+    {
+      "epoch": 33.63825363825364,
+      "grad_norm": 0.014399996958673,
+      "learning_rate": 3.239040871354885e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12640920,
+      "step": 16180
+    },
+    {
+      "epoch": 33.648648648648646,
+      "grad_norm": 0.0013115191832184792,
+      "learning_rate": 3.2381029441918596e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12644792,
+      "step": 16185
+    },
+    {
+      "epoch": 33.65904365904366,
+      "grad_norm": 0.0016239642864093184,
+      "learning_rate": 3.2371649032040845e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12648696,
+      "step": 16190
+    },
+    {
+      "epoch": 33.66943866943867,
+      "grad_norm": 0.007694312836974859,
+      "learning_rate": 3.2362267485362174e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 12652472,
+      "step": 16195
+    },
+    {
+      "epoch": 33.67983367983368,
+      "grad_norm": 0.00874379463493824,
+      "learning_rate": 3.235288480332934e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12656248,
+      "step": 16200
+    },
+    {
+      "epoch": 33.67983367983368,
+      "eval_loss": 0.3322441577911377,
+      "eval_runtime": 11.7289,
+      "eval_samples_per_second": 72.982,
+      "eval_steps_per_second": 18.246,
+      "num_input_tokens_seen": 12656248,
+      "step": 16200
+    },
+    {
+      "epoch": 33.69022869022869,
+      "grad_norm": 0.002198286121711135,
+      "learning_rate": 3.234350098738927e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12660152,
+      "step": 16205
+    },
+    {
+      "epoch": 33.7006237006237,
+      "grad_norm": 0.0008894646307453513,
+      "learning_rate": 3.233411603898906e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 12663864,
+      "step": 16210
+    },
+    {
+      "epoch": 33.71101871101871,
+      "grad_norm": 0.0010217411909252405,
+      "learning_rate": 3.232472995957599e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12667736,
+      "step": 16215
+    },
+    {
+      "epoch": 33.72141372141372,
+      "grad_norm": 0.046694859862327576,
+      "learning_rate": 3.231534275059751e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12671512,
+      "step": 16220
+    },
+    {
+      "epoch": 33.731808731808734,
+      "grad_norm": 0.0017392492154613137,
+      "learning_rate": 3.230595441350125e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12675480,
+      "step": 16225
+    },
+    {
+      "epoch": 33.74220374220374,
+      "grad_norm": 0.002412884496152401,
+      "learning_rate": 3.2296564949735e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12679416,
+      "step": 16230
+    },
+    {
+      "epoch": 33.75259875259875,
+      "grad_norm": 0.0011705493088811636,
+      "learning_rate": 3.228717436074675e-05,
+      "loss": 0.0065,
+      "num_input_tokens_seen": 12683352,
+      "step": 16235
+    },
+    {
+      "epoch": 33.762993762993766,
+      "grad_norm": 0.31528300046920776,
+      "learning_rate": 3.227778264798463e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 12687256,
+      "step": 16240
+    },
+    {
+      "epoch": 33.773388773388774,
+      "grad_norm": 0.0014004540862515569,
+      "learning_rate": 3.226838981289698e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12691224,
+      "step": 16245
+    },
+    {
+      "epoch": 33.78378378378378,
+      "grad_norm": 0.0005250251851975918,
+      "learning_rate": 3.225899585693227e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 12695128,
+      "step": 16250
+    },
+    {
+      "epoch": 33.79417879417879,
+      "grad_norm": 0.0007870589615777135,
+      "learning_rate": 3.224960078153918e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12698968,
+      "step": 16255
+    },
+    {
+      "epoch": 33.804573804573806,
+      "grad_norm": 0.0015756074571982026,
+      "learning_rate": 3.224020458816655e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 12702936,
+      "step": 16260
+    },
+    {
+      "epoch": 33.814968814968815,
+      "grad_norm": 0.0020835401955991983,
+      "learning_rate": 3.223080727826337e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12706712,
+      "step": 16265
+    },
+    {
+      "epoch": 33.82536382536382,
+      "grad_norm": 0.0005864507984369993,
+      "learning_rate": 3.222140885327885e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12710680,
+      "step": 16270
+    },
+    {
+      "epoch": 33.83575883575884,
+      "grad_norm": 0.0012295867782086134,
+      "learning_rate": 3.221200931466234e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 12714552,
+      "step": 16275
+    },
+    {
+      "epoch": 33.84615384615385,
+      "grad_norm": 0.0035383272916078568,
+      "learning_rate": 3.220260866386336e-05,
+      "loss": 0.006,
+      "num_input_tokens_seen": 12718520,
+      "step": 16280
+    },
+    {
+      "epoch": 33.856548856548855,
+      "grad_norm": 0.0007119208457879722,
+      "learning_rate": 3.21932069023316e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 12722520,
+      "step": 16285
+    },
+    {
+      "epoch": 33.86694386694387,
+      "grad_norm": 0.0011176763800904155,
+      "learning_rate": 3.218380403151695e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12726392,
+      "step": 16290
+    },
+    {
+      "epoch": 33.87733887733888,
+      "grad_norm": 0.002331357216462493,
+      "learning_rate": 3.217440005286943e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12730296,
+      "step": 16295
+    },
+    {
+      "epoch": 33.88773388773389,
+      "grad_norm": 0.0005158432177267969,
+      "learning_rate": 3.216499496783928e-05,
+      "loss": 0.0226,
+      "num_input_tokens_seen": 12734136,
+      "step": 16300
+    },
+    {
+      "epoch": 33.898128898128896,
+      "grad_norm": 0.02256474643945694,
+      "learning_rate": 3.2155588777876856e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12737976,
+      "step": 16305
+    },
+    {
+      "epoch": 33.90852390852391,
+      "grad_norm": 0.019005300477147102,
+      "learning_rate": 3.214618148443273e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 12741880,
+      "step": 16310
+    },
+    {
+      "epoch": 33.91891891891892,
+      "grad_norm": 0.012199360877275467,
+      "learning_rate": 3.2136773088957595e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 12745752,
+      "step": 16315
+    },
+    {
+      "epoch": 33.92931392931393,
+      "grad_norm": 0.1464119404554367,
+      "learning_rate": 3.2127363592902374e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 12749688,
+      "step": 16320
+    },
+    {
+      "epoch": 33.93970893970894,
+      "grad_norm": 0.017948584631085396,
+      "learning_rate": 3.211795299771812e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 12753560,
+      "step": 16325
+    },
+    {
+      "epoch": 33.95010395010395,
+      "grad_norm": 0.007366211619228125,
+      "learning_rate": 3.210854130485605e-05,
+      "loss": 0.0271,
+      "num_input_tokens_seen": 12757400,
+      "step": 16330
+    },
+    {
+      "epoch": 33.96049896049896,
+      "grad_norm": 0.005061611533164978,
+      "learning_rate": 3.209912851576759e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 12761240,
+      "step": 16335
+    },
+    {
+      "epoch": 33.97089397089397,
+      "grad_norm": 0.01765366643667221,
+      "learning_rate": 3.208971463190431e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 12765080,
+      "step": 16340
+    },
+    {
+      "epoch": 33.981288981288984,
+      "grad_norm": 0.0013762374874204397,
+      "learning_rate": 3.208029965471793e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12768888,
+      "step": 16345
+    },
+    {
+      "epoch": 33.99168399168399,
+      "grad_norm": 0.019970014691352844,
+      "learning_rate": 3.2070883585660364e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12772760,
+      "step": 16350
+    },
+    {
+      "epoch": 34.002079002079,
+      "grad_norm": 0.023887787014245987,
+      "learning_rate": 3.20614664261837e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12776648,
+      "step": 16355
+    },
+    {
+      "epoch": 34.012474012474016,
+      "grad_norm": 0.010390275157988071,
+      "learning_rate": 3.205204817774016e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12780584,
+      "step": 16360
+    },
+    {
+      "epoch": 34.022869022869024,
+      "grad_norm": 0.008651173673570156,
+      "learning_rate": 3.204262884178218e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12784584,
+      "step": 16365
+    },
+    {
+      "epoch": 34.03326403326403,
+      "grad_norm": 0.016797160729765892,
+      "learning_rate": 3.2033208419762314e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12788488,
+      "step": 16370
+    },
+    {
+      "epoch": 34.04365904365904,
+      "grad_norm": 0.00430661765858531,
+      "learning_rate": 3.2023786913133344e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12792392,
+      "step": 16375
+    },
+    {
+      "epoch": 34.054054054054056,
+      "grad_norm": 0.026464272290468216,
+      "learning_rate": 3.201436432334816e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12796232,
+      "step": 16380
+    },
+    {
+      "epoch": 34.064449064449065,
+      "grad_norm": 0.012847346253693104,
+      "learning_rate": 3.2004940651859844e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12800168,
+      "step": 16385
+    },
+    {
+      "epoch": 34.07484407484407,
+      "grad_norm": 0.0002442557015456259,
+      "learning_rate": 3.1995515900121655e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12804072,
+      "step": 16390
+    },
+    {
+      "epoch": 34.08523908523909,
+      "grad_norm": 0.0013922769576311111,
+      "learning_rate": 3.1986090069587e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12807912,
+      "step": 16395
+    },
+    {
+      "epoch": 34.0956340956341,
+      "grad_norm": 0.008318210020661354,
+      "learning_rate": 3.1976663161709466e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12811752,
+      "step": 16400
+    },
+    {
+      "epoch": 34.0956340956341,
+      "eval_loss": 0.3600994348526001,
+      "eval_runtime": 11.7656,
+      "eval_samples_per_second": 72.755,
+      "eval_steps_per_second": 18.189,
+      "num_input_tokens_seen": 12811752,
+      "step": 16400
+    },
+    {
+      "epoch": 34.106029106029105,
+      "grad_norm": 0.0003448236675467342,
+      "learning_rate": 3.196723517794279e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12815656,
+      "step": 16405
+    },
+    {
+      "epoch": 34.11642411642411,
+      "grad_norm": 0.0006635615136474371,
+      "learning_rate": 3.19578061197409e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12819592,
+      "step": 16410
+    },
+    {
+      "epoch": 34.12681912681913,
+      "grad_norm": 0.0002751847787294537,
+      "learning_rate": 3.194837598855787e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12823400,
+      "step": 16415
+    },
+    {
+      "epoch": 34.13721413721414,
+      "grad_norm": 4.67432975769043,
+      "learning_rate": 3.193894478584794e-05,
+      "loss": 0.0273,
+      "num_input_tokens_seen": 12827336,
+      "step": 16420
+    },
+    {
+      "epoch": 34.147609147609145,
+      "grad_norm": 0.00020852005400229245,
+      "learning_rate": 3.192951251306553e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12831240,
+      "step": 16425
+    },
+    {
+      "epoch": 34.15800415800416,
+      "grad_norm": 0.0009858449921011925,
+      "learning_rate": 3.192007917166521e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12835112,
+      "step": 16430
+    },
+    {
+      "epoch": 34.16839916839917,
+      "grad_norm": 0.012904543429613113,
+      "learning_rate": 3.191064476310171e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12839016,
+      "step": 16435
+    },
+    {
+      "epoch": 34.17879417879418,
+      "grad_norm": 0.004919559694826603,
+      "learning_rate": 3.1901209288829944e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12843048,
+      "step": 16440
+    },
+    {
+      "epoch": 34.189189189189186,
+      "grad_norm": 0.008398856036365032,
+      "learning_rate": 3.1891772750304985e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12847144,
+      "step": 16445
+    },
+    {
+      "epoch": 34.1995841995842,
+      "grad_norm": 0.0016466659726575017,
+      "learning_rate": 3.188233514898206e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 12850824,
+      "step": 16450
+    },
+    {
+      "epoch": 34.20997920997921,
+      "grad_norm": 0.006467035505920649,
+      "learning_rate": 3.187289648631657e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 12854792,
+      "step": 16455
+    },
+    {
+      "epoch": 34.22037422037422,
+      "grad_norm": 0.0040069143287837505,
+      "learning_rate": 3.186345676376406e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12858856,
+      "step": 16460
+    },
+    {
+      "epoch": 34.23076923076923,
+      "grad_norm": 0.005845969542860985,
+      "learning_rate": 3.1854015982780275e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12862632,
+      "step": 16465
+    },
+    {
+      "epoch": 34.24116424116424,
+      "grad_norm": 0.22399666905403137,
+      "learning_rate": 3.1844574144821084e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 12866600,
+      "step": 16470
+    },
+    {
+      "epoch": 34.25155925155925,
+      "grad_norm": 0.0006692269816994667,
+      "learning_rate": 3.1835131251342554e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12870472,
+      "step": 16475
+    },
+    {
+      "epoch": 34.26195426195426,
+      "grad_norm": 0.006773102562874556,
+      "learning_rate": 3.182568730380089e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12874440,
+      "step": 16480
+    },
+    {
+      "epoch": 34.272349272349274,
+      "grad_norm": 0.003221791237592697,
+      "learning_rate": 3.181624230365245e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 12878376,
+      "step": 16485
+    },
+    {
+      "epoch": 34.28274428274428,
+      "grad_norm": 0.0009739366360008717,
+      "learning_rate": 3.180679625235381e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12882216,
+      "step": 16490
+    },
+    {
+      "epoch": 34.29313929313929,
+      "grad_norm": 0.0014944231370463967,
+      "learning_rate": 3.1797349151361646e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 12886152,
+      "step": 16495
+    },
+    {
+      "epoch": 34.303534303534306,
+      "grad_norm": 0.0021672837901860476,
+      "learning_rate": 3.178790100213281e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12889992,
+      "step": 16500
+    },
+    {
+      "epoch": 34.313929313929314,
+      "grad_norm": 0.00286037172190845,
+      "learning_rate": 3.1778451806124346e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12893896,
+      "step": 16505
+    },
+    {
+      "epoch": 34.32432432432432,
+      "grad_norm": 0.005299687385559082,
+      "learning_rate": 3.176900156479342e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12897992,
+      "step": 16510
+    },
+    {
+      "epoch": 34.33471933471934,
+      "grad_norm": 0.045155372470617294,
+      "learning_rate": 3.17595502795974e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12901864,
+      "step": 16515
+    },
+    {
+      "epoch": 34.34511434511435,
+      "grad_norm": 0.0036400086246430874,
+      "learning_rate": 3.175009795199377e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12905736,
+      "step": 16520
+    },
+    {
+      "epoch": 34.355509355509355,
+      "grad_norm": 0.0005822427337989211,
+      "learning_rate": 3.1740644583440224e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12909704,
+      "step": 16525
+    },
+    {
+      "epoch": 34.36590436590436,
+      "grad_norm": 0.0009109693928621709,
+      "learning_rate": 3.173119017539457e-05,
+      "loss": 0.0047,
+      "num_input_tokens_seen": 12913640,
+      "step": 16530
+    },
+    {
+      "epoch": 34.37629937629938,
+      "grad_norm": 0.0001312764798058197,
+      "learning_rate": 3.172173472931479e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 12917640,
+      "step": 16535
+    },
+    {
+      "epoch": 34.38669438669439,
+      "grad_norm": 0.0005295681185089052,
+      "learning_rate": 3.1712278246659055e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12921512,
+      "step": 16540
+    },
+    {
+      "epoch": 34.397089397089395,
+      "grad_norm": 0.14427796006202698,
+      "learning_rate": 3.170282072888566e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 12925544,
+      "step": 16545
+    },
+    {
+      "epoch": 34.40748440748441,
+      "grad_norm": 0.00039236817974597216,
+      "learning_rate": 3.169336217745307e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 12929448,
+      "step": 16550
+    },
+    {
+      "epoch": 34.41787941787942,
+      "grad_norm": 0.0011002123355865479,
+      "learning_rate": 3.1683902593819924e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12933384,
+      "step": 16555
+    },
+    {
+      "epoch": 34.42827442827443,
+      "grad_norm": 0.221657395362854,
+      "learning_rate": 3.1674441979445e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 12937224,
+      "step": 16560
+    },
+    {
+      "epoch": 34.438669438669436,
+      "grad_norm": 0.00021330913295969367,
+      "learning_rate": 3.166498033578725e-05,
+      "loss": 0.03,
+      "num_input_tokens_seen": 12941096,
+      "step": 16565
+    },
+    {
+      "epoch": 34.44906444906445,
+      "grad_norm": 0.002520560519769788,
+      "learning_rate": 3.165551766430578e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12944936,
+      "step": 16570
+    },
+    {
+      "epoch": 34.45945945945946,
+      "grad_norm": 0.011694083921611309,
+      "learning_rate": 3.164605396645984e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 12948712,
+      "step": 16575
+    },
+    {
+      "epoch": 34.46985446985447,
+      "grad_norm": 0.0021592131815850735,
+      "learning_rate": 3.163658924370886e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12952488,
+      "step": 16580
+    },
+    {
+      "epoch": 34.48024948024948,
+      "grad_norm": 0.004726263228803873,
+      "learning_rate": 3.1627123497512415e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 12956456,
+      "step": 16585
+    },
+    {
+      "epoch": 34.49064449064449,
+      "grad_norm": 0.004078907426446676,
+      "learning_rate": 3.1617656729330245e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12960392,
+      "step": 16590
+    },
+    {
+      "epoch": 34.5010395010395,
+      "grad_norm": 0.002355274511501193,
+      "learning_rate": 3.1608188940622255e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 12964200,
+      "step": 16595
+    },
+    {
+      "epoch": 34.51143451143451,
+      "grad_norm": 0.007496857549995184,
+      "learning_rate": 3.159872013284847e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12968104,
+      "step": 16600
+    },
+    {
+      "epoch": 34.51143451143451,
+      "eval_loss": 0.3240512013435364,
+      "eval_runtime": 11.7811,
+      "eval_samples_per_second": 72.659,
+      "eval_steps_per_second": 18.165,
+      "num_input_tokens_seen": 12968104,
+      "step": 16600
+    },
+    {
+      "epoch": 34.521829521829524,
+      "grad_norm": 0.004634789656847715,
+      "learning_rate": 3.1589250307469134e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12972136,
+      "step": 16605
+    },
+    {
+      "epoch": 34.53222453222453,
+      "grad_norm": 0.0014058751985430717,
+      "learning_rate": 3.1579779465944586e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12976072,
+      "step": 16610
+    },
+    {
+      "epoch": 34.54261954261954,
+      "grad_norm": 0.0044890278950333595,
+      "learning_rate": 3.1570307609735363e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 12980008,
+      "step": 16615
+    },
+    {
+      "epoch": 34.553014553014556,
+      "grad_norm": 0.0009496125276200473,
+      "learning_rate": 3.156083474030213e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 12984008,
+      "step": 16620
+    },
+    {
+      "epoch": 34.563409563409564,
+      "grad_norm": 0.004111850168555975,
+      "learning_rate": 3.155136085910573e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 12988072,
+      "step": 16625
+    },
+    {
+      "epoch": 34.57380457380457,
+      "grad_norm": 0.0016637654043734074,
+      "learning_rate": 3.154188596760717e-05,
+      "loss": 0.0068,
+      "num_input_tokens_seen": 12992168,
+      "step": 16630
+    },
+    {
+      "epoch": 34.58419958419958,
+      "grad_norm": 18.426021575927734,
+      "learning_rate": 3.153241006726757e-05,
+      "loss": 0.0187,
+      "num_input_tokens_seen": 12996072,
+      "step": 16635
+    },
+    {
+      "epoch": 34.5945945945946,
+      "grad_norm": 0.000459986156783998,
+      "learning_rate": 3.152293315954825e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 13000200,
+      "step": 16640
+    },
+    {
+      "epoch": 34.604989604989605,
+      "grad_norm": 0.006780960597097874,
+      "learning_rate": 3.1513455245910666e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13004168,
+      "step": 16645
+    },
+    {
+      "epoch": 34.61538461538461,
+      "grad_norm": 0.1350301206111908,
+      "learning_rate": 3.150397632781643e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13008104,
+      "step": 16650
+    },
+    {
+      "epoch": 34.62577962577963,
+      "grad_norm": 0.0010910176206380129,
+      "learning_rate": 3.149449640672731e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13012040,
+      "step": 16655
+    },
+    {
+      "epoch": 34.63617463617464,
+      "grad_norm": 0.0019994887989014387,
+      "learning_rate": 3.148501548410523e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13015912,
+      "step": 16660
+    },
+    {
+      "epoch": 34.646569646569645,
+      "grad_norm": 0.0050387452356517315,
+      "learning_rate": 3.1475533561412256e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 13019816,
+      "step": 16665
+    },
+    {
+      "epoch": 34.656964656964654,
+      "grad_norm": 0.2854677736759186,
+      "learning_rate": 3.146605064011065e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 13023688,
+      "step": 16670
+    },
+    {
+      "epoch": 34.66735966735967,
+      "grad_norm": 0.001774379750713706,
+      "learning_rate": 3.145656672166277e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13027592,
+      "step": 16675
+    },
+    {
+      "epoch": 34.67775467775468,
+      "grad_norm": 0.0024247621186077595,
+      "learning_rate": 3.144708180753116e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13031400,
+      "step": 16680
+    },
+    {
+      "epoch": 34.688149688149686,
+      "grad_norm": 0.00449402304366231,
+      "learning_rate": 3.143759589917851e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 13035304,
+      "step": 16685
+    },
+    {
+      "epoch": 34.6985446985447,
+      "grad_norm": 0.0027929204516112804,
+      "learning_rate": 3.142810899806768e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13039112,
+      "step": 16690
+    },
+    {
+      "epoch": 34.70893970893971,
+      "grad_norm": 0.003002435201779008,
+      "learning_rate": 3.141862110566166e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13042952,
+      "step": 16695
+    },
+    {
+      "epoch": 34.71933471933472,
+      "grad_norm": 0.0013008499518036842,
+      "learning_rate": 3.1409132223423606e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13046760,
+      "step": 16700
+    },
+    {
+      "epoch": 34.729729729729726,
+      "grad_norm": 0.0004857135354541242,
+      "learning_rate": 3.139964235281682e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13050696,
+      "step": 16705
+    },
+    {
+      "epoch": 34.74012474012474,
+      "grad_norm": 0.0014802527148276567,
+      "learning_rate": 3.139015149530476e-05,
+      "loss": 0.0088,
+      "num_input_tokens_seen": 13054600,
+      "step": 16710
+    },
+    {
+      "epoch": 34.75051975051975,
+      "grad_norm": 0.0004781310854014009,
+      "learning_rate": 3.1380659652351034e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13058472,
+      "step": 16715
+    },
+    {
+      "epoch": 34.76091476091476,
+      "grad_norm": 0.0002660271420609206,
+      "learning_rate": 3.137116682541941e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13062280,
+      "step": 16720
+    },
+    {
+      "epoch": 34.771309771309774,
+      "grad_norm": 0.0006376301753334701,
+      "learning_rate": 3.136167301597379e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13066120,
+      "step": 16725
+    },
+    {
+      "epoch": 34.78170478170478,
+      "grad_norm": 0.004493085667490959,
+      "learning_rate": 3.1352178225478254e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13070088,
+      "step": 16730
+    },
+    {
+      "epoch": 34.79209979209979,
+      "grad_norm": 0.000260471657384187,
+      "learning_rate": 3.1342682455396996e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13073864,
+      "step": 16735
+    },
+    {
+      "epoch": 34.802494802494806,
+      "grad_norm": 0.011810233816504478,
+      "learning_rate": 3.133318570719441e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13077768,
+      "step": 16740
+    },
+    {
+      "epoch": 34.812889812889814,
+      "grad_norm": 0.0009449340286664665,
+      "learning_rate": 3.132368798233499e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13081736,
+      "step": 16745
+    },
+    {
+      "epoch": 34.82328482328482,
+      "grad_norm": 0.06392056494951248,
+      "learning_rate": 3.131418928228342e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13085544,
+      "step": 16750
+    },
+    {
+      "epoch": 34.83367983367983,
+      "grad_norm": 0.0021204915829002857,
+      "learning_rate": 3.1304689608504514e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13089320,
+      "step": 16755
+    },
+    {
+      "epoch": 34.84407484407485,
+      "grad_norm": 0.0006485505728051066,
+      "learning_rate": 3.129518896246324e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13093256,
+      "step": 16760
+    },
+    {
+      "epoch": 34.854469854469855,
+      "grad_norm": 0.004933089483529329,
+      "learning_rate": 3.128568734562472e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13097128,
+      "step": 16765
+    },
+    {
+      "epoch": 34.86486486486486,
+      "grad_norm": 0.001007548999041319,
+      "learning_rate": 3.127618475945421e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 13101032,
+      "step": 16770
+    },
+    {
+      "epoch": 34.87525987525988,
+      "grad_norm": 0.00011318695032969117,
+      "learning_rate": 3.126668120541715e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13104776,
+      "step": 16775
+    },
+    {
+      "epoch": 34.88565488565489,
+      "grad_norm": 0.11786258220672607,
+      "learning_rate": 3.1257176684979096e-05,
+      "loss": 0.0208,
+      "num_input_tokens_seen": 13108744,
+      "step": 16780
+    },
+    {
+      "epoch": 34.896049896049895,
+      "grad_norm": 0.00021185041987337172,
+      "learning_rate": 3.124767119960576e-05,
+      "loss": 0.0066,
+      "num_input_tokens_seen": 13112616,
+      "step": 16785
+    },
+    {
+      "epoch": 34.906444906444904,
+      "grad_norm": 0.000422989105572924,
+      "learning_rate": 3.123816475076301e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13116520,
+      "step": 16790
+    },
+    {
+      "epoch": 34.91683991683992,
+      "grad_norm": 0.005269061774015427,
+      "learning_rate": 3.122865733991687e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13120392,
+      "step": 16795
+    },
+    {
+      "epoch": 34.92723492723493,
+      "grad_norm": 0.0024324366822838783,
+      "learning_rate": 3.1219148968533486e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13124392,
+      "step": 16800
+    },
+    {
+      "epoch": 34.92723492723493,
+      "eval_loss": 0.3507760763168335,
+      "eval_runtime": 11.786,
+      "eval_samples_per_second": 72.628,
+      "eval_steps_per_second": 18.157,
+      "num_input_tokens_seen": 13124392,
+      "step": 16800
+    },
+    {
+      "epoch": 34.937629937629936,
+      "grad_norm": 0.06136063113808632,
+      "learning_rate": 3.120963963807918e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 13128232,
+      "step": 16805
+    },
+    {
+      "epoch": 34.94802494802495,
+      "grad_norm": 0.0003824520972557366,
+      "learning_rate": 3.12001293500204e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 13132104,
+      "step": 16810
+    },
+    {
+      "epoch": 34.95841995841996,
+      "grad_norm": 0.0445302315056324,
+      "learning_rate": 3.1190618105823765e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13136072,
+      "step": 16815
+    },
+    {
+      "epoch": 34.96881496881497,
+      "grad_norm": 0.0011084609432145953,
+      "learning_rate": 3.118110590695603e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13139912,
+      "step": 16820
+    },
+    {
+      "epoch": 34.979209979209976,
+      "grad_norm": 0.00032370458939112723,
+      "learning_rate": 3.117159275488407e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13143848,
+      "step": 16825
+    },
+    {
+      "epoch": 34.98960498960499,
+      "grad_norm": 0.00498083233833313,
+      "learning_rate": 3.1162078651074956e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13147784,
+      "step": 16830
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 0.0012537671718746424,
+      "learning_rate": 3.1152563596995885e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13151608,
+      "step": 16835
+    },
+    {
+      "epoch": 35.01039501039501,
+      "grad_norm": 0.002351682400330901,
+      "learning_rate": 3.1143047594114186e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13155512,
+      "step": 16840
+    },
+    {
+      "epoch": 35.020790020790024,
+      "grad_norm": 0.00016342928574886173,
+      "learning_rate": 3.113353064389734e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13159416,
+      "step": 16845
+    },
+    {
+      "epoch": 35.03118503118503,
+      "grad_norm": 0.00013956906332168728,
+      "learning_rate": 3.1124012747812993e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13163256,
+      "step": 16850
+    },
+    {
+      "epoch": 35.04158004158004,
+      "grad_norm": 0.001157280057668686,
+      "learning_rate": 3.1114493907328936e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 13167448,
+      "step": 16855
+    },
+    {
+      "epoch": 35.05197505197505,
+      "grad_norm": 3.508746522129513e-05,
+      "learning_rate": 3.110497412391306e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 13171384,
+      "step": 16860
+    },
+    {
+      "epoch": 35.062370062370064,
+      "grad_norm": 0.000140528311021626,
+      "learning_rate": 3.1095453399033466e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13175256,
+      "step": 16865
+    },
+    {
+      "epoch": 35.07276507276507,
+      "grad_norm": 35.56916809082031,
+      "learning_rate": 3.108593173415835e-05,
+      "loss": 0.0055,
+      "num_input_tokens_seen": 13179256,
+      "step": 16870
+    },
+    {
+      "epoch": 35.08316008316008,
+      "grad_norm": 0.007019090466201305,
+      "learning_rate": 3.107640913075609e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13183288,
+      "step": 16875
+    },
+    {
+      "epoch": 35.093555093555096,
+      "grad_norm": 0.001164576387964189,
+      "learning_rate": 3.106688559029517e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13187224,
+      "step": 16880
+    },
+    {
+      "epoch": 35.103950103950105,
+      "grad_norm": 0.0006896232371218503,
+      "learning_rate": 3.105736111424425e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 13191192,
+      "step": 16885
+    },
+    {
+      "epoch": 35.11434511434511,
+      "grad_norm": 0.001457841950468719,
+      "learning_rate": 3.1047835704072136e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13195096,
+      "step": 16890
+    },
+    {
+      "epoch": 35.12474012474012,
+      "grad_norm": 0.0019527594558894634,
+      "learning_rate": 3.103830936124775e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13199064,
+      "step": 16895
+    },
+    {
+      "epoch": 35.13513513513514,
+      "grad_norm": 0.002025145571678877,
+      "learning_rate": 3.102878208724018e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13203032,
+      "step": 16900
+    },
+    {
+      "epoch": 35.145530145530145,
+      "grad_norm": 0.0018443074077367783,
+      "learning_rate": 3.101925388351865e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13206936,
+      "step": 16905
+    },
+    {
+      "epoch": 35.15592515592515,
+      "grad_norm": 0.002648394089192152,
+      "learning_rate": 3.1009724751552515e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13210872,
+      "step": 16910
+    },
+    {
+      "epoch": 35.16632016632017,
+      "grad_norm": 0.0056763687171041965,
+      "learning_rate": 3.100019469281131e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13214904,
+      "step": 16915
+    },
+    {
+      "epoch": 35.17671517671518,
+      "grad_norm": 0.006043433677405119,
+      "learning_rate": 3.0990663708764685e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13218744,
+      "step": 16920
+    },
+    {
+      "epoch": 35.187110187110186,
+      "grad_norm": 0.0008356276666745543,
+      "learning_rate": 3.098113180088243e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 13222712,
+      "step": 16925
+    },
+    {
+      "epoch": 35.197505197505194,
+      "grad_norm": 0.0012964545749127865,
+      "learning_rate": 3.097159897063448e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13226456,
+      "step": 16930
+    },
+    {
+      "epoch": 35.20790020790021,
+      "grad_norm": 0.0019520395435392857,
+      "learning_rate": 3.096206521949094e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13230296,
+      "step": 16935
+    },
+    {
+      "epoch": 35.21829521829522,
+      "grad_norm": 4.0276274376083165e-05,
+      "learning_rate": 3.0952530548922006e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13234168,
+      "step": 16940
+    },
+    {
+      "epoch": 35.228690228690226,
+      "grad_norm": 0.011213792487978935,
+      "learning_rate": 3.0942994960398064e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13238008,
+      "step": 16945
+    },
+    {
+      "epoch": 35.23908523908524,
+      "grad_norm": 0.00022439064923673868,
+      "learning_rate": 3.093345845538961e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13241880,
+      "step": 16950
+    },
+    {
+      "epoch": 35.24948024948025,
+      "grad_norm": 0.007839543744921684,
+      "learning_rate": 3.09239210353673e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13245784,
+      "step": 16955
+    },
+    {
+      "epoch": 35.25987525987526,
+      "grad_norm": 0.00015915153198875487,
+      "learning_rate": 3.0914382701801926e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 13249656,
+      "step": 16960
+    },
+    {
+      "epoch": 35.270270270270274,
+      "grad_norm": 0.0024280142970383167,
+      "learning_rate": 3.090484345616441e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13253528,
+      "step": 16965
+    },
+    {
+      "epoch": 35.28066528066528,
+      "grad_norm": 7.71781063079834,
+      "learning_rate": 3.0895303299925825e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 13257592,
+      "step": 16970
+    },
+    {
+      "epoch": 35.29106029106029,
+      "grad_norm": 0.003639420261606574,
+      "learning_rate": 3.0885762234557393e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13261464,
+      "step": 16975
+    },
+    {
+      "epoch": 35.3014553014553,
+      "grad_norm": 0.0006342986598610878,
+      "learning_rate": 3.087622026153045e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13265336,
+      "step": 16980
+    },
+    {
+      "epoch": 35.311850311850314,
+      "grad_norm": 9.019027493195608e-05,
+      "learning_rate": 3.086667738231651e-05,
+      "loss": 0.0098,
+      "num_input_tokens_seen": 13269144,
+      "step": 16985
+    },
+    {
+      "epoch": 35.32224532224532,
+      "grad_norm": 0.0007295747054740787,
+      "learning_rate": 3.085713359838718e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13273112,
+      "step": 16990
+    },
+    {
+      "epoch": 35.33264033264033,
+      "grad_norm": 0.0001226935419254005,
+      "learning_rate": 3.084758891121425e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13277144,
+      "step": 16995
+    },
+    {
+      "epoch": 35.343035343035346,
+      "grad_norm": 0.0009216155158355832,
+      "learning_rate": 3.083804332226963e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13281144,
+      "step": 17000
+    },
+    {
+      "epoch": 35.343035343035346,
+      "eval_loss": 0.45685330033302307,
+      "eval_runtime": 11.7174,
+      "eval_samples_per_second": 73.054,
+      "eval_steps_per_second": 18.263,
+      "num_input_tokens_seen": 13281144,
+      "step": 17000
+    },
+    {
+      "epoch": 35.353430353430355,
+      "grad_norm": 0.0001538443029858172,
+      "learning_rate": 3.082849683302536e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 13285048,
+      "step": 17005
+    },
+    {
+      "epoch": 35.36382536382536,
+      "grad_norm": 8.784292003838345e-05,
+      "learning_rate": 3.081894944495363e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13288984,
+      "step": 17010
+    },
+    {
+      "epoch": 35.37422037422037,
+      "grad_norm": 0.00024164372007362545,
+      "learning_rate": 3.080940115952677e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 13292792,
+      "step": 17015
+    },
+    {
+      "epoch": 35.38461538461539,
+      "grad_norm": 5.801392762805335e-05,
+      "learning_rate": 3.0799851978217245e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 13296568,
+      "step": 17020
+    },
+    {
+      "epoch": 35.395010395010395,
+      "grad_norm": 0.00017037914949469268,
+      "learning_rate": 3.0790301902497666e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13300376,
+      "step": 17025
+    },
+    {
+      "epoch": 35.4054054054054,
+      "grad_norm": 0.00010124386608367786,
+      "learning_rate": 3.078075093384076e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13304344,
+      "step": 17030
+    },
+    {
+      "epoch": 35.41580041580042,
+      "grad_norm": 5.632597458316013e-05,
+      "learning_rate": 3.077119907371942e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13308280,
+      "step": 17035
+    },
+    {
+      "epoch": 35.42619542619543,
+      "grad_norm": 0.0006122209597378969,
+      "learning_rate": 3.076164632360666e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13312184,
+      "step": 17040
+    },
+    {
+      "epoch": 35.436590436590436,
+      "grad_norm": 0.14843147993087769,
+      "learning_rate": 3.075209268497563e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 13316056,
+      "step": 17045
+    },
+    {
+      "epoch": 35.446985446985444,
+      "grad_norm": 0.00013982472592033446,
+      "learning_rate": 3.074253815929961e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 13319992,
+      "step": 17050
+    },
+    {
+      "epoch": 35.45738045738046,
+      "grad_norm": 0.00028103188378736377,
+      "learning_rate": 3.0732982748052054e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13323992,
+      "step": 17055
+    },
+    {
+      "epoch": 35.46777546777547,
+      "grad_norm": 0.000158995550009422,
+      "learning_rate": 3.072342645270651e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13327896,
+      "step": 17060
+    },
+    {
+      "epoch": 35.478170478170476,
+      "grad_norm": 0.00041752337710931897,
+      "learning_rate": 3.071386927473668e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 13331864,
+      "step": 17065
+    },
+    {
+      "epoch": 35.48856548856549,
+      "grad_norm": 0.00011586988694034517,
+      "learning_rate": 3.0704311215616404e-05,
+      "loss": 0.0086,
+      "num_input_tokens_seen": 13335704,
+      "step": 17070
+    },
+    {
+      "epoch": 35.4989604989605,
+      "grad_norm": 0.00012428763147909194,
+      "learning_rate": 3.0694752276819656e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13339800,
+      "step": 17075
+    },
+    {
+      "epoch": 35.50935550935551,
+      "grad_norm": 0.0002638477017171681,
+      "learning_rate": 3.068519245982054e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 13343672,
+      "step": 17080
+    },
+    {
+      "epoch": 35.51975051975052,
+      "grad_norm": 0.00012005022290395573,
+      "learning_rate": 3.0675631766093304e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 13347704,
+      "step": 17085
+    },
+    {
+      "epoch": 35.53014553014553,
+      "grad_norm": 0.00027653801953420043,
+      "learning_rate": 3.066607019711232e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 13351736,
+      "step": 17090
+    },
+    {
+      "epoch": 35.54054054054054,
+      "grad_norm": 0.00042574532562866807,
+      "learning_rate": 3.065650775435211e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13355640,
+      "step": 17095
+    },
+    {
+      "epoch": 35.55093555093555,
+      "grad_norm": 0.00025310920318588614,
+      "learning_rate": 3.0646944439287326e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13359640,
+      "step": 17100
+    },
+    {
+      "epoch": 35.561330561330564,
+      "grad_norm": 0.000125138511066325,
+      "learning_rate": 3.0637380253392736e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13363640,
+      "step": 17105
+    },
+    {
+      "epoch": 35.57172557172557,
+      "grad_norm": 9.309384768130258e-05,
+      "learning_rate": 3.062781519814327e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 13367448,
+      "step": 17110
+    },
+    {
+      "epoch": 35.58212058212058,
+      "grad_norm": 9.177230094792321e-05,
+      "learning_rate": 3.0618249275013985e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13371352,
+      "step": 17115
+    },
+    {
+      "epoch": 35.59251559251559,
+      "grad_norm": 0.00014257486327551305,
+      "learning_rate": 3.060868248548005e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 13375256,
+      "step": 17120
+    },
+    {
+      "epoch": 35.602910602910605,
+      "grad_norm": 0.000349770940374583,
+      "learning_rate": 3.0599114831016796e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13379224,
+      "step": 17125
+    },
+    {
+      "epoch": 35.61330561330561,
+      "grad_norm": 0.0001753423421178013,
+      "learning_rate": 3.0589546313099666e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13383032,
+      "step": 17130
+    },
+    {
+      "epoch": 35.62370062370062,
+      "grad_norm": 0.0025761083234101534,
+      "learning_rate": 3.0579976933204255e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13386872,
+      "step": 17135
+    },
+    {
+      "epoch": 35.63409563409564,
+      "grad_norm": 8.442177932010964e-05,
+      "learning_rate": 3.0570406692806284e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13390712,
+      "step": 17140
+    },
+    {
+      "epoch": 35.644490644490645,
+      "grad_norm": 0.00012751456233672798,
+      "learning_rate": 3.05608355933816e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13394616,
+      "step": 17145
+    },
+    {
+      "epoch": 35.65488565488565,
+      "grad_norm": 0.00011009017180185765,
+      "learning_rate": 3.055126363640618e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13398584,
+      "step": 17150
+    },
+    {
+      "epoch": 35.66528066528066,
+      "grad_norm": 6.35967226116918e-05,
+      "learning_rate": 3.0541690823356146e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13402616,
+      "step": 17155
+    },
+    {
+      "epoch": 35.67567567567568,
+      "grad_norm": 6.422158912755549e-05,
+      "learning_rate": 3.053211715570775e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 13406488,
+      "step": 17160
+    },
+    {
+      "epoch": 35.686070686070686,
+      "grad_norm": 6.33431991445832e-05,
+      "learning_rate": 3.052254263493736e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13410424,
+      "step": 17165
+    },
+    {
+      "epoch": 35.696465696465694,
+      "grad_norm": 0.00016697369574103504,
+      "learning_rate": 3.0512967262521498e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13414264,
+      "step": 17170
+    },
+    {
+      "epoch": 35.70686070686071,
+      "grad_norm": 0.00012770321336574852,
+      "learning_rate": 3.0503391039936803e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 13418232,
+      "step": 17175
+    },
+    {
+      "epoch": 35.71725571725572,
+      "grad_norm": 0.0003351036866661161,
+      "learning_rate": 3.0493813968660056e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13422232,
+      "step": 17180
+    },
+    {
+      "epoch": 35.727650727650726,
+      "grad_norm": 0.00011907632870133966,
+      "learning_rate": 3.0484236050168153e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13426264,
+      "step": 17185
+    },
+    {
+      "epoch": 35.73804573804574,
+      "grad_norm": 7.9815108620096e-05,
+      "learning_rate": 3.0474657285938123e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13430040,
+      "step": 17190
+    },
+    {
+      "epoch": 35.74844074844075,
+      "grad_norm": 0.0008574778330512345,
+      "learning_rate": 3.046507767744715e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13433848,
+      "step": 17195
+    },
+    {
+      "epoch": 35.75883575883576,
+      "grad_norm": 0.00025170273147523403,
+      "learning_rate": 3.045549722617252e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13437720,
+      "step": 17200
+    },
+    {
+      "epoch": 35.75883575883576,
+      "eval_loss": 0.4359787702560425,
+      "eval_runtime": 11.757,
+      "eval_samples_per_second": 72.808,
+      "eval_steps_per_second": 18.202,
+      "num_input_tokens_seen": 13437720,
+      "step": 17200
+    },
+    {
+      "epoch": 35.76923076923077,
+      "grad_norm": 0.00016415039135608822,
+      "learning_rate": 3.0445915933591658e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13441656,
+      "step": 17205
+    },
+    {
+      "epoch": 35.77962577962578,
+      "grad_norm": 0.0008528147591277957,
+      "learning_rate": 3.0436333801182114e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 13445624,
+      "step": 17210
+    },
+    {
+      "epoch": 35.79002079002079,
+      "grad_norm": 0.00016739187412895262,
+      "learning_rate": 3.0426750830421596e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13449496,
+      "step": 17215
+    },
+    {
+      "epoch": 35.8004158004158,
+      "grad_norm": 0.0004993913462385535,
+      "learning_rate": 3.0417167022787897e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13453336,
+      "step": 17220
+    },
+    {
+      "epoch": 35.810810810810814,
+      "grad_norm": 0.0022125549148768187,
+      "learning_rate": 3.0407582379758966e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13457176,
+      "step": 17225
+    },
+    {
+      "epoch": 35.82120582120582,
+      "grad_norm": 0.00010936758917523548,
+      "learning_rate": 3.039799690281287e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13460952,
+      "step": 17230
+    },
+    {
+      "epoch": 35.83160083160083,
+      "grad_norm": 0.00017237670545000583,
+      "learning_rate": 3.0388410593427823e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 13464888,
+      "step": 17235
+    },
+    {
+      "epoch": 35.84199584199584,
+      "grad_norm": 9.142955968854949e-05,
+      "learning_rate": 3.0378823453082146e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 13468920,
+      "step": 17240
+    },
+    {
+      "epoch": 35.852390852390855,
+      "grad_norm": 0.00015176359738688916,
+      "learning_rate": 3.03692354832543e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13472920,
+      "step": 17245
+    },
+    {
+      "epoch": 35.86278586278586,
+      "grad_norm": 0.00016362531459890306,
+      "learning_rate": 3.0359646685422865e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13476696,
+      "step": 17250
+    },
+    {
+      "epoch": 35.87318087318087,
+      "grad_norm": 0.000784229370765388,
+      "learning_rate": 3.035005706106656e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13480728,
+      "step": 17255
+    },
+    {
+      "epoch": 35.88357588357589,
+      "grad_norm": 0.00020775612210854888,
+      "learning_rate": 3.034046661166422e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13484568,
+      "step": 17260
+    },
+    {
+      "epoch": 35.893970893970895,
+      "grad_norm": 0.00019233394414186478,
+      "learning_rate": 3.033087533869482e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13488312,
+      "step": 17265
+    },
+    {
+      "epoch": 35.9043659043659,
+      "grad_norm": 0.009973638691008091,
+      "learning_rate": 3.0321283243637444e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13492152,
+      "step": 17270
+    },
+    {
+      "epoch": 35.91476091476091,
+      "grad_norm": 5.4817035561427474e-05,
+      "learning_rate": 3.0311690327971326e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13496152,
+      "step": 17275
+    },
+    {
+      "epoch": 35.92515592515593,
+      "grad_norm": 0.04482612386345863,
+      "learning_rate": 3.030209659317581e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13500120,
+      "step": 17280
+    },
+    {
+      "epoch": 35.935550935550935,
+      "grad_norm": 0.00023789067927282304,
+      "learning_rate": 3.0292502040730362e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13503960,
+      "step": 17285
+    },
+    {
+      "epoch": 35.945945945945944,
+      "grad_norm": 0.0001338545698672533,
+      "learning_rate": 3.0282906672114597e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13507928,
+      "step": 17290
+    },
+    {
+      "epoch": 35.95634095634096,
+      "grad_norm": 0.00011318015458527952,
+      "learning_rate": 3.027331048880823e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13511896,
+      "step": 17295
+    },
+    {
+      "epoch": 35.96673596673597,
+      "grad_norm": 7.904368976596743e-05,
+      "learning_rate": 3.0263713492291123e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13515864,
+      "step": 17300
+    },
+    {
+      "epoch": 35.977130977130976,
+      "grad_norm": 0.0004942790255881846,
+      "learning_rate": 3.0254115684043242e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13519704,
+      "step": 17305
+    },
+    {
+      "epoch": 35.987525987525984,
+      "grad_norm": 0.00011050152534153312,
+      "learning_rate": 3.024451706554469e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13523640,
+      "step": 17310
+    },
+    {
+      "epoch": 35.997920997921,
+      "grad_norm": 0.00016651824989821762,
+      "learning_rate": 3.0234917638275705e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13527512,
+      "step": 17315
+    },
+    {
+      "epoch": 36.00831600831601,
+      "grad_norm": 6.897203275002539e-05,
+      "learning_rate": 3.0225317403716635e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13531216,
+      "step": 17320
+    },
+    {
+      "epoch": 36.018711018711016,
+      "grad_norm": 7.819651364115998e-05,
+      "learning_rate": 3.0215716363347956e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13535152,
+      "step": 17325
+    },
+    {
+      "epoch": 36.02910602910603,
+      "grad_norm": 0.0003460766456555575,
+      "learning_rate": 3.0206114518650275e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13539184,
+      "step": 17330
+    },
+    {
+      "epoch": 36.03950103950104,
+      "grad_norm": 0.00047457736218348145,
+      "learning_rate": 3.0196511871104304e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13542992,
+      "step": 17335
+    },
+    {
+      "epoch": 36.04989604989605,
+      "grad_norm": 0.0004102666862308979,
+      "learning_rate": 3.01869084221909e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13546928,
+      "step": 17340
+    },
+    {
+      "epoch": 36.06029106029106,
+      "grad_norm": 9.234292519977316e-05,
+      "learning_rate": 3.0177304173391037e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13550928,
+      "step": 17345
+    },
+    {
+      "epoch": 36.07068607068607,
+      "grad_norm": 0.0003158719919156283,
+      "learning_rate": 3.01676991261858e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13554768,
+      "step": 17350
+    },
+    {
+      "epoch": 36.08108108108108,
+      "grad_norm": 5.438017615233548e-05,
+      "learning_rate": 3.015809328205642e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 13558896,
+      "step": 17355
+    },
+    {
+      "epoch": 36.09147609147609,
+      "grad_norm": 0.0002088343317154795,
+      "learning_rate": 3.0148486642484248e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 13562800,
+      "step": 17360
+    },
+    {
+      "epoch": 36.101871101871104,
+      "grad_norm": 7.530041330028325e-05,
+      "learning_rate": 3.0138879208950722e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13566832,
+      "step": 17365
+    },
+    {
+      "epoch": 36.11226611226611,
+      "grad_norm": 0.0001159085295512341,
+      "learning_rate": 3.012927098293744e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13570800,
+      "step": 17370
+    },
+    {
+      "epoch": 36.12266112266112,
+      "grad_norm": 7.556709897471592e-05,
+      "learning_rate": 3.0119661965926123e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 13574800,
+      "step": 17375
+    },
+    {
+      "epoch": 36.13305613305613,
+      "grad_norm": 6.478973955381662e-05,
+      "learning_rate": 3.0110052159398587e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13578704,
+      "step": 17380
+    },
+    {
+      "epoch": 36.143451143451145,
+      "grad_norm": 0.00011365805403329432,
+      "learning_rate": 3.0100441564836802e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13582608,
+      "step": 17385
+    },
+    {
+      "epoch": 36.15384615384615,
+      "grad_norm": 0.0003130184777546674,
+      "learning_rate": 3.0090830183722817e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13586480,
+      "step": 17390
+    },
+    {
+      "epoch": 36.16424116424116,
+      "grad_norm": 0.000472166168037802,
+      "learning_rate": 3.0081218017538852e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13590544,
+      "step": 17395
+    },
+    {
+      "epoch": 36.17463617463618,
+      "grad_norm": 0.00010229231702396646,
+      "learning_rate": 3.0071605067767212e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 13594448,
+      "step": 17400
+    },
+    {
+      "epoch": 36.17463617463618,
+      "eval_loss": 0.44813454151153564,
+      "eval_runtime": 11.6982,
+      "eval_samples_per_second": 73.173,
+      "eval_steps_per_second": 18.293,
+      "num_input_tokens_seen": 13594448,
+      "step": 17400
+    },
+    {
+      "epoch": 36.185031185031185,
+      "grad_norm": 0.0003904421755578369,
+      "learning_rate": 3.006199133589034e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13598480,
+      "step": 17405
+    },
+    {
+      "epoch": 36.195426195426194,
+      "grad_norm": 0.0002532803046051413,
+      "learning_rate": 3.005237682339079e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13602416,
+      "step": 17410
+    },
+    {
+      "epoch": 36.20582120582121,
+      "grad_norm": 0.0002598919963929802,
+      "learning_rate": 3.0042761531751228e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 13606384,
+      "step": 17415
+    },
+    {
+      "epoch": 36.21621621621622,
+      "grad_norm": 7.00896853231825e-05,
+      "learning_rate": 3.0033145462454482e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 13610096,
+      "step": 17420
+    },
+    {
+      "epoch": 36.226611226611226,
+      "grad_norm": 0.0001690156350377947,
+      "learning_rate": 3.002352861698345e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13614096,
+      "step": 17425
+    },
+    {
+      "epoch": 36.237006237006234,
+      "grad_norm": 0.0010510009014979005,
+      "learning_rate": 3.0013910996821178e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13618096,
+      "step": 17430
+    },
+    {
+      "epoch": 36.24740124740125,
+      "grad_norm": 0.3653600513935089,
+      "learning_rate": 3.0004292603450817e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 13621936,
+      "step": 17435
+    },
+    {
+      "epoch": 36.25779625779626,
+      "grad_norm": 0.009764228947460651,
+      "learning_rate": 2.9994673438355653e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13626000,
+      "step": 17440
+    },
+    {
+      "epoch": 36.268191268191266,
+      "grad_norm": 0.21316806972026825,
+      "learning_rate": 2.9985053503019078e-05,
+      "loss": 0.0071,
+      "num_input_tokens_seen": 13629936,
+      "step": 17445
+    },
+    {
+      "epoch": 36.27858627858628,
+      "grad_norm": 0.00029260144219733775,
+      "learning_rate": 2.99754327989246e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13633936,
+      "step": 17450
+    },
+    {
+      "epoch": 36.28898128898129,
+      "grad_norm": 0.15685239434242249,
+      "learning_rate": 2.9965811327555864e-05,
+      "loss": 0.006,
+      "num_input_tokens_seen": 13637808,
+      "step": 17455
+    },
+    {
+      "epoch": 36.2993762993763,
+      "grad_norm": 0.00405800249427557,
+      "learning_rate": 2.995618909039662e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13641776,
+      "step": 17460
+    },
+    {
+      "epoch": 36.30977130977131,
+      "grad_norm": 0.10523105412721634,
+      "learning_rate": 2.9946566088930727e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13645616,
+      "step": 17465
+    },
+    {
+      "epoch": 36.32016632016632,
+      "grad_norm": 0.00014407490380108356,
+      "learning_rate": 2.9936942324642192e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13649424,
+      "step": 17470
+    },
+    {
+      "epoch": 36.33056133056133,
+      "grad_norm": 0.0004294767859391868,
+      "learning_rate": 2.9927317799015097e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 13653264,
+      "step": 17475
+    },
+    {
+      "epoch": 36.34095634095634,
+      "grad_norm": 0.000981958582997322,
+      "learning_rate": 2.9917692513533685e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13657136,
+      "step": 17480
+    },
+    {
+      "epoch": 36.351351351351354,
+      "grad_norm": 0.004131304565817118,
+      "learning_rate": 2.990806646968229e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13661136,
+      "step": 17485
+    },
+    {
+      "epoch": 36.36174636174636,
+      "grad_norm": 0.0024032522924244404,
+      "learning_rate": 2.989843966894536e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 13665200,
+      "step": 17490
+    },
+    {
+      "epoch": 36.37214137214137,
+      "grad_norm": 0.00011660126619972289,
+      "learning_rate": 2.9888812112807472e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13668912,
+      "step": 17495
+    },
+    {
+      "epoch": 36.38253638253638,
+      "grad_norm": 0.00026454811450093985,
+      "learning_rate": 2.987918380275333e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13672880,
+      "step": 17500
+    },
+    {
+      "epoch": 36.392931392931395,
+      "grad_norm": 0.004020621068775654,
+      "learning_rate": 2.9869554740267724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13676816,
+      "step": 17505
+    },
+    {
+      "epoch": 36.4033264033264,
+      "grad_norm": 0.0005820782971568406,
+      "learning_rate": 2.9859924926835585e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13680656,
+      "step": 17510
+    },
+    {
+      "epoch": 36.41372141372141,
+      "grad_norm": 0.00025373109383508563,
+      "learning_rate": 2.9850294363941944e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13684720,
+      "step": 17515
+    },
+    {
+      "epoch": 36.42411642411643,
+      "grad_norm": 0.0008166201296262443,
+      "learning_rate": 2.9840663053071967e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13688592,
+      "step": 17520
+    },
+    {
+      "epoch": 36.434511434511435,
+      "grad_norm": 0.11744668334722519,
+      "learning_rate": 2.983103099571091e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 13692400,
+      "step": 17525
+    },
+    {
+      "epoch": 36.444906444906444,
+      "grad_norm": 0.0006883711903356016,
+      "learning_rate": 2.9821398193344164e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13696272,
+      "step": 17530
+    },
+    {
+      "epoch": 36.45530145530145,
+      "grad_norm": 4.515674299909733e-05,
+      "learning_rate": 2.9811764647457226e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13700080,
+      "step": 17535
+    },
+    {
+      "epoch": 36.46569646569647,
+      "grad_norm": 4.6399723942158744e-05,
+      "learning_rate": 2.9802130359535714e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13703952,
+      "step": 17540
+    },
+    {
+      "epoch": 36.476091476091476,
+      "grad_norm": 2.7962982130702585e-05,
+      "learning_rate": 2.979249533106535e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13707856,
+      "step": 17545
+    },
+    {
+      "epoch": 36.486486486486484,
+      "grad_norm": 0.0030155405402183533,
+      "learning_rate": 2.9782859563531986e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13711728,
+      "step": 17550
+    },
+    {
+      "epoch": 36.4968814968815,
+      "grad_norm": 0.00047994405031204224,
+      "learning_rate": 2.977322305842156e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13715632,
+      "step": 17555
+    },
+    {
+      "epoch": 36.50727650727651,
+      "grad_norm": 1.6788944776635617e-05,
+      "learning_rate": 2.9763585817220162e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13719536,
+      "step": 17560
+    },
+    {
+      "epoch": 36.517671517671516,
+      "grad_norm": 0.004108342342078686,
+      "learning_rate": 2.975394784141397e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13723472,
+      "step": 17565
+    },
+    {
+      "epoch": 36.528066528066525,
+      "grad_norm": 2.6552244889899157e-05,
+      "learning_rate": 2.974430913248928e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 13727376,
+      "step": 17570
+    },
+    {
+      "epoch": 36.53846153846154,
+      "grad_norm": 0.0004000810731668025,
+      "learning_rate": 2.9734669691932497e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13731344,
+      "step": 17575
+    },
+    {
+      "epoch": 36.54885654885655,
+      "grad_norm": 0.0002995759423356503,
+      "learning_rate": 2.9725029521230147e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13735088,
+      "step": 17580
+    },
+    {
+      "epoch": 36.55925155925156,
+      "grad_norm": 0.004784012213349342,
+      "learning_rate": 2.9715388621868873e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13738992,
+      "step": 17585
+    },
+    {
+      "epoch": 36.56964656964657,
+      "grad_norm": 0.0006144788931123912,
+      "learning_rate": 2.970574699533541e-05,
+      "loss": 0.0047,
+      "num_input_tokens_seen": 13742832,
+      "step": 17590
+    },
+    {
+      "epoch": 36.58004158004158,
+      "grad_norm": 0.0007678499678149819,
+      "learning_rate": 2.969610464311662e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13746704,
+      "step": 17595
+    },
+    {
+      "epoch": 36.59043659043659,
+      "grad_norm": 4.292961602914147e-05,
+      "learning_rate": 2.9686461566699487e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13750544,
+      "step": 17600
+    },
+    {
+      "epoch": 36.59043659043659,
+      "eval_loss": 0.4794686436653137,
+      "eval_runtime": 11.6962,
+      "eval_samples_per_second": 73.186,
+      "eval_steps_per_second": 18.296,
+      "num_input_tokens_seen": 13750544,
+      "step": 17600
+    },
+    {
+      "epoch": 36.6008316008316,
+      "grad_norm": 2.1514009858947247e-05,
+      "learning_rate": 2.9676817767571086e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13754544,
+      "step": 17605
+    },
+    {
+      "epoch": 36.61122661122661,
+      "grad_norm": 0.0005610433872789145,
+      "learning_rate": 2.966717324721861e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 13758608,
+      "step": 17610
+    },
+    {
+      "epoch": 36.62162162162162,
+      "grad_norm": 2.9658218409167603e-05,
+      "learning_rate": 2.9657528007129366e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13762480,
+      "step": 17615
+    },
+    {
+      "epoch": 36.63201663201663,
+      "grad_norm": 2.427029721729923e-05,
+      "learning_rate": 2.9647882048790777e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13766256,
+      "step": 17620
+    },
+    {
+      "epoch": 36.642411642411645,
+      "grad_norm": 7.02160396031104e-05,
+      "learning_rate": 2.963823537369037e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 13770160,
+      "step": 17625
+    },
+    {
+      "epoch": 36.65280665280665,
+      "grad_norm": 0.0006363879656419158,
+      "learning_rate": 2.9628587983315775e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13774064,
+      "step": 17630
+    },
+    {
+      "epoch": 36.66320166320166,
+      "grad_norm": 0.00010333996033295989,
+      "learning_rate": 2.9618939879154746e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13777936,
+      "step": 17635
+    },
+    {
+      "epoch": 36.67359667359668,
+      "grad_norm": 0.00014181900769472122,
+      "learning_rate": 2.9609291062695143e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13781840,
+      "step": 17640
+    },
+    {
+      "epoch": 36.683991683991685,
+      "grad_norm": 0.8029877543449402,
+      "learning_rate": 2.9599641535424938e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 13785840,
+      "step": 17645
+    },
+    {
+      "epoch": 36.694386694386694,
+      "grad_norm": 0.00014565963647328317,
+      "learning_rate": 2.9589991298832202e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13789712,
+      "step": 17650
+    },
+    {
+      "epoch": 36.7047817047817,
+      "grad_norm": 0.0035215311218053102,
+      "learning_rate": 2.958034035440513e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13793488,
+      "step": 17655
+    },
+    {
+      "epoch": 36.71517671517672,
+      "grad_norm": 0.00038963829865679145,
+      "learning_rate": 2.957068870363201e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13797488,
+      "step": 17660
+    },
+    {
+      "epoch": 36.725571725571726,
+      "grad_norm": 0.0020385871175676584,
+      "learning_rate": 2.956103634800126e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13801328,
+      "step": 17665
+    },
+    {
+      "epoch": 36.735966735966734,
+      "grad_norm": 0.002559288637712598,
+      "learning_rate": 2.9551383289001384e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 13805232,
+      "step": 17670
+    },
+    {
+      "epoch": 36.74636174636175,
+      "grad_norm": 0.0032127108424901962,
+      "learning_rate": 2.9541729528121005e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13809264,
+      "step": 17675
+    },
+    {
+      "epoch": 36.75675675675676,
+      "grad_norm": 0.0016325335018336773,
+      "learning_rate": 2.9532075066848856e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 13813040,
+      "step": 17680
+    },
+    {
+      "epoch": 36.767151767151766,
+      "grad_norm": 0.0030751493759453297,
+      "learning_rate": 2.9522419906673786e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13816848,
+      "step": 17685
+    },
+    {
+      "epoch": 36.777546777546775,
+      "grad_norm": 0.0003635768953245133,
+      "learning_rate": 2.951276404908474e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13820656,
+      "step": 17690
+    },
+    {
+      "epoch": 36.78794178794179,
+      "grad_norm": 0.0012160009937360883,
+      "learning_rate": 2.9503107495570752e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13824464,
+      "step": 17695
+    },
+    {
+      "epoch": 36.7983367983368,
+      "grad_norm": 0.0005643287440761924,
+      "learning_rate": 2.9493450247621003e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13828464,
+      "step": 17700
+    },
+    {
+      "epoch": 36.80873180873181,
+      "grad_norm": 0.0005346128600649536,
+      "learning_rate": 2.948379230672476e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 13832432,
+      "step": 17705
+    },
+    {
+      "epoch": 36.81912681912682,
+      "grad_norm": 0.00015451267245225608,
+      "learning_rate": 2.9474133674371396e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13836400,
+      "step": 17710
+    },
+    {
+      "epoch": 36.82952182952183,
+      "grad_norm": 0.0004979278310202062,
+      "learning_rate": 2.9464474352050387e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13840496,
+      "step": 17715
+    },
+    {
+      "epoch": 36.83991683991684,
+      "grad_norm": 0.0005444091511890292,
+      "learning_rate": 2.9454814341251336e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13844400,
+      "step": 17720
+    },
+    {
+      "epoch": 36.85031185031185,
+      "grad_norm": 0.0003210220602340996,
+      "learning_rate": 2.9445153643463942e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 13848208,
+      "step": 17725
+    },
+    {
+      "epoch": 36.86070686070686,
+      "grad_norm": 9.686841076472774e-05,
+      "learning_rate": 2.943549226017798e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13852336,
+      "step": 17730
+    },
+    {
+      "epoch": 36.87110187110187,
+      "grad_norm": 8.678133599460125e-05,
+      "learning_rate": 2.942583019288337e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13856272,
+      "step": 17735
+    },
+    {
+      "epoch": 36.88149688149688,
+      "grad_norm": 0.004628194496035576,
+      "learning_rate": 2.9416167443070132e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13860016,
+      "step": 17740
+    },
+    {
+      "epoch": 36.891891891891895,
+      "grad_norm": 0.00013757136184722185,
+      "learning_rate": 2.9406504012228375e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13863888,
+      "step": 17745
+    },
+    {
+      "epoch": 36.9022869022869,
+      "grad_norm": 3.6651650589192286e-05,
+      "learning_rate": 2.939683990184832e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13867728,
+      "step": 17750
+    },
+    {
+      "epoch": 36.91268191268191,
+      "grad_norm": 0.0007856474258005619,
+      "learning_rate": 2.93871751134203e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 13871760,
+      "step": 17755
+    },
+    {
+      "epoch": 36.92307692307692,
+      "grad_norm": 0.004979207646101713,
+      "learning_rate": 2.9377509648434752e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13875632,
+      "step": 17760
+    },
+    {
+      "epoch": 36.933471933471935,
+      "grad_norm": 0.0002716601302381605,
+      "learning_rate": 2.9367843508382203e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13879440,
+      "step": 17765
+    },
+    {
+      "epoch": 36.943866943866944,
+      "grad_norm": 0.00045719632180407643,
+      "learning_rate": 2.9358176694753293e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13883280,
+      "step": 17770
+    },
+    {
+      "epoch": 36.95426195426195,
+      "grad_norm": 7.141811511246487e-05,
+      "learning_rate": 2.9348509209038766e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13887152,
+      "step": 17775
+    },
+    {
+      "epoch": 36.96465696465697,
+      "grad_norm": 4.620255276677199e-05,
+      "learning_rate": 2.933884105272947e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13890864,
+      "step": 17780
+    },
+    {
+      "epoch": 36.975051975051976,
+      "grad_norm": 0.0001510274305474013,
+      "learning_rate": 2.9329172227316366e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13894640,
+      "step": 17785
+    },
+    {
+      "epoch": 36.985446985446984,
+      "grad_norm": 0.00200596172362566,
+      "learning_rate": 2.93195027342905e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13898576,
+      "step": 17790
+    },
+    {
+      "epoch": 36.99584199584199,
+      "grad_norm": 0.0001429947733413428,
+      "learning_rate": 2.9309832575143024e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13902576,
+      "step": 17795
+    },
+    {
+      "epoch": 37.00623700623701,
+      "grad_norm": 0.00012227060506120324,
+      "learning_rate": 2.930016175136521e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 13906304,
+      "step": 17800
+    },
+    {
+      "epoch": 37.00623700623701,
+      "eval_loss": 0.47043538093566895,
+      "eval_runtime": 11.697,
+      "eval_samples_per_second": 73.181,
+      "eval_steps_per_second": 18.295,
+      "num_input_tokens_seen": 13906304,
+      "step": 17800
+    },
+    {
+      "epoch": 37.016632016632016,
+      "grad_norm": 0.0014924807474017143,
+      "learning_rate": 2.9290490264448412e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13910176,
+      "step": 17805
+    },
+    {
+      "epoch": 37.027027027027025,
+      "grad_norm": 0.000160578332724981,
+      "learning_rate": 2.9280818115884094e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13914048,
+      "step": 17810
+    },
+    {
+      "epoch": 37.03742203742204,
+      "grad_norm": 0.0006973114213906229,
+      "learning_rate": 2.9271145307163828e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13918016,
+      "step": 17815
+    },
+    {
+      "epoch": 37.04781704781705,
+      "grad_norm": 4.023642395623028e-05,
+      "learning_rate": 2.9261471839779287e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13922016,
+      "step": 17820
+    },
+    {
+      "epoch": 37.05821205821206,
+      "grad_norm": 0.0006723172846250236,
+      "learning_rate": 2.925179771522223e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13925888,
+      "step": 17825
+    },
+    {
+      "epoch": 37.06860706860707,
+      "grad_norm": 0.00028692890191450715,
+      "learning_rate": 2.9242122934984535e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 13929696,
+      "step": 17830
+    },
+    {
+      "epoch": 37.07900207900208,
+      "grad_norm": 0.00012977764708921313,
+      "learning_rate": 2.9232447500558176e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13933568,
+      "step": 17835
+    },
+    {
+      "epoch": 37.08939708939709,
+      "grad_norm": 0.0005242000333964825,
+      "learning_rate": 2.9222771413435225e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13937376,
+      "step": 17840
+    },
+    {
+      "epoch": 37.0997920997921,
+      "grad_norm": 0.00017074336938094348,
+      "learning_rate": 2.9213094675107848e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13941280,
+      "step": 17845
+    },
+    {
+      "epoch": 37.11018711018711,
+      "grad_norm": 0.00021043122978881001,
+      "learning_rate": 2.9203417287068335e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13945184,
+      "step": 17850
+    },
+    {
+      "epoch": 37.12058212058212,
+      "grad_norm": 0.00022233273193705827,
+      "learning_rate": 2.9193739250809042e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13949312,
+      "step": 17855
+    },
+    {
+      "epoch": 37.13097713097713,
+      "grad_norm": 8.448653534287587e-05,
+      "learning_rate": 2.9184060567822463e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13953216,
+      "step": 17860
+    },
+    {
+      "epoch": 37.141372141372145,
+      "grad_norm": 7.810819806763902e-05,
+      "learning_rate": 2.9174381239601166e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13957248,
+      "step": 17865
+    },
+    {
+      "epoch": 37.15176715176715,
+      "grad_norm": 0.00010649619798641652,
+      "learning_rate": 2.916470126763783e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13961312,
+      "step": 17870
+    },
+    {
+      "epoch": 37.16216216216216,
+      "grad_norm": 0.0007011841516941786,
+      "learning_rate": 2.9155020653425203e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 13965248,
+      "step": 17875
+    },
+    {
+      "epoch": 37.17255717255717,
+      "grad_norm": 2.878488157875836e-05,
+      "learning_rate": 2.9145339398456184e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13969184,
+      "step": 17880
+    },
+    {
+      "epoch": 37.182952182952185,
+      "grad_norm": 0.00020068598678335547,
+      "learning_rate": 2.913565750422374e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13972896,
+      "step": 17885
+    },
+    {
+      "epoch": 37.19334719334719,
+      "grad_norm": 0.0005069992039352655,
+      "learning_rate": 2.9125974972220938e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 13976768,
+      "step": 17890
+    },
+    {
+      "epoch": 37.2037422037422,
+      "grad_norm": 0.0002692264097277075,
+      "learning_rate": 2.9116291803940932e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 13980704,
+      "step": 17895
+    },
+    {
+      "epoch": 37.21413721413722,
+      "grad_norm": 0.0001622752461116761,
+      "learning_rate": 2.910660800087701e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13984704,
+      "step": 17900
+    },
+    {
+      "epoch": 37.224532224532226,
+      "grad_norm": 1.8542683392297477e-05,
+      "learning_rate": 2.909692356452254e-05,
+      "loss": 0.0061,
+      "num_input_tokens_seen": 13988768,
+      "step": 17905
+    },
+    {
+      "epoch": 37.234927234927234,
+      "grad_norm": 7.240205741254613e-05,
+      "learning_rate": 2.9087238496370962e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 13992640,
+      "step": 17910
+    },
+    {
+      "epoch": 37.24532224532224,
+      "grad_norm": 0.0005842349492013454,
+      "learning_rate": 2.907755279791583e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 13996512,
+      "step": 17915
+    },
+    {
+      "epoch": 37.25571725571726,
+      "grad_norm": 7.287641346920282e-05,
+      "learning_rate": 2.906786647065083e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14000416,
+      "step": 17920
+    },
+    {
+      "epoch": 37.266112266112266,
+      "grad_norm": 5.3801479225512594e-05,
+      "learning_rate": 2.9058179516069695e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14004384,
+      "step": 17925
+    },
+    {
+      "epoch": 37.276507276507274,
+      "grad_norm": 4.33320710726548e-05,
+      "learning_rate": 2.9048491935666282e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14008288,
+      "step": 17930
+    },
+    {
+      "epoch": 37.28690228690229,
+      "grad_norm": 5.532135401153937e-05,
+      "learning_rate": 2.9038803730934534e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14012192,
+      "step": 17935
+    },
+    {
+      "epoch": 37.2972972972973,
+      "grad_norm": 0.00023906338901724666,
+      "learning_rate": 2.9029114903368503e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14016128,
+      "step": 17940
+    },
+    {
+      "epoch": 37.30769230769231,
+      "grad_norm": 0.004563345108181238,
+      "learning_rate": 2.9019425454462318e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14020096,
+      "step": 17945
+    },
+    {
+      "epoch": 37.318087318087315,
+      "grad_norm": 7.609012391185388e-05,
+      "learning_rate": 2.9009735385710212e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14023840,
+      "step": 17950
+    },
+    {
+      "epoch": 37.32848232848233,
+      "grad_norm": 3.9351270970655605e-05,
+      "learning_rate": 2.900004469860652e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14027648,
+      "step": 17955
+    },
+    {
+      "epoch": 37.33887733887734,
+      "grad_norm": 0.000144250676385127,
+      "learning_rate": 2.8990353394645668e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 14031520,
+      "step": 17960
+    },
+    {
+      "epoch": 37.34927234927235,
+      "grad_norm": 4.648109097615816e-05,
+      "learning_rate": 2.8980661475322186e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14035392,
+      "step": 17965
+    },
+    {
+      "epoch": 37.35966735966736,
+      "grad_norm": 0.0006870568613521755,
+      "learning_rate": 2.897096894213067e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 14039296,
+      "step": 17970
+    },
+    {
+      "epoch": 37.37006237006237,
+      "grad_norm": 0.1399753838777542,
+      "learning_rate": 2.8961275796565845e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 14043104,
+      "step": 17975
+    },
+    {
+      "epoch": 37.38045738045738,
+      "grad_norm": 9.693729225546122e-05,
+      "learning_rate": 2.8951582040122517e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14047104,
+      "step": 17980
+    },
+    {
+      "epoch": 37.39085239085239,
+      "grad_norm": 0.00010231769556412473,
+      "learning_rate": 2.894188767429557e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 14051072,
+      "step": 17985
+    },
+    {
+      "epoch": 37.4012474012474,
+      "grad_norm": 8.372770389541984e-05,
+      "learning_rate": 2.8932192700580014e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14054976,
+      "step": 17990
+    },
+    {
+      "epoch": 37.41164241164241,
+      "grad_norm": 3.0956902503967285,
+      "learning_rate": 2.8922497120470916e-05,
+      "loss": 0.0348,
+      "num_input_tokens_seen": 14058912,
+      "step": 17995
+    },
+    {
+      "epoch": 37.42203742203742,
+      "grad_norm": 6.584256334463134e-05,
+      "learning_rate": 2.891280093546348e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14062784,
+      "step": 18000
+    },
+    {
+      "epoch": 37.42203742203742,
+      "eval_loss": 0.39015039801597595,
+      "eval_runtime": 11.6857,
+      "eval_samples_per_second": 73.252,
+      "eval_steps_per_second": 18.313,
+      "num_input_tokens_seen": 14062784,
+      "step": 18000
+    },
+    {
+      "epoch": 37.432432432432435,
+      "grad_norm": 0.006829886697232723,
+      "learning_rate": 2.890310414705297e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14066784,
+      "step": 18005
+    },
+    {
+      "epoch": 37.44282744282744,
+      "grad_norm": 0.01576889492571354,
+      "learning_rate": 2.8893406756734742e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 14070720,
+      "step": 18010
+    },
+    {
+      "epoch": 37.45322245322245,
+      "grad_norm": 0.14528755843639374,
+      "learning_rate": 2.888370876600427e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 14074784,
+      "step": 18015
+    },
+    {
+      "epoch": 37.46361746361746,
+      "grad_norm": 0.004521089605987072,
+      "learning_rate": 2.8874010176357104e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 14078784,
+      "step": 18020
+    },
+    {
+      "epoch": 37.474012474012476,
+      "grad_norm": 0.0006730987224727869,
+      "learning_rate": 2.886431098928888e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 14082528,
+      "step": 18025
+    },
+    {
+      "epoch": 37.484407484407484,
+      "grad_norm": 0.0002923270803876221,
+      "learning_rate": 2.885461120629534e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14086464,
+      "step": 18030
+    },
+    {
+      "epoch": 37.49480249480249,
+      "grad_norm": 0.00015976728172972798,
+      "learning_rate": 2.8844910828872317e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14090336,
+      "step": 18035
+    },
+    {
+      "epoch": 37.50519750519751,
+      "grad_norm": 0.0009739818633534014,
+      "learning_rate": 2.8835209858515715e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14094336,
+      "step": 18040
+    },
+    {
+      "epoch": 37.515592515592516,
+      "grad_norm": 0.0001482696970924735,
+      "learning_rate": 2.8825508296721566e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 14098304,
+      "step": 18045
+    },
+    {
+      "epoch": 37.525987525987524,
+      "grad_norm": 0.0005195132689550519,
+      "learning_rate": 2.881580614498596e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14102176,
+      "step": 18050
+    },
+    {
+      "epoch": 37.53638253638254,
+      "grad_norm": 0.003775889752432704,
+      "learning_rate": 2.8806103404805103e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14106144,
+      "step": 18055
+    },
+    {
+      "epoch": 37.54677754677755,
+      "grad_norm": 0.0001741406013024971,
+      "learning_rate": 2.8796400077675257e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 14110208,
+      "step": 18060
+    },
+    {
+      "epoch": 37.55717255717256,
+      "grad_norm": 0.00012469099601730704,
+      "learning_rate": 2.8786696165092812e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14114016,
+      "step": 18065
+    },
+    {
+      "epoch": 37.567567567567565,
+      "grad_norm": 0.0005225223721936345,
+      "learning_rate": 2.8776991668554236e-05,
+      "loss": 0.0089,
+      "num_input_tokens_seen": 14117824,
+      "step": 18070
+    },
+    {
+      "epoch": 37.57796257796258,
+      "grad_norm": 0.0004657044482883066,
+      "learning_rate": 2.876728658955608e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14121760,
+      "step": 18075
+    },
+    {
+      "epoch": 37.58835758835759,
+      "grad_norm": 0.0022629061713814735,
+      "learning_rate": 2.8757580929594986e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 14125600,
+      "step": 18080
+    },
+    {
+      "epoch": 37.5987525987526,
+      "grad_norm": 9.66729421634227e-05,
+      "learning_rate": 2.87478746901677e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14129312,
+      "step": 18085
+    },
+    {
+      "epoch": 37.60914760914761,
+      "grad_norm": 0.000821197871118784,
+      "learning_rate": 2.873816787277103e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14133184,
+      "step": 18090
+    },
+    {
+      "epoch": 37.61954261954262,
+      "grad_norm": 7.066036050673574e-05,
+      "learning_rate": 2.8728460478901903e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14137120,
+      "step": 18095
+    },
+    {
+      "epoch": 37.62993762993763,
+      "grad_norm": 0.00022460044419858605,
+      "learning_rate": 2.8718752510057307e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14141088,
+      "step": 18100
+    },
+    {
+      "epoch": 37.64033264033264,
+      "grad_norm": 0.0013480279594659805,
+      "learning_rate": 2.870904396773435e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14145120,
+      "step": 18105
+    },
+    {
+      "epoch": 37.65072765072765,
+      "grad_norm": 0.0002367514098295942,
+      "learning_rate": 2.86993348534302e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14148960,
+      "step": 18110
+    },
+    {
+      "epoch": 37.66112266112266,
+      "grad_norm": 0.0002395766496192664,
+      "learning_rate": 2.868962516864212e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14152896,
+      "step": 18115
+    },
+    {
+      "epoch": 37.67151767151767,
+      "grad_norm": 0.0002632838150020689,
+      "learning_rate": 2.8679914914867477e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14156800,
+      "step": 18120
+    },
+    {
+      "epoch": 37.681912681912685,
+      "grad_norm": 0.0010888942051678896,
+      "learning_rate": 2.8670204093603713e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14160704,
+      "step": 18125
+    },
+    {
+      "epoch": 37.69230769230769,
+      "grad_norm": 0.0006279794033616781,
+      "learning_rate": 2.8660492706348357e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14164576,
+      "step": 18130
+    },
+    {
+      "epoch": 37.7027027027027,
+      "grad_norm": 0.001130963210016489,
+      "learning_rate": 2.8650780754599022e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14168480,
+      "step": 18135
+    },
+    {
+      "epoch": 37.71309771309771,
+      "grad_norm": 0.003262902610003948,
+      "learning_rate": 2.8641068239853407e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14172448,
+      "step": 18140
+    },
+    {
+      "epoch": 37.723492723492726,
+      "grad_norm": 0.003171959426254034,
+      "learning_rate": 2.863135516360932e-05,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 14176352,
+      "step": 18145
+    },
+    {
+      "epoch": 37.733887733887734,
+      "grad_norm": 9.178261097986251e-05,
+      "learning_rate": 2.8621641527364633e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14180224,
+      "step": 18150
+    },
+    {
+      "epoch": 37.74428274428274,
+      "grad_norm": 0.00031358300475403666,
+      "learning_rate": 2.8611927332617313e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14184096,
+      "step": 18155
+    },
+    {
+      "epoch": 37.75467775467776,
+      "grad_norm": 0.0007381524192169309,
+      "learning_rate": 2.8602212580865405e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14188096,
+      "step": 18160
+    },
+    {
+      "epoch": 37.765072765072766,
+      "grad_norm": 0.0009369780891574919,
+      "learning_rate": 2.859249727360705e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 14192160,
+      "step": 18165
+    },
+    {
+      "epoch": 37.775467775467774,
+      "grad_norm": 0.003846742445603013,
+      "learning_rate": 2.8582781412340465e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14195968,
+      "step": 18170
+    },
+    {
+      "epoch": 37.78586278586278,
+      "grad_norm": 0.0002559203130658716,
+      "learning_rate": 2.857306499856397e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14199872,
+      "step": 18175
+    },
+    {
+      "epoch": 37.7962577962578,
+      "grad_norm": 0.0004959040088579059,
+      "learning_rate": 2.856334803377594e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14203648,
+      "step": 18180
+    },
+    {
+      "epoch": 37.80665280665281,
+      "grad_norm": 0.00010024299990618601,
+      "learning_rate": 2.8553630519474867e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 14207424,
+      "step": 18185
+    },
+    {
+      "epoch": 37.817047817047815,
+      "grad_norm": 0.0005270784022286534,
+      "learning_rate": 2.8543912457159317e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14211328,
+      "step": 18190
+    },
+    {
+      "epoch": 37.82744282744283,
+      "grad_norm": 0.002818794921040535,
+      "learning_rate": 2.853419384832792e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14215232,
+      "step": 18195
+    },
+    {
+      "epoch": 37.83783783783784,
+      "grad_norm": 0.00011119317059637979,
+      "learning_rate": 2.8524474694479423e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14219168,
+      "step": 18200
+    },
+    {
+      "epoch": 37.83783783783784,
+      "eval_loss": 0.4358499050140381,
+      "eval_runtime": 11.6935,
+      "eval_samples_per_second": 73.203,
+      "eval_steps_per_second": 18.301,
+      "num_input_tokens_seen": 14219168,
+      "step": 18200
+    },
+    {
+      "epoch": 37.84823284823285,
+      "grad_norm": 0.000946188229136169,
+      "learning_rate": 2.851475499711264e-05,
+      "loss": 0.018,
+      "num_input_tokens_seen": 14223040,
+      "step": 18205
+    },
+    {
+      "epoch": 37.858627858627855,
+      "grad_norm": 5.7517205277690664e-05,
+      "learning_rate": 2.8505034757726468e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14226976,
+      "step": 18210
+    },
+    {
+      "epoch": 37.86902286902287,
+      "grad_norm": 0.0003575643931981176,
+      "learning_rate": 2.8495313977819886e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 14230848,
+      "step": 18215
+    },
+    {
+      "epoch": 37.87941787941788,
+      "grad_norm": 0.12316498905420303,
+      "learning_rate": 2.8485592658891956e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 14234816,
+      "step": 18220
+    },
+    {
+      "epoch": 37.88981288981289,
+      "grad_norm": 0.301857054233551,
+      "learning_rate": 2.8475870802441844e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14238656,
+      "step": 18225
+    },
+    {
+      "epoch": 37.9002079002079,
+      "grad_norm": 0.0013071176363155246,
+      "learning_rate": 2.8466148409968774e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14242560,
+      "step": 18230
+    },
+    {
+      "epoch": 37.91060291060291,
+      "grad_norm": 0.00027913396479561925,
+      "learning_rate": 2.8456425482972067e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 14246528,
+      "step": 18235
+    },
+    {
+      "epoch": 37.92099792099792,
+      "grad_norm": 5.611907545244321e-05,
+      "learning_rate": 2.84467020229511e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14250240,
+      "step": 18240
+    },
+    {
+      "epoch": 37.931392931392935,
+      "grad_norm": 0.0017570200143381953,
+      "learning_rate": 2.8436978031405375e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14254016,
+      "step": 18245
+    },
+    {
+      "epoch": 37.94178794178794,
+      "grad_norm": 0.00014574271335732192,
+      "learning_rate": 2.842725350983445e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14257888,
+      "step": 18250
+    },
+    {
+      "epoch": 37.95218295218295,
+      "grad_norm": 0.00023738201707601547,
+      "learning_rate": 2.8417528459737957e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14261952,
+      "step": 18255
+    },
+    {
+      "epoch": 37.96257796257796,
+      "grad_norm": 0.00015755236381664872,
+      "learning_rate": 2.8407802882615624e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14265792,
+      "step": 18260
+    },
+    {
+      "epoch": 37.972972972972975,
+      "grad_norm": 0.0012987665832042694,
+      "learning_rate": 2.8398076779967277e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14269632,
+      "step": 18265
+    },
+    {
+      "epoch": 37.983367983367984,
+      "grad_norm": 7.285497849807143e-05,
+      "learning_rate": 2.8388350153292774e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14273472,
+      "step": 18270
+    },
+    {
+      "epoch": 37.99376299376299,
+      "grad_norm": 0.0003221695660613477,
+      "learning_rate": 2.8378623004092103e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14277312,
+      "step": 18275
+    },
+    {
+      "epoch": 38.00415800415801,
+      "grad_norm": 0.0002299432671861723,
+      "learning_rate": 2.8368895333865302e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 14281232,
+      "step": 18280
+    },
+    {
+      "epoch": 38.014553014553016,
+      "grad_norm": 0.13911522924900055,
+      "learning_rate": 2.835916714411251e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 14285104,
+      "step": 18285
+    },
+    {
+      "epoch": 38.024948024948024,
+      "grad_norm": 1.7972288333112374e-05,
+      "learning_rate": 2.8349438436333926e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14288944,
+      "step": 18290
+    },
+    {
+      "epoch": 38.03534303534303,
+      "grad_norm": 9.579512698110193e-05,
+      "learning_rate": 2.833970921202984e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14292848,
+      "step": 18295
+    },
+    {
+      "epoch": 38.04573804573805,
+      "grad_norm": 0.0008987009641714394,
+      "learning_rate": 2.8329979472700628e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14296816,
+      "step": 18300
+    },
+    {
+      "epoch": 38.056133056133056,
+      "grad_norm": 0.00032773238490335643,
+      "learning_rate": 2.832024921984674e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14300816,
+      "step": 18305
+    },
+    {
+      "epoch": 38.066528066528065,
+      "grad_norm": 0.0005956653621979058,
+      "learning_rate": 2.8310518454968693e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14304592,
+      "step": 18310
+    },
+    {
+      "epoch": 38.07692307692308,
+      "grad_norm": 0.0008557001128792763,
+      "learning_rate": 2.8300787179567095e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 14308432,
+      "step": 18315
+    },
+    {
+      "epoch": 38.08731808731809,
+      "grad_norm": 7.407024531858042e-05,
+      "learning_rate": 2.8291055395142636e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14312368,
+      "step": 18320
+    },
+    {
+      "epoch": 38.0977130977131,
+      "grad_norm": 8.006991993170232e-05,
+      "learning_rate": 2.8281323103196073e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 14316208,
+      "step": 18325
+    },
+    {
+      "epoch": 38.108108108108105,
+      "grad_norm": 0.0007671688217669725,
+      "learning_rate": 2.8271590305228256e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14320144,
+      "step": 18330
+    },
+    {
+      "epoch": 38.11850311850312,
+      "grad_norm": 5.6384211347904056e-05,
+      "learning_rate": 2.82618570027401e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14324144,
+      "step": 18335
+    },
+    {
+      "epoch": 38.12889812889813,
+      "grad_norm": 0.0008092052303254604,
+      "learning_rate": 2.8252123197232604e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14328112,
+      "step": 18340
+    },
+    {
+      "epoch": 38.13929313929314,
+      "grad_norm": 0.0008560745045542717,
+      "learning_rate": 2.8242388890206843e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14331920,
+      "step": 18345
+    },
+    {
+      "epoch": 38.14968814968815,
+      "grad_norm": 0.1057448759675026,
+      "learning_rate": 2.8232654083163967e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14335696,
+      "step": 18350
+    },
+    {
+      "epoch": 38.16008316008316,
+      "grad_norm": 0.0001665350719122216,
+      "learning_rate": 2.822291877760521e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14339632,
+      "step": 18355
+    },
+    {
+      "epoch": 38.17047817047817,
+      "grad_norm": 0.0007422614726237953,
+      "learning_rate": 2.8213182975031864e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14343632,
+      "step": 18360
+    },
+    {
+      "epoch": 38.18087318087318,
+      "grad_norm": 4.146387072978541e-05,
+      "learning_rate": 2.8203446676945337e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14347472,
+      "step": 18365
+    },
+    {
+      "epoch": 38.19126819126819,
+      "grad_norm": 4.6839115384500474e-05,
+      "learning_rate": 2.8193709884847075e-05,
+      "loss": 0.0045,
+      "num_input_tokens_seen": 14351280,
+      "step": 18370
+    },
+    {
+      "epoch": 38.2016632016632,
+      "grad_norm": 0.0001919236674439162,
+      "learning_rate": 2.8183972600238605e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14355280,
+      "step": 18375
+    },
+    {
+      "epoch": 38.21205821205821,
+      "grad_norm": 0.00033286443795077503,
+      "learning_rate": 2.817423482462156e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 14359248,
+      "step": 18380
+    },
+    {
+      "epoch": 38.222453222453225,
+      "grad_norm": 0.0018096743151545525,
+      "learning_rate": 2.8164496559497605e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14363120,
+      "step": 18385
+    },
+    {
+      "epoch": 38.232848232848234,
+      "grad_norm": 0.0004173617926426232,
+      "learning_rate": 2.815475780636852e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 14367216,
+      "step": 18390
+    },
+    {
+      "epoch": 38.24324324324324,
+      "grad_norm": 0.00350003014318645,
+      "learning_rate": 2.814501856673613e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14371152,
+      "step": 18395
+    },
+    {
+      "epoch": 38.25363825363825,
+      "grad_norm": 0.00010663764260243624,
+      "learning_rate": 2.8135278842102353e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14375024,
+      "step": 18400
+    },
+    {
+      "epoch": 38.25363825363825,
+      "eval_loss": 0.4856085479259491,
+      "eval_runtime": 11.7047,
+      "eval_samples_per_second": 73.133,
+      "eval_steps_per_second": 18.283,
+      "num_input_tokens_seen": 14375024,
+      "step": 18400
+    },
+    {
+      "epoch": 38.264033264033266,
+      "grad_norm": 0.0001398419262841344,
+      "learning_rate": 2.8125538633969183e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14378864,
+      "step": 18405
+    },
+    {
+      "epoch": 38.274428274428274,
+      "grad_norm": 2.6038655050797388e-05,
+      "learning_rate": 2.8115797943838677e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14382896,
+      "step": 18410
+    },
+    {
+      "epoch": 38.28482328482328,
+      "grad_norm": 0.00015786837320774794,
+      "learning_rate": 2.810605677321298e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14386864,
+      "step": 18415
+    },
+    {
+      "epoch": 38.2952182952183,
+      "grad_norm": 0.00011184514733031392,
+      "learning_rate": 2.809631512359428e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14390736,
+      "step": 18420
+    },
+    {
+      "epoch": 38.305613305613306,
+      "grad_norm": 0.0007220740662887692,
+      "learning_rate": 2.8086572996484884e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14394768,
+      "step": 18425
+    },
+    {
+      "epoch": 38.316008316008315,
+      "grad_norm": 4.719903154182248e-05,
+      "learning_rate": 2.8076830393387143e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14398672,
+      "step": 18430
+    },
+    {
+      "epoch": 38.32640332640332,
+      "grad_norm": 0.0002659767633304,
+      "learning_rate": 2.8067087315803497e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14402448,
+      "step": 18435
+    },
+    {
+      "epoch": 38.33679833679834,
+      "grad_norm": 9.212831355398521e-05,
+      "learning_rate": 2.8057343765236433e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14406288,
+      "step": 18440
+    },
+    {
+      "epoch": 38.34719334719335,
+      "grad_norm": 0.10091795027256012,
+      "learning_rate": 2.804759974318854e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 14410128,
+      "step": 18445
+    },
+    {
+      "epoch": 38.357588357588355,
+      "grad_norm": 9.51511028688401e-05,
+      "learning_rate": 2.8037855251162482e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14413968,
+      "step": 18450
+    },
+    {
+      "epoch": 38.36798336798337,
+      "grad_norm": 0.00019378152501303703,
+      "learning_rate": 2.802811029066096e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 14417872,
+      "step": 18455
+    },
+    {
+      "epoch": 38.37837837837838,
+      "grad_norm": 0.00011920429824385792,
+      "learning_rate": 2.8018364863186764e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14421712,
+      "step": 18460
+    },
+    {
+      "epoch": 38.38877338877339,
+      "grad_norm": 74.1795883178711,
+      "learning_rate": 2.800861897024279e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 14425616,
+      "step": 18465
+    },
+    {
+      "epoch": 38.3991683991684,
+      "grad_norm": 8.791544678388163e-05,
+      "learning_rate": 2.799887261333196e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14429552,
+      "step": 18470
+    },
+    {
+      "epoch": 38.40956340956341,
+      "grad_norm": 0.00023116712691262364,
+      "learning_rate": 2.798912579395728e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14433456,
+      "step": 18475
+    },
+    {
+      "epoch": 38.41995841995842,
+      "grad_norm": 2.7163659979123622e-05,
+      "learning_rate": 2.797937851362185e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14437232,
+      "step": 18480
+    },
+    {
+      "epoch": 38.43035343035343,
+      "grad_norm": 0.00199466897174716,
+      "learning_rate": 2.7969630773828802e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 14441136,
+      "step": 18485
+    },
+    {
+      "epoch": 38.44074844074844,
+      "grad_norm": 5.3290110372472554e-05,
+      "learning_rate": 2.7959882576081382e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14444880,
+      "step": 18490
+    },
+    {
+      "epoch": 38.45114345114345,
+      "grad_norm": 0.006551511585712433,
+      "learning_rate": 2.795013392188286e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 14448784,
+      "step": 18495
+    },
+    {
+      "epoch": 38.46153846153846,
+      "grad_norm": 3.5635555832413957e-05,
+      "learning_rate": 2.7940384812736614e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14452688,
+      "step": 18500
+    },
+    {
+      "epoch": 38.471933471933475,
+      "grad_norm": 0.00013964246318209916,
+      "learning_rate": 2.7930635250146087e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 14456656,
+      "step": 18505
+    },
+    {
+      "epoch": 38.482328482328484,
+      "grad_norm": 5.401599992183037e-05,
+      "learning_rate": 2.792088523561477e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14460624,
+      "step": 18510
+    },
+    {
+      "epoch": 38.49272349272349,
+      "grad_norm": 0.00010239979019388556,
+      "learning_rate": 2.7911134770646246e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14464496,
+      "step": 18515
+    },
+    {
+      "epoch": 38.5031185031185,
+      "grad_norm": 0.00011728092795237899,
+      "learning_rate": 2.7901383856744157e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14468432,
+      "step": 18520
+    },
+    {
+      "epoch": 38.513513513513516,
+      "grad_norm": 0.0009292506729252636,
+      "learning_rate": 2.7891632495412217e-05,
+      "loss": 0.0322,
+      "num_input_tokens_seen": 14472304,
+      "step": 18525
+    },
+    {
+      "epoch": 38.523908523908524,
+      "grad_norm": 0.00026752808480523527,
+      "learning_rate": 2.7881880688154205e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 14476272,
+      "step": 18530
+    },
+    {
+      "epoch": 38.53430353430353,
+      "grad_norm": 0.23965856432914734,
+      "learning_rate": 2.7872128436473977e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 14480208,
+      "step": 18535
+    },
+    {
+      "epoch": 38.54469854469855,
+      "grad_norm": 0.29733675718307495,
+      "learning_rate": 2.7862375741875448e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 14484144,
+      "step": 18540
+    },
+    {
+      "epoch": 38.555093555093556,
+      "grad_norm": 0.00013785011833533645,
+      "learning_rate": 2.785262260586261e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 14488048,
+      "step": 18545
+    },
+    {
+      "epoch": 38.565488565488565,
+      "grad_norm": 0.0011730899568647146,
+      "learning_rate": 2.7842869029939517e-05,
+      "loss": 0.009,
+      "num_input_tokens_seen": 14492016,
+      "step": 18550
+    },
+    {
+      "epoch": 38.57588357588357,
+      "grad_norm": 0.0011266242945566773,
+      "learning_rate": 2.7833115015610296e-05,
+      "loss": 0.034,
+      "num_input_tokens_seen": 14495888,
+      "step": 18555
+    },
+    {
+      "epoch": 38.58627858627859,
+      "grad_norm": 2.573843002319336,
+      "learning_rate": 2.7823360564379136e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 14499824,
+      "step": 18560
+    },
+    {
+      "epoch": 38.5966735966736,
+      "grad_norm": 0.014244379475712776,
+      "learning_rate": 2.7813605677750297e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14503728,
+      "step": 18565
+    },
+    {
+      "epoch": 38.607068607068605,
+      "grad_norm": 0.0029765432700514793,
+      "learning_rate": 2.7803850357228102e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 14507600,
+      "step": 18570
+    },
+    {
+      "epoch": 38.61746361746362,
+      "grad_norm": 0.0035077042412012815,
+      "learning_rate": 2.779409460431695e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14511536,
+      "step": 18575
+    },
+    {
+      "epoch": 38.62785862785863,
+      "grad_norm": 0.0012730876915156841,
+      "learning_rate": 2.778433842052129e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14515440,
+      "step": 18580
+    },
+    {
+      "epoch": 38.63825363825364,
+      "grad_norm": 0.019090162590146065,
+      "learning_rate": 2.7774581807345664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14519280,
+      "step": 18585
+    },
+    {
+      "epoch": 38.648648648648646,
+      "grad_norm": 0.0008092496427707374,
+      "learning_rate": 2.776482476629465e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14523088,
+      "step": 18590
+    },
+    {
+      "epoch": 38.65904365904366,
+      "grad_norm": 0.0002354128228034824,
+      "learning_rate": 2.7755067298872924e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14526960,
+      "step": 18595
+    },
+    {
+      "epoch": 38.66943866943867,
+      "grad_norm": 0.07045325636863708,
+      "learning_rate": 2.774530940658518e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 14530800,
+      "step": 18600
+    },
+    {
+      "epoch": 38.66943866943867,
+      "eval_loss": 0.3732648491859436,
+      "eval_runtime": 11.6843,
+      "eval_samples_per_second": 73.261,
+      "eval_steps_per_second": 18.315,
+      "num_input_tokens_seen": 14530800,
+      "step": 18600
+    },
+    {
+      "epoch": 38.67983367983368,
+      "grad_norm": 0.0025774617679417133,
+      "learning_rate": 2.7735551090936236e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14534672,
+      "step": 18605
+    },
+    {
+      "epoch": 38.69022869022869,
+      "grad_norm": 0.17132465541362762,
+      "learning_rate": 2.7725792353430934e-05,
+      "loss": 0.0073,
+      "num_input_tokens_seen": 14538480,
+      "step": 18610
+    },
+    {
+      "epoch": 38.7006237006237,
+      "grad_norm": 0.000321213185088709,
+      "learning_rate": 2.77160331955742e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14542480,
+      "step": 18615
+    },
+    {
+      "epoch": 38.71101871101871,
+      "grad_norm": 0.0006285730632953346,
+      "learning_rate": 2.7706273618871008e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14546480,
+      "step": 18620
+    },
+    {
+      "epoch": 38.72141372141372,
+      "grad_norm": 0.002042429754510522,
+      "learning_rate": 2.769651362482642e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14550448,
+      "step": 18625
+    },
+    {
+      "epoch": 38.731808731808734,
+      "grad_norm": 0.0009693122701719403,
+      "learning_rate": 2.768675321494555e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14554384,
+      "step": 18630
+    },
+    {
+      "epoch": 38.74220374220374,
+      "grad_norm": 0.0023413123562932014,
+      "learning_rate": 2.7676992390733565e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14558352,
+      "step": 18635
+    },
+    {
+      "epoch": 38.75259875259875,
+      "grad_norm": 0.0002533749211579561,
+      "learning_rate": 2.766723115369571e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14562160,
+      "step": 18640
+    },
+    {
+      "epoch": 38.762993762993766,
+      "grad_norm": 0.000477102177683264,
+      "learning_rate": 2.765746950533729e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14566096,
+      "step": 18645
+    },
+    {
+      "epoch": 38.773388773388774,
+      "grad_norm": 0.004135078750550747,
+      "learning_rate": 2.7647707447163684e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14570032,
+      "step": 18650
+    },
+    {
+      "epoch": 38.78378378378378,
+      "grad_norm": 0.00030244726804085076,
+      "learning_rate": 2.7637944980680315e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14574064,
+      "step": 18655
+    },
+    {
+      "epoch": 38.79417879417879,
+      "grad_norm": 0.0016814526170492172,
+      "learning_rate": 2.762818210739268e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14577968,
+      "step": 18660
+    },
+    {
+      "epoch": 38.804573804573806,
+      "grad_norm": 0.001456936472095549,
+      "learning_rate": 2.7618418828806332e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14582000,
+      "step": 18665
+    },
+    {
+      "epoch": 38.814968814968815,
+      "grad_norm": 0.002588148694485426,
+      "learning_rate": 2.76086551464269e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14585904,
+      "step": 18670
+    },
+    {
+      "epoch": 38.82536382536382,
+      "grad_norm": 0.0007764542242512107,
+      "learning_rate": 2.759889106176006e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14589840,
+      "step": 18675
+    },
+    {
+      "epoch": 38.83575883575884,
+      "grad_norm": 0.00035538809606805444,
+      "learning_rate": 2.758912657631156e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14593936,
+      "step": 18680
+    },
+    {
+      "epoch": 38.84615384615385,
+      "grad_norm": 0.0017344767693430185,
+      "learning_rate": 2.7579361691587198e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14597744,
+      "step": 18685
+    },
+    {
+      "epoch": 38.856548856548855,
+      "grad_norm": 29.8558406829834,
+      "learning_rate": 2.756959640909285e-05,
+      "loss": 0.0224,
+      "num_input_tokens_seen": 14601616,
+      "step": 18690
+    },
+    {
+      "epoch": 38.86694386694387,
+      "grad_norm": 0.004166330210864544,
+      "learning_rate": 2.7559830730334452e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14605840,
+      "step": 18695
+    },
+    {
+      "epoch": 38.87733887733888,
+      "grad_norm": 0.0011580680729821324,
+      "learning_rate": 2.7550064656817988e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14609776,
+      "step": 18700
+    },
+    {
+      "epoch": 38.88773388773389,
+      "grad_norm": 0.0015251798322424293,
+      "learning_rate": 2.7540298190049503e-05,
+      "loss": 0.0453,
+      "num_input_tokens_seen": 14613680,
+      "step": 18705
+    },
+    {
+      "epoch": 38.898128898128896,
+      "grad_norm": 0.005875087343156338,
+      "learning_rate": 2.7530531331535107e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14617584,
+      "step": 18710
+    },
+    {
+      "epoch": 38.90852390852391,
+      "grad_norm": 0.10216621309518814,
+      "learning_rate": 2.752076408278099e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 14621488,
+      "step": 18715
+    },
+    {
+      "epoch": 38.91891891891892,
+      "grad_norm": 0.031667113304138184,
+      "learning_rate": 2.751099644529337e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14625520,
+      "step": 18720
+    },
+    {
+      "epoch": 38.92931392931393,
+      "grad_norm": 0.11888834089040756,
+      "learning_rate": 2.7501228420578533e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 14629520,
+      "step": 18725
+    },
+    {
+      "epoch": 38.93970893970894,
+      "grad_norm": 0.003107032971456647,
+      "learning_rate": 2.7491460010142857e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14633360,
+      "step": 18730
+    },
+    {
+      "epoch": 38.95010395010395,
+      "grad_norm": 0.0397786945104599,
+      "learning_rate": 2.7481691215492727e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 14637264,
+      "step": 18735
+    },
+    {
+      "epoch": 38.96049896049896,
+      "grad_norm": 0.004868383053690195,
+      "learning_rate": 2.747192203813463e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 14641136,
+      "step": 18740
+    },
+    {
+      "epoch": 38.97089397089397,
+      "grad_norm": 0.001592302811332047,
+      "learning_rate": 2.7462152479575087e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14645008,
+      "step": 18745
+    },
+    {
+      "epoch": 38.981288981288984,
+      "grad_norm": 0.31468263268470764,
+      "learning_rate": 2.7452382541320697e-05,
+      "loss": 0.0067,
+      "num_input_tokens_seen": 14648752,
+      "step": 18750
+    },
+    {
+      "epoch": 38.99168399168399,
+      "grad_norm": 0.0019476424204185605,
+      "learning_rate": 2.7442612224878096e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 14652656,
+      "step": 18755
+    },
+    {
+      "epoch": 39.002079002079,
+      "grad_norm": 0.0006600210908800364,
+      "learning_rate": 2.7432841531753994e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14656640,
+      "step": 18760
+    },
+    {
+      "epoch": 39.012474012474016,
+      "grad_norm": 0.0011323858052492142,
+      "learning_rate": 2.7423070463455147e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14660544,
+      "step": 18765
+    },
+    {
+      "epoch": 39.022869022869024,
+      "grad_norm": 1.3399173021316528,
+      "learning_rate": 2.7413299021488397e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 14664480,
+      "step": 18770
+    },
+    {
+      "epoch": 39.03326403326403,
+      "grad_norm": 24.42351722717285,
+      "learning_rate": 2.7403527207360615e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 14668256,
+      "step": 18775
+    },
+    {
+      "epoch": 39.04365904365904,
+      "grad_norm": 0.001558577991090715,
+      "learning_rate": 2.7393755022578722e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14672192,
+      "step": 18780
+    },
+    {
+      "epoch": 39.054054054054056,
+      "grad_norm": 1.2755738496780396,
+      "learning_rate": 2.7383982468649714e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 14676160,
+      "step": 18785
+    },
+    {
+      "epoch": 39.064449064449065,
+      "grad_norm": 0.0003842535661533475,
+      "learning_rate": 2.7374209547080665e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14680000,
+      "step": 18790
+    },
+    {
+      "epoch": 39.07484407484407,
+      "grad_norm": 0.15172527730464935,
+      "learning_rate": 2.7364436259378663e-05,
+      "loss": 0.0059,
+      "num_input_tokens_seen": 14683904,
+      "step": 18795
+    },
+    {
+      "epoch": 39.08523908523909,
+      "grad_norm": 6.700149970129132e-05,
+      "learning_rate": 2.735466260705088e-05,
+      "loss": 0.0365,
+      "num_input_tokens_seen": 14687808,
+      "step": 18800
+    },
+    {
+      "epoch": 39.08523908523909,
+      "eval_loss": 0.3963184654712677,
+      "eval_runtime": 11.6914,
+      "eval_samples_per_second": 73.216,
+      "eval_steps_per_second": 18.304,
+      "num_input_tokens_seen": 14687808,
+      "step": 18800
+    },
+    {
+      "epoch": 39.0956340956341,
+      "grad_norm": 0.00016155581397470087,
+      "learning_rate": 2.7344888591604524e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14691520,
+      "step": 18805
+    },
+    {
+      "epoch": 39.106029106029105,
+      "grad_norm": 0.0005563784507103264,
+      "learning_rate": 2.7335114214546893e-05,
+      "loss": 0.0347,
+      "num_input_tokens_seen": 14695232,
+      "step": 18810
+    },
+    {
+      "epoch": 39.11642411642411,
+      "grad_norm": 0.009834839031100273,
+      "learning_rate": 2.7325339477385293e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14699168,
+      "step": 18815
+    },
+    {
+      "epoch": 39.12681912681913,
+      "grad_norm": 0.20932476222515106,
+      "learning_rate": 2.7315564381627128e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 14703072,
+      "step": 18820
+    },
+    {
+      "epoch": 39.13721413721414,
+      "grad_norm": 0.018707746639847755,
+      "learning_rate": 2.7305788928779835e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14706816,
+      "step": 18825
+    },
+    {
+      "epoch": 39.147609147609145,
+      "grad_norm": 0.010379645973443985,
+      "learning_rate": 2.729601312035091e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 14710752,
+      "step": 18830
+    },
+    {
+      "epoch": 39.15800415800416,
+      "grad_norm": 0.0021621703635901213,
+      "learning_rate": 2.7286236957847915e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14714592,
+      "step": 18835
+    },
+    {
+      "epoch": 39.16839916839917,
+      "grad_norm": 0.009802047163248062,
+      "learning_rate": 2.7276460442778446e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14718432,
+      "step": 18840
+    },
+    {
+      "epoch": 39.17879417879418,
+      "grad_norm": 0.003065102733671665,
+      "learning_rate": 2.726668357665017e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14722336,
+      "step": 18845
+    },
+    {
+      "epoch": 39.189189189189186,
+      "grad_norm": 0.026503097265958786,
+      "learning_rate": 2.7256906360970808e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14726304,
+      "step": 18850
+    },
+    {
+      "epoch": 39.1995841995842,
+      "grad_norm": 0.000290060619590804,
+      "learning_rate": 2.7247128797248117e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14730208,
+      "step": 18855
+    },
+    {
+      "epoch": 39.20997920997921,
+      "grad_norm": 0.14477013051509857,
+      "learning_rate": 2.7237350886989925e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 14734080,
+      "step": 18860
+    },
+    {
+      "epoch": 39.22037422037422,
+      "grad_norm": 0.002090044319629669,
+      "learning_rate": 2.7227572631704107e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14738048,
+      "step": 18865
+    },
+    {
+      "epoch": 39.23076923076923,
+      "grad_norm": 0.0007314856629818678,
+      "learning_rate": 2.7217794032898596e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14741984,
+      "step": 18870
+    },
+    {
+      "epoch": 39.24116424116424,
+      "grad_norm": 0.0012602807255461812,
+      "learning_rate": 2.7208015092081384e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14745856,
+      "step": 18875
+    },
+    {
+      "epoch": 39.25155925155925,
+      "grad_norm": 0.0033101008739322424,
+      "learning_rate": 2.719823581076049e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14749792,
+      "step": 18880
+    },
+    {
+      "epoch": 39.26195426195426,
+      "grad_norm": 0.0002595220285002142,
+      "learning_rate": 2.718845619044401e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14753696,
+      "step": 18885
+    },
+    {
+      "epoch": 39.272349272349274,
+      "grad_norm": 0.0007005718071013689,
+      "learning_rate": 2.7178676232640088e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14757504,
+      "step": 18890
+    },
+    {
+      "epoch": 39.28274428274428,
+      "grad_norm": 0.00045667015365324914,
+      "learning_rate": 2.716889593885691e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14761504,
+      "step": 18895
+    },
+    {
+      "epoch": 39.29313929313929,
+      "grad_norm": 0.000507210788782686,
+      "learning_rate": 2.7159115310602716e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14765408,
+      "step": 18900
+    },
+    {
+      "epoch": 39.303534303534306,
+      "grad_norm": 0.0007395817083306611,
+      "learning_rate": 2.7149334349385814e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14769280,
+      "step": 18905
+    },
+    {
+      "epoch": 39.313929313929314,
+      "grad_norm": 0.0006537990993820131,
+      "learning_rate": 2.713955305671454e-05,
+      "loss": 0.0161,
+      "num_input_tokens_seen": 14773088,
+      "step": 18910
+    },
+    {
+      "epoch": 39.32432432432432,
+      "grad_norm": 0.0038857911713421345,
+      "learning_rate": 2.71297714340973e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 14777024,
+      "step": 18915
+    },
+    {
+      "epoch": 39.33471933471934,
+      "grad_norm": 0.0011376622132956982,
+      "learning_rate": 2.7119989483042545e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 14780992,
+      "step": 18920
+    },
+    {
+      "epoch": 39.34511434511435,
+      "grad_norm": 0.004113975912332535,
+      "learning_rate": 2.7110207205058768e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 14784864,
+      "step": 18925
+    },
+    {
+      "epoch": 39.355509355509355,
+      "grad_norm": 0.14696930348873138,
+      "learning_rate": 2.7100424601654517e-05,
+      "loss": 0.0059,
+      "num_input_tokens_seen": 14788736,
+      "step": 18930
+    },
+    {
+      "epoch": 39.36590436590436,
+      "grad_norm": 0.00013321838923729956,
+      "learning_rate": 2.7090641674338403e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14792544,
+      "step": 18935
+    },
+    {
+      "epoch": 39.37629937629938,
+      "grad_norm": 7.278979319380596e-05,
+      "learning_rate": 2.7080858424619072e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14796416,
+      "step": 18940
+    },
+    {
+      "epoch": 39.38669438669439,
+      "grad_norm": 0.01070529967546463,
+      "learning_rate": 2.707107485400521e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14800448,
+      "step": 18945
+    },
+    {
+      "epoch": 39.397089397089395,
+      "grad_norm": 0.19171765446662903,
+      "learning_rate": 2.7061290964005586e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 14804288,
+      "step": 18950
+    },
+    {
+      "epoch": 39.40748440748441,
+      "grad_norm": 0.002140770899131894,
+      "learning_rate": 2.7051506756129e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 14808000,
+      "step": 18955
+    },
+    {
+      "epoch": 39.41787941787942,
+      "grad_norm": 0.007777233142405748,
+      "learning_rate": 2.704172223188428e-05,
+      "loss": 0.0069,
+      "num_input_tokens_seen": 14811840,
+      "step": 18960
+    },
+    {
+      "epoch": 39.42827442827443,
+      "grad_norm": 0.0035748507361859083,
+      "learning_rate": 2.7031937392780334e-05,
+      "loss": 0.019,
+      "num_input_tokens_seen": 14815712,
+      "step": 18965
+    },
+    {
+      "epoch": 39.438669438669436,
+      "grad_norm": 0.11386627703905106,
+      "learning_rate": 2.702215224032611e-05,
+      "loss": 0.0141,
+      "num_input_tokens_seen": 14819648,
+      "step": 18970
+    },
+    {
+      "epoch": 39.44906444906445,
+      "grad_norm": 0.28505975008010864,
+      "learning_rate": 2.70123667760306e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 14823552,
+      "step": 18975
+    },
+    {
+      "epoch": 39.45945945945946,
+      "grad_norm": 0.0005353609449230134,
+      "learning_rate": 2.7002581001402845e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 14827456,
+      "step": 18980
+    },
+    {
+      "epoch": 39.46985446985447,
+      "grad_norm": 0.0010412537958472967,
+      "learning_rate": 2.6992794917951923e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14831328,
+      "step": 18985
+    },
+    {
+      "epoch": 39.48024948024948,
+      "grad_norm": 0.0012916376581415534,
+      "learning_rate": 2.6983008527187e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14835424,
+      "step": 18990
+    },
+    {
+      "epoch": 39.49064449064449,
+      "grad_norm": 11.576411247253418,
+      "learning_rate": 2.697322183061723e-05,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 14839488,
+      "step": 18995
+    },
+    {
+      "epoch": 39.5010395010395,
+      "grad_norm": 0.01122965943068266,
+      "learning_rate": 2.696343482975186e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14843360,
+      "step": 19000
+    },
+    {
+      "epoch": 39.5010395010395,
+      "eval_loss": 0.39717480540275574,
+      "eval_runtime": 11.7028,
+      "eval_samples_per_second": 73.145,
+      "eval_steps_per_second": 18.286,
+      "num_input_tokens_seen": 14843360,
+      "step": 19000
+    },
+    {
+      "epoch": 39.51143451143451,
+      "grad_norm": 0.000526862742844969,
+      "learning_rate": 2.695364752610016e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 14847296,
+      "step": 19005
+    },
+    {
+      "epoch": 39.521829521829524,
+      "grad_norm": 0.00110554916318506,
+      "learning_rate": 2.6943859921171467e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 14851200,
+      "step": 19010
+    },
+    {
+      "epoch": 39.53222453222453,
+      "grad_norm": 0.012099450454115868,
+      "learning_rate": 2.6934072016475143e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14855168,
+      "step": 19015
+    },
+    {
+      "epoch": 39.54261954261954,
+      "grad_norm": 0.002799599664285779,
+      "learning_rate": 2.6924283813520606e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14859136,
+      "step": 19020
+    },
+    {
+      "epoch": 39.553014553014556,
+      "grad_norm": 0.008630353957414627,
+      "learning_rate": 2.691449531381733e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14863040,
+      "step": 19025
+    },
+    {
+      "epoch": 39.563409563409564,
+      "grad_norm": 0.00034761425922624767,
+      "learning_rate": 2.6904706518874816e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14867200,
+      "step": 19030
+    },
+    {
+      "epoch": 39.57380457380457,
+      "grad_norm": 0.0005964431911706924,
+      "learning_rate": 2.6894917430202615e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14870912,
+      "step": 19035
+    },
+    {
+      "epoch": 39.58419958419958,
+      "grad_norm": 0.0002286272356286645,
+      "learning_rate": 2.6885128049310343e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14874880,
+      "step": 19040
+    },
+    {
+      "epoch": 39.5945945945946,
+      "grad_norm": 0.009897160343825817,
+      "learning_rate": 2.687533837770762e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14878656,
+      "step": 19045
+    },
+    {
+      "epoch": 39.604989604989605,
+      "grad_norm": 0.0008818591013550758,
+      "learning_rate": 2.6865548416904162e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14882336,
+      "step": 19050
+    },
+    {
+      "epoch": 39.61538461538461,
+      "grad_norm": 0.002358754863962531,
+      "learning_rate": 2.68557581684097e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 14886336,
+      "step": 19055
+    },
+    {
+      "epoch": 39.62577962577963,
+      "grad_norm": 0.0004522472445387393,
+      "learning_rate": 2.6845967633733998e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14890144,
+      "step": 19060
+    },
+    {
+      "epoch": 39.63617463617464,
+      "grad_norm": 0.00046664345427416265,
+      "learning_rate": 2.683617681438689e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14894144,
+      "step": 19065
+    },
+    {
+      "epoch": 39.646569646569645,
+      "grad_norm": 0.000574543490074575,
+      "learning_rate": 2.682638571187825e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14898016,
+      "step": 19070
+    },
+    {
+      "epoch": 39.656964656964654,
+      "grad_norm": 0.00019474154396448284,
+      "learning_rate": 2.6816594327717976e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 14901920,
+      "step": 19075
+    },
+    {
+      "epoch": 39.66735966735967,
+      "grad_norm": 4.3610463762888685e-05,
+      "learning_rate": 2.680680266341603e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14905696,
+      "step": 19080
+    },
+    {
+      "epoch": 39.67775467775468,
+      "grad_norm": 5.9452486311784014e-05,
+      "learning_rate": 2.67970107204824e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14909632,
+      "step": 19085
+    },
+    {
+      "epoch": 39.688149688149686,
+      "grad_norm": 0.0002516386448405683,
+      "learning_rate": 2.6787218500427142e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14913696,
+      "step": 19090
+    },
+    {
+      "epoch": 39.6985446985447,
+      "grad_norm": 0.21667858958244324,
+      "learning_rate": 2.6777426004760332e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 14917536,
+      "step": 19095
+    },
+    {
+      "epoch": 39.70893970893971,
+      "grad_norm": 0.00032056710915639997,
+      "learning_rate": 2.6767633234992094e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14921472,
+      "step": 19100
+    },
+    {
+      "epoch": 39.71933471933472,
+      "grad_norm": 0.0007148553850129247,
+      "learning_rate": 2.6757840192632598e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 14925280,
+      "step": 19105
+    },
+    {
+      "epoch": 39.729729729729726,
+      "grad_norm": 0.002199756447225809,
+      "learning_rate": 2.6748046879192052e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14929216,
+      "step": 19110
+    },
+    {
+      "epoch": 39.74012474012474,
+      "grad_norm": 0.00014634434774052352,
+      "learning_rate": 2.673825329618071e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14933088,
+      "step": 19115
+    },
+    {
+      "epoch": 39.75051975051975,
+      "grad_norm": 0.00014509468746837229,
+      "learning_rate": 2.6728459445108866e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14937088,
+      "step": 19120
+    },
+    {
+      "epoch": 39.76091476091476,
+      "grad_norm": 6.148157262941822e-05,
+      "learning_rate": 2.6718665327486854e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14940896,
+      "step": 19125
+    },
+    {
+      "epoch": 39.771309771309774,
+      "grad_norm": 4.3313975766068324e-05,
+      "learning_rate": 2.6708870944825048e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 14944736,
+      "step": 19130
+    },
+    {
+      "epoch": 39.78170478170478,
+      "grad_norm": 0.00034632484312169254,
+      "learning_rate": 2.6699076298633874e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14948576,
+      "step": 19135
+    },
+    {
+      "epoch": 39.79209979209979,
+      "grad_norm": 0.00010767814092105255,
+      "learning_rate": 2.6689281390423788e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14952384,
+      "step": 19140
+    },
+    {
+      "epoch": 39.802494802494806,
+      "grad_norm": 0.0014267939841374755,
+      "learning_rate": 2.667948622170527e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14956448,
+      "step": 19145
+    },
+    {
+      "epoch": 39.812889812889814,
+      "grad_norm": 0.0004067688714712858,
+      "learning_rate": 2.6669690793988873e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 14960384,
+      "step": 19150
+    },
+    {
+      "epoch": 39.82328482328482,
+      "grad_norm": 0.002224275376647711,
+      "learning_rate": 2.665989510878518e-05,
+      "loss": 0.0156,
+      "num_input_tokens_seen": 14964288,
+      "step": 19155
+    },
+    {
+      "epoch": 39.83367983367983,
+      "grad_norm": 0.0013988894643262029,
+      "learning_rate": 2.6650099167604793e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14968256,
+      "step": 19160
+    },
+    {
+      "epoch": 39.84407484407485,
+      "grad_norm": 0.0002721311757341027,
+      "learning_rate": 2.6640302971958376e-05,
+      "loss": 0.0389,
+      "num_input_tokens_seen": 14972192,
+      "step": 19165
+    },
+    {
+      "epoch": 39.854469854469855,
+      "grad_norm": 0.0014739721082150936,
+      "learning_rate": 2.6630506523356635e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 14976256,
+      "step": 19170
+    },
+    {
+      "epoch": 39.86486486486486,
+      "grad_norm": 0.032408457249403,
+      "learning_rate": 2.6620709823310297e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 14980160,
+      "step": 19175
+    },
+    {
+      "epoch": 39.87525987525988,
+      "grad_norm": 0.0051488932222127914,
+      "learning_rate": 2.661091287333014e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14984064,
+      "step": 19180
+    },
+    {
+      "epoch": 39.88565488565489,
+      "grad_norm": 0.013889730907976627,
+      "learning_rate": 2.660111567492696e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14988000,
+      "step": 19185
+    },
+    {
+      "epoch": 39.896049896049895,
+      "grad_norm": 0.10628770291805267,
+      "learning_rate": 2.6591318229611635e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14991872,
+      "step": 19190
+    },
+    {
+      "epoch": 39.906444906444904,
+      "grad_norm": 0.0041453903540968895,
+      "learning_rate": 2.6581520538895037e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14995776,
+      "step": 19195
+    },
+    {
+      "epoch": 39.91683991683992,
+      "grad_norm": 0.0023924452252686024,
+      "learning_rate": 2.6571722604288102e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 14999808,
+      "step": 19200
+    },
+    {
+      "epoch": 39.91683991683992,
+      "eval_loss": 0.30934712290763855,
+      "eval_runtime": 11.7185,
+      "eval_samples_per_second": 73.047,
+      "eval_steps_per_second": 18.262,
+      "num_input_tokens_seen": 14999808,
+      "step": 19200
+    },
+    {
+      "epoch": 39.92723492723493,
+      "grad_norm": 0.0065239910036325455,
+      "learning_rate": 2.656192442730179e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15003712,
+      "step": 19205
+    },
+    {
+      "epoch": 39.937629937629936,
+      "grad_norm": 0.0021896250545978546,
+      "learning_rate": 2.6552126009447098e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 15007584,
+      "step": 19210
+    },
+    {
+      "epoch": 39.94802494802495,
+      "grad_norm": 0.009074578061699867,
+      "learning_rate": 2.654232735223507e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15011488,
+      "step": 19215
+    },
+    {
+      "epoch": 39.95841995841996,
+      "grad_norm": 0.003733016550540924,
+      "learning_rate": 2.6532528457176787e-05,
+      "loss": 0.0273,
+      "num_input_tokens_seen": 15015360,
+      "step": 19220
+    },
+    {
+      "epoch": 39.96881496881497,
+      "grad_norm": 0.005751362536102533,
+      "learning_rate": 2.6522729325783348e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15019136,
+      "step": 19225
+    },
+    {
+      "epoch": 39.979209979209976,
+      "grad_norm": 0.005490249488502741,
+      "learning_rate": 2.6512929959565914e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 15023200,
+      "step": 19230
+    },
+    {
+      "epoch": 39.98960498960499,
+      "grad_norm": 0.006430812180042267,
+      "learning_rate": 2.6503130360035673e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 15027104,
+      "step": 19235
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 0.0028944399673491716,
+      "learning_rate": 2.6493330528703835e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15031048,
+      "step": 19240
+    },
+    {
+      "epoch": 40.01039501039501,
+      "grad_norm": 0.0015539639862254262,
+      "learning_rate": 2.648353046708167e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15034984,
+      "step": 19245
+    },
+    {
+      "epoch": 40.020790020790024,
+      "grad_norm": 0.0007876884192228317,
+      "learning_rate": 2.647373017668046e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15038888,
+      "step": 19250
+    },
+    {
+      "epoch": 40.03118503118503,
+      "grad_norm": 0.1638832986354828,
+      "learning_rate": 2.6463929659011537e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 15042696,
+      "step": 19255
+    },
+    {
+      "epoch": 40.04158004158004,
+      "grad_norm": 0.0007958455244079232,
+      "learning_rate": 2.6454128915586262e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15046568,
+      "step": 19260
+    },
+    {
+      "epoch": 40.05197505197505,
+      "grad_norm": 0.0022073849104344845,
+      "learning_rate": 2.6444327947916036e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15050248,
+      "step": 19265
+    },
+    {
+      "epoch": 40.062370062370064,
+      "grad_norm": 0.0007983542745932937,
+      "learning_rate": 2.6434526757512292e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15054088,
+      "step": 19270
+    },
+    {
+      "epoch": 40.07276507276507,
+      "grad_norm": 0.00232068938203156,
+      "learning_rate": 2.6424725345886486e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15057928,
+      "step": 19275
+    },
+    {
+      "epoch": 40.08316008316008,
+      "grad_norm": 0.0004586240975186229,
+      "learning_rate": 2.641492371455014e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15061960,
+      "step": 19280
+    },
+    {
+      "epoch": 40.093555093555096,
+      "grad_norm": 0.0019555960316210985,
+      "learning_rate": 2.640512186501477e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 15065896,
+      "step": 19285
+    },
+    {
+      "epoch": 40.103950103950105,
+      "grad_norm": 0.00037496027653105557,
+      "learning_rate": 2.639531979879195e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15069672,
+      "step": 19290
+    },
+    {
+      "epoch": 40.11434511434511,
+      "grad_norm": 0.0004079672507941723,
+      "learning_rate": 2.638551751739328e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 15073480,
+      "step": 19295
+    },
+    {
+      "epoch": 40.12474012474012,
+      "grad_norm": 0.0011991914361715317,
+      "learning_rate": 2.6375715022330404e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15077512,
+      "step": 19300
+    },
+    {
+      "epoch": 40.13513513513514,
+      "grad_norm": 0.0011441146489232779,
+      "learning_rate": 2.6365912315114976e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 15081256,
+      "step": 19305
+    },
+    {
+      "epoch": 40.145530145530145,
+      "grad_norm": 0.001008369610644877,
+      "learning_rate": 2.6356109397258704e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15085192,
+      "step": 19310
+    },
+    {
+      "epoch": 40.15592515592515,
+      "grad_norm": 0.14944767951965332,
+      "learning_rate": 2.6346306270273325e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15089064,
+      "step": 19315
+    },
+    {
+      "epoch": 40.16632016632017,
+      "grad_norm": 0.14617642760276794,
+      "learning_rate": 2.6336502935670608e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 15093000,
+      "step": 19320
+    },
+    {
+      "epoch": 40.17671517671518,
+      "grad_norm": 0.18110144138336182,
+      "learning_rate": 2.6326699394962333e-05,
+      "loss": 0.0059,
+      "num_input_tokens_seen": 15096936,
+      "step": 19325
+    },
+    {
+      "epoch": 40.187110187110186,
+      "grad_norm": 0.00046247828868217766,
+      "learning_rate": 2.6316895649660334e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15100776,
+      "step": 19330
+    },
+    {
+      "epoch": 40.197505197505194,
+      "grad_norm": 0.0004720792348962277,
+      "learning_rate": 2.6307091701276486e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15104776,
+      "step": 19335
+    },
+    {
+      "epoch": 40.20790020790021,
+      "grad_norm": 0.005728157237172127,
+      "learning_rate": 2.629728755132267e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15108552,
+      "step": 19340
+    },
+    {
+      "epoch": 40.21829521829522,
+      "grad_norm": 0.19630686938762665,
+      "learning_rate": 2.628748320131081e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 15112488,
+      "step": 19345
+    },
+    {
+      "epoch": 40.228690228690226,
+      "grad_norm": 0.0006146032828837633,
+      "learning_rate": 2.6277678652752856e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 15116584,
+      "step": 19350
+    },
+    {
+      "epoch": 40.23908523908524,
+      "grad_norm": 0.09655628353357315,
+      "learning_rate": 2.6267873907160807e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 15120296,
+      "step": 19355
+    },
+    {
+      "epoch": 40.24948024948025,
+      "grad_norm": 0.0023384017404168844,
+      "learning_rate": 2.6258068966046668e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15124296,
+      "step": 19360
+    },
+    {
+      "epoch": 40.25987525987526,
+      "grad_norm": 0.0007649700855836272,
+      "learning_rate": 2.6248263830922475e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15128200,
+      "step": 19365
+    },
+    {
+      "epoch": 40.270270270270274,
+      "grad_norm": 0.0005072163767181337,
+      "learning_rate": 2.6238458503300318e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15132200,
+      "step": 19370
+    },
+    {
+      "epoch": 40.28066528066528,
+      "grad_norm": 1.5762439966201782,
+      "learning_rate": 2.6228652984692292e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 15136104,
+      "step": 19375
+    },
+    {
+      "epoch": 40.29106029106029,
+      "grad_norm": 0.00021309926523827016,
+      "learning_rate": 2.621884727661054e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15139848,
+      "step": 19380
+    },
+    {
+      "epoch": 40.3014553014553,
+      "grad_norm": 0.0009096472640521824,
+      "learning_rate": 2.6209041380567222e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 15143880,
+      "step": 19385
+    },
+    {
+      "epoch": 40.311850311850314,
+      "grad_norm": 0.00038681962178088725,
+      "learning_rate": 2.6199235298074527e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15147784,
+      "step": 19390
+    },
+    {
+      "epoch": 40.32224532224532,
+      "grad_norm": 0.000633061514236033,
+      "learning_rate": 2.618942903064468e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15151496,
+      "step": 19395
+    },
+    {
+      "epoch": 40.33264033264033,
+      "grad_norm": 0.0020938972011208534,
+      "learning_rate": 2.6179622579789932e-05,
+      "loss": 0.042,
+      "num_input_tokens_seen": 15155496,
+      "step": 19400
+    },
+    {
+      "epoch": 40.33264033264033,
+      "eval_loss": 0.33849257230758667,
+      "eval_runtime": 11.7184,
+      "eval_samples_per_second": 73.048,
+      "eval_steps_per_second": 18.262,
+      "num_input_tokens_seen": 15155496,
+      "step": 19400
+    },
+    {
+      "epoch": 40.343035343035346,
+      "grad_norm": 0.011577815748751163,
+      "learning_rate": 2.6169815947022553e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15159304,
+      "step": 19405
+    },
+    {
+      "epoch": 40.353430353430355,
+      "grad_norm": 0.004315183497965336,
+      "learning_rate": 2.6160009133854853e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15163400,
+      "step": 19410
+    },
+    {
+      "epoch": 40.36382536382536,
+      "grad_norm": 17.718791961669922,
+      "learning_rate": 2.6150202141799168e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 15167432,
+      "step": 19415
+    },
+    {
+      "epoch": 40.37422037422037,
+      "grad_norm": 0.02364109829068184,
+      "learning_rate": 2.614039497236786e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15171240,
+      "step": 19420
+    },
+    {
+      "epoch": 40.38461538461539,
+      "grad_norm": 0.00216110423207283,
+      "learning_rate": 2.6130587627073315e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 15175080,
+      "step": 19425
+    },
+    {
+      "epoch": 40.395010395010395,
+      "grad_norm": 0.0008927633753046393,
+      "learning_rate": 2.6120780107427956e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15179048,
+      "step": 19430
+    },
+    {
+      "epoch": 40.4054054054054,
+      "grad_norm": 0.00366004416719079,
+      "learning_rate": 2.6110972414944214e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 15182984,
+      "step": 19435
+    },
+    {
+      "epoch": 40.41580041580042,
+      "grad_norm": 0.0011777938343584538,
+      "learning_rate": 2.6101164551134565e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15186952,
+      "step": 19440
+    },
+    {
+      "epoch": 40.42619542619543,
+      "grad_norm": 0.001007233397103846,
+      "learning_rate": 2.6091356517511505e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15190792,
+      "step": 19445
+    },
+    {
+      "epoch": 40.436590436590436,
+      "grad_norm": 0.0009607000392861664,
+      "learning_rate": 2.608154831558755e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15194664,
+      "step": 19450
+    },
+    {
+      "epoch": 40.446985446985444,
+      "grad_norm": 2.200681447982788,
+      "learning_rate": 2.607173994687526e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 15198632,
+      "step": 19455
+    },
+    {
+      "epoch": 40.45738045738046,
+      "grad_norm": 0.0019472696585580707,
+      "learning_rate": 2.6061931412887196e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15202664,
+      "step": 19460
+    },
+    {
+      "epoch": 40.46777546777547,
+      "grad_norm": 0.0018819918623194098,
+      "learning_rate": 2.6052122715135973e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15206568,
+      "step": 19465
+    },
+    {
+      "epoch": 40.478170478170476,
+      "grad_norm": 0.0021605060901492834,
+      "learning_rate": 2.60423138551342e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15210408,
+      "step": 19470
+    },
+    {
+      "epoch": 40.48856548856549,
+      "grad_norm": 0.0006518399459309876,
+      "learning_rate": 2.6032504834394527e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 15214184,
+      "step": 19475
+    },
+    {
+      "epoch": 40.4989604989605,
+      "grad_norm": 0.0030579783488065004,
+      "learning_rate": 2.602269565442964e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15218056,
+      "step": 19480
+    },
+    {
+      "epoch": 40.50935550935551,
+      "grad_norm": 0.0005411073798313737,
+      "learning_rate": 2.6012886316752227e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15221928,
+      "step": 19485
+    },
+    {
+      "epoch": 40.51975051975052,
+      "grad_norm": 0.0014075929066166282,
+      "learning_rate": 2.6003076822875018e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15225864,
+      "step": 19490
+    },
+    {
+      "epoch": 40.53014553014553,
+      "grad_norm": 0.0004688511835411191,
+      "learning_rate": 2.5993267174310755e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15229768,
+      "step": 19495
+    },
+    {
+      "epoch": 40.54054054054054,
+      "grad_norm": 0.0004720921569969505,
+      "learning_rate": 2.5983457372572218e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15233640,
+      "step": 19500
+    },
+    {
+      "epoch": 40.55093555093555,
+      "grad_norm": 0.00045240408508107066,
+      "learning_rate": 2.597364741917219e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15237608,
+      "step": 19505
+    },
+    {
+      "epoch": 40.561330561330564,
+      "grad_norm": 0.0002942961291410029,
+      "learning_rate": 2.5963837315623492e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15241512,
+      "step": 19510
+    },
+    {
+      "epoch": 40.57172557172557,
+      "grad_norm": 0.0002122103760484606,
+      "learning_rate": 2.595402706343897e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 15245384,
+      "step": 19515
+    },
+    {
+      "epoch": 40.58212058212058,
+      "grad_norm": 0.0005762209766544402,
+      "learning_rate": 2.594421666413148e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15249480,
+      "step": 19520
+    },
+    {
+      "epoch": 40.59251559251559,
+      "grad_norm": 0.0027091794181615114,
+      "learning_rate": 2.5934406119213928e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15253480,
+      "step": 19525
+    },
+    {
+      "epoch": 40.602910602910605,
+      "grad_norm": 0.00293094664812088,
+      "learning_rate": 2.5924595430199193e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15257448,
+      "step": 19530
+    },
+    {
+      "epoch": 40.61330561330561,
+      "grad_norm": 0.04045235365629196,
+      "learning_rate": 2.5914784598600238e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15261288,
+      "step": 19535
+    },
+    {
+      "epoch": 40.62370062370062,
+      "grad_norm": 7.5608110427856445,
+      "learning_rate": 2.5904973625930002e-05,
+      "loss": 0.0163,
+      "num_input_tokens_seen": 15265096,
+      "step": 19540
+    },
+    {
+      "epoch": 40.63409563409564,
+      "grad_norm": 0.0027225161902606487,
+      "learning_rate": 2.5895162513701456e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15268872,
+      "step": 19545
+    },
+    {
+      "epoch": 40.644490644490645,
+      "grad_norm": 0.01021570060402155,
+      "learning_rate": 2.5885351263427593e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15272872,
+      "step": 19550
+    },
+    {
+      "epoch": 40.65488565488565,
+      "grad_norm": 0.15441863238811493,
+      "learning_rate": 2.5875539876621448e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 15276776,
+      "step": 19555
+    },
+    {
+      "epoch": 40.66528066528066,
+      "grad_norm": 0.0014281789772212505,
+      "learning_rate": 2.586572835479605e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15280648,
+      "step": 19560
+    },
+    {
+      "epoch": 40.67567567567568,
+      "grad_norm": 0.0007496661855839193,
+      "learning_rate": 2.585591669946446e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15284520,
+      "step": 19565
+    },
+    {
+      "epoch": 40.686070686070686,
+      "grad_norm": 0.00198157224804163,
+      "learning_rate": 2.5846104912139756e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15288296,
+      "step": 19570
+    },
+    {
+      "epoch": 40.696465696465694,
+      "grad_norm": 0.0012856846442446113,
+      "learning_rate": 2.583629299433505e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15292136,
+      "step": 19575
+    },
+    {
+      "epoch": 40.70686070686071,
+      "grad_norm": 0.001892540603876114,
+      "learning_rate": 2.582648094756345e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 15296040,
+      "step": 19580
+    },
+    {
+      "epoch": 40.71725571725572,
+      "grad_norm": 0.0015033468371257186,
+      "learning_rate": 2.5816668773338098e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15299880,
+      "step": 19585
+    },
+    {
+      "epoch": 40.727650727650726,
+      "grad_norm": 0.46302106976509094,
+      "learning_rate": 2.580685647317216e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15303912,
+      "step": 19590
+    },
+    {
+      "epoch": 40.73804573804574,
+      "grad_norm": 0.12079153954982758,
+      "learning_rate": 2.5797044048578818e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 15307912,
+      "step": 19595
+    },
+    {
+      "epoch": 40.74844074844075,
+      "grad_norm": 0.004087117966264486,
+      "learning_rate": 2.5787231501071262e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15311688,
+      "step": 19600
+    },
+    {
+      "epoch": 40.74844074844075,
+      "eval_loss": 0.36300429701805115,
+      "eval_runtime": 11.6954,
+      "eval_samples_per_second": 73.191,
+      "eval_steps_per_second": 18.298,
+      "num_input_tokens_seen": 15311688,
+      "step": 19600
+    },
+    {
+      "epoch": 40.75883575883576,
+      "grad_norm": 0.0032758917659521103,
+      "learning_rate": 2.577741883216272e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15315656,
+      "step": 19605
+    },
+    {
+      "epoch": 40.76923076923077,
+      "grad_norm": 0.0014241052558645606,
+      "learning_rate": 2.576760604336642e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15319752,
+      "step": 19610
+    },
+    {
+      "epoch": 40.77962577962578,
+      "grad_norm": 0.0013096719048917294,
+      "learning_rate": 2.575779313619563e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15323752,
+      "step": 19615
+    },
+    {
+      "epoch": 40.79002079002079,
+      "grad_norm": 0.0021408398170024157,
+      "learning_rate": 2.5747980112163605e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15327848,
+      "step": 19620
+    },
+    {
+      "epoch": 40.8004158004158,
+      "grad_norm": 0.0004841337213292718,
+      "learning_rate": 2.5738166972783656e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15331784,
+      "step": 19625
+    },
+    {
+      "epoch": 40.810810810810814,
+      "grad_norm": 0.0018808944150805473,
+      "learning_rate": 2.5728353719569075e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15335784,
+      "step": 19630
+    },
+    {
+      "epoch": 40.82120582120582,
+      "grad_norm": 0.0009331804467365146,
+      "learning_rate": 2.57185403540332e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 15339688,
+      "step": 19635
+    },
+    {
+      "epoch": 40.83160083160083,
+      "grad_norm": 0.0017681237077340484,
+      "learning_rate": 2.5708726877689375e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15343464,
+      "step": 19640
+    },
+    {
+      "epoch": 40.84199584199584,
+      "grad_norm": 0.0007460681372322142,
+      "learning_rate": 2.5698913292050964e-05,
+      "loss": 0.0069,
+      "num_input_tokens_seen": 15347368,
+      "step": 19645
+    },
+    {
+      "epoch": 40.852390852390855,
+      "grad_norm": 0.00020703785412479192,
+      "learning_rate": 2.568909959863133e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15351208,
+      "step": 19650
+    },
+    {
+      "epoch": 40.86278586278586,
+      "grad_norm": 0.00038249927456490695,
+      "learning_rate": 2.5679285798943887e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15355080,
+      "step": 19655
+    },
+    {
+      "epoch": 40.87318087318087,
+      "grad_norm": 0.0012328897137194872,
+      "learning_rate": 2.5669471894502035e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15359016,
+      "step": 19660
+    },
+    {
+      "epoch": 40.88357588357589,
+      "grad_norm": 0.0005671063554473221,
+      "learning_rate": 2.56596578868192e-05,
+      "loss": 0.0047,
+      "num_input_tokens_seen": 15362856,
+      "step": 19665
+    },
+    {
+      "epoch": 40.893970893970895,
+      "grad_norm": 0.010464992374181747,
+      "learning_rate": 2.564984377740883e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15366664,
+      "step": 19670
+    },
+    {
+      "epoch": 40.9043659043659,
+      "grad_norm": 0.002484408672899008,
+      "learning_rate": 2.564002956778438e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15370600,
+      "step": 19675
+    },
+    {
+      "epoch": 40.91476091476091,
+      "grad_norm": 0.0023382683284580708,
+      "learning_rate": 2.563021525945934e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15374408,
+      "step": 19680
+    },
+    {
+      "epoch": 40.92515592515593,
+      "grad_norm": 0.0002534711384214461,
+      "learning_rate": 2.562040085394718e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15378312,
+      "step": 19685
+    },
+    {
+      "epoch": 40.935550935550935,
+      "grad_norm": 0.00014709253446199,
+      "learning_rate": 2.56105863527614e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 15382248,
+      "step": 19690
+    },
+    {
+      "epoch": 40.945945945945944,
+      "grad_norm": 0.0007090037106536329,
+      "learning_rate": 2.5600771757415548e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15386280,
+      "step": 19695
+    },
+    {
+      "epoch": 40.95634095634096,
+      "grad_norm": 0.0005415303749032319,
+      "learning_rate": 2.5590957069423134e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15390312,
+      "step": 19700
+    },
+    {
+      "epoch": 40.96673596673597,
+      "grad_norm": 0.0005693057901225984,
+      "learning_rate": 2.5581142290297716e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15394184,
+      "step": 19705
+    },
+    {
+      "epoch": 40.977130977130976,
+      "grad_norm": 0.0015496687265112996,
+      "learning_rate": 2.557132742155285e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15398088,
+      "step": 19710
+    },
+    {
+      "epoch": 40.987525987525984,
+      "grad_norm": 0.0003332363849040121,
+      "learning_rate": 2.556151246470212e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15402024,
+      "step": 19715
+    },
+    {
+      "epoch": 40.997920997921,
+      "grad_norm": 0.0002476558438502252,
+      "learning_rate": 2.5551697421259114e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15406024,
+      "step": 19720
+    },
+    {
+      "epoch": 41.00831600831601,
+      "grad_norm": 0.000557642662897706,
+      "learning_rate": 2.554188229273743e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15409864,
+      "step": 19725
+    },
+    {
+      "epoch": 41.018711018711016,
+      "grad_norm": 9.279174264520407e-05,
+      "learning_rate": 2.5532067080650678e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15413576,
+      "step": 19730
+    },
+    {
+      "epoch": 41.02910602910603,
+      "grad_norm": 0.001921175979077816,
+      "learning_rate": 2.55222517865125e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15417576,
+      "step": 19735
+    },
+    {
+      "epoch": 41.03950103950104,
+      "grad_norm": 0.002705150516703725,
+      "learning_rate": 2.5512436411836538e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15421384,
+      "step": 19740
+    },
+    {
+      "epoch": 41.04989604989605,
+      "grad_norm": 6.337297963909805e-05,
+      "learning_rate": 2.5502620958136443e-05,
+      "loss": 0.0055,
+      "num_input_tokens_seen": 15425320,
+      "step": 19745
+    },
+    {
+      "epoch": 41.06029106029106,
+      "grad_norm": 0.1299017369747162,
+      "learning_rate": 2.5492805426925874e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 15429224,
+      "step": 19750
+    },
+    {
+      "epoch": 41.07068607068607,
+      "grad_norm": 0.00043784850277006626,
+      "learning_rate": 2.5482989819718523e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 15433000,
+      "step": 19755
+    },
+    {
+      "epoch": 41.08108108108108,
+      "grad_norm": 0.0012464715400710702,
+      "learning_rate": 2.5473174138028065e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15436840,
+      "step": 19760
+    },
+    {
+      "epoch": 41.09147609147609,
+      "grad_norm": 4.796228313352913e-05,
+      "learning_rate": 2.5463358383368212e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 15440680,
+      "step": 19765
+    },
+    {
+      "epoch": 41.101871101871104,
+      "grad_norm": 0.00020346646488178521,
+      "learning_rate": 2.545354255725267e-05,
+      "loss": 0.0195,
+      "num_input_tokens_seen": 15444520,
+      "step": 19770
+    },
+    {
+      "epoch": 41.11226611226611,
+      "grad_norm": 0.0006916631245985627,
+      "learning_rate": 2.5443726661195165e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15448520,
+      "step": 19775
+    },
+    {
+      "epoch": 41.12266112266112,
+      "grad_norm": 0.00048696622252464294,
+      "learning_rate": 2.543391069670944e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15452552,
+      "step": 19780
+    },
+    {
+      "epoch": 41.13305613305613,
+      "grad_norm": 0.003772723488509655,
+      "learning_rate": 2.5424094665309228e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15456456,
+      "step": 19785
+    },
+    {
+      "epoch": 41.143451143451145,
+      "grad_norm": 0.0015307108405977488,
+      "learning_rate": 2.5414278568508292e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15460392,
+      "step": 19790
+    },
+    {
+      "epoch": 41.15384615384615,
+      "grad_norm": 0.002845014911144972,
+      "learning_rate": 2.540446240782039e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 15464520,
+      "step": 19795
+    },
+    {
+      "epoch": 41.16424116424116,
+      "grad_norm": 0.000593795906752348,
+      "learning_rate": 2.5394646184759307e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15468264,
+      "step": 19800
+    },
+    {
+      "epoch": 41.16424116424116,
+      "eval_loss": 0.3730482757091522,
+      "eval_runtime": 11.6961,
+      "eval_samples_per_second": 73.187,
+      "eval_steps_per_second": 18.297,
+      "num_input_tokens_seen": 15468264,
+      "step": 19800
+    },
+    {
+      "epoch": 41.17463617463618,
+      "grad_norm": 0.0016106627881526947,
+      "learning_rate": 2.538482990083882e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15472264,
+      "step": 19805
+    },
+    {
+      "epoch": 41.185031185031185,
+      "grad_norm": 0.00034869005321525037,
+      "learning_rate": 2.5375013557572725e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 15476136,
+      "step": 19810
+    },
+    {
+      "epoch": 41.195426195426194,
+      "grad_norm": 0.0010966552654281259,
+      "learning_rate": 2.536519715647483e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15480040,
+      "step": 19815
+    },
+    {
+      "epoch": 41.20582120582121,
+      "grad_norm": 0.0009634891175664961,
+      "learning_rate": 2.535538069905894e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15483848,
+      "step": 19820
+    },
+    {
+      "epoch": 41.21621621621622,
+      "grad_norm": 0.2513203024864197,
+      "learning_rate": 2.534556418683888e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 15487784,
+      "step": 19825
+    },
+    {
+      "epoch": 41.226611226611226,
+      "grad_norm": 0.0006391438655555248,
+      "learning_rate": 2.5335747621328486e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 15491752,
+      "step": 19830
+    },
+    {
+      "epoch": 41.237006237006234,
+      "grad_norm": 0.00016560197400394827,
+      "learning_rate": 2.5325931004041586e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15495592,
+      "step": 19835
+    },
+    {
+      "epoch": 41.24740124740125,
+      "grad_norm": 5.9872087149415165e-05,
+      "learning_rate": 2.5316114336492032e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15499528,
+      "step": 19840
+    },
+    {
+      "epoch": 41.25779625779626,
+      "grad_norm": 0.001318295719102025,
+      "learning_rate": 2.530629762019367e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15503528,
+      "step": 19845
+    },
+    {
+      "epoch": 41.268191268191266,
+      "grad_norm": 0.001812561647966504,
+      "learning_rate": 2.5296480856660364e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15507336,
+      "step": 19850
+    },
+    {
+      "epoch": 41.27858627858628,
+      "grad_norm": 0.0009242283413186669,
+      "learning_rate": 2.528666404740599e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15511176,
+      "step": 19855
+    },
+    {
+      "epoch": 41.28898128898129,
+      "grad_norm": 0.0004879616608377546,
+      "learning_rate": 2.527684719394442e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15515080,
+      "step": 19860
+    },
+    {
+      "epoch": 41.2993762993763,
+      "grad_norm": 0.00014803015801589936,
+      "learning_rate": 2.526703029778953e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15518888,
+      "step": 19865
+    },
+    {
+      "epoch": 41.30977130977131,
+      "grad_norm": 0.0015923893079161644,
+      "learning_rate": 2.5257213360455208e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15522728,
+      "step": 19870
+    },
+    {
+      "epoch": 41.32016632016632,
+      "grad_norm": 0.0007058817427605391,
+      "learning_rate": 2.5247396383455353e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15526568,
+      "step": 19875
+    },
+    {
+      "epoch": 41.33056133056133,
+      "grad_norm": 0.0005633067921735346,
+      "learning_rate": 2.523757936830387e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15530472,
+      "step": 19880
+    },
+    {
+      "epoch": 41.34095634095634,
+      "grad_norm": 0.00019922560022678226,
+      "learning_rate": 2.5227762316514662e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15534408,
+      "step": 19885
+    },
+    {
+      "epoch": 41.351351351351354,
+      "grad_norm": 0.00010631595068844035,
+      "learning_rate": 2.5217945229601648e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15538312,
+      "step": 19890
+    },
+    {
+      "epoch": 41.36174636174636,
+      "grad_norm": 7.836112490622327e-05,
+      "learning_rate": 2.5208128109078738e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15542280,
+      "step": 19895
+    },
+    {
+      "epoch": 41.37214137214137,
+      "grad_norm": 0.00011420881492085755,
+      "learning_rate": 2.5198310956459853e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15546344,
+      "step": 19900
+    },
+    {
+      "epoch": 41.38253638253638,
+      "grad_norm": 0.0001495929463999346,
+      "learning_rate": 2.518849377325893e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15550248,
+      "step": 19905
+    },
+    {
+      "epoch": 41.392931392931395,
+      "grad_norm": 0.0003261232632212341,
+      "learning_rate": 2.51786765609899e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15553928,
+      "step": 19910
+    },
+    {
+      "epoch": 41.4033264033264,
+      "grad_norm": 0.00018256678595207632,
+      "learning_rate": 2.5168859321166694e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15557768,
+      "step": 19915
+    },
+    {
+      "epoch": 41.41372141372141,
+      "grad_norm": 7.248680776683614e-05,
+      "learning_rate": 2.515904205530326e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 15561640,
+      "step": 19920
+    },
+    {
+      "epoch": 41.42411642411643,
+      "grad_norm": 9.213522571371868e-05,
+      "learning_rate": 2.514922476491355e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15565576,
+      "step": 19925
+    },
+    {
+      "epoch": 41.434511434511435,
+      "grad_norm": 6.619276973651722e-05,
+      "learning_rate": 2.51394074515115e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 15569512,
+      "step": 19930
+    },
+    {
+      "epoch": 41.444906444906444,
+      "grad_norm": 8.191059896489605e-05,
+      "learning_rate": 2.5129590116611067e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15573416,
+      "step": 19935
+    },
+    {
+      "epoch": 41.45530145530145,
+      "grad_norm": 6.163645593915135e-05,
+      "learning_rate": 2.5119772761726212e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15577288,
+      "step": 19940
+    },
+    {
+      "epoch": 41.46569646569647,
+      "grad_norm": 0.00015970206004567444,
+      "learning_rate": 2.5109955388370893e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15581256,
+      "step": 19945
+    },
+    {
+      "epoch": 41.476091476091476,
+      "grad_norm": 0.0004217509995214641,
+      "learning_rate": 2.510013799805907e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15585192,
+      "step": 19950
+    },
+    {
+      "epoch": 41.486486486486484,
+      "grad_norm": 0.0003437822451815009,
+      "learning_rate": 2.5090320592304706e-05,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 15589064,
+      "step": 19955
+    },
+    {
+      "epoch": 41.4968814968815,
+      "grad_norm": 9.096925350604579e-05,
+      "learning_rate": 2.5080503172621777e-05,
+      "loss": 0.0375,
+      "num_input_tokens_seen": 15592872,
+      "step": 19960
+    },
+    {
+      "epoch": 41.50727650727651,
+      "grad_norm": 0.0010779905132949352,
+      "learning_rate": 2.5070685740524246e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15596840,
+      "step": 19965
+    },
+    {
+      "epoch": 41.517671517671516,
+      "grad_norm": 0.001991451717913151,
+      "learning_rate": 2.5060868297526084e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15600680,
+      "step": 19970
+    },
+    {
+      "epoch": 41.528066528066525,
+      "grad_norm": 0.00968247465789318,
+      "learning_rate": 2.5051050845141267e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 15604520,
+      "step": 19975
+    },
+    {
+      "epoch": 41.53846153846154,
+      "grad_norm": 0.007882979698479176,
+      "learning_rate": 2.5041233384883765e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 15608456,
+      "step": 19980
+    },
+    {
+      "epoch": 41.54885654885655,
+      "grad_norm": 0.04582761973142624,
+      "learning_rate": 2.5031415918267564e-05,
+      "loss": 0.0207,
+      "num_input_tokens_seen": 15612296,
+      "step": 19985
+    },
+    {
+      "epoch": 41.55925155925156,
+      "grad_norm": 0.0028019356541335583,
+      "learning_rate": 2.5021598446806626e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 15616136,
+      "step": 19990
+    },
+    {
+      "epoch": 41.56964656964657,
+      "grad_norm": 0.00294914236292243,
+      "learning_rate": 2.5011780972014937e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 15620200,
+      "step": 19995
+    },
+    {
+      "epoch": 41.58004158004158,
+      "grad_norm": 0.26056554913520813,
+      "learning_rate": 2.5001963495406478e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 15624072,
+      "step": 20000
+    },
+    {
+      "epoch": 41.58004158004158,
+      "eval_loss": 0.35231125354766846,
+      "eval_runtime": 11.6977,
+      "eval_samples_per_second": 73.177,
+      "eval_steps_per_second": 18.294,
+      "num_input_tokens_seen": 15624072,
+      "step": 20000
+    },
+    {
+      "epoch": 41.59043659043659,
+      "grad_norm": 0.0006975905271247029,
+      "learning_rate": 2.499214601849522e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15627912,
+      "step": 20005
+    },
+    {
+      "epoch": 41.6008316008316,
+      "grad_norm": 0.0007423459319397807,
+      "learning_rate": 2.4982328542795148e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15631944,
+      "step": 20010
+    },
+    {
+      "epoch": 41.61122661122661,
+      "grad_norm": 0.003919398412108421,
+      "learning_rate": 2.497251106982024e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 15635912,
+      "step": 20015
+    },
+    {
+      "epoch": 41.62162162162162,
+      "grad_norm": 0.0004791999817825854,
+      "learning_rate": 2.4962693601084458e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15639784,
+      "step": 20020
+    },
+    {
+      "epoch": 41.63201663201663,
+      "grad_norm": 0.0013055065646767616,
+      "learning_rate": 2.4952876138101794e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15643592,
+      "step": 20025
+    },
+    {
+      "epoch": 41.642411642411645,
+      "grad_norm": 0.0011355100432410836,
+      "learning_rate": 2.4943058682386233e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15647496,
+      "step": 20030
+    },
+    {
+      "epoch": 41.65280665280665,
+      "grad_norm": 0.0002559924032539129,
+      "learning_rate": 2.493324123545173e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 15651528,
+      "step": 20035
+    },
+    {
+      "epoch": 41.66320166320166,
+      "grad_norm": 0.0032568711321800947,
+      "learning_rate": 2.4923423798812272e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15655400,
+      "step": 20040
+    },
+    {
+      "epoch": 41.67359667359668,
+      "grad_norm": 0.0007413168787024915,
+      "learning_rate": 2.4913606373981825e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15659080,
+      "step": 20045
+    },
+    {
+      "epoch": 41.683991683991685,
+      "grad_norm": 0.0006223480449989438,
+      "learning_rate": 2.4903788962474357e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 15662888,
+      "step": 20050
+    },
+    {
+      "epoch": 41.694386694386694,
+      "grad_norm": 0.000514441286213696,
+      "learning_rate": 2.489397156580385e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 15666792,
+      "step": 20055
+    },
+    {
+      "epoch": 41.7047817047817,
+      "grad_norm": 0.0010525810066610575,
+      "learning_rate": 2.4884154185484246e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15670856,
+      "step": 20060
+    },
+    {
+      "epoch": 41.71517671517672,
+      "grad_norm": 0.0015639675548300147,
+      "learning_rate": 2.4874336823029526e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15674792,
+      "step": 20065
+    },
+    {
+      "epoch": 41.725571725571726,
+      "grad_norm": 0.00020934996427968144,
+      "learning_rate": 2.4864519479953656e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 15678504,
+      "step": 20070
+    },
+    {
+      "epoch": 41.735966735966734,
+      "grad_norm": 0.0021203490905463696,
+      "learning_rate": 2.485470215777058e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15682632,
+      "step": 20075
+    },
+    {
+      "epoch": 41.74636174636175,
+      "grad_norm": 0.0010922455694526434,
+      "learning_rate": 2.4844884857994258e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15686568,
+      "step": 20080
+    },
+    {
+      "epoch": 41.75675675675676,
+      "grad_norm": 0.0010074219899252057,
+      "learning_rate": 2.4835067582138638e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15690408,
+      "step": 20085
+    },
+    {
+      "epoch": 41.767151767151766,
+      "grad_norm": 0.0005437919171527028,
+      "learning_rate": 2.4825250331717666e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15694376,
+      "step": 20090
+    },
+    {
+      "epoch": 41.777546777546775,
+      "grad_norm": 0.0002920369151979685,
+      "learning_rate": 2.4815433108245298e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15698376,
+      "step": 20095
+    },
+    {
+      "epoch": 41.78794178794179,
+      "grad_norm": 0.004157998599112034,
+      "learning_rate": 2.4805615913235456e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15702248,
+      "step": 20100
+    },
+    {
+      "epoch": 41.7983367983368,
+      "grad_norm": 0.00109457119833678,
+      "learning_rate": 2.479579874820208e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15706120,
+      "step": 20105
+    },
+    {
+      "epoch": 41.80873180873181,
+      "grad_norm": 0.0005420264787971973,
+      "learning_rate": 2.4785981614659115e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 15710024,
+      "step": 20110
+    },
+    {
+      "epoch": 41.81912681912682,
+      "grad_norm": 6.551392289111391e-05,
+      "learning_rate": 2.477616451412047e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15713960,
+      "step": 20115
+    },
+    {
+      "epoch": 41.82952182952183,
+      "grad_norm": 0.0007670401246286929,
+      "learning_rate": 2.476634744810007e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15717896,
+      "step": 20120
+    },
+    {
+      "epoch": 41.83991683991684,
+      "grad_norm": 0.002445520367473364,
+      "learning_rate": 2.475653041811183e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15721768,
+      "step": 20125
+    },
+    {
+      "epoch": 41.85031185031185,
+      "grad_norm": 0.0019770702347159386,
+      "learning_rate": 2.4746713425669652e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15725576,
+      "step": 20130
+    },
+    {
+      "epoch": 41.86070686070686,
+      "grad_norm": 0.000504054594784975,
+      "learning_rate": 2.4736896472287458e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15729672,
+      "step": 20135
+    },
+    {
+      "epoch": 41.87110187110187,
+      "grad_norm": 0.00011463448754511774,
+      "learning_rate": 2.4727079559479124e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15733576,
+      "step": 20140
+    },
+    {
+      "epoch": 41.88149688149688,
+      "grad_norm": 0.001044674776494503,
+      "learning_rate": 2.4717262688758557e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 15737448,
+      "step": 20145
+    },
+    {
+      "epoch": 41.891891891891895,
+      "grad_norm": 0.0014425700064748526,
+      "learning_rate": 2.4707445861639637e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15741416,
+      "step": 20150
+    },
+    {
+      "epoch": 41.9022869022869,
+      "grad_norm": 0.001637310371734202,
+      "learning_rate": 2.4697629079636244e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15745320,
+      "step": 20155
+    },
+    {
+      "epoch": 41.91268191268191,
+      "grad_norm": 0.00048373351455666125,
+      "learning_rate": 2.4687812344262244e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15749096,
+      "step": 20160
+    },
+    {
+      "epoch": 41.92307692307692,
+      "grad_norm": 0.0027380662504583597,
+      "learning_rate": 2.46779956570315e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15753000,
+      "step": 20165
+    },
+    {
+      "epoch": 41.933471933471935,
+      "grad_norm": 0.002597673563286662,
+      "learning_rate": 2.466817901945787e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15756776,
+      "step": 20170
+    },
+    {
+      "epoch": 41.943866943866944,
+      "grad_norm": 7.136188651202247e-05,
+      "learning_rate": 2.4658362433055217e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 15760616,
+      "step": 20175
+    },
+    {
+      "epoch": 41.95426195426195,
+      "grad_norm": 0.0010777947027236223,
+      "learning_rate": 2.4648545899337356e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 15764520,
+      "step": 20180
+    },
+    {
+      "epoch": 41.96465696465697,
+      "grad_norm": 0.0005380319198593497,
+      "learning_rate": 2.4638729419818143e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15768648,
+      "step": 20185
+    },
+    {
+      "epoch": 41.975051975051976,
+      "grad_norm": 0.00041885170503519475,
+      "learning_rate": 2.46289129960114e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15772616,
+      "step": 20190
+    },
+    {
+      "epoch": 41.985446985446984,
+      "grad_norm": 0.0006579115288332105,
+      "learning_rate": 2.4619096629430924e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15776616,
+      "step": 20195
+    },
+    {
+      "epoch": 41.99584199584199,
+      "grad_norm": 0.00011185410403413698,
+      "learning_rate": 2.4609280321590543e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15780456,
+      "step": 20200
+    },
+    {
+      "epoch": 41.99584199584199,
+      "eval_loss": 0.4160495698451996,
+      "eval_runtime": 11.6921,
+      "eval_samples_per_second": 73.212,
+      "eval_steps_per_second": 18.303,
+      "num_input_tokens_seen": 15780456,
+      "step": 20200
+    },
+    {
+      "epoch": 42.00623700623701,
+      "grad_norm": 0.0010749701177701354,
+      "learning_rate": 2.4599464074004037e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 15784368,
+      "step": 20205
+    },
+    {
+      "epoch": 42.016632016632016,
+      "grad_norm": 0.001509393216110766,
+      "learning_rate": 2.4589647888185204e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15788304,
+      "step": 20210
+    },
+    {
+      "epoch": 42.027027027027025,
+      "grad_norm": 0.00015687820268794894,
+      "learning_rate": 2.4579831765647836e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15792112,
+      "step": 20215
+    },
+    {
+      "epoch": 42.03742203742204,
+      "grad_norm": 0.0028200331144034863,
+      "learning_rate": 2.4570015707905676e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 15795984,
+      "step": 20220
+    },
+    {
+      "epoch": 42.04781704781705,
+      "grad_norm": 0.007994499988853931,
+      "learning_rate": 2.4560199716472508e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15799984,
+      "step": 20225
+    },
+    {
+      "epoch": 42.05821205821206,
+      "grad_norm": 34.82675552368164,
+      "learning_rate": 2.455038379286207e-05,
+      "loss": 0.0071,
+      "num_input_tokens_seen": 15803952,
+      "step": 20230
+    },
+    {
+      "epoch": 42.06860706860707,
+      "grad_norm": 0.0001823818893171847,
+      "learning_rate": 2.4540567938588095e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15807760,
+      "step": 20235
+    },
+    {
+      "epoch": 42.07900207900208,
+      "grad_norm": 7.306822953978553e-05,
+      "learning_rate": 2.4530752155164328e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15811856,
+      "step": 20240
+    },
+    {
+      "epoch": 42.08939708939709,
+      "grad_norm": 0.0007886007660999894,
+      "learning_rate": 2.4520936444104463e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15815728,
+      "step": 20245
+    },
+    {
+      "epoch": 42.0997920997921,
+      "grad_norm": 0.0003381914575584233,
+      "learning_rate": 2.4511120806922218e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15819568,
+      "step": 20250
+    },
+    {
+      "epoch": 42.11018711018711,
+      "grad_norm": 0.0006187465623952448,
+      "learning_rate": 2.45013052451313e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 15823376,
+      "step": 20255
+    },
+    {
+      "epoch": 42.12058212058212,
+      "grad_norm": 0.0003419234708417207,
+      "learning_rate": 2.4491489760245376e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15827280,
+      "step": 20260
+    },
+    {
+      "epoch": 42.13097713097713,
+      "grad_norm": 0.00048070173943415284,
+      "learning_rate": 2.4481674353778115e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 15831216,
+      "step": 20265
+    },
+    {
+      "epoch": 42.141372141372145,
+      "grad_norm": 0.00037174904718995094,
+      "learning_rate": 2.447185902724319e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15835088,
+      "step": 20270
+    },
+    {
+      "epoch": 42.15176715176715,
+      "grad_norm": 0.0006403488223440945,
+      "learning_rate": 2.4462043782154233e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15839056,
+      "step": 20275
+    },
+    {
+      "epoch": 42.16216216216216,
+      "grad_norm": 0.12653225660324097,
+      "learning_rate": 2.4452228620024895e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15842960,
+      "step": 20280
+    },
+    {
+      "epoch": 42.17255717255717,
+      "grad_norm": 0.002133456524461508,
+      "learning_rate": 2.4442413542368776e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 15847024,
+      "step": 20285
+    },
+    {
+      "epoch": 42.182952182952185,
+      "grad_norm": 0.0001998131483560428,
+      "learning_rate": 2.4432598550699502e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15850992,
+      "step": 20290
+    },
+    {
+      "epoch": 42.19334719334719,
+      "grad_norm": 0.0001556073984829709,
+      "learning_rate": 2.4422783646530663e-05,
+      "loss": 0.0153,
+      "num_input_tokens_seen": 15854768,
+      "step": 20295
+    },
+    {
+      "epoch": 42.2037422037422,
+      "grad_norm": 0.0012634146260097623,
+      "learning_rate": 2.441296883137584e-05,
+      "loss": 0.012,
+      "num_input_tokens_seen": 15858640,
+      "step": 20300
+    },
+    {
+      "epoch": 42.21413721413722,
+      "grad_norm": 0.007299724966287613,
+      "learning_rate": 2.4403154106748592e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 15862512,
+      "step": 20305
+    },
+    {
+      "epoch": 42.224532224532226,
+      "grad_norm": 0.005817460361868143,
+      "learning_rate": 2.4393339474162494e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15866416,
+      "step": 20310
+    },
+    {
+      "epoch": 42.234927234927234,
+      "grad_norm": 0.012736319564282894,
+      "learning_rate": 2.4383524935131062e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 15870448,
+      "step": 20315
+    },
+    {
+      "epoch": 42.24532224532224,
+      "grad_norm": 0.008905544877052307,
+      "learning_rate": 2.437371049116784e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 15874384,
+      "step": 20320
+    },
+    {
+      "epoch": 42.25571725571726,
+      "grad_norm": 0.0009572419803589582,
+      "learning_rate": 2.436389614378632e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15878192,
+      "step": 20325
+    },
+    {
+      "epoch": 42.266112266112266,
+      "grad_norm": 0.0007836163276806474,
+      "learning_rate": 2.435408189450002e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15882096,
+      "step": 20330
+    },
+    {
+      "epoch": 42.276507276507274,
+      "grad_norm": 0.00019901100313290954,
+      "learning_rate": 2.4344267744822406e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15885872,
+      "step": 20335
+    },
+    {
+      "epoch": 42.28690228690229,
+      "grad_norm": 0.0005666622309945524,
+      "learning_rate": 2.4334453696266944e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15889872,
+      "step": 20340
+    },
+    {
+      "epoch": 42.2972972972973,
+      "grad_norm": 0.0009065307676792145,
+      "learning_rate": 2.432463975034708e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15893712,
+      "step": 20345
+    },
+    {
+      "epoch": 42.30769230769231,
+      "grad_norm": 0.00010661953274393454,
+      "learning_rate": 2.4314825908576265e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15897648,
+      "step": 20350
+    },
+    {
+      "epoch": 42.318087318087315,
+      "grad_norm": 4.956878547091037e-05,
+      "learning_rate": 2.4305012172467897e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15901424,
+      "step": 20355
+    },
+    {
+      "epoch": 42.32848232848233,
+      "grad_norm": 0.0002652436960488558,
+      "learning_rate": 2.4295198543535393e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15905296,
+      "step": 20360
+    },
+    {
+      "epoch": 42.33887733887734,
+      "grad_norm": 0.0006517907604575157,
+      "learning_rate": 2.4285385023292124e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15909232,
+      "step": 20365
+    },
+    {
+      "epoch": 42.34927234927235,
+      "grad_norm": 0.00014258723240345716,
+      "learning_rate": 2.427557161325147e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15913168,
+      "step": 20370
+    },
+    {
+      "epoch": 42.35966735966736,
+      "grad_norm": 0.00014572932559531182,
+      "learning_rate": 2.4265758314926778e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15917168,
+      "step": 20375
+    },
+    {
+      "epoch": 42.37006237006237,
+      "grad_norm": 0.0005490597686730325,
+      "learning_rate": 2.4255945129831373e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15920912,
+      "step": 20380
+    },
+    {
+      "epoch": 42.38045738045738,
+      "grad_norm": 0.00012775263166986406,
+      "learning_rate": 2.4246132059478578e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 15924720,
+      "step": 20385
+    },
+    {
+      "epoch": 42.39085239085239,
+      "grad_norm": 0.0003656811604741961,
+      "learning_rate": 2.4236319105381706e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 15928560,
+      "step": 20390
+    },
+    {
+      "epoch": 42.4012474012474,
+      "grad_norm": 0.00010815545101650059,
+      "learning_rate": 2.422650626905401e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15932464,
+      "step": 20395
+    },
+    {
+      "epoch": 42.41164241164241,
+      "grad_norm": 0.00020391438738442957,
+      "learning_rate": 2.4216693552008785e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15936432,
+      "step": 20400
+    },
+    {
+      "epoch": 42.41164241164241,
+      "eval_loss": 0.4530101716518402,
+      "eval_runtime": 11.69,
+      "eval_samples_per_second": 73.225,
+      "eval_steps_per_second": 18.306,
+      "num_input_tokens_seen": 15936432,
+      "step": 20400
+    },
+    {
+      "epoch": 42.42203742203742,
+      "grad_norm": 0.0008404789841733873,
+      "learning_rate": 2.4206880955759247e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15940400,
+      "step": 20405
+    },
+    {
+      "epoch": 42.432432432432435,
+      "grad_norm": 0.00286590657196939,
+      "learning_rate": 2.419706848181863e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15944336,
+      "step": 20410
+    },
+    {
+      "epoch": 42.44282744282744,
+      "grad_norm": 0.004327325616031885,
+      "learning_rate": 2.4187256131700153e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 15948368,
+      "step": 20415
+    },
+    {
+      "epoch": 42.45322245322245,
+      "grad_norm": 0.0013223864370957017,
+      "learning_rate": 2.4177443906916985e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 15952176,
+      "step": 20420
+    },
+    {
+      "epoch": 42.46361746361746,
+      "grad_norm": 1.4238255062082317e-05,
+      "learning_rate": 2.4167631808982303e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 15956112,
+      "step": 20425
+    },
+    {
+      "epoch": 42.474012474012476,
+      "grad_norm": 8.614963735453784e-05,
+      "learning_rate": 2.4157819839409264e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15960112,
+      "step": 20430
+    },
+    {
+      "epoch": 42.484407484407484,
+      "grad_norm": 0.00013529823627322912,
+      "learning_rate": 2.414800799971098e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15964016,
+      "step": 20435
+    },
+    {
+      "epoch": 42.49480249480249,
+      "grad_norm": 0.00010413950803922489,
+      "learning_rate": 2.4138196291400582e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15967888,
+      "step": 20440
+    },
+    {
+      "epoch": 42.50519750519751,
+      "grad_norm": 8.320940105477348e-05,
+      "learning_rate": 2.412838471599114e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15971792,
+      "step": 20445
+    },
+    {
+      "epoch": 42.515592515592516,
+      "grad_norm": 0.00010351278615416959,
+      "learning_rate": 2.411857327499572e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15975824,
+      "step": 20450
+    },
+    {
+      "epoch": 42.525987525987524,
+      "grad_norm": 0.029509078711271286,
+      "learning_rate": 2.410876196992739e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15979792,
+      "step": 20455
+    },
+    {
+      "epoch": 42.53638253638254,
+      "grad_norm": 0.000323368760291487,
+      "learning_rate": 2.4098950802299156e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15983728,
+      "step": 20460
+    },
+    {
+      "epoch": 42.54677754677755,
+      "grad_norm": 0.0010530998697504401,
+      "learning_rate": 2.4089139773624027e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15987504,
+      "step": 20465
+    },
+    {
+      "epoch": 42.55717255717256,
+      "grad_norm": 5.919565592193976e-05,
+      "learning_rate": 2.4079328885415007e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15991440,
+      "step": 20470
+    },
+    {
+      "epoch": 42.567567567567565,
+      "grad_norm": 0.007028836756944656,
+      "learning_rate": 2.4069518139185036e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15995376,
+      "step": 20475
+    },
+    {
+      "epoch": 42.57796257796258,
+      "grad_norm": 0.002006805967539549,
+      "learning_rate": 2.405970753644706e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 15999152,
+      "step": 20480
+    },
+    {
+      "epoch": 42.58835758835759,
+      "grad_norm": 0.0004598653467837721,
+      "learning_rate": 2.4049897078714e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 16003056,
+      "step": 20485
+    },
+    {
+      "epoch": 42.5987525987526,
+      "grad_norm": 0.0001999890519073233,
+      "learning_rate": 2.404008676749874e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16006992,
+      "step": 20490
+    },
+    {
+      "epoch": 42.60914760914761,
+      "grad_norm": 9.796230733627453e-05,
+      "learning_rate": 2.403027660431418e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16010768,
+      "step": 20495
+    },
+    {
+      "epoch": 42.61954261954262,
+      "grad_norm": 4.308269126340747e-05,
+      "learning_rate": 2.402046659067314e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16014704,
+      "step": 20500
+    },
+    {
+      "epoch": 42.62993762993763,
+      "grad_norm": 0.00025219115195795894,
+      "learning_rate": 2.401065672808847e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16018640,
+      "step": 20505
+    },
+    {
+      "epoch": 42.64033264033264,
+      "grad_norm": 0.018069937825202942,
+      "learning_rate": 2.400084701807296e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16022576,
+      "step": 20510
+    },
+    {
+      "epoch": 42.65072765072765,
+      "grad_norm": 6.68342036078684e-05,
+      "learning_rate": 2.39910374621394e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16026416,
+      "step": 20515
+    },
+    {
+      "epoch": 42.66112266112266,
+      "grad_norm": 4.59220573247876e-05,
+      "learning_rate": 2.3981228061800544e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16030192,
+      "step": 20520
+    },
+    {
+      "epoch": 42.67151767151767,
+      "grad_norm": 0.00010087050759466365,
+      "learning_rate": 2.3971418818569115e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16034064,
+      "step": 20525
+    },
+    {
+      "epoch": 42.681912681912685,
+      "grad_norm": 0.0002644479973241687,
+      "learning_rate": 2.3961609733957832e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 16037872,
+      "step": 20530
+    },
+    {
+      "epoch": 42.69230769230769,
+      "grad_norm": 0.00010957532504107803,
+      "learning_rate": 2.395180080947939e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 16041744,
+      "step": 20535
+    },
+    {
+      "epoch": 42.7027027027027,
+      "grad_norm": 0.0003947852528654039,
+      "learning_rate": 2.394199204664642e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 16045616,
+      "step": 20540
+    },
+    {
+      "epoch": 42.71309771309771,
+      "grad_norm": 1.4359360648086295e-05,
+      "learning_rate": 2.3932183446971583e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16049456,
+      "step": 20545
+    },
+    {
+      "epoch": 42.723492723492726,
+      "grad_norm": 18.411800384521484,
+      "learning_rate": 2.3922375011967473e-05,
+      "loss": 0.0199,
+      "num_input_tokens_seen": 16053296,
+      "step": 20550
+    },
+    {
+      "epoch": 42.733887733887734,
+      "grad_norm": 0.0003688051365315914,
+      "learning_rate": 2.3912566743146676e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16057136,
+      "step": 20555
+    },
+    {
+      "epoch": 42.74428274428274,
+      "grad_norm": 0.00017576251411810517,
+      "learning_rate": 2.390275864202176e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16061008,
+      "step": 20560
+    },
+    {
+      "epoch": 42.75467775467776,
+      "grad_norm": 0.0013504315866157413,
+      "learning_rate": 2.3892950710105243e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16065168,
+      "step": 20565
+    },
+    {
+      "epoch": 42.765072765072766,
+      "grad_norm": 0.000596087658777833,
+      "learning_rate": 2.3883142948909635e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 16069040,
+      "step": 20570
+    },
+    {
+      "epoch": 42.775467775467774,
+      "grad_norm": 0.0007634523790329695,
+      "learning_rate": 2.3873335359947433e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16072784,
+      "step": 20575
+    },
+    {
+      "epoch": 42.78586278586278,
+      "grad_norm": 0.0013042689533904195,
+      "learning_rate": 2.3863527944731066e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16076592,
+      "step": 20580
+    },
+    {
+      "epoch": 42.7962577962578,
+      "grad_norm": 0.00019410178356338292,
+      "learning_rate": 2.385372070477298e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16080528,
+      "step": 20585
+    },
+    {
+      "epoch": 42.80665280665281,
+      "grad_norm": 0.0019454647554084659,
+      "learning_rate": 2.384391364158556e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16084528,
+      "step": 20590
+    },
+    {
+      "epoch": 42.817047817047815,
+      "grad_norm": 0.0011863256804645061,
+      "learning_rate": 2.3834106756681185e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16088400,
+      "step": 20595
+    },
+    {
+      "epoch": 42.82744282744283,
+      "grad_norm": 0.00502532534301281,
+      "learning_rate": 2.3824300051572206e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 16092272,
+      "step": 20600
+    },
+    {
+      "epoch": 42.82744282744283,
+      "eval_loss": 0.4244086742401123,
+      "eval_runtime": 11.6942,
+      "eval_samples_per_second": 73.198,
+      "eval_steps_per_second": 18.3,
+      "num_input_tokens_seen": 16092272,
+      "step": 20600
+    },
+    {
+      "epoch": 42.83783783783784,
+      "grad_norm": 0.0005089411279186606,
+      "learning_rate": 2.3814493527770923e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16096176,
+      "step": 20605
+    },
+    {
+      "epoch": 42.84823284823285,
+      "grad_norm": 0.00039865353028289974,
+      "learning_rate": 2.3804687186789637e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 16100208,
+      "step": 20610
+    },
+    {
+      "epoch": 42.858627858627855,
+      "grad_norm": 0.0012574605643749237,
+      "learning_rate": 2.379488103014062e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 16104176,
+      "step": 20615
+    },
+    {
+      "epoch": 42.86902286902287,
+      "grad_norm": 0.00044609044562093914,
+      "learning_rate": 2.3785075059336086e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16108144,
+      "step": 20620
+    },
+    {
+      "epoch": 42.87941787941788,
+      "grad_norm": 9.071978274732828e-05,
+      "learning_rate": 2.3775269275888248e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16112048,
+      "step": 20625
+    },
+    {
+      "epoch": 42.88981288981289,
+      "grad_norm": 0.0007968292338773608,
+      "learning_rate": 2.3765463681309274e-05,
+      "loss": 0.0361,
+      "num_input_tokens_seen": 16116016,
+      "step": 20630
+    },
+    {
+      "epoch": 42.9002079002079,
+      "grad_norm": 0.002103642327710986,
+      "learning_rate": 2.3755658277111313e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16120112,
+      "step": 20635
+    },
+    {
+      "epoch": 42.91060291060291,
+      "grad_norm": 0.0009070364758372307,
+      "learning_rate": 2.374585306480649e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16124016,
+      "step": 20640
+    },
+    {
+      "epoch": 42.92099792099792,
+      "grad_norm": 0.0011426984565332532,
+      "learning_rate": 2.3736048045906877e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16127920,
+      "step": 20645
+    },
+    {
+      "epoch": 42.931392931392935,
+      "grad_norm": 0.18585839867591858,
+      "learning_rate": 2.372624322192454e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 16131760,
+      "step": 20650
+    },
+    {
+      "epoch": 42.94178794178794,
+      "grad_norm": 0.001772083342075348,
+      "learning_rate": 2.3716438594371516e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16135728,
+      "step": 20655
+    },
+    {
+      "epoch": 42.95218295218295,
+      "grad_norm": 0.0018146125366911292,
+      "learning_rate": 2.3706634164759784e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16139600,
+      "step": 20660
+    },
+    {
+      "epoch": 42.96257796257796,
+      "grad_norm": 0.0006465365295298398,
+      "learning_rate": 2.3696829934601323e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16143568,
+      "step": 20665
+    },
+    {
+      "epoch": 42.972972972972975,
+      "grad_norm": 0.00020426098490133882,
+      "learning_rate": 2.3687025905408053e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16147536,
+      "step": 20670
+    },
+    {
+      "epoch": 42.983367983367984,
+      "grad_norm": 0.000732767628505826,
+      "learning_rate": 2.3677222078691886e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16151376,
+      "step": 20675
+    },
+    {
+      "epoch": 42.99376299376299,
+      "grad_norm": 0.14691023528575897,
+      "learning_rate": 2.366741845596471e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 16155312,
+      "step": 20680
+    },
+    {
+      "epoch": 43.00415800415801,
+      "grad_norm": 0.006579479668289423,
+      "learning_rate": 2.3657615038738343e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16159224,
+      "step": 20685
+    },
+    {
+      "epoch": 43.014553014553016,
+      "grad_norm": 0.002914112526923418,
+      "learning_rate": 2.3647811828524614e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16163160,
+      "step": 20690
+    },
+    {
+      "epoch": 43.024948024948024,
+      "grad_norm": 0.9068381190299988,
+      "learning_rate": 2.363800882683529e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 16167096,
+      "step": 20695
+    },
+    {
+      "epoch": 43.03534303534303,
+      "grad_norm": 0.0014504495775327086,
+      "learning_rate": 2.3628206035182125e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 16171064,
+      "step": 20700
+    },
+    {
+      "epoch": 43.04573804573805,
+      "grad_norm": 0.0005934596410952508,
+      "learning_rate": 2.361840345507683e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16174904,
+      "step": 20705
+    },
+    {
+      "epoch": 43.056133056133056,
+      "grad_norm": 0.0019272859208285809,
+      "learning_rate": 2.3608601088031073e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16178744,
+      "step": 20710
+    },
+    {
+      "epoch": 43.066528066528065,
+      "grad_norm": 0.00020069706079084426,
+      "learning_rate": 2.3598798935556516e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16182680,
+      "step": 20715
+    },
+    {
+      "epoch": 43.07692307692308,
+      "grad_norm": 0.00014204718172550201,
+      "learning_rate": 2.3588996999164784e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16186616,
+      "step": 20720
+    },
+    {
+      "epoch": 43.08731808731809,
+      "grad_norm": 0.0004661197599489242,
+      "learning_rate": 2.3579195280367434e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16190424,
+      "step": 20725
+    },
+    {
+      "epoch": 43.0977130977131,
+      "grad_norm": 0.0001789227972039953,
+      "learning_rate": 2.356939378067603e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16194360,
+      "step": 20730
+    },
+    {
+      "epoch": 43.108108108108105,
+      "grad_norm": 0.001264728605747223,
+      "learning_rate": 2.3559592501602092e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16198328,
+      "step": 20735
+    },
+    {
+      "epoch": 43.11850311850312,
+      "grad_norm": 0.00025076259043999016,
+      "learning_rate": 2.3549791444657076e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16202200,
+      "step": 20740
+    },
+    {
+      "epoch": 43.12889812889813,
+      "grad_norm": 0.00042946074972860515,
+      "learning_rate": 2.353999061135246e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16206040,
+      "step": 20745
+    },
+    {
+      "epoch": 43.13929313929314,
+      "grad_norm": 0.0003230917500331998,
+      "learning_rate": 2.3530190003199626e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16210008,
+      "step": 20750
+    },
+    {
+      "epoch": 43.14968814968815,
+      "grad_norm": 0.0003692317695822567,
+      "learning_rate": 2.3520389621709965e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16213848,
+      "step": 20755
+    },
+    {
+      "epoch": 43.16008316008316,
+      "grad_norm": 1.8565354347229004,
+      "learning_rate": 2.351058946839483e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 16217688,
+      "step": 20760
+    },
+    {
+      "epoch": 43.17047817047817,
+      "grad_norm": 0.000595595280174166,
+      "learning_rate": 2.350078954476551e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 16221496,
+      "step": 20765
+    },
+    {
+      "epoch": 43.18087318087318,
+      "grad_norm": 0.0005349307903088629,
+      "learning_rate": 2.3490989852333272e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16225432,
+      "step": 20770
+    },
+    {
+      "epoch": 43.19126819126819,
+      "grad_norm": 0.0021819970570504665,
+      "learning_rate": 2.3481190392609377e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16229336,
+      "step": 20775
+    },
+    {
+      "epoch": 43.2016632016632,
+      "grad_norm": 0.0011935236398130655,
+      "learning_rate": 2.3471391167105e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16233176,
+      "step": 20780
+    },
+    {
+      "epoch": 43.21205821205821,
+      "grad_norm": 0.13994193077087402,
+      "learning_rate": 2.3461592177331325e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 16237272,
+      "step": 20785
+    },
+    {
+      "epoch": 43.222453222453225,
+      "grad_norm": 0.0008431302849203348,
+      "learning_rate": 2.345179342479946e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16241112,
+      "step": 20790
+    },
+    {
+      "epoch": 43.232848232848234,
+      "grad_norm": 0.2114466428756714,
+      "learning_rate": 2.3441994911020503e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 16245080,
+      "step": 20795
+    },
+    {
+      "epoch": 43.24324324324324,
+      "grad_norm": 0.020069794729351997,
+      "learning_rate": 2.3432196637505522e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16249048,
+      "step": 20800
+    },
+    {
+      "epoch": 43.24324324324324,
+      "eval_loss": 0.45716020464897156,
+      "eval_runtime": 11.7045,
+      "eval_samples_per_second": 73.134,
+      "eval_steps_per_second": 18.284,
+      "num_input_tokens_seen": 16249048,
+      "step": 20800
+    },
+    {
+      "epoch": 43.25363825363825,
+      "grad_norm": 0.0006179632036946714,
+      "learning_rate": 2.3422398605765515e-05,
+      "loss": 0.008,
+      "num_input_tokens_seen": 16252856,
+      "step": 20805
+    },
+    {
+      "epoch": 43.264033264033266,
+      "grad_norm": 0.16802337765693665,
+      "learning_rate": 2.3412600817311462e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 16256760,
+      "step": 20810
+    },
+    {
+      "epoch": 43.274428274428274,
+      "grad_norm": 0.14923346042633057,
+      "learning_rate": 2.3402803273654326e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 16260696,
+      "step": 20815
+    },
+    {
+      "epoch": 43.28482328482328,
+      "grad_norm": 0.0006100265891291201,
+      "learning_rate": 2.3393005976304983e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16264728,
+      "step": 20820
+    },
+    {
+      "epoch": 43.2952182952183,
+      "grad_norm": 0.00022311277280095965,
+      "learning_rate": 2.338320892677432e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16268632,
+      "step": 20825
+    },
+    {
+      "epoch": 43.305613305613306,
+      "grad_norm": 0.00019659523968584836,
+      "learning_rate": 2.3373412126573155e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16272440,
+      "step": 20830
+    },
+    {
+      "epoch": 43.316008316008315,
+      "grad_norm": 0.0011362145887687802,
+      "learning_rate": 2.3363615577212285e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16276440,
+      "step": 20835
+    },
+    {
+      "epoch": 43.32640332640332,
+      "grad_norm": 0.001742366119287908,
+      "learning_rate": 2.3353819280202455e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16280248,
+      "step": 20840
+    },
+    {
+      "epoch": 43.33679833679834,
+      "grad_norm": 0.0001748505310388282,
+      "learning_rate": 2.334402323705438e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16284120,
+      "step": 20845
+    },
+    {
+      "epoch": 43.34719334719335,
+      "grad_norm": 0.0001484177482780069,
+      "learning_rate": 2.3334227449278725e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 16288056,
+      "step": 20850
+    },
+    {
+      "epoch": 43.357588357588355,
+      "grad_norm": 0.0008238742593675852,
+      "learning_rate": 2.3324431918386143e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16291928,
+      "step": 20855
+    },
+    {
+      "epoch": 43.36798336798337,
+      "grad_norm": 4.987459659576416,
+      "learning_rate": 2.3314636645887207e-05,
+      "loss": 0.0457,
+      "num_input_tokens_seen": 16296056,
+      "step": 20860
+    },
+    {
+      "epoch": 43.37837837837838,
+      "grad_norm": 0.000860365922562778,
+      "learning_rate": 2.3304841633292487e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16299960,
+      "step": 20865
+    },
+    {
+      "epoch": 43.38877338877339,
+      "grad_norm": 0.00036710381391458213,
+      "learning_rate": 2.329504688211248e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16303768,
+      "step": 20870
+    },
+    {
+      "epoch": 43.3991683991684,
+      "grad_norm": 0.09492574632167816,
+      "learning_rate": 2.3285252393857677e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 16307672,
+      "step": 20875
+    },
+    {
+      "epoch": 43.40956340956341,
+      "grad_norm": 0.0035492200404405594,
+      "learning_rate": 2.327545817003851e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16311576,
+      "step": 20880
+    },
+    {
+      "epoch": 43.41995841995842,
+      "grad_norm": 0.0010640786495059729,
+      "learning_rate": 2.326566421216535e-05,
+      "loss": 0.0061,
+      "num_input_tokens_seen": 16315448,
+      "step": 20885
+    },
+    {
+      "epoch": 43.43035343035343,
+      "grad_norm": 0.0007823574705980718,
+      "learning_rate": 2.3255870521748565e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 16319320,
+      "step": 20890
+    },
+    {
+      "epoch": 43.44074844074844,
+      "grad_norm": 0.0015277586644515395,
+      "learning_rate": 2.3246077100298474e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16323160,
+      "step": 20895
+    },
+    {
+      "epoch": 43.45114345114345,
+      "grad_norm": 0.0013684754958376288,
+      "learning_rate": 2.3236283949325328e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16327032,
+      "step": 20900
+    },
+    {
+      "epoch": 43.46153846153846,
+      "grad_norm": 0.001533312606625259,
+      "learning_rate": 2.3226491070339368e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16330872,
+      "step": 20905
+    },
+    {
+      "epoch": 43.471933471933475,
+      "grad_norm": 0.00031090047559700906,
+      "learning_rate": 2.3216698464850762e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16334776,
+      "step": 20910
+    },
+    {
+      "epoch": 43.482328482328484,
+      "grad_norm": 0.0003848281630780548,
+      "learning_rate": 2.320690613436967e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 16338744,
+      "step": 20915
+    },
+    {
+      "epoch": 43.49272349272349,
+      "grad_norm": 0.002602728782221675,
+      "learning_rate": 2.3197114080406192e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16342712,
+      "step": 20920
+    },
+    {
+      "epoch": 43.5031185031185,
+      "grad_norm": 0.0010092881275340915,
+      "learning_rate": 2.3187322304470365e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16346616,
+      "step": 20925
+    },
+    {
+      "epoch": 43.513513513513516,
+      "grad_norm": 0.08550321310758591,
+      "learning_rate": 2.3177530808072222e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16350488,
+      "step": 20930
+    },
+    {
+      "epoch": 43.523908523908524,
+      "grad_norm": 0.00034836100530810654,
+      "learning_rate": 2.316773959272174e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 16354520,
+      "step": 20935
+    },
+    {
+      "epoch": 43.53430353430353,
+      "grad_norm": 0.0008592975209467113,
+      "learning_rate": 2.3157948659928823e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16358392,
+      "step": 20940
+    },
+    {
+      "epoch": 43.54469854469855,
+      "grad_norm": 0.0008040325483307242,
+      "learning_rate": 2.3148158011203388e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16362520,
+      "step": 20945
+    },
+    {
+      "epoch": 43.555093555093556,
+      "grad_norm": 0.0002649552479851991,
+      "learning_rate": 2.3138367648055253e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16366488,
+      "step": 20950
+    },
+    {
+      "epoch": 43.565488565488565,
+      "grad_norm": 0.0011269205715507269,
+      "learning_rate": 2.312857757199422e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16370424,
+      "step": 20955
+    },
+    {
+      "epoch": 43.57588357588357,
+      "grad_norm": 0.0010838838061317801,
+      "learning_rate": 2.3118787784530048e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16374360,
+      "step": 20960
+    },
+    {
+      "epoch": 43.58627858627859,
+      "grad_norm": 0.0005141847068443894,
+      "learning_rate": 2.310899828717243e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16378232,
+      "step": 20965
+    },
+    {
+      "epoch": 43.5966735966736,
+      "grad_norm": 0.0006281076348386705,
+      "learning_rate": 2.309920908143104e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16382200,
+      "step": 20970
+    },
+    {
+      "epoch": 43.607068607068605,
+      "grad_norm": 0.00014750596892554313,
+      "learning_rate": 2.308942016881551e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16385944,
+      "step": 20975
+    },
+    {
+      "epoch": 43.61746361746362,
+      "grad_norm": 0.0005438126972876489,
+      "learning_rate": 2.307963155083539e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16389784,
+      "step": 20980
+    },
+    {
+      "epoch": 43.62785862785863,
+      "grad_norm": 0.00012784010323230177,
+      "learning_rate": 2.306984322900022e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16393720,
+      "step": 20985
+    },
+    {
+      "epoch": 43.63825363825364,
+      "grad_norm": 0.00039807800203561783,
+      "learning_rate": 2.3060055204819482e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16397656,
+      "step": 20990
+    },
+    {
+      "epoch": 43.648648648648646,
+      "grad_norm": 17.354202270507812,
+      "learning_rate": 2.3050267479802604e-05,
+      "loss": 0.0272,
+      "num_input_tokens_seen": 16401432,
+      "step": 20995
+    },
+    {
+      "epoch": 43.65904365904366,
+      "grad_norm": 0.000853641249705106,
+      "learning_rate": 2.304048005545899e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16405368,
+      "step": 21000
+    },
+    {
+      "epoch": 43.65904365904366,
+      "eval_loss": 0.3547861576080322,
+      "eval_runtime": 11.6629,
+      "eval_samples_per_second": 73.395,
+      "eval_steps_per_second": 18.349,
+      "num_input_tokens_seen": 16405368,
+      "step": 21000
+    },
+    {
+      "epoch": 43.66943866943867,
+      "grad_norm": 0.17259985208511353,
+      "learning_rate": 2.3030692933297972e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 16409176,
+      "step": 21005
+    },
+    {
+      "epoch": 43.67983367983368,
+      "grad_norm": 0.0002078032266581431,
+      "learning_rate": 2.3020906114828843e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16413112,
+      "step": 21010
+    },
+    {
+      "epoch": 43.69022869022869,
+      "grad_norm": 10.368476867675781,
+      "learning_rate": 2.301111960156088e-05,
+      "loss": 0.0014,
+      "num_input_tokens_seen": 16417048,
+      "step": 21015
+    },
+    {
+      "epoch": 43.7006237006237,
+      "grad_norm": 0.0017174235545098782,
+      "learning_rate": 2.300133339500326e-05,
+      "loss": 0.0267,
+      "num_input_tokens_seen": 16420952,
+      "step": 21020
+    },
+    {
+      "epoch": 43.71101871101871,
+      "grad_norm": 0.0014577835099771619,
+      "learning_rate": 2.2991547496665148e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16424920,
+      "step": 21025
+    },
+    {
+      "epoch": 43.72141372141372,
+      "grad_norm": 0.00294153718277812,
+      "learning_rate": 2.298176190805565e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 16428760,
+      "step": 21030
+    },
+    {
+      "epoch": 43.731808731808734,
+      "grad_norm": 0.005749391857534647,
+      "learning_rate": 2.2971976630683826e-05,
+      "loss": 0.0005,
+      "num_input_tokens_seen": 16432568,
+      "step": 21035
+    },
+    {
+      "epoch": 43.74220374220374,
+      "grad_norm": 0.00356932170689106,
+      "learning_rate": 2.29621916660587e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16436312,
+      "step": 21040
+    },
+    {
+      "epoch": 43.75259875259875,
+      "grad_norm": 0.0003373512881807983,
+      "learning_rate": 2.295240701568922e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16440184,
+      "step": 21045
+    },
+    {
+      "epoch": 43.762993762993766,
+      "grad_norm": 0.1277269572019577,
+      "learning_rate": 2.2942622681084312e-05,
+      "loss": 0.0175,
+      "num_input_tokens_seen": 16443832,
+      "step": 21050
+    },
+    {
+      "epoch": 43.773388773388774,
+      "grad_norm": 0.00020646025950554758,
+      "learning_rate": 2.293283866375284e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16447832,
+      "step": 21055
+    },
+    {
+      "epoch": 43.78378378378378,
+      "grad_norm": 0.0013060923665761948,
+      "learning_rate": 2.2923054965203627e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16451672,
+      "step": 21060
+    },
+    {
+      "epoch": 43.79417879417879,
+      "grad_norm": 0.002223026705905795,
+      "learning_rate": 2.2913271586945443e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16455480,
+      "step": 21065
+    },
+    {
+      "epoch": 43.804573804573806,
+      "grad_norm": 0.21471332013607025,
+      "learning_rate": 2.290348853048699e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 16459320,
+      "step": 21070
+    },
+    {
+      "epoch": 43.814968814968815,
+      "grad_norm": 0.021518848836421967,
+      "learning_rate": 2.2893705797336956e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16463224,
+      "step": 21075
+    },
+    {
+      "epoch": 43.82536382536382,
+      "grad_norm": 0.013551289215683937,
+      "learning_rate": 2.288392338900397e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16467192,
+      "step": 21080
+    },
+    {
+      "epoch": 43.83575883575884,
+      "grad_norm": 0.0042636822909116745,
+      "learning_rate": 2.2874141306996576e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16471224,
+      "step": 21085
+    },
+    {
+      "epoch": 43.84615384615385,
+      "grad_norm": 0.03576645255088806,
+      "learning_rate": 2.2864359552823312e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 16475064,
+      "step": 21090
+    },
+    {
+      "epoch": 43.856548856548855,
+      "grad_norm": 0.0002953230286948383,
+      "learning_rate": 2.2854578127992648e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 16478936,
+      "step": 21095
+    },
+    {
+      "epoch": 43.86694386694387,
+      "grad_norm": 0.0012243862729519606,
+      "learning_rate": 2.2844797034012988e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 16482808,
+      "step": 21100
+    },
+    {
+      "epoch": 43.87733887733888,
+      "grad_norm": 0.00043986961827613413,
+      "learning_rate": 2.2835016272392722e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 16486808,
+      "step": 21105
+    },
+    {
+      "epoch": 43.88773388773389,
+      "grad_norm": 0.0005565420724451542,
+      "learning_rate": 2.2825235844640142e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16490744,
+      "step": 21110
+    },
+    {
+      "epoch": 43.898128898128896,
+      "grad_norm": 0.0004119938821531832,
+      "learning_rate": 2.2815455752263522e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 16494744,
+      "step": 21115
+    },
+    {
+      "epoch": 43.90852390852391,
+      "grad_norm": 0.0003641632793005556,
+      "learning_rate": 2.2805675996771092e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16498520,
+      "step": 21120
+    },
+    {
+      "epoch": 43.91891891891892,
+      "grad_norm": 0.0009628223488107324,
+      "learning_rate": 2.2795896579670987e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16502584,
+      "step": 21125
+    },
+    {
+      "epoch": 43.92931392931393,
+      "grad_norm": 0.0008763039950281382,
+      "learning_rate": 2.2786117502471337e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16506712,
+      "step": 21130
+    },
+    {
+      "epoch": 43.93970893970894,
+      "grad_norm": 0.00029960719984956086,
+      "learning_rate": 2.2776338766680185e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16510744,
+      "step": 21135
+    },
+    {
+      "epoch": 43.95010395010395,
+      "grad_norm": 0.00013196143845561892,
+      "learning_rate": 2.2766560373805533e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16514616,
+      "step": 21140
+    },
+    {
+      "epoch": 43.96049896049896,
+      "grad_norm": 0.00012334060738794506,
+      "learning_rate": 2.2756782325355353e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16518552,
+      "step": 21145
+    },
+    {
+      "epoch": 43.97089397089397,
+      "grad_norm": 0.00039526837645098567,
+      "learning_rate": 2.2747004622837514e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16522456,
+      "step": 21150
+    },
+    {
+      "epoch": 43.981288981288984,
+      "grad_norm": 0.0006496057030744851,
+      "learning_rate": 2.2737227267759878e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16526424,
+      "step": 21155
+    },
+    {
+      "epoch": 43.99168399168399,
+      "grad_norm": 0.0003457559214439243,
+      "learning_rate": 2.272745026163024e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16530296,
+      "step": 21160
+    },
+    {
+      "epoch": 44.002079002079,
+      "grad_norm": 0.00022672518389299512,
+      "learning_rate": 2.271767360595633e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16534088,
+      "step": 21165
+    },
+    {
+      "epoch": 44.012474012474016,
+      "grad_norm": 0.0002098531840601936,
+      "learning_rate": 2.270789730224583e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16537928,
+      "step": 21170
+    },
+    {
+      "epoch": 44.022869022869024,
+      "grad_norm": 0.0005066643934696913,
+      "learning_rate": 2.2698121352006367e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 16541800,
+      "step": 21175
+    },
+    {
+      "epoch": 44.03326403326403,
+      "grad_norm": 0.0003789167676586658,
+      "learning_rate": 2.2688345756745517e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 16545736,
+      "step": 21180
+    },
+    {
+      "epoch": 44.04365904365904,
+      "grad_norm": 9.201901411870494e-05,
+      "learning_rate": 2.267857051797081e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16549576,
+      "step": 21185
+    },
+    {
+      "epoch": 44.054054054054056,
+      "grad_norm": 0.0005577909178100526,
+      "learning_rate": 2.2668795637189695e-05,
+      "loss": 0.0085,
+      "num_input_tokens_seen": 16553320,
+      "step": 21190
+    },
+    {
+      "epoch": 44.064449064449065,
+      "grad_norm": 0.0010099033825099468,
+      "learning_rate": 2.2659021115909586e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16557128,
+      "step": 21195
+    },
+    {
+      "epoch": 44.07484407484407,
+      "grad_norm": 0.0014952565543353558,
+      "learning_rate": 2.2649246955637847e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 16561000,
+      "step": 21200
+    },
+    {
+      "epoch": 44.07484407484407,
+      "eval_loss": 0.3601955473423004,
+      "eval_runtime": 11.6976,
+      "eval_samples_per_second": 73.177,
+      "eval_steps_per_second": 18.294,
+      "num_input_tokens_seen": 16561000,
+      "step": 21200
+    },
+    {
+      "epoch": 44.08523908523909,
+      "grad_norm": 0.010155725292861462,
+      "learning_rate": 2.2639473157881766e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16565000,
+      "step": 21205
+    },
+    {
+      "epoch": 44.0956340956341,
+      "grad_norm": 0.007360816467553377,
+      "learning_rate": 2.2629699724148594e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16569192,
+      "step": 21210
+    },
+    {
+      "epoch": 44.106029106029105,
+      "grad_norm": 0.0013755448162555695,
+      "learning_rate": 2.26199266559455e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16573096,
+      "step": 21215
+    },
+    {
+      "epoch": 44.11642411642411,
+      "grad_norm": 0.0047738212160766125,
+      "learning_rate": 2.2610153954779625e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16577096,
+      "step": 21220
+    },
+    {
+      "epoch": 44.12681912681913,
+      "grad_norm": 0.006541479378938675,
+      "learning_rate": 2.2600381622158056e-05,
+      "loss": 0.001,
+      "num_input_tokens_seen": 16581096,
+      "step": 21225
+    },
+    {
+      "epoch": 44.13721413721414,
+      "grad_norm": 0.006792145781219006,
+      "learning_rate": 2.2590609659587783e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16584904,
+      "step": 21230
+    },
+    {
+      "epoch": 44.147609147609145,
+      "grad_norm": 0.001185899367555976,
+      "learning_rate": 2.2580838068575787e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16588840,
+      "step": 21235
+    },
+    {
+      "epoch": 44.15800415800416,
+      "grad_norm": 1.1928044557571411,
+      "learning_rate": 2.257106685062896e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 16592840,
+      "step": 21240
+    },
+    {
+      "epoch": 44.16839916839917,
+      "grad_norm": 0.011668382212519646,
+      "learning_rate": 2.256129600725415e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16596680,
+      "step": 21245
+    },
+    {
+      "epoch": 44.17879417879418,
+      "grad_norm": 0.0007650284678675234,
+      "learning_rate": 2.2551525539958145e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16600648,
+      "step": 21250
+    },
+    {
+      "epoch": 44.189189189189186,
+      "grad_norm": 0.01291921827942133,
+      "learning_rate": 2.2541755450247663e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16604616,
+      "step": 21255
+    },
+    {
+      "epoch": 44.1995841995842,
+      "grad_norm": 0.0021016329992562532,
+      "learning_rate": 2.2531985739629382e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16608552,
+      "step": 21260
+    },
+    {
+      "epoch": 44.20997920997921,
+      "grad_norm": 0.0007266281172633171,
+      "learning_rate": 2.2522216409609924e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16612488,
+      "step": 21265
+    },
+    {
+      "epoch": 44.22037422037422,
+      "grad_norm": 15.104486465454102,
+      "learning_rate": 2.2512447461695826e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 16616296,
+      "step": 21270
+    },
+    {
+      "epoch": 44.23076923076923,
+      "grad_norm": 0.013948577456176281,
+      "learning_rate": 2.2502678897393593e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 16620232,
+      "step": 21275
+    },
+    {
+      "epoch": 44.24116424116424,
+      "grad_norm": 0.003115355968475342,
+      "learning_rate": 2.2492910718209665e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16624200,
+      "step": 21280
+    },
+    {
+      "epoch": 44.25155925155925,
+      "grad_norm": 0.0014203224563971162,
+      "learning_rate": 2.2483142925650398e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16628040,
+      "step": 21285
+    },
+    {
+      "epoch": 44.26195426195426,
+      "grad_norm": 0.0009515886195003986,
+      "learning_rate": 2.247337552122213e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16631912,
+      "step": 21290
+    },
+    {
+      "epoch": 44.272349272349274,
+      "grad_norm": 0.003344462253153324,
+      "learning_rate": 2.24636085064311e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16635848,
+      "step": 21295
+    },
+    {
+      "epoch": 44.28274428274428,
+      "grad_norm": 0.002072685630992055,
+      "learning_rate": 2.245384188278351e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16639880,
+      "step": 21300
+    },
+    {
+      "epoch": 44.29313929313929,
+      "grad_norm": 0.003531254827976227,
+      "learning_rate": 2.2444075651785513e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 16643720,
+      "step": 21305
+    },
+    {
+      "epoch": 44.303534303534306,
+      "grad_norm": 0.00040230818558484316,
+      "learning_rate": 2.243430981494316e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 16647560,
+      "step": 21310
+    },
+    {
+      "epoch": 44.313929313929314,
+      "grad_norm": 0.003849542699754238,
+      "learning_rate": 2.2424544373762475e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16651432,
+      "step": 21315
+    },
+    {
+      "epoch": 44.32432432432432,
+      "grad_norm": 0.004628174938261509,
+      "learning_rate": 2.2414779329749418e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16655400,
+      "step": 21320
+    },
+    {
+      "epoch": 44.33471933471934,
+      "grad_norm": 0.03541455417871475,
+      "learning_rate": 2.2405014684409873e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 16659304,
+      "step": 21325
+    },
+    {
+      "epoch": 44.34511434511435,
+      "grad_norm": 0.0014108283212408423,
+      "learning_rate": 2.239525043924968e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16663080,
+      "step": 21330
+    },
+    {
+      "epoch": 44.355509355509355,
+      "grad_norm": 0.0031769946217536926,
+      "learning_rate": 2.2385486595774592e-05,
+      "loss": 0.0129,
+      "num_input_tokens_seen": 16667016,
+      "step": 21335
+    },
+    {
+      "epoch": 44.36590436590436,
+      "grad_norm": 0.00023576209787279367,
+      "learning_rate": 2.237572315549033e-05,
+      "loss": 0.0166,
+      "num_input_tokens_seen": 16671048,
+      "step": 21340
+    },
+    {
+      "epoch": 44.37629937629938,
+      "grad_norm": 0.000680162978824228,
+      "learning_rate": 2.2365960119902545e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16675080,
+      "step": 21345
+    },
+    {
+      "epoch": 44.38669438669439,
+      "grad_norm": 0.00026351053384132683,
+      "learning_rate": 2.2356197490516806e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16679112,
+      "step": 21350
+    },
+    {
+      "epoch": 44.397089397089395,
+      "grad_norm": 0.00027201903867535293,
+      "learning_rate": 2.234643526883863e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 16683048,
+      "step": 21355
+    },
+    {
+      "epoch": 44.40748440748441,
+      "grad_norm": 5.2663472160929814e-05,
+      "learning_rate": 2.2336673456373497e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 16686920,
+      "step": 21360
+    },
+    {
+      "epoch": 44.41787941787942,
+      "grad_norm": 5.257022348814644e-05,
+      "learning_rate": 2.2326912054626772e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16690792,
+      "step": 21365
+    },
+    {
+      "epoch": 44.42827442827443,
+      "grad_norm": 0.00020247026986908168,
+      "learning_rate": 2.2317151065103813e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16694568,
+      "step": 21370
+    },
+    {
+      "epoch": 44.438669438669436,
+      "grad_norm": 0.1420183628797531,
+      "learning_rate": 2.2307390489309865e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 16698536,
+      "step": 21375
+    },
+    {
+      "epoch": 44.44906444906445,
+      "grad_norm": 0.0027140388265252113,
+      "learning_rate": 2.2297630328750146e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16702568,
+      "step": 21380
+    },
+    {
+      "epoch": 44.45945945945946,
+      "grad_norm": 0.003788984613493085,
+      "learning_rate": 2.228787058492979e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 16706600,
+      "step": 21385
+    },
+    {
+      "epoch": 44.46985446985447,
+      "grad_norm": 0.00013727330951951444,
+      "learning_rate": 2.2278111259353875e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16710440,
+      "step": 21390
+    },
+    {
+      "epoch": 44.48024948024948,
+      "grad_norm": 0.0003741932741831988,
+      "learning_rate": 2.2268352353527395e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16714376,
+      "step": 21395
+    },
+    {
+      "epoch": 44.49064449064449,
+      "grad_norm": 0.0005932883359491825,
+      "learning_rate": 2.225859386895533e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16718312,
+      "step": 21400
+    },
+    {
+      "epoch": 44.49064449064449,
+      "eval_loss": 0.42835769057273865,
+      "eval_runtime": 11.6766,
+      "eval_samples_per_second": 73.309,
+      "eval_steps_per_second": 18.327,
+      "num_input_tokens_seen": 16718312,
+      "step": 21400
+    },
+    {
+      "epoch": 44.5010395010395,
+      "grad_norm": 7.070948777254671e-05,
+      "learning_rate": 2.2248835807142525e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16722312,
+      "step": 21405
+    },
+    {
+      "epoch": 44.51143451143451,
+      "grad_norm": 0.00045730359852313995,
+      "learning_rate": 2.2239078169593826e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16726152,
+      "step": 21410
+    },
+    {
+      "epoch": 44.521829521829524,
+      "grad_norm": 0.00016503382357768714,
+      "learning_rate": 2.222932095781396e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16730088,
+      "step": 21415
+    },
+    {
+      "epoch": 44.53222453222453,
+      "grad_norm": 0.00020172391668893397,
+      "learning_rate": 2.221956417330762e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16733896,
+      "step": 21420
+    },
+    {
+      "epoch": 44.54261954261954,
+      "grad_norm": 0.00045673910062760115,
+      "learning_rate": 2.2209807817579438e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16737672,
+      "step": 21425
+    },
+    {
+      "epoch": 44.553014553014556,
+      "grad_norm": 0.00014753377763554454,
+      "learning_rate": 2.220005189213394e-05,
+      "loss": 0.0225,
+      "num_input_tokens_seen": 16741608,
+      "step": 21430
+    },
+    {
+      "epoch": 44.563409563409564,
+      "grad_norm": 0.00012318839435465634,
+      "learning_rate": 2.2190296398475624e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16745544,
+      "step": 21435
+    },
+    {
+      "epoch": 44.57380457380457,
+      "grad_norm": 0.0012117564911022782,
+      "learning_rate": 2.2180541338108926e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16749448,
+      "step": 21440
+    },
+    {
+      "epoch": 44.58419958419958,
+      "grad_norm": 0.0003279683005530387,
+      "learning_rate": 2.2170786712538176e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16753288,
+      "step": 21445
+    },
+    {
+      "epoch": 44.5945945945946,
+      "grad_norm": 0.006731686647981405,
+      "learning_rate": 2.216103252326768e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 16757192,
+      "step": 21450
+    },
+    {
+      "epoch": 44.604989604989605,
+      "grad_norm": 0.000445123907411471,
+      "learning_rate": 2.2151278771801635e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 16761096,
+      "step": 21455
+    },
+    {
+      "epoch": 44.61538461538461,
+      "grad_norm": 0.0006128852255642414,
+      "learning_rate": 2.21415254596442e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16764904,
+      "step": 21460
+    },
+    {
+      "epoch": 44.62577962577963,
+      "grad_norm": 0.0008705764776095748,
+      "learning_rate": 2.213177258829947e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16768840,
+      "step": 21465
+    },
+    {
+      "epoch": 44.63617463617464,
+      "grad_norm": 4.99073600769043,
+      "learning_rate": 2.2122020159271445e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 16772616,
+      "step": 21470
+    },
+    {
+      "epoch": 44.646569646569645,
+      "grad_norm": 1.0532747507095337,
+      "learning_rate": 2.2112268174064075e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16776360,
+      "step": 21475
+    },
+    {
+      "epoch": 44.656964656964654,
+      "grad_norm": 0.000841858854983002,
+      "learning_rate": 2.2102516634181253e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16780296,
+      "step": 21480
+    },
+    {
+      "epoch": 44.66735966735967,
+      "grad_norm": 0.00032284791814163327,
+      "learning_rate": 2.209276554112677e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16784072,
+      "step": 21485
+    },
+    {
+      "epoch": 44.67775467775468,
+      "grad_norm": 0.00020107410091441125,
+      "learning_rate": 2.2083014896404384e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16788136,
+      "step": 21490
+    },
+    {
+      "epoch": 44.688149688149686,
+      "grad_norm": 0.00022395637643057853,
+      "learning_rate": 2.207326470151775e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 16791976,
+      "step": 21495
+    },
+    {
+      "epoch": 44.6985446985447,
+      "grad_norm": 0.0008758327458053827,
+      "learning_rate": 2.2063514957970477e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16795976,
+      "step": 21500
+    },
+    {
+      "epoch": 44.70893970893971,
+      "grad_norm": 0.0002540031273383647,
+      "learning_rate": 2.205376566726611e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16799912,
+      "step": 21505
+    },
+    {
+      "epoch": 44.71933471933472,
+      "grad_norm": 0.0004052610311191529,
+      "learning_rate": 2.204401683090809e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16803976,
+      "step": 21510
+    },
+    {
+      "epoch": 44.729729729729726,
+      "grad_norm": 0.002271627075970173,
+      "learning_rate": 2.203426845039982e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16808072,
+      "step": 21515
+    },
+    {
+      "epoch": 44.74012474012474,
+      "grad_norm": 0.00021137023577466607,
+      "learning_rate": 2.202452052724464e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16812040,
+      "step": 21520
+    },
+    {
+      "epoch": 44.75051975051975,
+      "grad_norm": 0.0042997063137590885,
+      "learning_rate": 2.2014773062945777e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16815944,
+      "step": 21525
+    },
+    {
+      "epoch": 44.76091476091476,
+      "grad_norm": 0.00027936333208344877,
+      "learning_rate": 2.2005026059006427e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 16820104,
+      "step": 21530
+    },
+    {
+      "epoch": 44.771309771309774,
+      "grad_norm": 0.0005010418244637549,
+      "learning_rate": 2.1995279516929695e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16824104,
+      "step": 21535
+    },
+    {
+      "epoch": 44.78170478170478,
+      "grad_norm": 0.0015682792291045189,
+      "learning_rate": 2.1985533438218613e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16828040,
+      "step": 21540
+    },
+    {
+      "epoch": 44.79209979209979,
+      "grad_norm": 0.0003172227879986167,
+      "learning_rate": 2.197578782437617e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16831912,
+      "step": 21545
+    },
+    {
+      "epoch": 44.802494802494806,
+      "grad_norm": 0.0005640472518280149,
+      "learning_rate": 2.196604267690524e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16835848,
+      "step": 21550
+    },
+    {
+      "epoch": 44.812889812889814,
+      "grad_norm": 0.011373632587492466,
+      "learning_rate": 2.195629799730865e-05,
+      "loss": 0.012,
+      "num_input_tokens_seen": 16839720,
+      "step": 21555
+    },
+    {
+      "epoch": 44.82328482328482,
+      "grad_norm": 0.000283221248537302,
+      "learning_rate": 2.1946553787089173e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16843560,
+      "step": 21560
+    },
+    {
+      "epoch": 44.83367983367983,
+      "grad_norm": 0.00034561080974526703,
+      "learning_rate": 2.193681004774947e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 16847592,
+      "step": 21565
+    },
+    {
+      "epoch": 44.84407484407485,
+      "grad_norm": 5.701817035675049,
+      "learning_rate": 2.1927066780792154e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 16851496,
+      "step": 21570
+    },
+    {
+      "epoch": 44.854469854469855,
+      "grad_norm": 0.05701056122779846,
+      "learning_rate": 2.191732398771975e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 16855368,
+      "step": 21575
+    },
+    {
+      "epoch": 44.86486486486486,
+      "grad_norm": 0.00025376438861712813,
+      "learning_rate": 2.1907581670034725e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16859336,
+      "step": 21580
+    },
+    {
+      "epoch": 44.87525987525988,
+      "grad_norm": 0.00012226907711010426,
+      "learning_rate": 2.189783982923948e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16863080,
+      "step": 21585
+    },
+    {
+      "epoch": 44.88565488565489,
+      "grad_norm": 0.18374599516391754,
+      "learning_rate": 2.1888098466836303e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 16866856,
+      "step": 21590
+    },
+    {
+      "epoch": 44.896049896049895,
+      "grad_norm": 0.00010256265522912145,
+      "learning_rate": 2.1878357584327457e-05,
+      "loss": 0.0315,
+      "num_input_tokens_seen": 16870632,
+      "step": 21595
+    },
+    {
+      "epoch": 44.906444906444904,
+      "grad_norm": 0.00031515525188297033,
+      "learning_rate": 2.1868617183215103e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16874632,
+      "step": 21600
+    },
+    {
+      "epoch": 44.906444906444904,
+      "eval_loss": 0.4154350757598877,
+      "eval_runtime": 11.6876,
+      "eval_samples_per_second": 73.24,
+      "eval_steps_per_second": 18.31,
+      "num_input_tokens_seen": 16874632,
+      "step": 21600
+    },
+    {
+      "epoch": 44.91683991683992,
+      "grad_norm": 0.00010140907397726551,
+      "learning_rate": 2.1858877265001327e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16878568,
+      "step": 21605
+    },
+    {
+      "epoch": 44.92723492723493,
+      "grad_norm": 0.00015195923333521932,
+      "learning_rate": 2.184913783118816e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16882408,
+      "step": 21610
+    },
+    {
+      "epoch": 44.937629937629936,
+      "grad_norm": 0.002424131613224745,
+      "learning_rate": 2.1839398883277522e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 16886280,
+      "step": 21615
+    },
+    {
+      "epoch": 44.94802494802495,
+      "grad_norm": 2.8027812732034363e-05,
+      "learning_rate": 2.182966042277129e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16890312,
+      "step": 21620
+    },
+    {
+      "epoch": 44.95841995841996,
+      "grad_norm": 0.11930251866579056,
+      "learning_rate": 2.181992245117128e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 16894216,
+      "step": 21625
+    },
+    {
+      "epoch": 44.96881496881497,
+      "grad_norm": 3.725696660694666e-05,
+      "learning_rate": 2.181018496997918e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16898120,
+      "step": 21630
+    },
+    {
+      "epoch": 44.979209979209976,
+      "grad_norm": 5.525549204321578e-05,
+      "learning_rate": 2.1800447980696648e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16902088,
+      "step": 21635
+    },
+    {
+      "epoch": 44.98960498960499,
+      "grad_norm": 0.0001695760729489848,
+      "learning_rate": 2.1790711484825248e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 16905896,
+      "step": 21640
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 0.0007360472809523344,
+      "learning_rate": 2.178097548386646e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 16909856,
+      "step": 21645
+    },
+    {
+      "epoch": 45.01039501039501,
+      "grad_norm": 0.00017812965961638838,
+      "learning_rate": 2.1771239979321712e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16913760,
+      "step": 21650
+    },
+    {
+      "epoch": 45.020790020790024,
+      "grad_norm": 0.00015810044715180993,
+      "learning_rate": 2.1761504972692327e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16917632,
+      "step": 21655
+    },
+    {
+      "epoch": 45.03118503118503,
+      "grad_norm": 0.001021243049763143,
+      "learning_rate": 2.1751770465479572e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16921632,
+      "step": 21660
+    },
+    {
+      "epoch": 45.04158004158004,
+      "grad_norm": 0.0001643935829633847,
+      "learning_rate": 2.174203645918464e-05,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 16925536,
+      "step": 21665
+    },
+    {
+      "epoch": 45.05197505197505,
+      "grad_norm": 0.00012747515575028956,
+      "learning_rate": 2.1732302955308624e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 16929440,
+      "step": 21670
+    },
+    {
+      "epoch": 45.062370062370064,
+      "grad_norm": 0.0008057176019065082,
+      "learning_rate": 2.172256995535255e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16933312,
+      "step": 21675
+    },
+    {
+      "epoch": 45.07276507276507,
+      "grad_norm": 4.137010546401143e-05,
+      "learning_rate": 2.171283746081739e-05,
+      "loss": 0.0117,
+      "num_input_tokens_seen": 16937280,
+      "step": 21680
+    },
+    {
+      "epoch": 45.08316008316008,
+      "grad_norm": 0.00044369263923726976,
+      "learning_rate": 2.1703105473203988e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16941312,
+      "step": 21685
+    },
+    {
+      "epoch": 45.093555093555096,
+      "grad_norm": 0.000302577274851501,
+      "learning_rate": 2.1693373994013168e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16945280,
+      "step": 21690
+    },
+    {
+      "epoch": 45.103950103950105,
+      "grad_norm": 0.0007005178485997021,
+      "learning_rate": 2.168364302474562e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16949184,
+      "step": 21695
+    },
+    {
+      "epoch": 45.11434511434511,
+      "grad_norm": 0.002494456944987178,
+      "learning_rate": 2.167391256690199e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16953216,
+      "step": 21700
+    },
+    {
+      "epoch": 45.12474012474012,
+      "grad_norm": 0.0006915337289683521,
+      "learning_rate": 2.1664182621982855e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16957152,
+      "step": 21705
+    },
+    {
+      "epoch": 45.13513513513514,
+      "grad_norm": 0.00017493109044153243,
+      "learning_rate": 2.1654453191488673e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 16961120,
+      "step": 21710
+    },
+    {
+      "epoch": 45.145530145530145,
+      "grad_norm": 0.0007055007736198604,
+      "learning_rate": 2.1644724276919846e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16964960,
+      "step": 21715
+    },
+    {
+      "epoch": 45.15592515592515,
+      "grad_norm": 0.002687404165044427,
+      "learning_rate": 2.1634995879776715e-05,
+      "loss": 0.007,
+      "num_input_tokens_seen": 16968896,
+      "step": 21720
+    },
+    {
+      "epoch": 45.16632016632017,
+      "grad_norm": 0.007541349157691002,
+      "learning_rate": 2.162526800155949e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16972736,
+      "step": 21725
+    },
+    {
+      "epoch": 45.17671517671518,
+      "grad_norm": 0.0010953597957268357,
+      "learning_rate": 2.1615540643768363e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16976608,
+      "step": 21730
+    },
+    {
+      "epoch": 45.187110187110186,
+      "grad_norm": 0.0022138592321425676,
+      "learning_rate": 2.160581380790339e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 16980576,
+      "step": 21735
+    },
+    {
+      "epoch": 45.197505197505194,
+      "grad_norm": 0.0008484123973175883,
+      "learning_rate": 2.1596087495464586e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16984416,
+      "step": 21740
+    },
+    {
+      "epoch": 45.20790020790021,
+      "grad_norm": 0.00018943999020848423,
+      "learning_rate": 2.1586361707951866e-05,
+      "loss": 0.0074,
+      "num_input_tokens_seen": 16988192,
+      "step": 21745
+    },
+    {
+      "epoch": 45.21829521829522,
+      "grad_norm": 0.0012969373492524028,
+      "learning_rate": 2.157663644686507e-05,
+      "loss": 0.0006,
+      "num_input_tokens_seen": 16992128,
+      "step": 21750
+    },
+    {
+      "epoch": 45.228690228690226,
+      "grad_norm": 0.00026516718207858503,
+      "learning_rate": 2.156691171370396e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 16996160,
+      "step": 21755
+    },
+    {
+      "epoch": 45.23908523908524,
+      "grad_norm": 0.001626188401132822,
+      "learning_rate": 2.1557187509968195e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 17000064,
+      "step": 21760
+    },
+    {
+      "epoch": 45.24948024948025,
+      "grad_norm": 0.00041352983680553734,
+      "learning_rate": 2.1547463837157382e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17004096,
+      "step": 21765
+    },
+    {
+      "epoch": 45.25987525987526,
+      "grad_norm": 0.00013040391786489636,
+      "learning_rate": 2.1537740696771045e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17007904,
+      "step": 21770
+    },
+    {
+      "epoch": 45.270270270270274,
+      "grad_norm": 0.0004566200659610331,
+      "learning_rate": 2.1528018090308587e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17011872,
+      "step": 21775
+    },
+    {
+      "epoch": 45.28066528066528,
+      "grad_norm": 0.0006969392998144031,
+      "learning_rate": 2.151829601926938e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17016000,
+      "step": 21780
+    },
+    {
+      "epoch": 45.29106029106029,
+      "grad_norm": 0.00010761382145574316,
+      "learning_rate": 2.1508574485152684e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17019776,
+      "step": 21785
+    },
+    {
+      "epoch": 45.3014553014553,
+      "grad_norm": 0.0003633410087786615,
+      "learning_rate": 2.1498853489457667e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 17023776,
+      "step": 21790
+    },
+    {
+      "epoch": 45.311850311850314,
+      "grad_norm": 0.006307696457952261,
+      "learning_rate": 2.1489133033683455e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17027648,
+      "step": 21795
+    },
+    {
+      "epoch": 45.32224532224532,
+      "grad_norm": 0.00021563727932516485,
+      "learning_rate": 2.1479413119329038e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17031680,
+      "step": 21800
+    },
+    {
+      "epoch": 45.32224532224532,
+      "eval_loss": 0.4509370028972626,
+      "eval_runtime": 11.6898,
+      "eval_samples_per_second": 73.226,
+      "eval_steps_per_second": 18.307,
+      "num_input_tokens_seen": 17031680,
+      "step": 21800
+    },
+    {
+      "epoch": 45.33264033264033,
+      "grad_norm": 0.00016149439034052193,
+      "learning_rate": 2.1469693747893355e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 17035808,
+      "step": 21805
+    },
+    {
+      "epoch": 45.343035343035346,
+      "grad_norm": 0.0002542481233831495,
+      "learning_rate": 2.1459974920875274e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17039776,
+      "step": 21810
+    },
+    {
+      "epoch": 45.353430353430355,
+      "grad_norm": 0.00042736949399113655,
+      "learning_rate": 2.145025663977354e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17043552,
+      "step": 21815
+    },
+    {
+      "epoch": 45.36382536382536,
+      "grad_norm": 0.0044482676312327385,
+      "learning_rate": 2.1440538906086844e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17047456,
+      "step": 21820
+    },
+    {
+      "epoch": 45.37422037422037,
+      "grad_norm": 0.00016355862317141145,
+      "learning_rate": 2.1430821721313782e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17051200,
+      "step": 21825
+    },
+    {
+      "epoch": 45.38461538461539,
+      "grad_norm": 0.00036660669138655066,
+      "learning_rate": 2.142110508695286e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17055104,
+      "step": 21830
+    },
+    {
+      "epoch": 45.395010395010395,
+      "grad_norm": 0.0002879988751374185,
+      "learning_rate": 2.1411389004502515e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17058912,
+      "step": 21835
+    },
+    {
+      "epoch": 45.4054054054054,
+      "grad_norm": 9.760905231814831e-05,
+      "learning_rate": 2.140167347546107e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 17062912,
+      "step": 21840
+    },
+    {
+      "epoch": 45.41580041580042,
+      "grad_norm": 0.0006169471889734268,
+      "learning_rate": 2.1391958501326793e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 17066880,
+      "step": 21845
+    },
+    {
+      "epoch": 45.42619542619543,
+      "grad_norm": 0.00028177336207590997,
+      "learning_rate": 2.1382244083597873e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17070816,
+      "step": 21850
+    },
+    {
+      "epoch": 45.436590436590436,
+      "grad_norm": 0.0005228181835263968,
+      "learning_rate": 2.137253022377237e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 17074688,
+      "step": 21855
+    },
+    {
+      "epoch": 45.446985446985444,
+      "grad_norm": 0.0005424906848929822,
+      "learning_rate": 2.136281692334829e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17078688,
+      "step": 21860
+    },
+    {
+      "epoch": 45.45738045738046,
+      "grad_norm": 0.00013733626110479236,
+      "learning_rate": 2.135310418382356e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17082592,
+      "step": 21865
+    },
+    {
+      "epoch": 45.46777546777547,
+      "grad_norm": 0.0001020021372823976,
+      "learning_rate": 2.134339200669598e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17086496,
+      "step": 21870
+    },
+    {
+      "epoch": 45.478170478170476,
+      "grad_norm": 0.0002977371623273939,
+      "learning_rate": 2.133368039346331e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17090496,
+      "step": 21875
+    },
+    {
+      "epoch": 45.48856548856549,
+      "grad_norm": 7.045474194455892e-05,
+      "learning_rate": 2.1323969345623195e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17094656,
+      "step": 21880
+    },
+    {
+      "epoch": 45.4989604989605,
+      "grad_norm": 5.101331407786347e-05,
+      "learning_rate": 2.1314258864673207e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17098528,
+      "step": 21885
+    },
+    {
+      "epoch": 45.50935550935551,
+      "grad_norm": 0.00554515328258276,
+      "learning_rate": 2.130454895211082e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17102528,
+      "step": 21890
+    },
+    {
+      "epoch": 45.51975051975052,
+      "grad_norm": 9.332630725111812e-05,
+      "learning_rate": 2.129483960943342e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17106400,
+      "step": 21895
+    },
+    {
+      "epoch": 45.53014553014553,
+      "grad_norm": 2.7999949452350847e-05,
+      "learning_rate": 2.128513083813831e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17110112,
+      "step": 21900
+    },
+    {
+      "epoch": 45.54054054054054,
+      "grad_norm": 0.00036137716961093247,
+      "learning_rate": 2.1275422639722724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17114080,
+      "step": 21905
+    },
+    {
+      "epoch": 45.55093555093555,
+      "grad_norm": 0.004016959108412266,
+      "learning_rate": 2.126571501568376e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17118016,
+      "step": 21910
+    },
+    {
+      "epoch": 45.561330561330564,
+      "grad_norm": 0.00010347493662266061,
+      "learning_rate": 2.1256007967518478e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17121856,
+      "step": 21915
+    },
+    {
+      "epoch": 45.57172557172557,
+      "grad_norm": 0.0008717462769709527,
+      "learning_rate": 2.124630149672381e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17125888,
+      "step": 21920
+    },
+    {
+      "epoch": 45.58212058212058,
+      "grad_norm": 0.00011781774810515344,
+      "learning_rate": 2.1236595604796624e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17129760,
+      "step": 21925
+    },
+    {
+      "epoch": 45.59251559251559,
+      "grad_norm": 0.0025370761286467314,
+      "learning_rate": 2.1226890293233693e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17133600,
+      "step": 21930
+    },
+    {
+      "epoch": 45.602910602910605,
+      "grad_norm": 3.079451926168986e-05,
+      "learning_rate": 2.1217185563531694e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 17137408,
+      "step": 21935
+    },
+    {
+      "epoch": 45.61330561330561,
+      "grad_norm": 0.00013584572297986597,
+      "learning_rate": 2.120748141718721e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17141344,
+      "step": 21940
+    },
+    {
+      "epoch": 45.62370062370062,
+      "grad_norm": 0.003888380480930209,
+      "learning_rate": 2.1197777855696765e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17145152,
+      "step": 21945
+    },
+    {
+      "epoch": 45.63409563409564,
+      "grad_norm": 0.00015623362560290843,
+      "learning_rate": 2.1188074880556746e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17149120,
+      "step": 21950
+    },
+    {
+      "epoch": 45.644490644490645,
+      "grad_norm": 0.00017940899124369025,
+      "learning_rate": 2.1178372493263495e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17152992,
+      "step": 21955
+    },
+    {
+      "epoch": 45.65488565488565,
+      "grad_norm": 9.47152148000896e-05,
+      "learning_rate": 2.116867069531322e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17156960,
+      "step": 21960
+    },
+    {
+      "epoch": 45.66528066528066,
+      "grad_norm": 0.13927777111530304,
+      "learning_rate": 2.1158969488202073e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 17161024,
+      "step": 21965
+    },
+    {
+      "epoch": 45.67567567567568,
+      "grad_norm": 0.00021011792705394328,
+      "learning_rate": 2.114926887342611e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 17164864,
+      "step": 21970
+    },
+    {
+      "epoch": 45.686070686070686,
+      "grad_norm": 0.000153936751303263,
+      "learning_rate": 2.113956885248127e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17168704,
+      "step": 21975
+    },
+    {
+      "epoch": 45.696465696465694,
+      "grad_norm": 4.2876785300904885e-05,
+      "learning_rate": 2.112986942686342e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17172512,
+      "step": 21980
+    },
+    {
+      "epoch": 45.70686070686071,
+      "grad_norm": 0.0014333154540508986,
+      "learning_rate": 2.112017059806835e-05,
+      "loss": 0.0169,
+      "num_input_tokens_seen": 17176448,
+      "step": 21985
+    },
+    {
+      "epoch": 45.71725571725572,
+      "grad_norm": 0.0022358056157827377,
+      "learning_rate": 2.1110472367591724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17180448,
+      "step": 21990
+    },
+    {
+      "epoch": 45.727650727650726,
+      "grad_norm": 4.635427103494294e-05,
+      "learning_rate": 2.1100774736929145e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17184288,
+      "step": 21995
+    },
+    {
+      "epoch": 45.73804573804574,
+      "grad_norm": 0.0025484280195087194,
+      "learning_rate": 2.10910777075761e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17188288,
+      "step": 22000
+    },
+    {
+      "epoch": 45.73804573804574,
+      "eval_loss": 0.4369456470012665,
+      "eval_runtime": 11.7137,
+      "eval_samples_per_second": 73.077,
+      "eval_steps_per_second": 18.269,
+      "num_input_tokens_seen": 17188288,
+      "step": 22000
+    },
+    {
+      "epoch": 45.74844074844075,
+      "grad_norm": 0.00014791334979236126,
+      "learning_rate": 2.108138128102799e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17192064,
+      "step": 22005
+    },
+    {
+      "epoch": 45.75883575883576,
+      "grad_norm": 1.4160209502733778e-05,
+      "learning_rate": 2.107168545878014e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17196128,
+      "step": 22010
+    },
+    {
+      "epoch": 45.76923076923077,
+      "grad_norm": 0.0001980109082069248,
+      "learning_rate": 2.106199024232775e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17200032,
+      "step": 22015
+    },
+    {
+      "epoch": 45.77962577962578,
+      "grad_norm": 2.6707330107456073e-05,
+      "learning_rate": 2.105229563316595e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17203968,
+      "step": 22020
+    },
+    {
+      "epoch": 45.79002079002079,
+      "grad_norm": 0.00017502847185824066,
+      "learning_rate": 2.1042601632789784e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 17207968,
+      "step": 22025
+    },
+    {
+      "epoch": 45.8004158004158,
+      "grad_norm": 0.00014455665950663388,
+      "learning_rate": 2.103290824269417e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17212000,
+      "step": 22030
+    },
+    {
+      "epoch": 45.810810810810814,
+      "grad_norm": 0.0009767632000148296,
+      "learning_rate": 2.1023215464373965e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17215904,
+      "step": 22035
+    },
+    {
+      "epoch": 45.82120582120582,
+      "grad_norm": 0.00010659085819497705,
+      "learning_rate": 2.1013523299323908e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17219744,
+      "step": 22040
+    },
+    {
+      "epoch": 45.83160083160083,
+      "grad_norm": 0.12021437287330627,
+      "learning_rate": 2.1003831749038654e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 17223808,
+      "step": 22045
+    },
+    {
+      "epoch": 45.84199584199584,
+      "grad_norm": 0.0035987647715955973,
+      "learning_rate": 2.099414081501277e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17227712,
+      "step": 22050
+    },
+    {
+      "epoch": 45.852390852390855,
+      "grad_norm": 3.0094670364633203e-05,
+      "learning_rate": 2.09844504987407e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17231616,
+      "step": 22055
+    },
+    {
+      "epoch": 45.86278586278586,
+      "grad_norm": 0.0001599674578756094,
+      "learning_rate": 2.097476080171683e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17235488,
+      "step": 22060
+    },
+    {
+      "epoch": 45.87318087318087,
+      "grad_norm": 3.9122693124227226e-05,
+      "learning_rate": 2.0965071725435436e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17239392,
+      "step": 22065
+    },
+    {
+      "epoch": 45.88357588357589,
+      "grad_norm": 7.423113129334524e-05,
+      "learning_rate": 2.0955383271390684e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17243264,
+      "step": 22070
+    },
+    {
+      "epoch": 45.893970893970895,
+      "grad_norm": 0.00048299613990820944,
+      "learning_rate": 2.094569544107666e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17247168,
+      "step": 22075
+    },
+    {
+      "epoch": 45.9043659043659,
+      "grad_norm": 0.003933286760002375,
+      "learning_rate": 2.093600823598735e-05,
+      "loss": 0.0016,
+      "num_input_tokens_seen": 17251104,
+      "step": 22080
+    },
+    {
+      "epoch": 45.91476091476091,
+      "grad_norm": 0.0006282374961301684,
+      "learning_rate": 2.092632165761663e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17255040,
+      "step": 22085
+    },
+    {
+      "epoch": 45.92515592515593,
+      "grad_norm": 8.145610627252609e-05,
+      "learning_rate": 2.091663570745832e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17259104,
+      "step": 22090
+    },
+    {
+      "epoch": 45.935550935550935,
+      "grad_norm": 2.9307862860150635e-05,
+      "learning_rate": 2.0906950387006086e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17262944,
+      "step": 22095
+    },
+    {
+      "epoch": 45.945945945945944,
+      "grad_norm": 0.0004686945176217705,
+      "learning_rate": 2.0897265697753543e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 17266784,
+      "step": 22100
+    },
+    {
+      "epoch": 45.95634095634096,
+      "grad_norm": 2.2613268811255693e-05,
+      "learning_rate": 2.088758164119419e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 17270656,
+      "step": 22105
+    },
+    {
+      "epoch": 45.96673596673597,
+      "grad_norm": 7.396617002086714e-05,
+      "learning_rate": 2.0877898218821428e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17274496,
+      "step": 22110
+    },
+    {
+      "epoch": 45.977130977130976,
+      "grad_norm": 7.019987242529169e-05,
+      "learning_rate": 2.0868215432128565e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17278368,
+      "step": 22115
+    },
+    {
+      "epoch": 45.987525987525984,
+      "grad_norm": 0.004932828713208437,
+      "learning_rate": 2.0858533282608796e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17282176,
+      "step": 22120
+    },
+    {
+      "epoch": 45.997920997921,
+      "grad_norm": 0.000545042974408716,
+      "learning_rate": 2.084885177175524e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17286016,
+      "step": 22125
+    },
+    {
+      "epoch": 46.00831600831601,
+      "grad_norm": 0.001012853579595685,
+      "learning_rate": 2.0839170901060917e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17289912,
+      "step": 22130
+    },
+    {
+      "epoch": 46.018711018711016,
+      "grad_norm": 0.00014478586672339588,
+      "learning_rate": 2.082949067201872e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17293848,
+      "step": 22135
+    },
+    {
+      "epoch": 46.02910602910603,
+      "grad_norm": 0.00037781844730488956,
+      "learning_rate": 2.0819811086121475e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 17297816,
+      "step": 22140
+    },
+    {
+      "epoch": 46.03950103950104,
+      "grad_norm": 5.384023461374454e-05,
+      "learning_rate": 2.08101321448619e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17301816,
+      "step": 22145
+    },
+    {
+      "epoch": 46.04989604989605,
+      "grad_norm": 0.0004442780918907374,
+      "learning_rate": 2.080045384973259e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17305656,
+      "step": 22150
+    },
+    {
+      "epoch": 46.06029106029106,
+      "grad_norm": 0.06067771092057228,
+      "learning_rate": 2.0790776202226082e-05,
+      "loss": 0.0007,
+      "num_input_tokens_seen": 17309528,
+      "step": 22155
+    },
+    {
+      "epoch": 46.07068607068607,
+      "grad_norm": 4.5731001591775566e-05,
+      "learning_rate": 2.078109920383477e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 17313528,
+      "step": 22160
+    },
+    {
+      "epoch": 46.08108108108108,
+      "grad_norm": 0.28359636664390564,
+      "learning_rate": 2.0771422856050978e-05,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 17317560,
+      "step": 22165
+    },
+    {
+      "epoch": 46.09147609147609,
+      "grad_norm": 6.543874042108655e-05,
+      "learning_rate": 2.076174716036693e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17321368,
+      "step": 22170
+    },
+    {
+      "epoch": 46.101871101871104,
+      "grad_norm": 0.004627638962119818,
+      "learning_rate": 2.075207211827472e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17325336,
+      "step": 22175
+    },
+    {
+      "epoch": 46.11226611226611,
+      "grad_norm": 0.0001294106914428994,
+      "learning_rate": 2.074239773126638e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17329272,
+      "step": 22180
+    },
+    {
+      "epoch": 46.12266112266112,
+      "grad_norm": 0.0008755957242101431,
+      "learning_rate": 2.073272400083382e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17333208,
+      "step": 22185
+    },
+    {
+      "epoch": 46.13305613305613,
+      "grad_norm": 0.001678756671026349,
+      "learning_rate": 2.072305092846883e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 17337016,
+      "step": 22190
+    },
+    {
+      "epoch": 46.143451143451145,
+      "grad_norm": 0.0009821596322581172,
+      "learning_rate": 2.0713378515663152e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17341112,
+      "step": 22195
+    },
+    {
+      "epoch": 46.15384615384615,
+      "grad_norm": 0.00014464334526564926,
+      "learning_rate": 2.070370676390836e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17345048,
+      "step": 22200
+    },
+    {
+      "epoch": 46.15384615384615,
+      "eval_loss": 0.5120408535003662,
+      "eval_runtime": 11.7187,
+      "eval_samples_per_second": 73.046,
+      "eval_steps_per_second": 18.261,
+      "num_input_tokens_seen": 17345048,
+      "step": 22200
+    },
+    {
+      "epoch": 46.16424116424116,
+      "grad_norm": 21.3585147857666,
+      "learning_rate": 2.0694035674695974e-05,
+      "loss": 0.0384,
+      "num_input_tokens_seen": 17348920,
+      "step": 22205
+    },
+    {
+      "epoch": 46.17463617463618,
+      "grad_norm": 0.0003941717150155455,
+      "learning_rate": 2.0684365249517416e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17352984,
+      "step": 22210
+    },
+    {
+      "epoch": 46.185031185031185,
+      "grad_norm": 0.00011299164179945365,
+      "learning_rate": 2.067469548986396e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17356920,
+      "step": 22215
+    },
+    {
+      "epoch": 46.195426195426194,
+      "grad_norm": 0.008597766980528831,
+      "learning_rate": 2.066502639722681e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17360760,
+      "step": 22220
+    },
+    {
+      "epoch": 46.20582120582121,
+      "grad_norm": 3.2148022000910714e-05,
+      "learning_rate": 2.065535797309708e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 17364504,
+      "step": 22225
+    },
+    {
+      "epoch": 46.21621621621622,
+      "grad_norm": 0.0009251331794075668,
+      "learning_rate": 2.0645690218965736e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17368440,
+      "step": 22230
+    },
+    {
+      "epoch": 46.226611226611226,
+      "grad_norm": 0.0001060314680216834,
+      "learning_rate": 2.063602313632369e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 17372440,
+      "step": 22235
+    },
+    {
+      "epoch": 46.237006237006234,
+      "grad_norm": 0.00014165989705361426,
+      "learning_rate": 2.0626356726661704e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17376376,
+      "step": 22240
+    },
+    {
+      "epoch": 46.24740124740125,
+      "grad_norm": 0.00020899806986562908,
+      "learning_rate": 2.0616690991470477e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 17380312,
+      "step": 22245
+    },
+    {
+      "epoch": 46.25779625779626,
+      "grad_norm": 0.00048443302512168884,
+      "learning_rate": 2.0607025932240595e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17384376,
+      "step": 22250
+    },
+    {
+      "epoch": 46.268191268191266,
+      "grad_norm": 0.00013916939496994019,
+      "learning_rate": 2.059736155046251e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17388216,
+      "step": 22255
+    },
+    {
+      "epoch": 46.27858627858628,
+      "grad_norm": 6.259568908717483e-05,
+      "learning_rate": 2.0587697847626603e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17392088,
+      "step": 22260
+    },
+    {
+      "epoch": 46.28898128898129,
+      "grad_norm": 0.0006796177476644516,
+      "learning_rate": 2.057803482522314e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17395960,
+      "step": 22265
+    },
+    {
+      "epoch": 46.2993762993763,
+      "grad_norm": 6.277975626289845e-05,
+      "learning_rate": 2.056837248474227e-05,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 17399832,
+      "step": 22270
+    },
+    {
+      "epoch": 46.30977130977131,
+      "grad_norm": 0.000250400131335482,
+      "learning_rate": 2.0558710827674064e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 17403736,
+      "step": 22275
+    },
+    {
+      "epoch": 46.32016632016632,
+      "grad_norm": 5.196855272515677e-05,
+      "learning_rate": 2.054904985550845e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17407640,
+      "step": 22280
+    },
+    {
+      "epoch": 46.33056133056133,
+      "grad_norm": 0.00010980484512401745,
+      "learning_rate": 2.0539389569735287e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 17411608,
+      "step": 22285
+    },
+    {
+      "epoch": 46.34095634095634,
+      "grad_norm": 5.2475210395641625e-05,
+      "learning_rate": 2.052972997184431e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17415512,
+      "step": 22290
+    },
+    {
+      "epoch": 46.351351351351354,
+      "grad_norm": 5.719993350794539e-05,
+      "learning_rate": 2.0520071063325146e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17419320,
+      "step": 22295
+    },
+    {
+      "epoch": 46.36174636174636,
+      "grad_norm": 0.00036891462514176965,
+      "learning_rate": 2.051041284566732e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17423192,
+      "step": 22300
+    },
+    {
+      "epoch": 46.37214137214137,
+      "grad_norm": 0.00016399317246396095,
+      "learning_rate": 2.050075532036026e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 17427256,
+      "step": 22305
+    },
+    {
+      "epoch": 46.38253638253638,
+      "grad_norm": 0.0005037639057263732,
+      "learning_rate": 2.0491098488893264e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17431192,
+      "step": 22310
+    },
+    {
+      "epoch": 46.392931392931395,
+      "grad_norm": 0.0001075132458936423,
+      "learning_rate": 2.0481442352755546e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17435064,
+      "step": 22315
+    },
+    {
+      "epoch": 46.4033264033264,
+      "grad_norm": 0.00023783359210938215,
+      "learning_rate": 2.0471786913436198e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17438968,
+      "step": 22320
+    },
+    {
+      "epoch": 46.41372141372141,
+      "grad_norm": 8.000004891073331e-05,
+      "learning_rate": 2.0462132172424218e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17442840,
+      "step": 22325
+    },
+    {
+      "epoch": 46.42411642411643,
+      "grad_norm": 0.0028397399000823498,
+      "learning_rate": 2.0452478131208484e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17446712,
+      "step": 22330
+    },
+    {
+      "epoch": 46.434511434511435,
+      "grad_norm": 0.00023335030709858984,
+      "learning_rate": 2.0442824791277765e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17450648,
+      "step": 22335
+    },
+    {
+      "epoch": 46.444906444906444,
+      "grad_norm": 0.00023680638696532696,
+      "learning_rate": 2.0433172154120727e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17454616,
+      "step": 22340
+    },
+    {
+      "epoch": 46.45530145530145,
+      "grad_norm": 0.0002487488673068583,
+      "learning_rate": 2.0423520221225947e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17458488,
+      "step": 22345
+    },
+    {
+      "epoch": 46.46569646569647,
+      "grad_norm": 9.301076352130622e-05,
+      "learning_rate": 2.0413868994081848e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17462456,
+      "step": 22350
+    },
+    {
+      "epoch": 46.476091476091476,
+      "grad_norm": 0.00015086799976415932,
+      "learning_rate": 2.0404218474176795e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17466296,
+      "step": 22355
+    },
+    {
+      "epoch": 46.486486486486484,
+      "grad_norm": 2.357303128519561e-05,
+      "learning_rate": 2.0394568662999002e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17470104,
+      "step": 22360
+    },
+    {
+      "epoch": 46.4968814968815,
+      "grad_norm": 0.09325283020734787,
+      "learning_rate": 2.0384919562036593e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 17474040,
+      "step": 22365
+    },
+    {
+      "epoch": 46.50727650727651,
+      "grad_norm": 0.00017864926485344768,
+      "learning_rate": 2.0375271172777593e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17478008,
+      "step": 22370
+    },
+    {
+      "epoch": 46.517671517671516,
+      "grad_norm": 0.00013914886221755296,
+      "learning_rate": 2.0365623496709885e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17481880,
+      "step": 22375
+    },
+    {
+      "epoch": 46.528066528066525,
+      "grad_norm": 9.57100055529736e-05,
+      "learning_rate": 2.0355976535321283e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 17485816,
+      "step": 22380
+    },
+    {
+      "epoch": 46.53846153846154,
+      "grad_norm": 0.0006630280986428261,
+      "learning_rate": 2.034633029009945e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17489784,
+      "step": 22385
+    },
+    {
+      "epoch": 46.54885654885655,
+      "grad_norm": 5.270799738354981e-05,
+      "learning_rate": 2.0336684762531972e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17493656,
+      "step": 22390
+    },
+    {
+      "epoch": 46.55925155925156,
+      "grad_norm": 0.00010848091915249825,
+      "learning_rate": 2.032703995410631e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17497592,
+      "step": 22395
+    },
+    {
+      "epoch": 46.56964656964657,
+      "grad_norm": 7.002206257311627e-05,
+      "learning_rate": 2.031739586630981e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17501560,
+      "step": 22400
+    },
+    {
+      "epoch": 46.56964656964657,
+      "eval_loss": 0.4886241555213928,
+      "eval_runtime": 11.7009,
+      "eval_samples_per_second": 73.157,
+      "eval_steps_per_second": 18.289,
+      "num_input_tokens_seen": 17501560,
+      "step": 22400
+    },
+    {
+      "epoch": 46.58004158004158,
+      "grad_norm": 0.00018221483333036304,
+      "learning_rate": 2.0307752500629707e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17505624,
+      "step": 22405
+    },
+    {
+      "epoch": 46.59043659043659,
+      "grad_norm": 2.1368123270804062e-05,
+      "learning_rate": 2.0298109858553144e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17509496,
+      "step": 22410
+    },
+    {
+      "epoch": 46.6008316008316,
+      "grad_norm": 0.25658711791038513,
+      "learning_rate": 2.028846794156712e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 17513368,
+      "step": 22415
+    },
+    {
+      "epoch": 46.61122661122661,
+      "grad_norm": 4.452955909073353e-05,
+      "learning_rate": 2.027882675115856e-05,
+      "loss": 0.0061,
+      "num_input_tokens_seen": 17517304,
+      "step": 22420
+    },
+    {
+      "epoch": 46.62162162162162,
+      "grad_norm": 0.010932697914540768,
+      "learning_rate": 2.026918628881423e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17521208,
+      "step": 22425
+    },
+    {
+      "epoch": 46.63201663201663,
+      "grad_norm": 0.0003565509105101228,
+      "learning_rate": 2.0259546556020833e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17525016,
+      "step": 22430
+    },
+    {
+      "epoch": 46.642411642411645,
+      "grad_norm": 8.650534437038004e-05,
+      "learning_rate": 2.024990755426493e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17528888,
+      "step": 22435
+    },
+    {
+      "epoch": 46.65280665280665,
+      "grad_norm": 2.4507653506589122e-05,
+      "learning_rate": 2.0240269285032975e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 17532856,
+      "step": 22440
+    },
+    {
+      "epoch": 46.66320166320166,
+      "grad_norm": 6.504787597805262e-05,
+      "learning_rate": 2.0230631749811306e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17536664,
+      "step": 22445
+    },
+    {
+      "epoch": 46.67359667359668,
+      "grad_norm": 0.0005638066795654595,
+      "learning_rate": 2.0220994950086162e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17540568,
+      "step": 22450
+    },
+    {
+      "epoch": 46.683991683991685,
+      "grad_norm": 7.510359137086198e-05,
+      "learning_rate": 2.021135888734365e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17544408,
+      "step": 22455
+    },
+    {
+      "epoch": 46.694386694386694,
+      "grad_norm": 0.004764720797538757,
+      "learning_rate": 2.0201723563069783e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17548216,
+      "step": 22460
+    },
+    {
+      "epoch": 46.7047817047817,
+      "grad_norm": 3.412998194107786e-05,
+      "learning_rate": 2.0192088978750433e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17552120,
+      "step": 22465
+    },
+    {
+      "epoch": 46.71517671517672,
+      "grad_norm": 0.0025652172043919563,
+      "learning_rate": 2.0182455135871385e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17555928,
+      "step": 22470
+    },
+    {
+      "epoch": 46.725571725571726,
+      "grad_norm": 6.68809880153276e-05,
+      "learning_rate": 2.0172822035918305e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17559800,
+      "step": 22475
+    },
+    {
+      "epoch": 46.735966735966734,
+      "grad_norm": 0.0002684799546841532,
+      "learning_rate": 2.016318968037671e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17563896,
+      "step": 22480
+    },
+    {
+      "epoch": 46.74636174636175,
+      "grad_norm": 0.00017241215391550213,
+      "learning_rate": 2.015355807073206e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17567704,
+      "step": 22485
+    },
+    {
+      "epoch": 46.75675675675676,
+      "grad_norm": 0.004674103111028671,
+      "learning_rate": 2.0143927208469664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17571640,
+      "step": 22490
+    },
+    {
+      "epoch": 46.767151767151766,
+      "grad_norm": 0.11693982779979706,
+      "learning_rate": 2.0134297095074708e-05,
+      "loss": 0.0004,
+      "num_input_tokens_seen": 17575480,
+      "step": 22495
+    },
+    {
+      "epoch": 46.777546777546775,
+      "grad_norm": 5.734618389396928e-05,
+      "learning_rate": 2.0124667732032297e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 17579384,
+      "step": 22500
+    },
+    {
+      "epoch": 46.78794178794179,
+      "grad_norm": 0.00020167302864138037,
+      "learning_rate": 2.011503912082738e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17583288,
+      "step": 22505
+    },
+    {
+      "epoch": 46.7983367983368,
+      "grad_norm": 3.628330159699544e-05,
+      "learning_rate": 2.0105411262944823e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17587096,
+      "step": 22510
+    },
+    {
+      "epoch": 46.80873180873181,
+      "grad_norm": 0.00012383806461002678,
+      "learning_rate": 2.0095784159869366e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17591000,
+      "step": 22515
+    },
+    {
+      "epoch": 46.81912681912682,
+      "grad_norm": 0.0002904472639784217,
+      "learning_rate": 2.0086157813085608e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17594936,
+      "step": 22520
+    },
+    {
+      "epoch": 46.82952182952183,
+      "grad_norm": 0.003952148370444775,
+      "learning_rate": 2.0076532224078068e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 17598936,
+      "step": 22525
+    },
+    {
+      "epoch": 46.83991683991684,
+      "grad_norm": 0.00029844645177945495,
+      "learning_rate": 2.0066907394331142e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 17602744,
+      "step": 22530
+    },
+    {
+      "epoch": 46.85031185031185,
+      "grad_norm": 9.007832704810426e-05,
+      "learning_rate": 2.0057283325329077e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17606584,
+      "step": 22535
+    },
+    {
+      "epoch": 46.86070686070686,
+      "grad_norm": 5.551103095058352e-05,
+      "learning_rate": 2.0047660018556047e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17610392,
+      "step": 22540
+    },
+    {
+      "epoch": 46.87110187110187,
+      "grad_norm": 0.0008742660284042358,
+      "learning_rate": 2.0038037475496075e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 17614328,
+      "step": 22545
+    },
+    {
+      "epoch": 46.88149688149688,
+      "grad_norm": 6.915597623446956e-05,
+      "learning_rate": 2.0028415697633073e-05,
+      "loss": 0.0068,
+      "num_input_tokens_seen": 17618232,
+      "step": 22550
+    },
+    {
+      "epoch": 46.891891891891895,
+      "grad_norm": 3.902612661477178e-05,
+      "learning_rate": 2.0018794686450858e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17622072,
+      "step": 22555
+    },
+    {
+      "epoch": 46.9022869022869,
+      "grad_norm": 0.00031969172414392233,
+      "learning_rate": 2.0009174443433088e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17626008,
+      "step": 22560
+    },
+    {
+      "epoch": 46.91268191268191,
+      "grad_norm": 0.00015387946041300893,
+      "learning_rate": 1.999955497006334e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17629848,
+      "step": 22565
+    },
+    {
+      "epoch": 46.92307692307692,
+      "grad_norm": 0.0005416994099505246,
+      "learning_rate": 1.9989936267825067e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17633720,
+      "step": 22570
+    },
+    {
+      "epoch": 46.933471933471935,
+      "grad_norm": 0.003948270808905363,
+      "learning_rate": 1.9980318338201572e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17637784,
+      "step": 22575
+    },
+    {
+      "epoch": 46.943866943866944,
+      "grad_norm": 2.4741957531659864e-05,
+      "learning_rate": 1.997070118267607e-05,
+      "loss": 0.0003,
+      "num_input_tokens_seen": 17641624,
+      "step": 22580
+    },
+    {
+      "epoch": 46.95426195426195,
+      "grad_norm": 0.00021226101671345532,
+      "learning_rate": 1.9961084802731654e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17645400,
+      "step": 22585
+    },
+    {
+      "epoch": 46.96465696465697,
+      "grad_norm": 0.001214460120536387,
+      "learning_rate": 1.9951469199851273e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17649400,
+      "step": 22590
+    },
+    {
+      "epoch": 46.975051975051976,
+      "grad_norm": 0.00013326705084182322,
+      "learning_rate": 1.99418543755178e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 17653496,
+      "step": 22595
+    },
+    {
+      "epoch": 46.985446985446984,
+      "grad_norm": 1.7250429664272815e-05,
+      "learning_rate": 1.9932240331213936e-05,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 17657336,
+      "step": 22600
+    },
+    {
+      "epoch": 46.985446985446984,
+      "eval_loss": 0.5075212121009827,
+      "eval_runtime": 11.6969,
+      "eval_samples_per_second": 73.182,
+      "eval_steps_per_second": 18.295,
+      "num_input_tokens_seen": 17657336,
+      "step": 22600
+    },
+    {
+      "epoch": 46.99584199584199,
+      "grad_norm": 0.00017508056771475822,
+      "learning_rate": 1.9922627068422297e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17661272,
+      "step": 22605
+    },
+    {
+      "epoch": 47.00623700623701,
+      "grad_norm": 0.0003663891402538866,
+      "learning_rate": 1.991301458862538e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17665032,
+      "step": 22610
+    },
+    {
+      "epoch": 47.016632016632016,
+      "grad_norm": 0.00017397466581314802,
+      "learning_rate": 1.9903402893305536e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 17668968,
+      "step": 22615
+    },
+    {
+      "epoch": 47.027027027027025,
+      "grad_norm": 3.5976827348349616e-05,
+      "learning_rate": 1.9893791983945016e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17673000,
+      "step": 22620
+    },
+    {
+      "epoch": 47.03742203742204,
+      "grad_norm": 0.0008050675387494266,
+      "learning_rate": 1.988418186202594e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17676968,
+      "step": 22625
+    },
+    {
+      "epoch": 47.04781704781705,
+      "grad_norm": 0.0001437910395907238,
+      "learning_rate": 1.98745725290303e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17680904,
+      "step": 22630
+    },
+    {
+      "epoch": 47.05821205821206,
+      "grad_norm": 9.53032576944679e-05,
+      "learning_rate": 1.986496398644e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17684840,
+      "step": 22635
+    },
+    {
+      "epoch": 47.06860706860707,
+      "grad_norm": 0.0008304787334054708,
+      "learning_rate": 1.9855356235736777e-05,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 17688744,
+      "step": 22640
+    },
+    {
+      "epoch": 47.07900207900208,
+      "grad_norm": 5.125362076796591e-05,
+      "learning_rate": 1.9845749278402277e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17692712,
+      "step": 22645
+    },
+    {
+      "epoch": 47.08939708939709,
+      "grad_norm": 0.0001791425165720284,
+      "learning_rate": 1.9836143115918006e-05,
+      "loss": 0.012,
+      "num_input_tokens_seen": 17696616,
+      "step": 22650
+    },
+    {
+      "epoch": 47.0997920997921,
+      "grad_norm": 8.089235052466393e-05,
+      "learning_rate": 1.9826537749765367e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17700488,
+      "step": 22655
+    },
+    {
+      "epoch": 47.11018711018711,
+      "grad_norm": 0.00024530975497327745,
+      "learning_rate": 1.9816933181425625e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17704424,
+      "step": 22660
+    },
+    {
+      "epoch": 47.12058212058212,
+      "grad_norm": 2.3193630113382824e-05,
+      "learning_rate": 1.9807329412379903e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17708168,
+      "step": 22665
+    },
+    {
+      "epoch": 47.13097713097713,
+      "grad_norm": 0.0008051345939747989,
+      "learning_rate": 1.9797726444109247e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 17711944,
+      "step": 22670
+    },
+    {
+      "epoch": 47.141372141372145,
+      "grad_norm": 1.1289465427398682,
+      "learning_rate": 1.9788124278094557e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 17715848,
+      "step": 22675
+    },
+    {
+      "epoch": 47.15176715176715,
+      "grad_norm": 4.6811481297481805e-05,
+      "learning_rate": 1.9778522915816594e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17719912,
+      "step": 22680
+    },
+    {
+      "epoch": 47.16216216216216,
+      "grad_norm": 1.8547510990174487e-05,
+      "learning_rate": 1.9768922358756014e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17723784,
+      "step": 22685
+    },
+    {
+      "epoch": 47.17255717255717,
+      "grad_norm": 6.537731678690761e-05,
+      "learning_rate": 1.9759322608393353e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 17727656,
+      "step": 22690
+    },
+    {
+      "epoch": 47.182952182952185,
+      "grad_norm": 0.4451598525047302,
+      "learning_rate": 1.9749723666208992e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17731688,
+      "step": 22695
+    },
+    {
+      "epoch": 47.19334719334719,
+      "grad_norm": 0.011716783046722412,
+      "learning_rate": 1.9740125533683235e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17735624,
+      "step": 22700
+    },
+    {
+      "epoch": 47.2037422037422,
+      "grad_norm": 2.2450713004218414e-05,
+      "learning_rate": 1.9730528212296208e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17739656,
+      "step": 22705
+    },
+    {
+      "epoch": 47.21413721413722,
+      "grad_norm": 0.0005115432431921363,
+      "learning_rate": 1.9720931703527945e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 17743560,
+      "step": 22710
+    },
+    {
+      "epoch": 47.224532224532226,
+      "grad_norm": 0.00016512704314664006,
+      "learning_rate": 1.9711336008858373e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 17747496,
+      "step": 22715
+    },
+    {
+      "epoch": 47.234927234927234,
+      "grad_norm": 0.005755986552685499,
+      "learning_rate": 1.9701741129767233e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17751528,
+      "step": 22720
+    },
+    {
+      "epoch": 47.24532224532224,
+      "grad_norm": 0.00012748970766551793,
+      "learning_rate": 1.9692147067734202e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17755464,
+      "step": 22725
+    },
+    {
+      "epoch": 47.25571725571726,
+      "grad_norm": 0.00040572750731371343,
+      "learning_rate": 1.96825538242388e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17759464,
+      "step": 22730
+    },
+    {
+      "epoch": 47.266112266112266,
+      "grad_norm": 0.00013119670620653778,
+      "learning_rate": 1.967296140076041e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17763432,
+      "step": 22735
+    },
+    {
+      "epoch": 47.276507276507274,
+      "grad_norm": 5.8287794672651216e-05,
+      "learning_rate": 1.966336979877833e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17767304,
+      "step": 22740
+    },
+    {
+      "epoch": 47.28690228690229,
+      "grad_norm": 0.002838443499058485,
+      "learning_rate": 1.9653779019771678e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17771240,
+      "step": 22745
+    },
+    {
+      "epoch": 47.2972972972973,
+      "grad_norm": 0.00011183625611010939,
+      "learning_rate": 1.9644189065219488e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17775144,
+      "step": 22750
+    },
+    {
+      "epoch": 47.30769230769231,
+      "grad_norm": 1.8113085388904437e-05,
+      "learning_rate": 1.9634599936600655e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17778920,
+      "step": 22755
+    },
+    {
+      "epoch": 47.318087318087315,
+      "grad_norm": 3.013863715750631e-05,
+      "learning_rate": 1.9625011635393935e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17782760,
+      "step": 22760
+    },
+    {
+      "epoch": 47.32848232848233,
+      "grad_norm": 0.0001005881858873181,
+      "learning_rate": 1.9615424163077963e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17786568,
+      "step": 22765
+    },
+    {
+      "epoch": 47.33887733887734,
+      "grad_norm": 0.00010243827273370698,
+      "learning_rate": 1.9605837521131263e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 17790408,
+      "step": 22770
+    },
+    {
+      "epoch": 47.34927234927235,
+      "grad_norm": 1.683307345956564e-05,
+      "learning_rate": 1.9596251711032192e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17794408,
+      "step": 22775
+    },
+    {
+      "epoch": 47.35966735966736,
+      "grad_norm": 0.00029985548462718725,
+      "learning_rate": 1.958666673425903e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17798184,
+      "step": 22780
+    },
+    {
+      "epoch": 47.37006237006237,
+      "grad_norm": 1.6677931853337213e-05,
+      "learning_rate": 1.957708259228987e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17802024,
+      "step": 22785
+    },
+    {
+      "epoch": 47.38045738045738,
+      "grad_norm": 6.683650553895859e-06,
+      "learning_rate": 1.956749928660273e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17805800,
+      "step": 22790
+    },
+    {
+      "epoch": 47.39085239085239,
+      "grad_norm": 7.089578048180556e-06,
+      "learning_rate": 1.955791681867547e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17809576,
+      "step": 22795
+    },
+    {
+      "epoch": 47.4012474012474,
+      "grad_norm": 1.325780976912938e-05,
+      "learning_rate": 1.9548335189985824e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17813576,
+      "step": 22800
+    },
+    {
+      "epoch": 47.4012474012474,
+      "eval_loss": 0.5011555552482605,
+      "eval_runtime": 11.7272,
+      "eval_samples_per_second": 72.992,
+      "eval_steps_per_second": 18.248,
+      "num_input_tokens_seen": 17813576,
+      "step": 22800
+    },
+    {
+      "epoch": 47.41164241164241,
+      "grad_norm": 37.14784240722656,
+      "learning_rate": 1.9538754402011396e-05,
+      "loss": 0.0196,
+      "num_input_tokens_seen": 17817480,
+      "step": 22805
+    },
+    {
+      "epoch": 47.42203742203742,
+      "grad_norm": 0.2522355318069458,
+      "learning_rate": 1.952917445622968e-05,
+      "loss": 0.006,
+      "num_input_tokens_seen": 17821320,
+      "step": 22810
+    },
+    {
+      "epoch": 47.432432432432435,
+      "grad_norm": 3.167682370985858e-05,
+      "learning_rate": 1.9519595354118005e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17825160,
+      "step": 22815
+    },
+    {
+      "epoch": 47.44282744282744,
+      "grad_norm": 0.00016180532111320645,
+      "learning_rate": 1.951001709715361e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17829064,
+      "step": 22820
+    },
+    {
+      "epoch": 47.45322245322245,
+      "grad_norm": 3.249083965783939e-05,
+      "learning_rate": 1.9500439686813556e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17832904,
+      "step": 22825
+    },
+    {
+      "epoch": 47.46361746361746,
+      "grad_norm": 5.244701242190786e-05,
+      "learning_rate": 1.949086312457482e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 17836968,
+      "step": 22830
+    },
+    {
+      "epoch": 47.474012474012476,
+      "grad_norm": 0.0001903692027553916,
+      "learning_rate": 1.9481287411914223e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17841032,
+      "step": 22835
+    },
+    {
+      "epoch": 47.484407484407484,
+      "grad_norm": 4.9991092964774e-05,
+      "learning_rate": 1.9471712550308457e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17845000,
+      "step": 22840
+    },
+    {
+      "epoch": 47.49480249480249,
+      "grad_norm": 0.0003490489616524428,
+      "learning_rate": 1.946213854123409e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17849096,
+      "step": 22845
+    },
+    {
+      "epoch": 47.50519750519751,
+      "grad_norm": 0.0006364610162563622,
+      "learning_rate": 1.9452565386167554e-05,
+      "loss": 0.0015,
+      "num_input_tokens_seen": 17852872,
+      "step": 22850
+    },
+    {
+      "epoch": 47.515592515592516,
+      "grad_norm": 0.00011939080286538228,
+      "learning_rate": 1.9442993086585142e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17856840,
+      "step": 22855
+    },
+    {
+      "epoch": 47.525987525987524,
+      "grad_norm": 0.07823726534843445,
+      "learning_rate": 1.9433421643963043e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 17860680,
+      "step": 22860
+    },
+    {
+      "epoch": 47.53638253638254,
+      "grad_norm": 0.00017966995073948056,
+      "learning_rate": 1.942385105977727e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17864616,
+      "step": 22865
+    },
+    {
+      "epoch": 47.54677754677755,
+      "grad_norm": 0.00045225032954476774,
+      "learning_rate": 1.9414281335503743e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17868520,
+      "step": 22870
+    },
+    {
+      "epoch": 47.55717255717256,
+      "grad_norm": 0.00013090248103253543,
+      "learning_rate": 1.9404712472618232e-05,
+      "loss": 0.005,
+      "num_input_tokens_seen": 17872392,
+      "step": 22875
+    },
+    {
+      "epoch": 47.567567567567565,
+      "grad_norm": 0.005077596288174391,
+      "learning_rate": 1.939514447259636e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17876328,
+      "step": 22880
+    },
+    {
+      "epoch": 47.57796257796258,
+      "grad_norm": 3.681773159769364e-05,
+      "learning_rate": 1.938557733691365e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17880168,
+      "step": 22885
+    },
+    {
+      "epoch": 47.58835758835759,
+      "grad_norm": 0.010062659159302711,
+      "learning_rate": 1.9376011067045476e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17884168,
+      "step": 22890
+    },
+    {
+      "epoch": 47.5987525987526,
+      "grad_norm": 0.0006245516124181449,
+      "learning_rate": 1.9366445664467065e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17888104,
+      "step": 22895
+    },
+    {
+      "epoch": 47.60914760914761,
+      "grad_norm": 0.0011196625418961048,
+      "learning_rate": 1.9356881130653533e-05,
+      "loss": 0.0008,
+      "num_input_tokens_seen": 17891880,
+      "step": 22900
+    },
+    {
+      "epoch": 47.61954261954262,
+      "grad_norm": 0.5031498074531555,
+      "learning_rate": 1.9347317467079846e-05,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 17895880,
+      "step": 22905
+    },
+    {
+      "epoch": 47.62993762993763,
+      "grad_norm": 5.607548519037664e-05,
+      "learning_rate": 1.9337754675220836e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17899848,
+      "step": 22910
+    },
+    {
+      "epoch": 47.64033264033264,
+      "grad_norm": 0.024107003584504128,
+      "learning_rate": 1.9328192756551218e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17903816,
+      "step": 22915
+    },
+    {
+      "epoch": 47.65072765072765,
+      "grad_norm": 0.00028723853756673634,
+      "learning_rate": 1.931863171254555e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17907752,
+      "step": 22920
+    },
+    {
+      "epoch": 47.66112266112266,
+      "grad_norm": 1.1954268302361015e-05,
+      "learning_rate": 1.930907154467826e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17911496,
+      "step": 22925
+    },
+    {
+      "epoch": 47.67151767151767,
+      "grad_norm": 1.0296588698111009e-05,
+      "learning_rate": 1.9299512254423673e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17915336,
+      "step": 22930
+    },
+    {
+      "epoch": 47.681912681912685,
+      "grad_norm": 2.663316081452649e-05,
+      "learning_rate": 1.9289953843255914e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17919208,
+      "step": 22935
+    },
+    {
+      "epoch": 47.69230769230769,
+      "grad_norm": 2.0956595108145848e-05,
+      "learning_rate": 1.9280396312649048e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17923080,
+      "step": 22940
+    },
+    {
+      "epoch": 47.7027027027027,
+      "grad_norm": 0.0002497107780072838,
+      "learning_rate": 1.9270839664076936e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17926920,
+      "step": 22945
+    },
+    {
+      "epoch": 47.71309771309771,
+      "grad_norm": 0.000195358443306759,
+      "learning_rate": 1.9261283899013345e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17930920,
+      "step": 22950
+    },
+    {
+      "epoch": 47.723492723492726,
+      "grad_norm": 5.087560566607863e-05,
+      "learning_rate": 1.92517290189319e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17934824,
+      "step": 22955
+    },
+    {
+      "epoch": 47.733887733887734,
+      "grad_norm": 0.00011614750837907195,
+      "learning_rate": 1.924217502530607e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17938664,
+      "step": 22960
+    },
+    {
+      "epoch": 47.74428274428274,
+      "grad_norm": 1.6701207641744986e-05,
+      "learning_rate": 1.9232621919609207e-05,
+      "loss": 0.0278,
+      "num_input_tokens_seen": 17942728,
+      "step": 22965
+    },
+    {
+      "epoch": 47.75467775467776,
+      "grad_norm": 2.481622141203843e-05,
+      "learning_rate": 1.9223069703314534e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 17946696,
+      "step": 22970
+    },
+    {
+      "epoch": 47.765072765072766,
+      "grad_norm": 1.7002754248096608e-05,
+      "learning_rate": 1.92135183778951e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17950536,
+      "step": 22975
+    },
+    {
+      "epoch": 47.775467775467774,
+      "grad_norm": 7.317130803130567e-05,
+      "learning_rate": 1.9203967944823857e-05,
+      "loss": 0.0074,
+      "num_input_tokens_seen": 17954408,
+      "step": 22980
+    },
+    {
+      "epoch": 47.78586278586278,
+      "grad_norm": 0.0005342771764844656,
+      "learning_rate": 1.9194418405573588e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17958312,
+      "step": 22985
+    },
+    {
+      "epoch": 47.7962577962578,
+      "grad_norm": 0.0009529058588668704,
+      "learning_rate": 1.9184869761616954e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17962184,
+      "step": 22990
+    },
+    {
+      "epoch": 47.80665280665281,
+      "grad_norm": 0.001625004573725164,
+      "learning_rate": 1.9175322014426495e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 17965992,
+      "step": 22995
+    },
+    {
+      "epoch": 47.817047817047815,
+      "grad_norm": 0.00015840691048651934,
+      "learning_rate": 1.9165775165474565e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17970024,
+      "step": 23000
+    },
+    {
+      "epoch": 47.817047817047815,
+      "eval_loss": 0.4887303113937378,
+      "eval_runtime": 11.6774,
+      "eval_samples_per_second": 73.304,
+      "eval_steps_per_second": 18.326,
+      "num_input_tokens_seen": 17970024,
+      "step": 23000
+    },
+    {
+      "epoch": 47.82744282744283,
+      "grad_norm": 0.00013597037468571216,
+      "learning_rate": 1.9156229216233434e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17973960,
+      "step": 23005
+    },
+    {
+      "epoch": 47.83783783783784,
+      "grad_norm": 0.0022586809936910868,
+      "learning_rate": 1.9146684168175184e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17977800,
+      "step": 23010
+    },
+    {
+      "epoch": 47.84823284823285,
+      "grad_norm": 0.2043021321296692,
+      "learning_rate": 1.9137140022771796e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 17981576,
+      "step": 23015
+    },
+    {
+      "epoch": 47.858627858627855,
+      "grad_norm": 6.079965896788053e-05,
+      "learning_rate": 1.9127596781495103e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17985608,
+      "step": 23020
+    },
+    {
+      "epoch": 47.86902286902287,
+      "grad_norm": 5.5712676839903e-05,
+      "learning_rate": 1.9118054445816767e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17989544,
+      "step": 23025
+    },
+    {
+      "epoch": 47.87941787941788,
+      "grad_norm": 0.0005309552070684731,
+      "learning_rate": 1.9108513017208356e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 17993384,
+      "step": 23030
+    },
+    {
+      "epoch": 47.88981288981289,
+      "grad_norm": 4.9122681957669556e-05,
+      "learning_rate": 1.9098972497141287e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 17997192,
+      "step": 23035
+    },
+    {
+      "epoch": 47.9002079002079,
+      "grad_norm": 5.609843719867058e-05,
+      "learning_rate": 1.9089432887086806e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18001064,
+      "step": 23040
+    },
+    {
+      "epoch": 47.91060291060291,
+      "grad_norm": 0.00015566784713882953,
+      "learning_rate": 1.9079894188516056e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 18004968,
+      "step": 23045
+    },
+    {
+      "epoch": 47.92099792099792,
+      "grad_norm": 0.00017356264288537204,
+      "learning_rate": 1.907035640290002e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18008872,
+      "step": 23050
+    },
+    {
+      "epoch": 47.931392931392935,
+      "grad_norm": 0.00010890791600104421,
+      "learning_rate": 1.9060819531709534e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18012936,
+      "step": 23055
+    },
+    {
+      "epoch": 47.94178794178794,
+      "grad_norm": 0.0006500133313238621,
+      "learning_rate": 1.9051283576415325e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18016872,
+      "step": 23060
+    },
+    {
+      "epoch": 47.95218295218295,
+      "grad_norm": 0.00021246248797979206,
+      "learning_rate": 1.904174853848793e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18020712,
+      "step": 23065
+    },
+    {
+      "epoch": 47.96257796257796,
+      "grad_norm": 4.314256875659339e-05,
+      "learning_rate": 1.903221441939779e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18024616,
+      "step": 23070
+    },
+    {
+      "epoch": 47.972972972972975,
+      "grad_norm": 9.919074364006519e-05,
+      "learning_rate": 1.9022681220615194e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18028520,
+      "step": 23075
+    },
+    {
+      "epoch": 47.983367983367984,
+      "grad_norm": 0.003450539195910096,
+      "learning_rate": 1.9013148943610255e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 18032584,
+      "step": 23080
+    },
+    {
+      "epoch": 47.99376299376299,
+      "grad_norm": 7.08883089828305e-05,
+      "learning_rate": 1.9003617589852998e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18036552,
+      "step": 23085
+    },
+    {
+      "epoch": 48.00415800415801,
+      "grad_norm": 4.6587345423176885e-05,
+      "learning_rate": 1.899408716081326e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18040488,
+      "step": 23090
+    },
+    {
+      "epoch": 48.014553014553016,
+      "grad_norm": 0.00015644257655367255,
+      "learning_rate": 1.898455765796075e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18044456,
+      "step": 23095
+    },
+    {
+      "epoch": 48.024948024948024,
+      "grad_norm": 0.00011759501649066806,
+      "learning_rate": 1.8975029082765053e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18048488,
+      "step": 23100
+    },
+    {
+      "epoch": 48.03534303534303,
+      "grad_norm": 2.7889500415767543e-05,
+      "learning_rate": 1.8965501436695577e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18052392,
+      "step": 23105
+    },
+    {
+      "epoch": 48.04573804573805,
+      "grad_norm": 0.00021743755496572703,
+      "learning_rate": 1.895597472122161e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18056232,
+      "step": 23110
+    },
+    {
+      "epoch": 48.056133056133056,
+      "grad_norm": 0.00025736671523191035,
+      "learning_rate": 1.894644893781231e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18060104,
+      "step": 23115
+    },
+    {
+      "epoch": 48.066528066528065,
+      "grad_norm": 0.0007948701968416572,
+      "learning_rate": 1.893692408793665e-05,
+      "loss": 0.0011,
+      "num_input_tokens_seen": 18064008,
+      "step": 23120
+    },
+    {
+      "epoch": 48.07692307692308,
+      "grad_norm": 0.0033185030333697796,
+      "learning_rate": 1.8927400173063493e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18067816,
+      "step": 23125
+    },
+    {
+      "epoch": 48.08731808731809,
+      "grad_norm": 0.00038652773946523666,
+      "learning_rate": 1.891787719466154e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18071816,
+      "step": 23130
+    },
+    {
+      "epoch": 48.0977130977131,
+      "grad_norm": 3.139784894301556e-05,
+      "learning_rate": 1.8908355154199346e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 18075752,
+      "step": 23135
+    },
+    {
+      "epoch": 48.108108108108105,
+      "grad_norm": 5.932737622060813e-05,
+      "learning_rate": 1.8898834053145357e-05,
+      "loss": 0.006,
+      "num_input_tokens_seen": 18079752,
+      "step": 23140
+    },
+    {
+      "epoch": 48.11850311850312,
+      "grad_norm": 3.459229628788307e-05,
+      "learning_rate": 1.8889313892967813e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18083720,
+      "step": 23145
+    },
+    {
+      "epoch": 48.12889812889813,
+      "grad_norm": 0.14184918999671936,
+      "learning_rate": 1.8879794675134863e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 18087496,
+      "step": 23150
+    },
+    {
+      "epoch": 48.13929313929314,
+      "grad_norm": 1.503270959801739e-05,
+      "learning_rate": 1.8870276401114494e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18091336,
+      "step": 23155
+    },
+    {
+      "epoch": 48.14968814968815,
+      "grad_norm": 0.0005981053109280765,
+      "learning_rate": 1.886075907237453e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18095112,
+      "step": 23160
+    },
+    {
+      "epoch": 48.16008316008316,
+      "grad_norm": 0.0002470419858582318,
+      "learning_rate": 1.8851242690382672e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18099112,
+      "step": 23165
+    },
+    {
+      "epoch": 48.17047817047817,
+      "grad_norm": 0.0011883819242939353,
+      "learning_rate": 1.884172725660645e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18103016,
+      "step": 23170
+    },
+    {
+      "epoch": 48.18087318087318,
+      "grad_norm": 1.4137749531073496e-05,
+      "learning_rate": 1.8832212772513277e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 18106856,
+      "step": 23175
+    },
+    {
+      "epoch": 48.19126819126819,
+      "grad_norm": 0.09916135668754578,
+      "learning_rate": 1.8822699239570414e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 18110824,
+      "step": 23180
+    },
+    {
+      "epoch": 48.2016632016632,
+      "grad_norm": 3.417845073272474e-05,
+      "learning_rate": 1.8813186659244943e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18114728,
+      "step": 23185
+    },
+    {
+      "epoch": 48.21205821205821,
+      "grad_norm": 4.53219618066214e-05,
+      "learning_rate": 1.880367503300385e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18118664,
+      "step": 23190
+    },
+    {
+      "epoch": 48.222453222453225,
+      "grad_norm": 5.397202130552614e-06,
+      "learning_rate": 1.8794164362313927e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18122536,
+      "step": 23195
+    },
+    {
+      "epoch": 48.232848232848234,
+      "grad_norm": 7.598746833537007e-06,
+      "learning_rate": 1.878465464864185e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18126280,
+      "step": 23200
+    },
+    {
+      "epoch": 48.232848232848234,
+      "eval_loss": 0.5224213600158691,
+      "eval_runtime": 11.741,
+      "eval_samples_per_second": 72.907,
+      "eval_steps_per_second": 18.227,
+      "num_input_tokens_seen": 18126280,
+      "step": 23200
+    },
+    {
+      "epoch": 48.24324324324324,
+      "grad_norm": 0.00019667136075440794,
+      "learning_rate": 1.877514589345414e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18130184,
+      "step": 23205
+    },
+    {
+      "epoch": 48.25363825363825,
+      "grad_norm": 2.3637849153601564e-05,
+      "learning_rate": 1.876563809821715e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18134024,
+      "step": 23210
+    },
+    {
+      "epoch": 48.264033264033266,
+      "grad_norm": 0.0004043622757308185,
+      "learning_rate": 1.8756131264397106e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18138056,
+      "step": 23215
+    },
+    {
+      "epoch": 48.274428274428274,
+      "grad_norm": 0.00046428883797489107,
+      "learning_rate": 1.87466253934601e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18141896,
+      "step": 23220
+    },
+    {
+      "epoch": 48.28482328482328,
+      "grad_norm": 6.532065890496597e-05,
+      "learning_rate": 1.8737120486872033e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18145928,
+      "step": 23225
+    },
+    {
+      "epoch": 48.2952182952183,
+      "grad_norm": 1.175331726699369e-05,
+      "learning_rate": 1.8727616546098696e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18149832,
+      "step": 23230
+    },
+    {
+      "epoch": 48.305613305613306,
+      "grad_norm": 0.1520174890756607,
+      "learning_rate": 1.8718113572605716e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 18153768,
+      "step": 23235
+    },
+    {
+      "epoch": 48.316008316008315,
+      "grad_norm": 0.15273338556289673,
+      "learning_rate": 1.8708611567858554e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 18157640,
+      "step": 23240
+    },
+    {
+      "epoch": 48.32640332640332,
+      "grad_norm": 0.16201312839984894,
+      "learning_rate": 1.8699110533322565e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 18161512,
+      "step": 23245
+    },
+    {
+      "epoch": 48.33679833679834,
+      "grad_norm": 3.675209518405609e-05,
+      "learning_rate": 1.8689610470462897e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18165480,
+      "step": 23250
+    },
+    {
+      "epoch": 48.34719334719335,
+      "grad_norm": 8.66847112774849e-05,
+      "learning_rate": 1.8680111380744604e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18169544,
+      "step": 23255
+    },
+    {
+      "epoch": 48.357588357588355,
+      "grad_norm": 4.415732837514952e-05,
+      "learning_rate": 1.8670613265632564e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18173544,
+      "step": 23260
+    },
+    {
+      "epoch": 48.36798336798337,
+      "grad_norm": 2.929928814410232e-05,
+      "learning_rate": 1.866111612659149e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18177352,
+      "step": 23265
+    },
+    {
+      "epoch": 48.37837837837838,
+      "grad_norm": 0.18549282848834991,
+      "learning_rate": 1.8651619965085967e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 18181224,
+      "step": 23270
+    },
+    {
+      "epoch": 48.38877338877339,
+      "grad_norm": 0.00016255013179033995,
+      "learning_rate": 1.8642124782580433e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18185064,
+      "step": 23275
+    },
+    {
+      "epoch": 48.3991683991684,
+      "grad_norm": 0.0024176931474357843,
+      "learning_rate": 1.8632630580539144e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18188904,
+      "step": 23280
+    },
+    {
+      "epoch": 48.40956340956341,
+      "grad_norm": 3.471785021247342e-05,
+      "learning_rate": 1.862313736042625e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 18192776,
+      "step": 23285
+    },
+    {
+      "epoch": 48.41995841995842,
+      "grad_norm": 0.00012559938477352262,
+      "learning_rate": 1.8613645123705703e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18196680,
+      "step": 23290
+    },
+    {
+      "epoch": 48.43035343035343,
+      "grad_norm": 8.134434756357223e-05,
+      "learning_rate": 1.8604153871841328e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18200648,
+      "step": 23295
+    },
+    {
+      "epoch": 48.44074844074844,
+      "grad_norm": 0.00017207834753207862,
+      "learning_rate": 1.859466360629682e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18204456,
+      "step": 23300
+    },
+    {
+      "epoch": 48.45114345114345,
+      "grad_norm": 6.702930841129273e-05,
+      "learning_rate": 1.8585174328535666e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18208520,
+      "step": 23305
+    },
+    {
+      "epoch": 48.46153846153846,
+      "grad_norm": 3.4319797123316675e-05,
+      "learning_rate": 1.857568604002124e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18212584,
+      "step": 23310
+    },
+    {
+      "epoch": 48.471933471933475,
+      "grad_norm": 8.988335321191698e-05,
+      "learning_rate": 1.8566198742216774e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18216552,
+      "step": 23315
+    },
+    {
+      "epoch": 48.482328482328484,
+      "grad_norm": 2.7501584554556757e-05,
+      "learning_rate": 1.85567124365853e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18220488,
+      "step": 23320
+    },
+    {
+      "epoch": 48.49272349272349,
+      "grad_norm": 5.5363103456329554e-05,
+      "learning_rate": 1.854722712458975e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18224424,
+      "step": 23325
+    },
+    {
+      "epoch": 48.5031185031185,
+      "grad_norm": 0.00010783670586533844,
+      "learning_rate": 1.853774280769286e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18228296,
+      "step": 23330
+    },
+    {
+      "epoch": 48.513513513513516,
+      "grad_norm": 3.5432924050837755e-05,
+      "learning_rate": 1.852825948735724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18232264,
+      "step": 23335
+    },
+    {
+      "epoch": 48.523908523908524,
+      "grad_norm": 0.00012596204760484397,
+      "learning_rate": 1.851877716504534e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 18236200,
+      "step": 23340
+    },
+    {
+      "epoch": 48.53430353430353,
+      "grad_norm": 4.9048328946810216e-05,
+      "learning_rate": 1.8509295842219448e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18240040,
+      "step": 23345
+    },
+    {
+      "epoch": 48.54469854469855,
+      "grad_norm": 6.609917363675777e-06,
+      "learning_rate": 1.8499815520341697e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18243912,
+      "step": 23350
+    },
+    {
+      "epoch": 48.555093555093556,
+      "grad_norm": 3.498014120850712e-05,
+      "learning_rate": 1.8490336200874094e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18247688,
+      "step": 23355
+    },
+    {
+      "epoch": 48.565488565488565,
+      "grad_norm": 4.518833156907931e-05,
+      "learning_rate": 1.848085788527844e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18251560,
+      "step": 23360
+    },
+    {
+      "epoch": 48.57588357588357,
+      "grad_norm": 3.327256126794964e-05,
+      "learning_rate": 1.847138057501644e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18255496,
+      "step": 23365
+    },
+    {
+      "epoch": 48.58627858627859,
+      "grad_norm": 9.728327131597325e-05,
+      "learning_rate": 1.8461904271549582e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18259336,
+      "step": 23370
+    },
+    {
+      "epoch": 48.5966735966736,
+      "grad_norm": 0.00027633385616354644,
+      "learning_rate": 1.845242897633926e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18263176,
+      "step": 23375
+    },
+    {
+      "epoch": 48.607068607068605,
+      "grad_norm": 0.000227213284233585,
+      "learning_rate": 1.844295469084667e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18267208,
+      "step": 23380
+    },
+    {
+      "epoch": 48.61746361746362,
+      "grad_norm": 2.6613741283654235e-05,
+      "learning_rate": 1.843348141653286e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18271048,
+      "step": 23385
+    },
+    {
+      "epoch": 48.62785862785863,
+      "grad_norm": 7.860564801376313e-05,
+      "learning_rate": 1.842400915485874e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 18274920,
+      "step": 23390
+    },
+    {
+      "epoch": 48.63825363825364,
+      "grad_norm": 5.334526576916687e-05,
+      "learning_rate": 1.8414537907285053e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 18278792,
+      "step": 23395
+    },
+    {
+      "epoch": 48.648648648648646,
+      "grad_norm": 5.989719647914171e-05,
+      "learning_rate": 1.840506767527237e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 18282568,
+      "step": 23400
+    },
+    {
+      "epoch": 48.648648648648646,
+      "eval_loss": 0.5204456448554993,
+      "eval_runtime": 11.7245,
+      "eval_samples_per_second": 73.009,
+      "eval_steps_per_second": 18.252,
+      "num_input_tokens_seen": 18282568,
+      "step": 23400
+    },
+    {
+      "epoch": 48.65904365904366,
+      "grad_norm": 0.00021366353030316532,
+      "learning_rate": 1.8395598460281137e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18286792,
+      "step": 23405
+    },
+    {
+      "epoch": 48.66943866943867,
+      "grad_norm": 2.8520729756564833e-05,
+      "learning_rate": 1.838613026377161e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18290760,
+      "step": 23410
+    },
+    {
+      "epoch": 48.67983367983368,
+      "grad_norm": 7.24225610611029e-05,
+      "learning_rate": 1.8376663087203917e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18294536,
+      "step": 23415
+    },
+    {
+      "epoch": 48.69022869022869,
+      "grad_norm": 1.049715774570359e-05,
+      "learning_rate": 1.8367196932038014e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18298504,
+      "step": 23420
+    },
+    {
+      "epoch": 48.7006237006237,
+      "grad_norm": 8.687181980349123e-05,
+      "learning_rate": 1.8357731799733686e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18302472,
+      "step": 23425
+    },
+    {
+      "epoch": 48.71101871101871,
+      "grad_norm": 1.5669895219616592e-05,
+      "learning_rate": 1.8348267691750586e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18306408,
+      "step": 23430
+    },
+    {
+      "epoch": 48.72141372141372,
+      "grad_norm": 4.774280751007609e-05,
+      "learning_rate": 1.833880460954821e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18310312,
+      "step": 23435
+    },
+    {
+      "epoch": 48.731808731808734,
+      "grad_norm": 0.00016962691734079272,
+      "learning_rate": 1.8329342554585866e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18314152,
+      "step": 23440
+    },
+    {
+      "epoch": 48.74220374220374,
+      "grad_norm": 2.9601569622172974e-05,
+      "learning_rate": 1.8319881528322735e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18317992,
+      "step": 23445
+    },
+    {
+      "epoch": 48.75259875259875,
+      "grad_norm": 0.0006237992201931775,
+      "learning_rate": 1.8310421532217815e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18321928,
+      "step": 23450
+    },
+    {
+      "epoch": 48.762993762993766,
+      "grad_norm": 0.0007444994407705963,
+      "learning_rate": 1.8300962567729958e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18325928,
+      "step": 23455
+    },
+    {
+      "epoch": 48.773388773388774,
+      "grad_norm": 1.2598485227499623e-05,
+      "learning_rate": 1.8291504636317866e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18329704,
+      "step": 23460
+    },
+    {
+      "epoch": 48.78378378378378,
+      "grad_norm": 1.6359674191335216e-05,
+      "learning_rate": 1.8282047739440055e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18333576,
+      "step": 23465
+    },
+    {
+      "epoch": 48.79417879417879,
+      "grad_norm": 0.0004631706979125738,
+      "learning_rate": 1.8272591878554903e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 18337608,
+      "step": 23470
+    },
+    {
+      "epoch": 48.804573804573806,
+      "grad_norm": 0.10976377129554749,
+      "learning_rate": 1.8263137055120638e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 18341512,
+      "step": 23475
+    },
+    {
+      "epoch": 48.814968814968815,
+      "grad_norm": 0.001938387518748641,
+      "learning_rate": 1.8253683270595295e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 18345352,
+      "step": 23480
+    },
+    {
+      "epoch": 48.82536382536382,
+      "grad_norm": 8.720582263777032e-05,
+      "learning_rate": 1.824423052643677e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 18349192,
+      "step": 23485
+    },
+    {
+      "epoch": 48.83575883575884,
+      "grad_norm": 3.95796523662284e-05,
+      "learning_rate": 1.82347788241028e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18353032,
+      "step": 23490
+    },
+    {
+      "epoch": 48.84615384615385,
+      "grad_norm": 0.23272621631622314,
+      "learning_rate": 1.8225328165050942e-05,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 18357064,
+      "step": 23495
+    },
+    {
+      "epoch": 48.856548856548855,
+      "grad_norm": 0.0006219825008884072,
+      "learning_rate": 1.821587855073863e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18360968,
+      "step": 23500
+    },
+    {
+      "epoch": 48.86694386694387,
+      "grad_norm": 0.0010352300014346838,
+      "learning_rate": 1.8206429982623086e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18364872,
+      "step": 23505
+    },
+    {
+      "epoch": 48.87733887733888,
+      "grad_norm": 2.620505438244436e-05,
+      "learning_rate": 1.8196982462161416e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18368776,
+      "step": 23510
+    },
+    {
+      "epoch": 48.88773388773389,
+      "grad_norm": 0.00026731527759693563,
+      "learning_rate": 1.818753599081055e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18372584,
+      "step": 23515
+    },
+    {
+      "epoch": 48.898128898128896,
+      "grad_norm": 5.819413854624145e-05,
+      "learning_rate": 1.817809057002724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18376488,
+      "step": 23520
+    },
+    {
+      "epoch": 48.90852390852391,
+      "grad_norm": 1.985273411264643e-05,
+      "learning_rate": 1.8168646201268096e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 18380392,
+      "step": 23525
+    },
+    {
+      "epoch": 48.91891891891892,
+      "grad_norm": 1.530184999865014e-05,
+      "learning_rate": 1.8159202885989557e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18384296,
+      "step": 23530
+    },
+    {
+      "epoch": 48.92931392931393,
+      "grad_norm": 9.495672202319838e-06,
+      "learning_rate": 1.814976062564789e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18388200,
+      "step": 23535
+    },
+    {
+      "epoch": 48.93970893970894,
+      "grad_norm": 3.805133019341156e-05,
+      "learning_rate": 1.8140319421699234e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18392168,
+      "step": 23540
+    },
+    {
+      "epoch": 48.95010395010395,
+      "grad_norm": 2.1030900825280696e-05,
+      "learning_rate": 1.8130879275599515e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 18396104,
+      "step": 23545
+    },
+    {
+      "epoch": 48.96049896049896,
+      "grad_norm": 0.0016333397943526506,
+      "learning_rate": 1.8121440188804544e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18400008,
+      "step": 23550
+    },
+    {
+      "epoch": 48.97089397089397,
+      "grad_norm": 2.537048567319289e-05,
+      "learning_rate": 1.811200216276993e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18403880,
+      "step": 23555
+    },
+    {
+      "epoch": 48.981288981288984,
+      "grad_norm": 1.1496681509015616e-05,
+      "learning_rate": 1.810256519895115e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 18407848,
+      "step": 23560
+    },
+    {
+      "epoch": 48.99168399168399,
+      "grad_norm": 0.00022154061298351735,
+      "learning_rate": 1.8093129298803494e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18411624,
+      "step": 23565
+    },
+    {
+      "epoch": 49.002079002079,
+      "grad_norm": 2.4682547518750653e-05,
+      "learning_rate": 1.808369446378209e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 18415448,
+      "step": 23570
+    },
+    {
+      "epoch": 49.012474012474016,
+      "grad_norm": 8.609574433648959e-05,
+      "learning_rate": 1.8074260695341914e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18419352,
+      "step": 23575
+    },
+    {
+      "epoch": 49.022869022869024,
+      "grad_norm": 1.771956158336252e-05,
+      "learning_rate": 1.8064827994937782e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18423128,
+      "step": 23580
+    },
+    {
+      "epoch": 49.03326403326403,
+      "grad_norm": 2.6265508950018557e-06,
+      "learning_rate": 1.8055396364024317e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18427032,
+      "step": 23585
+    },
+    {
+      "epoch": 49.04365904365904,
+      "grad_norm": 7.438737520715222e-05,
+      "learning_rate": 1.804596580405601e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18431128,
+      "step": 23590
+    },
+    {
+      "epoch": 49.054054054054056,
+      "grad_norm": 1.557037467136979e-05,
+      "learning_rate": 1.8036536316487174e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18434968,
+      "step": 23595
+    },
+    {
+      "epoch": 49.064449064449065,
+      "grad_norm": 1.1113729669887107e-05,
+      "learning_rate": 1.802710790277193e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18438872,
+      "step": 23600
+    },
+    {
+      "epoch": 49.064449064449065,
+      "eval_loss": 0.5278869867324829,
+      "eval_runtime": 11.7215,
+      "eval_samples_per_second": 73.028,
+      "eval_steps_per_second": 18.257,
+      "num_input_tokens_seen": 18438872,
+      "step": 23600
+    },
+    {
+      "epoch": 49.07484407484407,
+      "grad_norm": 1.3451236554828938e-05,
+      "learning_rate": 1.801768056436429e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 18442776,
+      "step": 23605
+    },
+    {
+      "epoch": 49.08523908523909,
+      "grad_norm": 0.00035920957452617586,
+      "learning_rate": 1.8008254302718035e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18446744,
+      "step": 23610
+    },
+    {
+      "epoch": 49.0956340956341,
+      "grad_norm": 1.66279078257503e-05,
+      "learning_rate": 1.7998829119286837e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18450776,
+      "step": 23615
+    },
+    {
+      "epoch": 49.106029106029105,
+      "grad_norm": 2.1659230696968734e-05,
+      "learning_rate": 1.798940501552418e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18454616,
+      "step": 23620
+    },
+    {
+      "epoch": 49.11642411642411,
+      "grad_norm": 0.0002607192727737129,
+      "learning_rate": 1.797998199288336e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18458616,
+      "step": 23625
+    },
+    {
+      "epoch": 49.12681912681913,
+      "grad_norm": 3.7203273677732795e-05,
+      "learning_rate": 1.7970560052817543e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 18462520,
+      "step": 23630
+    },
+    {
+      "epoch": 49.13721413721414,
+      "grad_norm": 6.1291444581002e-05,
+      "learning_rate": 1.7961139196779702e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18466488,
+      "step": 23635
+    },
+    {
+      "epoch": 49.147609147609145,
+      "grad_norm": 2.4244089217972942e-05,
+      "learning_rate": 1.7951719426222647e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18470424,
+      "step": 23640
+    },
+    {
+      "epoch": 49.15800415800416,
+      "grad_norm": 1.2916459127154667e-05,
+      "learning_rate": 1.794230074259904e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18474456,
+      "step": 23645
+    },
+    {
+      "epoch": 49.16839916839917,
+      "grad_norm": 6.1458682466764e-05,
+      "learning_rate": 1.7932883147361336e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 18478456,
+      "step": 23650
+    },
+    {
+      "epoch": 49.17879417879418,
+      "grad_norm": 1.0466680578247178e-05,
+      "learning_rate": 1.7923466641961865e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18482328,
+      "step": 23655
+    },
+    {
+      "epoch": 49.189189189189186,
+      "grad_norm": 7.894321606727317e-05,
+      "learning_rate": 1.791405122785278e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18486264,
+      "step": 23660
+    },
+    {
+      "epoch": 49.1995841995842,
+      "grad_norm": 1.7850854419521056e-05,
+      "learning_rate": 1.7904636906486037e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18490200,
+      "step": 23665
+    },
+    {
+      "epoch": 49.20997920997921,
+      "grad_norm": 2.6218014681944624e-05,
+      "learning_rate": 1.7895223679313448e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 18494072,
+      "step": 23670
+    },
+    {
+      "epoch": 49.22037422037422,
+      "grad_norm": 3.158472100039944e-05,
+      "learning_rate": 1.7885811547786653e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 18498072,
+      "step": 23675
+    },
+    {
+      "epoch": 49.23076923076923,
+      "grad_norm": 3.7479712773347273e-05,
+      "learning_rate": 1.7876400513357115e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18501976,
+      "step": 23680
+    },
+    {
+      "epoch": 49.24116424116424,
+      "grad_norm": 0.0002222458424512297,
+      "learning_rate": 1.7866990577476146e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18505752,
+      "step": 23685
+    },
+    {
+      "epoch": 49.25155925155925,
+      "grad_norm": 0.0008142099832184613,
+      "learning_rate": 1.7857581741594863e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18509592,
+      "step": 23690
+    },
+    {
+      "epoch": 49.26195426195426,
+      "grad_norm": 0.0016858886228874326,
+      "learning_rate": 1.7848174007164237e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18513624,
+      "step": 23695
+    },
+    {
+      "epoch": 49.272349272349274,
+      "grad_norm": 7.327842467930168e-05,
+      "learning_rate": 1.7838767375635052e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18517560,
+      "step": 23700
+    },
+    {
+      "epoch": 49.28274428274428,
+      "grad_norm": 7.314842514460906e-05,
+      "learning_rate": 1.782936184845793e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18521464,
+      "step": 23705
+    },
+    {
+      "epoch": 49.29313929313929,
+      "grad_norm": 7.28085869923234e-05,
+      "learning_rate": 1.7819957427083334e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18525336,
+      "step": 23710
+    },
+    {
+      "epoch": 49.303534303534306,
+      "grad_norm": 2.7827292797155678e-05,
+      "learning_rate": 1.7810554112961516e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18529272,
+      "step": 23715
+    },
+    {
+      "epoch": 49.313929313929314,
+      "grad_norm": 0.0003824725281447172,
+      "learning_rate": 1.7801151907542607e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18533112,
+      "step": 23720
+    },
+    {
+      "epoch": 49.32432432432432,
+      "grad_norm": 7.005767201917479e-06,
+      "learning_rate": 1.7791750812276547e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18537016,
+      "step": 23725
+    },
+    {
+      "epoch": 49.33471933471934,
+      "grad_norm": 3.563404970918782e-05,
+      "learning_rate": 1.778235082861309e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 18540888,
+      "step": 23730
+    },
+    {
+      "epoch": 49.34511434511435,
+      "grad_norm": 8.938922837842256e-05,
+      "learning_rate": 1.777295195800184e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18544760,
+      "step": 23735
+    },
+    {
+      "epoch": 49.355509355509355,
+      "grad_norm": 0.00020838616183027625,
+      "learning_rate": 1.7763554201892215e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18548600,
+      "step": 23740
+    },
+    {
+      "epoch": 49.36590436590436,
+      "grad_norm": 0.000677417207043618,
+      "learning_rate": 1.7754157561733476e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 18552568,
+      "step": 23745
+    },
+    {
+      "epoch": 49.37629937629938,
+      "grad_norm": 5.324198355083354e-05,
+      "learning_rate": 1.7744762038974702e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18556568,
+      "step": 23750
+    },
+    {
+      "epoch": 49.38669438669439,
+      "grad_norm": 1.4688665032736026e-05,
+      "learning_rate": 1.7735367635064788e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18560504,
+      "step": 23755
+    },
+    {
+      "epoch": 49.397089397089395,
+      "grad_norm": 7.816320248821285e-06,
+      "learning_rate": 1.7725974351452474e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 18564280,
+      "step": 23760
+    },
+    {
+      "epoch": 49.40748440748441,
+      "grad_norm": 3.3422515116399154e-05,
+      "learning_rate": 1.771658218958634e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18568152,
+      "step": 23765
+    },
+    {
+      "epoch": 49.41787941787942,
+      "grad_norm": 6.527698860736564e-05,
+      "learning_rate": 1.770719115091475e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18572024,
+      "step": 23770
+    },
+    {
+      "epoch": 49.42827442827443,
+      "grad_norm": 7.703196752117947e-05,
+      "learning_rate": 1.7697801236885935e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18576056,
+      "step": 23775
+    },
+    {
+      "epoch": 49.438669438669436,
+      "grad_norm": 4.5150147343520075e-05,
+      "learning_rate": 1.7688412448947944e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 18579800,
+      "step": 23780
+    },
+    {
+      "epoch": 49.44906444906445,
+      "grad_norm": 2.7146286811330356e-05,
+      "learning_rate": 1.767902478854862e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18583672,
+      "step": 23785
+    },
+    {
+      "epoch": 49.45945945945946,
+      "grad_norm": 0.0001554125628899783,
+      "learning_rate": 1.766963825713569e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18587512,
+      "step": 23790
+    },
+    {
+      "epoch": 49.46985446985447,
+      "grad_norm": 0.00033190587419085205,
+      "learning_rate": 1.766025285615665e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18591480,
+      "step": 23795
+    },
+    {
+      "epoch": 49.48024948024948,
+      "grad_norm": 1.9885746951331384e-05,
+      "learning_rate": 1.7650868587058854e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18595416,
+      "step": 23800
+    },
+    {
+      "epoch": 49.48024948024948,
+      "eval_loss": 0.5578108429908752,
+      "eval_runtime": 11.7075,
+      "eval_samples_per_second": 73.116,
+      "eval_steps_per_second": 18.279,
+      "num_input_tokens_seen": 18595416,
+      "step": 23800
+    },
+    {
+      "epoch": 49.49064449064449,
+      "grad_norm": 3.5725950056075817e-06,
+      "learning_rate": 1.7641485451289484e-05,
+      "loss": 0.007,
+      "num_input_tokens_seen": 18599288,
+      "step": 23805
+    },
+    {
+      "epoch": 49.5010395010395,
+      "grad_norm": 1.624423202883918e-05,
+      "learning_rate": 1.7632103450295534e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18603384,
+      "step": 23810
+    },
+    {
+      "epoch": 49.51143451143451,
+      "grad_norm": 5.270133988233283e-05,
+      "learning_rate": 1.762272258552381e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18607320,
+      "step": 23815
+    },
+    {
+      "epoch": 49.521829521829524,
+      "grad_norm": 5.112049984745681e-05,
+      "learning_rate": 1.7613342858420988e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 18611192,
+      "step": 23820
+    },
+    {
+      "epoch": 49.53222453222453,
+      "grad_norm": 0.13736627995967865,
+      "learning_rate": 1.760396427043351e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 18615160,
+      "step": 23825
+    },
+    {
+      "epoch": 49.54261954261954,
+      "grad_norm": 4.716330295195803e-05,
+      "learning_rate": 1.7594586823007696e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18619128,
+      "step": 23830
+    },
+    {
+      "epoch": 49.553014553014556,
+      "grad_norm": 7.216800440801308e-05,
+      "learning_rate": 1.7585210517589646e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18623128,
+      "step": 23835
+    },
+    {
+      "epoch": 49.563409563409564,
+      "grad_norm": 1.965203182408004e-06,
+      "learning_rate": 1.7575835355625314e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18627064,
+      "step": 23840
+    },
+    {
+      "epoch": 49.57380457380457,
+      "grad_norm": 46.83247756958008,
+      "learning_rate": 1.756646133856048e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 18630840,
+      "step": 23845
+    },
+    {
+      "epoch": 49.58419958419958,
+      "grad_norm": 3.3664538932498544e-05,
+      "learning_rate": 1.7557088467840714e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18634680,
+      "step": 23850
+    },
+    {
+      "epoch": 49.5945945945946,
+      "grad_norm": 0.00017092286725528538,
+      "learning_rate": 1.7547716744911438e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18638552,
+      "step": 23855
+    },
+    {
+      "epoch": 49.604989604989605,
+      "grad_norm": 0.00017264849157072604,
+      "learning_rate": 1.7538346171217902e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18642424,
+      "step": 23860
+    },
+    {
+      "epoch": 49.61538461538461,
+      "grad_norm": 0.00013644543651025742,
+      "learning_rate": 1.7528976748205146e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 18646392,
+      "step": 23865
+    },
+    {
+      "epoch": 49.62577962577963,
+      "grad_norm": 3.4147273254347965e-05,
+      "learning_rate": 1.751960847731807e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 18650264,
+      "step": 23870
+    },
+    {
+      "epoch": 49.63617463617464,
+      "grad_norm": 1.8432547221891582e-05,
+      "learning_rate": 1.7510241360001362e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18654008,
+      "step": 23875
+    },
+    {
+      "epoch": 49.646569646569645,
+      "grad_norm": 1.3797697647532914e-05,
+      "learning_rate": 1.7500875397699562e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18657976,
+      "step": 23880
+    },
+    {
+      "epoch": 49.656964656964654,
+      "grad_norm": 0.0006736028008162975,
+      "learning_rate": 1.7491510591857015e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18661880,
+      "step": 23885
+    },
+    {
+      "epoch": 49.66735966735967,
+      "grad_norm": 2.454956302244682e-05,
+      "learning_rate": 1.7482146943917896e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18665880,
+      "step": 23890
+    },
+    {
+      "epoch": 49.67775467775468,
+      "grad_norm": 4.191500920569524e-06,
+      "learning_rate": 1.7472784455326185e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 18669656,
+      "step": 23895
+    },
+    {
+      "epoch": 49.688149688149686,
+      "grad_norm": 1.524004578590393,
+      "learning_rate": 1.746342312752572e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 18673496,
+      "step": 23900
+    },
+    {
+      "epoch": 49.6985446985447,
+      "grad_norm": 8.398877071158495e-06,
+      "learning_rate": 1.74540629619601e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18677272,
+      "step": 23905
+    },
+    {
+      "epoch": 49.70893970893971,
+      "grad_norm": 0.0038028694689273834,
+      "learning_rate": 1.7444703960072815e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18681144,
+      "step": 23910
+    },
+    {
+      "epoch": 49.71933471933472,
+      "grad_norm": 2.6177032850682735e-05,
+      "learning_rate": 1.7435346123307118e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18685112,
+      "step": 23915
+    },
+    {
+      "epoch": 49.729729729729726,
+      "grad_norm": 0.0001722809683997184,
+      "learning_rate": 1.742598945310611e-05,
+      "loss": 0.0519,
+      "num_input_tokens_seen": 18688856,
+      "step": 23920
+    },
+    {
+      "epoch": 49.74012474012474,
+      "grad_norm": 2.453915476507973e-05,
+      "learning_rate": 1.741663395091272e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18692760,
+      "step": 23925
+    },
+    {
+      "epoch": 49.75051975051975,
+      "grad_norm": 0.00042167925857938826,
+      "learning_rate": 1.7407279618169657e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18696664,
+      "step": 23930
+    },
+    {
+      "epoch": 49.76091476091476,
+      "grad_norm": 0.015479839406907558,
+      "learning_rate": 1.73979264563195e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18700664,
+      "step": 23935
+    },
+    {
+      "epoch": 49.771309771309774,
+      "grad_norm": 0.0020664616022258997,
+      "learning_rate": 1.7388574466804625e-05,
+      "loss": 0.0002,
+      "num_input_tokens_seen": 18704600,
+      "step": 23940
+    },
+    {
+      "epoch": 49.78170478170478,
+      "grad_norm": 0.00025550933787599206,
+      "learning_rate": 1.7379223651067207e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18708504,
+      "step": 23945
+    },
+    {
+      "epoch": 49.79209979209979,
+      "grad_norm": 0.00037920961040072143,
+      "learning_rate": 1.736987401054928e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18712504,
+      "step": 23950
+    },
+    {
+      "epoch": 49.802494802494806,
+      "grad_norm": 1.1551474017323926e-05,
+      "learning_rate": 1.736052554669266e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18716440,
+      "step": 23955
+    },
+    {
+      "epoch": 49.812889812889814,
+      "grad_norm": 4.162968252785504e-05,
+      "learning_rate": 1.7351178260939007e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18720280,
+      "step": 23960
+    },
+    {
+      "epoch": 49.82328482328482,
+      "grad_norm": 0.0008253369014710188,
+      "learning_rate": 1.7341832154729794e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18724312,
+      "step": 23965
+    },
+    {
+      "epoch": 49.83367983367983,
+      "grad_norm": 4.033024379168637e-05,
+      "learning_rate": 1.7332487229506286e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18728216,
+      "step": 23970
+    },
+    {
+      "epoch": 49.84407484407485,
+      "grad_norm": 0.00011368381092324853,
+      "learning_rate": 1.732314348670961e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18732120,
+      "step": 23975
+    },
+    {
+      "epoch": 49.854469854469855,
+      "grad_norm": 0.00025429483503103256,
+      "learning_rate": 1.7313800927780686e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18736056,
+      "step": 23980
+    },
+    {
+      "epoch": 49.86486486486486,
+      "grad_norm": 0.00017634428513702005,
+      "learning_rate": 1.7304459554160245e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18739960,
+      "step": 23985
+    },
+    {
+      "epoch": 49.87525987525988,
+      "grad_norm": 1.77591537067201e-05,
+      "learning_rate": 1.7295119367288853e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18743928,
+      "step": 23990
+    },
+    {
+      "epoch": 49.88565488565489,
+      "grad_norm": 0.0005036984221078455,
+      "learning_rate": 1.728578036860688e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18747800,
+      "step": 23995
+    },
+    {
+      "epoch": 49.896049896049895,
+      "grad_norm": 0.09137123823165894,
+      "learning_rate": 1.7276442559554513e-05,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 18751672,
+      "step": 24000
+    },
+    {
+      "epoch": 49.896049896049895,
+      "eval_loss": 0.4464492201805115,
+      "eval_runtime": 11.7261,
+      "eval_samples_per_second": 72.999,
+      "eval_steps_per_second": 18.25,
+      "num_input_tokens_seen": 18751672,
+      "step": 24000
+    },
+    {
+      "epoch": 49.906444906444904,
+      "grad_norm": 0.0005622492171823978,
+      "learning_rate": 1.726710594157177e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18755544,
+      "step": 24005
+    },
+    {
+      "epoch": 49.91683991683992,
+      "grad_norm": 0.00039234646828845143,
+      "learning_rate": 1.725777051609846e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 18759480,
+      "step": 24010
+    },
+    {
+      "epoch": 49.92723492723493,
+      "grad_norm": 6.830157508375123e-05,
+      "learning_rate": 1.7248436284574228e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 18763352,
+      "step": 24015
+    },
+    {
+      "epoch": 49.937629937629936,
+      "grad_norm": 7.393539272015914e-05,
+      "learning_rate": 1.723910324843855e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 18767160,
+      "step": 24020
+    },
+    {
+      "epoch": 49.94802494802495,
+      "grad_norm": 0.0017183022573590279,
+      "learning_rate": 1.722977140913067e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18770872,
+      "step": 24025
+    },
+    {
+      "epoch": 49.95841995841996,
+      "grad_norm": 0.0015798379899933934,
+      "learning_rate": 1.7220440768089688e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18774616,
+      "step": 24030
+    },
+    {
+      "epoch": 49.96881496881497,
+      "grad_norm": 0.016717582941055298,
+      "learning_rate": 1.7211111326754505e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18778680,
+      "step": 24035
+    },
+    {
+      "epoch": 49.979209979209976,
+      "grad_norm": 1.845407678047195e-05,
+      "learning_rate": 1.720178308656383e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18782552,
+      "step": 24040
+    },
+    {
+      "epoch": 49.98960498960499,
+      "grad_norm": 5.452830737340264e-05,
+      "learning_rate": 1.719245604895621e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 18786552,
+      "step": 24045
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 0.006155683659017086,
+      "learning_rate": 1.7183130215369972e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18790464,
+      "step": 24050
+    },
+    {
+      "epoch": 50.01039501039501,
+      "grad_norm": 0.0008199262665584683,
+      "learning_rate": 1.7173805587243292e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18794400,
+      "step": 24055
+    },
+    {
+      "epoch": 50.020790020790024,
+      "grad_norm": 0.007169474381953478,
+      "learning_rate": 1.7164482166014147e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18798336,
+      "step": 24060
+    },
+    {
+      "epoch": 50.03118503118503,
+      "grad_norm": 0.00014799137716181576,
+      "learning_rate": 1.7155159953120313e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 18802112,
+      "step": 24065
+    },
+    {
+      "epoch": 50.04158004158004,
+      "grad_norm": 9.399287955602631e-05,
+      "learning_rate": 1.714583894999941e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18805984,
+      "step": 24070
+    },
+    {
+      "epoch": 50.05197505197505,
+      "grad_norm": 0.0004952634335495532,
+      "learning_rate": 1.7136519158088826e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 18809984,
+      "step": 24075
+    },
+    {
+      "epoch": 50.062370062370064,
+      "grad_norm": 6.233627209439874e-05,
+      "learning_rate": 1.712720057882581e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18813856,
+      "step": 24080
+    },
+    {
+      "epoch": 50.07276507276507,
+      "grad_norm": 0.0002554133243393153,
+      "learning_rate": 1.7117883213647413e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18817792,
+      "step": 24085
+    },
+    {
+      "epoch": 50.08316008316008,
+      "grad_norm": 4.18501986132469e-05,
+      "learning_rate": 1.710856706399046e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18821568,
+      "step": 24090
+    },
+    {
+      "epoch": 50.093555093555096,
+      "grad_norm": 1.7640119040152058e-05,
+      "learning_rate": 1.7099252131291648e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18825408,
+      "step": 24095
+    },
+    {
+      "epoch": 50.103950103950105,
+      "grad_norm": 2.8865040803793818e-05,
+      "learning_rate": 1.708993841698744e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18829344,
+      "step": 24100
+    },
+    {
+      "epoch": 50.11434511434511,
+      "grad_norm": 0.0002973552036564797,
+      "learning_rate": 1.7080625922514132e-05,
+      "loss": 0.0184,
+      "num_input_tokens_seen": 18833216,
+      "step": 24105
+    },
+    {
+      "epoch": 50.12474012474012,
+      "grad_norm": 5.351654544938356e-05,
+      "learning_rate": 1.7071314649307836e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 18837216,
+      "step": 24110
+    },
+    {
+      "epoch": 50.13513513513514,
+      "grad_norm": 0.00038230957579798996,
+      "learning_rate": 1.7062004598804448e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18841184,
+      "step": 24115
+    },
+    {
+      "epoch": 50.145530145530145,
+      "grad_norm": 0.0001530943118268624,
+      "learning_rate": 1.7052695772439702e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 18845024,
+      "step": 24120
+    },
+    {
+      "epoch": 50.15592515592515,
+      "grad_norm": 0.0003382140421308577,
+      "learning_rate": 1.7043388171649154e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 18848832,
+      "step": 24125
+    },
+    {
+      "epoch": 50.16632016632017,
+      "grad_norm": 6.662511441390961e-05,
+      "learning_rate": 1.7034081797868127e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18852704,
+      "step": 24130
+    },
+    {
+      "epoch": 50.17671517671518,
+      "grad_norm": 0.0003982293710578233,
+      "learning_rate": 1.70247766525318e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18856512,
+      "step": 24135
+    },
+    {
+      "epoch": 50.187110187110186,
+      "grad_norm": 0.00017842241504695266,
+      "learning_rate": 1.701547273707514e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18860352,
+      "step": 24140
+    },
+    {
+      "epoch": 50.197505197505194,
+      "grad_norm": 0.00018989147793035954,
+      "learning_rate": 1.7006170052932916e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18864224,
+      "step": 24145
+    },
+    {
+      "epoch": 50.20790020790021,
+      "grad_norm": 0.17211249470710754,
+      "learning_rate": 1.6996868601539735e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 18868032,
+      "step": 24150
+    },
+    {
+      "epoch": 50.21829521829522,
+      "grad_norm": 1.8864393496187404e-05,
+      "learning_rate": 1.6987568384329977e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18871872,
+      "step": 24155
+    },
+    {
+      "epoch": 50.228690228690226,
+      "grad_norm": 4.238690962665714e-05,
+      "learning_rate": 1.6978269402737866e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18875712,
+      "step": 24160
+    },
+    {
+      "epoch": 50.23908523908524,
+      "grad_norm": 0.0006556712323799729,
+      "learning_rate": 1.696897165819743e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18879712,
+      "step": 24165
+    },
+    {
+      "epoch": 50.24948024948025,
+      "grad_norm": 3.353905049152672e-05,
+      "learning_rate": 1.6959675152142487e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18883680,
+      "step": 24170
+    },
+    {
+      "epoch": 50.25987525987526,
+      "grad_norm": 5.4734948207624257e-05,
+      "learning_rate": 1.6950379886006667e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18887616,
+      "step": 24175
+    },
+    {
+      "epoch": 50.270270270270274,
+      "grad_norm": 0.0008601880981586874,
+      "learning_rate": 1.6941085861223438e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18891648,
+      "step": 24180
+    },
+    {
+      "epoch": 50.28066528066528,
+      "grad_norm": 0.0001522926177131012,
+      "learning_rate": 1.6931793079226034e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18895360,
+      "step": 24185
+    },
+    {
+      "epoch": 50.29106029106029,
+      "grad_norm": 0.0014350449200719595,
+      "learning_rate": 1.692250154144754e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18899200,
+      "step": 24190
+    },
+    {
+      "epoch": 50.3014553014553,
+      "grad_norm": 0.002095304662361741,
+      "learning_rate": 1.6913211249320807e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18903040,
+      "step": 24195
+    },
+    {
+      "epoch": 50.311850311850314,
+      "grad_norm": 2.3682119717705064e-05,
+      "learning_rate": 1.6903922204278522e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18906848,
+      "step": 24200
+    },
+    {
+      "epoch": 50.311850311850314,
+      "eval_loss": 0.4530361592769623,
+      "eval_runtime": 11.6974,
+      "eval_samples_per_second": 73.178,
+      "eval_steps_per_second": 18.295,
+      "num_input_tokens_seen": 18906848,
+      "step": 24200
+    },
+    {
+      "epoch": 50.32224532224532,
+      "grad_norm": 4.0099672332871705e-05,
+      "learning_rate": 1.6894634407753186e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18910880,
+      "step": 24205
+    },
+    {
+      "epoch": 50.33264033264033,
+      "grad_norm": 5.406565833254717e-05,
+      "learning_rate": 1.6885347861177077e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18914944,
+      "step": 24210
+    },
+    {
+      "epoch": 50.343035343035346,
+      "grad_norm": 0.000444029486970976,
+      "learning_rate": 1.6876062565982298e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18918880,
+      "step": 24215
+    },
+    {
+      "epoch": 50.353430353430355,
+      "grad_norm": 3.620945062721148e-05,
+      "learning_rate": 1.6866778523600774e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18922656,
+      "step": 24220
+    },
+    {
+      "epoch": 50.36382536382536,
+      "grad_norm": 0.00011289483518339694,
+      "learning_rate": 1.6857495735464195e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18926720,
+      "step": 24225
+    },
+    {
+      "epoch": 50.37422037422037,
+      "grad_norm": 0.00042590891825966537,
+      "learning_rate": 1.6848214203004115e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18930624,
+      "step": 24230
+    },
+    {
+      "epoch": 50.38461538461539,
+      "grad_norm": 3.826674219453707e-05,
+      "learning_rate": 1.6838933927651835e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18934560,
+      "step": 24235
+    },
+    {
+      "epoch": 50.395010395010395,
+      "grad_norm": 0.1356697976589203,
+      "learning_rate": 1.6829654910838506e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 18938400,
+      "step": 24240
+    },
+    {
+      "epoch": 50.4054054054054,
+      "grad_norm": 0.15824513137340546,
+      "learning_rate": 1.6820377153995065e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 18942368,
+      "step": 24245
+    },
+    {
+      "epoch": 50.41580041580042,
+      "grad_norm": 0.0003070566162932664,
+      "learning_rate": 1.681110065855226e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18946336,
+      "step": 24250
+    },
+    {
+      "epoch": 50.42619542619543,
+      "grad_norm": 0.18826903402805328,
+      "learning_rate": 1.6801825425940642e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 18950304,
+      "step": 24255
+    },
+    {
+      "epoch": 50.436590436590436,
+      "grad_norm": 7.406516670016572e-05,
+      "learning_rate": 1.679255145759056e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18954304,
+      "step": 24260
+    },
+    {
+      "epoch": 50.446985446985444,
+      "grad_norm": 0.00018810995970852673,
+      "learning_rate": 1.6783278754932187e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18958112,
+      "step": 24265
+    },
+    {
+      "epoch": 50.45738045738046,
+      "grad_norm": 0.000978575088083744,
+      "learning_rate": 1.6774007319395496e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18961920,
+      "step": 24270
+    },
+    {
+      "epoch": 50.46777546777547,
+      "grad_norm": 1.2044565664837137e-05,
+      "learning_rate": 1.6764737152410243e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18965856,
+      "step": 24275
+    },
+    {
+      "epoch": 50.478170478170476,
+      "grad_norm": 0.00011942642595386133,
+      "learning_rate": 1.6755468255406016e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 18969696,
+      "step": 24280
+    },
+    {
+      "epoch": 50.48856548856549,
+      "grad_norm": 0.00037469383096322417,
+      "learning_rate": 1.674620062981219e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18973792,
+      "step": 24285
+    },
+    {
+      "epoch": 50.4989604989605,
+      "grad_norm": 1.0478056537976954e-05,
+      "learning_rate": 1.6736934277057947e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18977600,
+      "step": 24290
+    },
+    {
+      "epoch": 50.50935550935551,
+      "grad_norm": 8.559552952647209e-05,
+      "learning_rate": 1.6727669198572286e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18981568,
+      "step": 24295
+    },
+    {
+      "epoch": 50.51975051975052,
+      "grad_norm": 0.00010020992340287194,
+      "learning_rate": 1.6718405395783984e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 18985440,
+      "step": 24300
+    },
+    {
+      "epoch": 50.53014553014553,
+      "grad_norm": 8.512655040249228e-05,
+      "learning_rate": 1.6709142870121643e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 18989440,
+      "step": 24305
+    },
+    {
+      "epoch": 50.54054054054054,
+      "grad_norm": 1.1761820132960565e-05,
+      "learning_rate": 1.669988162301367e-05,
+      "loss": 0.0067,
+      "num_input_tokens_seen": 18993344,
+      "step": 24310
+    },
+    {
+      "epoch": 50.55093555093555,
+      "grad_norm": 0.15655384957790375,
+      "learning_rate": 1.6690621655888243e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 18997184,
+      "step": 24315
+    },
+    {
+      "epoch": 50.561330561330564,
+      "grad_norm": 8.254167914856225e-05,
+      "learning_rate": 1.6681362970173386e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 19001184,
+      "step": 24320
+    },
+    {
+      "epoch": 50.57172557172557,
+      "grad_norm": 5.7314820878673345e-05,
+      "learning_rate": 1.6672105567296904e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19005216,
+      "step": 24325
+    },
+    {
+      "epoch": 50.58212058212058,
+      "grad_norm": 0.00014788770931772888,
+      "learning_rate": 1.666284944868639e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19009216,
+      "step": 24330
+    },
+    {
+      "epoch": 50.59251559251559,
+      "grad_norm": 0.00012394029181450605,
+      "learning_rate": 1.665359461576927e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19013216,
+      "step": 24335
+    },
+    {
+      "epoch": 50.602910602910605,
+      "grad_norm": 0.0002557918487582356,
+      "learning_rate": 1.6644341069972736e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19017120,
+      "step": 24340
+    },
+    {
+      "epoch": 50.61330561330561,
+      "grad_norm": 5.196575784793822e-06,
+      "learning_rate": 1.6635088812723813e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19021088,
+      "step": 24345
+    },
+    {
+      "epoch": 50.62370062370062,
+      "grad_norm": 9.656909242039546e-05,
+      "learning_rate": 1.6625837845449328e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19025152,
+      "step": 24350
+    },
+    {
+      "epoch": 50.63409563409564,
+      "grad_norm": 6.069087976356968e-05,
+      "learning_rate": 1.6616588169575874e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19028864,
+      "step": 24355
+    },
+    {
+      "epoch": 50.644490644490645,
+      "grad_norm": 0.0016829530941322446,
+      "learning_rate": 1.6607339786529878e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19032800,
+      "step": 24360
+    },
+    {
+      "epoch": 50.65488565488565,
+      "grad_norm": 5.074846922070719e-05,
+      "learning_rate": 1.659809269773756e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19036672,
+      "step": 24365
+    },
+    {
+      "epoch": 50.66528066528066,
+      "grad_norm": 3.179465056746267e-05,
+      "learning_rate": 1.658884690462493e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19040544,
+      "step": 24370
+    },
+    {
+      "epoch": 50.67567567567568,
+      "grad_norm": 0.0025211048778146505,
+      "learning_rate": 1.6579602408617813e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19044544,
+      "step": 24375
+    },
+    {
+      "epoch": 50.686070686070686,
+      "grad_norm": 0.0004782506439369172,
+      "learning_rate": 1.657035921114181e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19048512,
+      "step": 24380
+    },
+    {
+      "epoch": 50.696465696465694,
+      "grad_norm": 3.1205505365505815e-05,
+      "learning_rate": 1.656111731362236e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 19052416,
+      "step": 24385
+    },
+    {
+      "epoch": 50.70686070686071,
+      "grad_norm": 0.0004036475729662925,
+      "learning_rate": 1.6551876717484666e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19056416,
+      "step": 24390
+    },
+    {
+      "epoch": 50.71725571725572,
+      "grad_norm": 0.00023797598259989172,
+      "learning_rate": 1.6542637424153752e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19060256,
+      "step": 24395
+    },
+    {
+      "epoch": 50.727650727650726,
+      "grad_norm": 5.417058855528012e-05,
+      "learning_rate": 1.6533399435054418e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19064192,
+      "step": 24400
+    },
+    {
+      "epoch": 50.727650727650726,
+      "eval_loss": 0.4825330972671509,
+      "eval_runtime": 11.7088,
+      "eval_samples_per_second": 73.107,
+      "eval_steps_per_second": 18.277,
+      "num_input_tokens_seen": 19064192,
+      "step": 24400
+    },
+    {
+      "epoch": 50.73804573804574,
+      "grad_norm": 1.2590202459250577e-05,
+      "learning_rate": 1.6524162751611304e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19068160,
+      "step": 24405
+    },
+    {
+      "epoch": 50.74844074844075,
+      "grad_norm": 1.472567419114057e-05,
+      "learning_rate": 1.6514927375248796e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19072064,
+      "step": 24410
+    },
+    {
+      "epoch": 50.75883575883576,
+      "grad_norm": 1.5618617908330634e-05,
+      "learning_rate": 1.6505693307391127e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19075968,
+      "step": 24415
+    },
+    {
+      "epoch": 50.76923076923077,
+      "grad_norm": 0.003326207399368286,
+      "learning_rate": 1.6496460549462288e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19079904,
+      "step": 24420
+    },
+    {
+      "epoch": 50.77962577962578,
+      "grad_norm": 0.00012773170601576567,
+      "learning_rate": 1.6487229102886097e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19083904,
+      "step": 24425
+    },
+    {
+      "epoch": 50.79002079002079,
+      "grad_norm": 0.0007392203551717103,
+      "learning_rate": 1.6477998969086155e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19087808,
+      "step": 24430
+    },
+    {
+      "epoch": 50.8004158004158,
+      "grad_norm": 0.00010354300320614129,
+      "learning_rate": 1.646877014948587e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19091744,
+      "step": 24435
+    },
+    {
+      "epoch": 50.810810810810814,
+      "grad_norm": 0.02315804362297058,
+      "learning_rate": 1.6459542645508433e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19095840,
+      "step": 24440
+    },
+    {
+      "epoch": 50.82120582120582,
+      "grad_norm": 2.4158545784302987e-05,
+      "learning_rate": 1.6450316458576852e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19099680,
+      "step": 24445
+    },
+    {
+      "epoch": 50.83160083160083,
+      "grad_norm": 0.00019999819051008672,
+      "learning_rate": 1.6441091590113912e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19103520,
+      "step": 24450
+    },
+    {
+      "epoch": 50.84199584199584,
+      "grad_norm": 0.00020316193695180118,
+      "learning_rate": 1.6431868041542213e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 19107392,
+      "step": 24455
+    },
+    {
+      "epoch": 50.852390852390855,
+      "grad_norm": 1.1194607395736966e-05,
+      "learning_rate": 1.6422645814284123e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19111232,
+      "step": 24460
+    },
+    {
+      "epoch": 50.86278586278586,
+      "grad_norm": 0.0002700036275200546,
+      "learning_rate": 1.6413424909761846e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19115168,
+      "step": 24465
+    },
+    {
+      "epoch": 50.87318087318087,
+      "grad_norm": 1.782187791832257e-05,
+      "learning_rate": 1.640420532939736e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19118976,
+      "step": 24470
+    },
+    {
+      "epoch": 50.88357588357589,
+      "grad_norm": 4.707912739831954e-05,
+      "learning_rate": 1.639498707461242e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19122784,
+      "step": 24475
+    },
+    {
+      "epoch": 50.893970893970895,
+      "grad_norm": 0.0002709157415665686,
+      "learning_rate": 1.6385770146828614e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 19126624,
+      "step": 24480
+    },
+    {
+      "epoch": 50.9043659043659,
+      "grad_norm": 0.00013787171337753534,
+      "learning_rate": 1.637655454746731e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19130464,
+      "step": 24485
+    },
+    {
+      "epoch": 50.91476091476091,
+      "grad_norm": 0.0005835352931171656,
+      "learning_rate": 1.6367340277949658e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19134176,
+      "step": 24490
+    },
+    {
+      "epoch": 50.92515592515593,
+      "grad_norm": 6.441454024752602e-05,
+      "learning_rate": 1.635812733969663e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19138176,
+      "step": 24495
+    },
+    {
+      "epoch": 50.935550935550935,
+      "grad_norm": 3.870148429996334e-05,
+      "learning_rate": 1.634891573412896e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19142048,
+      "step": 24500
+    },
+    {
+      "epoch": 50.945945945945944,
+      "grad_norm": 0.0036225193180143833,
+      "learning_rate": 1.6339705462667196e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19145952,
+      "step": 24505
+    },
+    {
+      "epoch": 50.95634095634096,
+      "grad_norm": 2.8279693651711568e-05,
+      "learning_rate": 1.633049652673169e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19149792,
+      "step": 24510
+    },
+    {
+      "epoch": 50.96673596673597,
+      "grad_norm": 1.1991761311946902e-05,
+      "learning_rate": 1.632128892774256e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19153568,
+      "step": 24515
+    },
+    {
+      "epoch": 50.977130977130976,
+      "grad_norm": 0.000374707393348217,
+      "learning_rate": 1.6312082667119737e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19157408,
+      "step": 24520
+    },
+    {
+      "epoch": 50.987525987525984,
+      "grad_norm": 8.223004988394678e-05,
+      "learning_rate": 1.630287774628296e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 19161376,
+      "step": 24525
+    },
+    {
+      "epoch": 50.997920997921,
+      "grad_norm": 0.1357693374156952,
+      "learning_rate": 1.6293674166651718e-05,
+      "loss": 0.0061,
+      "num_input_tokens_seen": 19165248,
+      "step": 24530
+    },
+    {
+      "epoch": 51.00831600831601,
+      "grad_norm": 0.00016208783199544996,
+      "learning_rate": 1.6284471929645338e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19169168,
+      "step": 24535
+    },
+    {
+      "epoch": 51.018711018711016,
+      "grad_norm": 3.866818587994203e-05,
+      "learning_rate": 1.627527103668291e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19173200,
+      "step": 24540
+    },
+    {
+      "epoch": 51.02910602910603,
+      "grad_norm": 4.8964069719659165e-05,
+      "learning_rate": 1.6266071489183327e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19177040,
+      "step": 24545
+    },
+    {
+      "epoch": 51.03950103950104,
+      "grad_norm": 8.357252954738215e-05,
+      "learning_rate": 1.6256873288565283e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19181008,
+      "step": 24550
+    },
+    {
+      "epoch": 51.04989604989605,
+      "grad_norm": 9.880377183435485e-05,
+      "learning_rate": 1.6247676436247245e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19185040,
+      "step": 24555
+    },
+    {
+      "epoch": 51.06029106029106,
+      "grad_norm": 0.00019356640405021608,
+      "learning_rate": 1.6238480933647486e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19188880,
+      "step": 24560
+    },
+    {
+      "epoch": 51.07068607068607,
+      "grad_norm": 0.13146018981933594,
+      "learning_rate": 1.6229286782184083e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 19192688,
+      "step": 24565
+    },
+    {
+      "epoch": 51.08108108108108,
+      "grad_norm": 0.10451217740774155,
+      "learning_rate": 1.622009398327487e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 19196528,
+      "step": 24570
+    },
+    {
+      "epoch": 51.09147609147609,
+      "grad_norm": 0.00019624881679192185,
+      "learning_rate": 1.6210902538337502e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19200432,
+      "step": 24575
+    },
+    {
+      "epoch": 51.101871101871104,
+      "grad_norm": 9.390337800141424e-05,
+      "learning_rate": 1.6201712448789413e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19204336,
+      "step": 24580
+    },
+    {
+      "epoch": 51.11226611226611,
+      "grad_norm": 0.00015143769269343466,
+      "learning_rate": 1.6192523716047827e-05,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 19208208,
+      "step": 24585
+    },
+    {
+      "epoch": 51.12266112266112,
+      "grad_norm": 0.00011082330456702039,
+      "learning_rate": 1.6183336341529776e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19212176,
+      "step": 24590
+    },
+    {
+      "epoch": 51.13305613305613,
+      "grad_norm": 0.00034761911956593394,
+      "learning_rate": 1.6174150326652047e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19216048,
+      "step": 24595
+    },
+    {
+      "epoch": 51.143451143451145,
+      "grad_norm": 9.270855662180111e-05,
+      "learning_rate": 1.6164965672831256e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19219856,
+      "step": 24600
+    },
+    {
+      "epoch": 51.143451143451145,
+      "eval_loss": 0.4913453757762909,
+      "eval_runtime": 11.7113,
+      "eval_samples_per_second": 73.092,
+      "eval_steps_per_second": 18.273,
+      "num_input_tokens_seen": 19219856,
+      "step": 24600
+    },
+    {
+      "epoch": 51.15384615384615,
+      "grad_norm": 6.967059016460553e-05,
+      "learning_rate": 1.6155782381483784e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19223792,
+      "step": 24605
+    },
+    {
+      "epoch": 51.16424116424116,
+      "grad_norm": 4.308867119107163e-06,
+      "learning_rate": 1.6146600454025813e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19227632,
+      "step": 24610
+    },
+    {
+      "epoch": 51.17463617463618,
+      "grad_norm": 8.439717930741608e-06,
+      "learning_rate": 1.6137419891873317e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19231536,
+      "step": 24615
+    },
+    {
+      "epoch": 51.185031185031185,
+      "grad_norm": 1.1606390216911677e-05,
+      "learning_rate": 1.6128240696442038e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19235568,
+      "step": 24620
+    },
+    {
+      "epoch": 51.195426195426194,
+      "grad_norm": 5.5880464060464874e-05,
+      "learning_rate": 1.611906286914753e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19239280,
+      "step": 24625
+    },
+    {
+      "epoch": 51.20582120582121,
+      "grad_norm": 0.0004107923014089465,
+      "learning_rate": 1.6109886411405144e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19242992,
+      "step": 24630
+    },
+    {
+      "epoch": 51.21621621621622,
+      "grad_norm": 3.7663568946300074e-05,
+      "learning_rate": 1.6100711324629985e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19246992,
+      "step": 24635
+    },
+    {
+      "epoch": 51.226611226611226,
+      "grad_norm": 3.335276778670959e-05,
+      "learning_rate": 1.609153761023698e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 19250960,
+      "step": 24640
+    },
+    {
+      "epoch": 51.237006237006234,
+      "grad_norm": 2.8801798180211335e-05,
+      "learning_rate": 1.608236526964083e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19254864,
+      "step": 24645
+    },
+    {
+      "epoch": 51.24740124740125,
+      "grad_norm": 2.8959681003470905e-05,
+      "learning_rate": 1.607319430425601e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19259056,
+      "step": 24650
+    },
+    {
+      "epoch": 51.25779625779626,
+      "grad_norm": 0.0024583416525274515,
+      "learning_rate": 1.606402471549682e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19262832,
+      "step": 24655
+    },
+    {
+      "epoch": 51.268191268191266,
+      "grad_norm": 0.0003965949872508645,
+      "learning_rate": 1.6054856504777312e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19266672,
+      "step": 24660
+    },
+    {
+      "epoch": 51.27858627858628,
+      "grad_norm": 0.005341534502804279,
+      "learning_rate": 1.6045689673511334e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19270608,
+      "step": 24665
+    },
+    {
+      "epoch": 51.28898128898129,
+      "grad_norm": 0.0014809112763032317,
+      "learning_rate": 1.6036524223112548e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19274608,
+      "step": 24670
+    },
+    {
+      "epoch": 51.2993762993763,
+      "grad_norm": 1.4668246876681224e-05,
+      "learning_rate": 1.602736015499436e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19278512,
+      "step": 24675
+    },
+    {
+      "epoch": 51.30977130977131,
+      "grad_norm": 3.2403593650087714e-05,
+      "learning_rate": 1.601819747057e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19282416,
+      "step": 24680
+    },
+    {
+      "epoch": 51.32016632016632,
+      "grad_norm": 0.00014363942318595946,
+      "learning_rate": 1.6009036171252465e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19286192,
+      "step": 24685
+    },
+    {
+      "epoch": 51.33056133056133,
+      "grad_norm": 0.0007600771496072412,
+      "learning_rate": 1.599987625845453e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19290256,
+      "step": 24690
+    },
+    {
+      "epoch": 51.34095634095634,
+      "grad_norm": 5.493756361829583e-06,
+      "learning_rate": 1.599071773358879e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19294192,
+      "step": 24695
+    },
+    {
+      "epoch": 51.351351351351354,
+      "grad_norm": 1.7072774198823026e-06,
+      "learning_rate": 1.598156059806758e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19297968,
+      "step": 24700
+    },
+    {
+      "epoch": 51.36174636174636,
+      "grad_norm": 6.356921949191019e-05,
+      "learning_rate": 1.5972404853303062e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19301840,
+      "step": 24705
+    },
+    {
+      "epoch": 51.37214137214137,
+      "grad_norm": 5.377722118282691e-06,
+      "learning_rate": 1.5963250500707172e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19305616,
+      "step": 24710
+    },
+    {
+      "epoch": 51.38253638253638,
+      "grad_norm": 4.126340718357824e-05,
+      "learning_rate": 1.5954097541691612e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19309584,
+      "step": 24715
+    },
+    {
+      "epoch": 51.392931392931395,
+      "grad_norm": 7.881220517447218e-05,
+      "learning_rate": 1.5944945977667884e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19313456,
+      "step": 24720
+    },
+    {
+      "epoch": 51.4033264033264,
+      "grad_norm": 3.714083140948787e-05,
+      "learning_rate": 1.593579581004729e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19317520,
+      "step": 24725
+    },
+    {
+      "epoch": 51.41372141372141,
+      "grad_norm": 0.0001926444674609229,
+      "learning_rate": 1.592664704024088e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19321424,
+      "step": 24730
+    },
+    {
+      "epoch": 51.42411642411643,
+      "grad_norm": 2.3670823793509044e-05,
+      "learning_rate": 1.591749966965953e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 19325392,
+      "step": 24735
+    },
+    {
+      "epoch": 51.434511434511435,
+      "grad_norm": 0.000509709760081023,
+      "learning_rate": 1.5908353699713856e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19329200,
+      "step": 24740
+    },
+    {
+      "epoch": 51.444906444906444,
+      "grad_norm": 0.007988718338310719,
+      "learning_rate": 1.5899209131814298e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19333200,
+      "step": 24745
+    },
+    {
+      "epoch": 51.45530145530145,
+      "grad_norm": 5.361893636290915e-05,
+      "learning_rate": 1.5890065967371067e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19337200,
+      "step": 24750
+    },
+    {
+      "epoch": 51.46569646569647,
+      "grad_norm": 0.00016752135707065463,
+      "learning_rate": 1.5880924207794144e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19341200,
+      "step": 24755
+    },
+    {
+      "epoch": 51.476091476091476,
+      "grad_norm": 9.191124991048127e-05,
+      "learning_rate": 1.5871783854493298e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19345200,
+      "step": 24760
+    },
+    {
+      "epoch": 51.486486486486484,
+      "grad_norm": 0.00017040970851667225,
+      "learning_rate": 1.5862644908878106e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19349136,
+      "step": 24765
+    },
+    {
+      "epoch": 51.4968814968815,
+      "grad_norm": 0.00020416783809196204,
+      "learning_rate": 1.5853507372357885e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19353040,
+      "step": 24770
+    },
+    {
+      "epoch": 51.50727650727651,
+      "grad_norm": 6.627016409765929e-05,
+      "learning_rate": 1.5844371246341776e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19356912,
+      "step": 24775
+    },
+    {
+      "epoch": 51.517671517671516,
+      "grad_norm": 1.0840996765182354e-05,
+      "learning_rate": 1.5835236532238674e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19360912,
+      "step": 24780
+    },
+    {
+      "epoch": 51.528066528066525,
+      "grad_norm": 3.7717392842750996e-05,
+      "learning_rate": 1.582610323145727e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19364816,
+      "step": 24785
+    },
+    {
+      "epoch": 51.53846153846154,
+      "grad_norm": 2.1910682335146703e-05,
+      "learning_rate": 1.5816971345406035e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19368688,
+      "step": 24790
+    },
+    {
+      "epoch": 51.54885654885655,
+      "grad_norm": 0.0006053996621631086,
+      "learning_rate": 1.5807840875493225e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19372624,
+      "step": 24795
+    },
+    {
+      "epoch": 51.55925155925156,
+      "grad_norm": 1.933919975272147e-06,
+      "learning_rate": 1.5798711823126854e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19376464,
+      "step": 24800
+    },
+    {
+      "epoch": 51.55925155925156,
+      "eval_loss": 0.5009268522262573,
+      "eval_runtime": 11.6893,
+      "eval_samples_per_second": 73.229,
+      "eval_steps_per_second": 18.307,
+      "num_input_tokens_seen": 19376464,
+      "step": 24800
+    },
+    {
+      "epoch": 51.56964656964657,
+      "grad_norm": 6.459219730459154e-05,
+      "learning_rate": 1.578958418971477e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19380240,
+      "step": 24805
+    },
+    {
+      "epoch": 51.58004158004158,
+      "grad_norm": 2.7568270525080152e-05,
+      "learning_rate": 1.578045797666453e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19384080,
+      "step": 24810
+    },
+    {
+      "epoch": 51.59043659043659,
+      "grad_norm": 0.0001217473327415064,
+      "learning_rate": 1.5771333185383548e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 19388112,
+      "step": 24815
+    },
+    {
+      "epoch": 51.6008316008316,
+      "grad_norm": 0.0010741878068074584,
+      "learning_rate": 1.576220981727895e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19391952,
+      "step": 24820
+    },
+    {
+      "epoch": 51.61122661122661,
+      "grad_norm": 5.542473081732169e-05,
+      "learning_rate": 1.575308787375769e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 19395792,
+      "step": 24825
+    },
+    {
+      "epoch": 51.62162162162162,
+      "grad_norm": 0.00033460353733971715,
+      "learning_rate": 1.5743967356226492e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19399696,
+      "step": 24830
+    },
+    {
+      "epoch": 51.63201663201663,
+      "grad_norm": 2.1969683075440116e-05,
+      "learning_rate": 1.5734848266091835e-05,
+      "loss": 0.0001,
+      "num_input_tokens_seen": 19403696,
+      "step": 24835
+    },
+    {
+      "epoch": 51.642411642411645,
+      "grad_norm": 0.10573029518127441,
+      "learning_rate": 1.572573060476001e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 19407664,
+      "step": 24840
+    },
+    {
+      "epoch": 51.65280665280665,
+      "grad_norm": 7.552927127107978e-05,
+      "learning_rate": 1.5716614373637085e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19411632,
+      "step": 24845
+    },
+    {
+      "epoch": 51.66320166320166,
+      "grad_norm": 0.000106684303318616,
+      "learning_rate": 1.570749957412887e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19415568,
+      "step": 24850
+    },
+    {
+      "epoch": 51.67359667359668,
+      "grad_norm": 0.0007586849387735128,
+      "learning_rate": 1.5698386207641013e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19419312,
+      "step": 24855
+    },
+    {
+      "epoch": 51.683991683991685,
+      "grad_norm": 0.00016211200272664428,
+      "learning_rate": 1.5689274275578884e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19423280,
+      "step": 24860
+    },
+    {
+      "epoch": 51.694386694386694,
+      "grad_norm": 5.9925801906501874e-05,
+      "learning_rate": 1.5680163779347667e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19427280,
+      "step": 24865
+    },
+    {
+      "epoch": 51.7047817047817,
+      "grad_norm": 3.084196214331314e-05,
+      "learning_rate": 1.5671054720352327e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19431280,
+      "step": 24870
+    },
+    {
+      "epoch": 51.71517671517672,
+      "grad_norm": 6.318907981039956e-05,
+      "learning_rate": 1.566194709999757e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 19435120,
+      "step": 24875
+    },
+    {
+      "epoch": 51.725571725571726,
+      "grad_norm": 2.022091757680755e-05,
+      "learning_rate": 1.5652840919687933e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19438928,
+      "step": 24880
+    },
+    {
+      "epoch": 51.735966735966734,
+      "grad_norm": 6.585849041584879e-05,
+      "learning_rate": 1.5643736180827676e-05,
+      "loss": 0.0048,
+      "num_input_tokens_seen": 19442960,
+      "step": 24885
+    },
+    {
+      "epoch": 51.74636174636175,
+      "grad_norm": 2.319141640327871e-05,
+      "learning_rate": 1.5634632884820878e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19446736,
+      "step": 24890
+    },
+    {
+      "epoch": 51.75675675675676,
+      "grad_norm": 1.303243789152475e-05,
+      "learning_rate": 1.5625531033071395e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 19450608,
+      "step": 24895
+    },
+    {
+      "epoch": 51.767151767151766,
+      "grad_norm": 7.10733438609168e-05,
+      "learning_rate": 1.5616430626982828e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19454512,
+      "step": 24900
+    },
+    {
+      "epoch": 51.777546777546775,
+      "grad_norm": 4.112203896511346e-05,
+      "learning_rate": 1.5607331667958575e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 19458448,
+      "step": 24905
+    },
+    {
+      "epoch": 51.78794178794179,
+      "grad_norm": 9.191443496092688e-06,
+      "learning_rate": 1.5598234157401824e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19462640,
+      "step": 24910
+    },
+    {
+      "epoch": 51.7983367983368,
+      "grad_norm": 4.671305214287713e-05,
+      "learning_rate": 1.5589138096715503e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 19466576,
+      "step": 24915
+    },
+    {
+      "epoch": 51.80873180873181,
+      "grad_norm": 0.007930667139589787,
+      "learning_rate": 1.5580043487302365e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19470512,
+      "step": 24920
+    },
+    {
+      "epoch": 51.81912681912682,
+      "grad_norm": 4.120919038541615e-05,
+      "learning_rate": 1.5570950330564888e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 19474512,
+      "step": 24925
+    },
+    {
+      "epoch": 51.82952182952183,
+      "grad_norm": 0.0001797623699530959,
+      "learning_rate": 1.5561858627905367e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19478448,
+      "step": 24930
+    },
+    {
+      "epoch": 51.83991683991684,
+      "grad_norm": 0.0010962719097733498,
+      "learning_rate": 1.5552768380725857e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 19482352,
+      "step": 24935
+    },
+    {
+      "epoch": 51.85031185031185,
+      "grad_norm": 2.0728144590975717e-05,
+      "learning_rate": 1.5543679590428183e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19486192,
+      "step": 24940
+    },
+    {
+      "epoch": 51.86070686070686,
+      "grad_norm": 4.732411616714671e-05,
+      "learning_rate": 1.5534592258413943e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19490096,
+      "step": 24945
+    },
+    {
+      "epoch": 51.87110187110187,
+      "grad_norm": 6.32482478977181e-05,
+      "learning_rate": 1.5525506386084538e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19493968,
+      "step": 24950
+    },
+    {
+      "epoch": 51.88149688149688,
+      "grad_norm": 0.00011055392678827047,
+      "learning_rate": 1.55164219748411e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19498000,
+      "step": 24955
+    },
+    {
+      "epoch": 51.891891891891895,
+      "grad_norm": 1.4204683793650474e-05,
+      "learning_rate": 1.550733902608459e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19501904,
+      "step": 24960
+    },
+    {
+      "epoch": 51.9022869022869,
+      "grad_norm": 2.1329229639377445e-05,
+      "learning_rate": 1.549825754121568e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19505712,
+      "step": 24965
+    },
+    {
+      "epoch": 51.91268191268191,
+      "grad_norm": 0.005506077315658331,
+      "learning_rate": 1.5489177521634864e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19509520,
+      "step": 24970
+    },
+    {
+      "epoch": 51.92307692307692,
+      "grad_norm": 0.00016556914488319308,
+      "learning_rate": 1.5480098968742402e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 19513360,
+      "step": 24975
+    },
+    {
+      "epoch": 51.933471933471935,
+      "grad_norm": 0.00012903503375127912,
+      "learning_rate": 1.5471021883938304e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19517200,
+      "step": 24980
+    },
+    {
+      "epoch": 51.943866943866944,
+      "grad_norm": 4.404856008477509e-05,
+      "learning_rate": 1.546194626862238e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19520912,
+      "step": 24985
+    },
+    {
+      "epoch": 51.95426195426195,
+      "grad_norm": 4.918393187836045e-06,
+      "learning_rate": 1.5452872124194216e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 19524720,
+      "step": 24990
+    },
+    {
+      "epoch": 51.96465696465697,
+      "grad_norm": 8.039517706492916e-05,
+      "learning_rate": 1.5443799452053136e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19528560,
+      "step": 24995
+    },
+    {
+      "epoch": 51.975051975051976,
+      "grad_norm": 2.242345544800628e-05,
+      "learning_rate": 1.543472825359828e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19532272,
+      "step": 25000
+    },
+    {
+      "epoch": 51.975051975051976,
+      "eval_loss": 0.5191696286201477,
+      "eval_runtime": 11.719,
+      "eval_samples_per_second": 73.044,
+      "eval_steps_per_second": 18.261,
+      "num_input_tokens_seen": 19532272,
+      "step": 25000
+    },
+    {
+      "epoch": 51.985446985446984,
+      "grad_norm": 2.3468002837034874e-05,
+      "learning_rate": 1.5425658530228522e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19536144,
+      "step": 25005
+    },
+    {
+      "epoch": 51.99584199584199,
+      "grad_norm": 3.054193439311348e-05,
+      "learning_rate": 1.5416590283342546e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19540272,
+      "step": 25010
+    },
+    {
+      "epoch": 52.00623700623701,
+      "grad_norm": 6.387617031577975e-05,
+      "learning_rate": 1.5407523514338783e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19544032,
+      "step": 25015
+    },
+    {
+      "epoch": 52.016632016632016,
+      "grad_norm": 3.683889372041449e-05,
+      "learning_rate": 1.539845822461543e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19548000,
+      "step": 25020
+    },
+    {
+      "epoch": 52.027027027027025,
+      "grad_norm": 1.6941106878221035e-05,
+      "learning_rate": 1.538939441557048e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19551744,
+      "step": 25025
+    },
+    {
+      "epoch": 52.03742203742204,
+      "grad_norm": 1.6796479030745104e-05,
+      "learning_rate": 1.5380332088601696e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19555744,
+      "step": 25030
+    },
+    {
+      "epoch": 52.04781704781705,
+      "grad_norm": 7.45799407013692e-05,
+      "learning_rate": 1.537127124510658e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19559680,
+      "step": 25035
+    },
+    {
+      "epoch": 52.05821205821206,
+      "grad_norm": 1.1059516509703826e-05,
+      "learning_rate": 1.5362211886482457e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19563488,
+      "step": 25040
+    },
+    {
+      "epoch": 52.06860706860707,
+      "grad_norm": 1.1458750122983474e-05,
+      "learning_rate": 1.5353154014126363e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19567264,
+      "step": 25045
+    },
+    {
+      "epoch": 52.07900207900208,
+      "grad_norm": 1.7411861335858703e-05,
+      "learning_rate": 1.534409762943515e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19571136,
+      "step": 25050
+    },
+    {
+      "epoch": 52.08939708939709,
+      "grad_norm": 8.86204288690351e-05,
+      "learning_rate": 1.5335042733805438e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19575264,
+      "step": 25055
+    },
+    {
+      "epoch": 52.0997920997921,
+      "grad_norm": 4.775008619617438e-06,
+      "learning_rate": 1.532598932863358e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19579136,
+      "step": 25060
+    },
+    {
+      "epoch": 52.11018711018711,
+      "grad_norm": 1.0272594408888835e-05,
+      "learning_rate": 1.531693741531574e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19583008,
+      "step": 25065
+    },
+    {
+      "epoch": 52.12058212058212,
+      "grad_norm": 0.000562442175578326,
+      "learning_rate": 1.5307886995247844e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19586816,
+      "step": 25070
+    },
+    {
+      "epoch": 52.13097713097713,
+      "grad_norm": 3.534054121701047e-05,
+      "learning_rate": 1.529883806982557e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19590720,
+      "step": 25075
+    },
+    {
+      "epoch": 52.141372141372145,
+      "grad_norm": 1.5920179066597484e-05,
+      "learning_rate": 1.5289790640444376e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19594656,
+      "step": 25080
+    },
+    {
+      "epoch": 52.15176715176715,
+      "grad_norm": 1.2734960364468861e-05,
+      "learning_rate": 1.5280744708499494e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 19598560,
+      "step": 25085
+    },
+    {
+      "epoch": 52.16216216216216,
+      "grad_norm": 7.755186379654333e-05,
+      "learning_rate": 1.527170027538591e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19602496,
+      "step": 25090
+    },
+    {
+      "epoch": 52.17255717255717,
+      "grad_norm": 0.00021814630599692464,
+      "learning_rate": 1.5262657342498407e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19606304,
+      "step": 25095
+    },
+    {
+      "epoch": 52.182952182952185,
+      "grad_norm": 1.85260232683504e-05,
+      "learning_rate": 1.52536159112315e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19610208,
+      "step": 25100
+    },
+    {
+      "epoch": 52.19334719334719,
+      "grad_norm": 0.0006525013013742864,
+      "learning_rate": 1.5244575982979497e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19614176,
+      "step": 25105
+    },
+    {
+      "epoch": 52.2037422037422,
+      "grad_norm": 9.351633707410656e-06,
+      "learning_rate": 1.5235537559136487e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19618048,
+      "step": 25110
+    },
+    {
+      "epoch": 52.21413721413722,
+      "grad_norm": 2.5366073259647237e-06,
+      "learning_rate": 1.5226500641096286e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19621888,
+      "step": 25115
+    },
+    {
+      "epoch": 52.224532224532226,
+      "grad_norm": 3.9214487514982466e-06,
+      "learning_rate": 1.5217465230252509e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19625760,
+      "step": 25120
+    },
+    {
+      "epoch": 52.234927234927234,
+      "grad_norm": 0.15611281991004944,
+      "learning_rate": 1.5208431327998523e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 19629600,
+      "step": 25125
+    },
+    {
+      "epoch": 52.24532224532224,
+      "grad_norm": 2.4768494768068194e-05,
+      "learning_rate": 1.5199398935727477e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19633472,
+      "step": 25130
+    },
+    {
+      "epoch": 52.25571725571726,
+      "grad_norm": 0.000208377925446257,
+      "learning_rate": 1.5190368054832282e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19637568,
+      "step": 25135
+    },
+    {
+      "epoch": 52.266112266112266,
+      "grad_norm": 9.853696246864274e-05,
+      "learning_rate": 1.5181338686705601e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19641408,
+      "step": 25140
+    },
+    {
+      "epoch": 52.276507276507274,
+      "grad_norm": 0.0005967452889308333,
+      "learning_rate": 1.5172310832739889e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 19645376,
+      "step": 25145
+    },
+    {
+      "epoch": 52.28690228690229,
+      "grad_norm": 3.2683245080988854e-05,
+      "learning_rate": 1.5163284494327346e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19649248,
+      "step": 25150
+    },
+    {
+      "epoch": 52.2972972972973,
+      "grad_norm": 7.283157628990011e-06,
+      "learning_rate": 1.5154259672859952e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19653216,
+      "step": 25155
+    },
+    {
+      "epoch": 52.30769230769231,
+      "grad_norm": 1.6672742276568897e-05,
+      "learning_rate": 1.5145236369729452e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19657184,
+      "step": 25160
+    },
+    {
+      "epoch": 52.318087318087315,
+      "grad_norm": 1.6600763046881184e-05,
+      "learning_rate": 1.5136214586327335e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19660928,
+      "step": 25165
+    },
+    {
+      "epoch": 52.32848232848233,
+      "grad_norm": 0.00020884584228042513,
+      "learning_rate": 1.5127194324044885e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19664896,
+      "step": 25170
+    },
+    {
+      "epoch": 52.33887733887734,
+      "grad_norm": 0.0001399814645992592,
+      "learning_rate": 1.5118175584273148e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19668800,
+      "step": 25175
+    },
+    {
+      "epoch": 52.34927234927235,
+      "grad_norm": 3.427914998610504e-05,
+      "learning_rate": 1.5109158368402909e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19672736,
+      "step": 25180
+    },
+    {
+      "epoch": 52.35966735966736,
+      "grad_norm": 0.00017105315055232495,
+      "learning_rate": 1.5100142677824753e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 19676608,
+      "step": 25185
+    },
+    {
+      "epoch": 52.37006237006237,
+      "grad_norm": 5.739998869103147e-06,
+      "learning_rate": 1.509112851392901e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19680544,
+      "step": 25190
+    },
+    {
+      "epoch": 52.38045738045738,
+      "grad_norm": 2.0516859876806848e-05,
+      "learning_rate": 1.5082115878105763e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19684416,
+      "step": 25195
+    },
+    {
+      "epoch": 52.39085239085239,
+      "grad_norm": 0.0010022163623943925,
+      "learning_rate": 1.5073104771744892e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 19688288,
+      "step": 25200
+    },
+    {
+      "epoch": 52.39085239085239,
+      "eval_loss": 0.5195792317390442,
+      "eval_runtime": 11.6947,
+      "eval_samples_per_second": 73.195,
+      "eval_steps_per_second": 18.299,
+      "num_input_tokens_seen": 19688288,
+      "step": 25200
+    },
+    {
+      "epoch": 52.4012474012474,
+      "grad_norm": 1.4095300684857648e-05,
+      "learning_rate": 1.5064095196236006e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 19692224,
+      "step": 25205
+    },
+    {
+      "epoch": 52.41164241164241,
+      "grad_norm": 1.8062896742776502e-06,
+      "learning_rate": 1.50550871529685e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 19696128,
+      "step": 25210
+    },
+    {
+      "epoch": 52.42203742203742,
+      "grad_norm": 3.6367146094562486e-05,
+      "learning_rate": 1.5046080643331546e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19700064,
+      "step": 25215
+    },
+    {
+      "epoch": 52.432432432432435,
+      "grad_norm": 5.402194074122235e-05,
+      "learning_rate": 1.5037075668714028e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19704000,
+      "step": 25220
+    },
+    {
+      "epoch": 52.44282744282744,
+      "grad_norm": 3.1908302844385616e-06,
+      "learning_rate": 1.5028072230504656e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19707840,
+      "step": 25225
+    },
+    {
+      "epoch": 52.45322245322245,
+      "grad_norm": 2.273040263389703e-05,
+      "learning_rate": 1.5019070330091861e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19711712,
+      "step": 25230
+    },
+    {
+      "epoch": 52.46361746361746,
+      "grad_norm": 0.0001558463554829359,
+      "learning_rate": 1.5010069968863843e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 19715776,
+      "step": 25235
+    },
+    {
+      "epoch": 52.474012474012476,
+      "grad_norm": 3.874272351822583e-06,
+      "learning_rate": 1.5001071148208584e-05,
+      "loss": 0.0046,
+      "num_input_tokens_seen": 19719584,
+      "step": 25240
+    },
+    {
+      "epoch": 52.484407484407484,
+      "grad_norm": 5.06705509906169e-05,
+      "learning_rate": 1.49920738695138e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19723520,
+      "step": 25245
+    },
+    {
+      "epoch": 52.49480249480249,
+      "grad_norm": 4.437856205186108e-06,
+      "learning_rate": 1.4983078134166995e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19727456,
+      "step": 25250
+    },
+    {
+      "epoch": 52.50519750519751,
+      "grad_norm": 1.1873336916323751e-05,
+      "learning_rate": 1.4974083943555428e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19731488,
+      "step": 25255
+    },
+    {
+      "epoch": 52.515592515592516,
+      "grad_norm": 4.023259407404112e-06,
+      "learning_rate": 1.496509129906611e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19735488,
+      "step": 25260
+    },
+    {
+      "epoch": 52.525987525987524,
+      "grad_norm": 2.6595382678351598e-06,
+      "learning_rate": 1.4956100202085809e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19739328,
+      "step": 25265
+    },
+    {
+      "epoch": 52.53638253638254,
+      "grad_norm": 2.5027138690347783e-06,
+      "learning_rate": 1.4947110654001093e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19743264,
+      "step": 25270
+    },
+    {
+      "epoch": 52.54677754677755,
+      "grad_norm": 2.057686106127221e-05,
+      "learning_rate": 1.4938122656198234e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19747136,
+      "step": 25275
+    },
+    {
+      "epoch": 52.55717255717256,
+      "grad_norm": 1.7287617083638906e-05,
+      "learning_rate": 1.4929136210063316e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19751104,
+      "step": 25280
+    },
+    {
+      "epoch": 52.567567567567565,
+      "grad_norm": 9.884679457172751e-05,
+      "learning_rate": 1.4920151316982146e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19755072,
+      "step": 25285
+    },
+    {
+      "epoch": 52.57796257796258,
+      "grad_norm": 2.4747776478761807e-05,
+      "learning_rate": 1.4911167978340312e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19758848,
+      "step": 25290
+    },
+    {
+      "epoch": 52.58835758835759,
+      "grad_norm": 6.247931196412537e-06,
+      "learning_rate": 1.4902186195523166e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19762752,
+      "step": 25295
+    },
+    {
+      "epoch": 52.5987525987526,
+      "grad_norm": 1.5762496332172304e-05,
+      "learning_rate": 1.4893205969915805e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19766624,
+      "step": 25300
+    },
+    {
+      "epoch": 52.60914760914761,
+      "grad_norm": 0.00024687600671313703,
+      "learning_rate": 1.4884227302903086e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19770496,
+      "step": 25305
+    },
+    {
+      "epoch": 52.61954261954262,
+      "grad_norm": 8.385935871046968e-06,
+      "learning_rate": 1.4875250195869653e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19774336,
+      "step": 25310
+    },
+    {
+      "epoch": 52.62993762993763,
+      "grad_norm": 1.9765344404731877e-05,
+      "learning_rate": 1.4866274650199862e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19778240,
+      "step": 25315
+    },
+    {
+      "epoch": 52.64033264033264,
+      "grad_norm": 0.0005193252582103014,
+      "learning_rate": 1.485730066727788e-05,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 19782240,
+      "step": 25320
+    },
+    {
+      "epoch": 52.65072765072765,
+      "grad_norm": 3.446054688538425e-05,
+      "learning_rate": 1.4848328248487586e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 19786144,
+      "step": 25325
+    },
+    {
+      "epoch": 52.66112266112266,
+      "grad_norm": 8.050037649809383e-06,
+      "learning_rate": 1.4839357395212656e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19789984,
+      "step": 25330
+    },
+    {
+      "epoch": 52.67151767151767,
+      "grad_norm": 7.611668115714565e-05,
+      "learning_rate": 1.4830388108836502e-05,
+      "loss": 0.0098,
+      "num_input_tokens_seen": 19793984,
+      "step": 25335
+    },
+    {
+      "epoch": 52.681912681912685,
+      "grad_norm": 8.363992492377292e-06,
+      "learning_rate": 1.4821420390742299e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19797888,
+      "step": 25340
+    },
+    {
+      "epoch": 52.69230769230769,
+      "grad_norm": 0.0035357936285436153,
+      "learning_rate": 1.4812454242312979e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19801824,
+      "step": 25345
+    },
+    {
+      "epoch": 52.7027027027027,
+      "grad_norm": 0.0010829203529283404,
+      "learning_rate": 1.4803489664931253e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19805888,
+      "step": 25350
+    },
+    {
+      "epoch": 52.71309771309771,
+      "grad_norm": 4.042047294205986e-05,
+      "learning_rate": 1.4794526659979544e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19809792,
+      "step": 25355
+    },
+    {
+      "epoch": 52.723492723492726,
+      "grad_norm": 0.0019854377023875713,
+      "learning_rate": 1.4785565228840086e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19813760,
+      "step": 25360
+    },
+    {
+      "epoch": 52.733887733887734,
+      "grad_norm": 5.919835530221462e-05,
+      "learning_rate": 1.4776605372894819e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19817664,
+      "step": 25365
+    },
+    {
+      "epoch": 52.74428274428274,
+      "grad_norm": 0.14753121137619019,
+      "learning_rate": 1.4767647093525488e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 19821632,
+      "step": 25370
+    },
+    {
+      "epoch": 52.75467775467776,
+      "grad_norm": 7.804558117641136e-05,
+      "learning_rate": 1.4758690392113566e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19825472,
+      "step": 25375
+    },
+    {
+      "epoch": 52.765072765072766,
+      "grad_norm": 1.114716815209249e-05,
+      "learning_rate": 1.4749735270040276e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19829376,
+      "step": 25380
+    },
+    {
+      "epoch": 52.775467775467774,
+      "grad_norm": 0.0001266059698536992,
+      "learning_rate": 1.4740781728686623e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19833216,
+      "step": 25385
+    },
+    {
+      "epoch": 52.78586278586278,
+      "grad_norm": 1.0738266610132996e-05,
+      "learning_rate": 1.4731829769433358e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19837056,
+      "step": 25390
+    },
+    {
+      "epoch": 52.7962577962578,
+      "grad_norm": 8.188795618480071e-06,
+      "learning_rate": 1.4722879393660976e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19840896,
+      "step": 25395
+    },
+    {
+      "epoch": 52.80665280665281,
+      "grad_norm": 0.0001764265907695517,
+      "learning_rate": 1.4713930602749748e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19844672,
+      "step": 25400
+    },
+    {
+      "epoch": 52.80665280665281,
+      "eval_loss": 0.5195030570030212,
+      "eval_runtime": 11.686,
+      "eval_samples_per_second": 73.25,
+      "eval_steps_per_second": 18.313,
+      "num_input_tokens_seen": 19844672,
+      "step": 25400
+    },
+    {
+      "epoch": 52.817047817047815,
+      "grad_norm": 0.005748292896896601,
+      "learning_rate": 1.470498339807968e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 19848704,
+      "step": 25405
+    },
+    {
+      "epoch": 52.82744282744283,
+      "grad_norm": 0.0013281063875183463,
+      "learning_rate": 1.4696037781030542e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19852608,
+      "step": 25410
+    },
+    {
+      "epoch": 52.83783783783784,
+      "grad_norm": 6.799814582336694e-05,
+      "learning_rate": 1.4687093752981876e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19856448,
+      "step": 25415
+    },
+    {
+      "epoch": 52.84823284823285,
+      "grad_norm": 0.00012058308493578807,
+      "learning_rate": 1.4678151315312943e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19860224,
+      "step": 25420
+    },
+    {
+      "epoch": 52.858627858627855,
+      "grad_norm": 4.593388894136297e-06,
+      "learning_rate": 1.4669210469402789e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19864288,
+      "step": 25425
+    },
+    {
+      "epoch": 52.86902286902287,
+      "grad_norm": 0.00011223143519600853,
+      "learning_rate": 1.4660271216630218e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19868224,
+      "step": 25430
+    },
+    {
+      "epoch": 52.87941787941788,
+      "grad_norm": 1.7946853404282592e-05,
+      "learning_rate": 1.4651333558373748e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19872224,
+      "step": 25435
+    },
+    {
+      "epoch": 52.88981288981289,
+      "grad_norm": 1.1220158739888575e-05,
+      "learning_rate": 1.4642397496011707e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19876256,
+      "step": 25440
+    },
+    {
+      "epoch": 52.9002079002079,
+      "grad_norm": 0.00024714789469726384,
+      "learning_rate": 1.4633463030922129e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19880288,
+      "step": 25445
+    },
+    {
+      "epoch": 52.91060291060291,
+      "grad_norm": 2.205192686233204e-05,
+      "learning_rate": 1.462453016448282e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19884256,
+      "step": 25450
+    },
+    {
+      "epoch": 52.92099792099792,
+      "grad_norm": 2.1915704564889893e-05,
+      "learning_rate": 1.4615598898071354e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19888000,
+      "step": 25455
+    },
+    {
+      "epoch": 52.931392931392935,
+      "grad_norm": 0.0002727980609051883,
+      "learning_rate": 1.4606669233065026e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19891968,
+      "step": 25460
+    },
+    {
+      "epoch": 52.94178794178794,
+      "grad_norm": 1.4380085303855594e-05,
+      "learning_rate": 1.4597741170840914e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 19895936,
+      "step": 25465
+    },
+    {
+      "epoch": 52.95218295218295,
+      "grad_norm": 0.12748661637306213,
+      "learning_rate": 1.4588814712775853e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 19899808,
+      "step": 25470
+    },
+    {
+      "epoch": 52.96257796257796,
+      "grad_norm": 1.377418539050268e-05,
+      "learning_rate": 1.4579889860246382e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19903744,
+      "step": 25475
+    },
+    {
+      "epoch": 52.972972972972975,
+      "grad_norm": 7.597828516736627e-05,
+      "learning_rate": 1.457096661462885e-05,
+      "loss": 0.0074,
+      "num_input_tokens_seen": 19907680,
+      "step": 25480
+    },
+    {
+      "epoch": 52.983367983367984,
+      "grad_norm": 0.00011736094893421978,
+      "learning_rate": 1.4562044977299322e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19911488,
+      "step": 25485
+    },
+    {
+      "epoch": 52.99376299376299,
+      "grad_norm": 0.14197395741939545,
+      "learning_rate": 1.4553124949633623e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 19915392,
+      "step": 25490
+    },
+    {
+      "epoch": 53.00415800415801,
+      "grad_norm": 6.242563540581614e-05,
+      "learning_rate": 1.4544206533007354e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19919248,
+      "step": 25495
+    },
+    {
+      "epoch": 53.014553014553016,
+      "grad_norm": 4.5485335249395575e-06,
+      "learning_rate": 1.4535289728795821e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19923088,
+      "step": 25500
+    },
+    {
+      "epoch": 53.024948024948024,
+      "grad_norm": 1.2754134331771638e-05,
+      "learning_rate": 1.4526374538374132e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19927056,
+      "step": 25505
+    },
+    {
+      "epoch": 53.03534303534303,
+      "grad_norm": 2.4444145310553722e-05,
+      "learning_rate": 1.4517460963117097e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19931152,
+      "step": 25510
+    },
+    {
+      "epoch": 53.04573804573805,
+      "grad_norm": 0.00020800794300157577,
+      "learning_rate": 1.4508549004399314e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 19934928,
+      "step": 25515
+    },
+    {
+      "epoch": 53.056133056133056,
+      "grad_norm": 1.3220678738434799e-05,
+      "learning_rate": 1.449963866359513e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19938800,
+      "step": 25520
+    },
+    {
+      "epoch": 53.066528066528065,
+      "grad_norm": 0.00019995507318526506,
+      "learning_rate": 1.4490729942078607e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19942672,
+      "step": 25525
+    },
+    {
+      "epoch": 53.07692307692308,
+      "grad_norm": 0.11731410026550293,
+      "learning_rate": 1.4481822841223608e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 19946512,
+      "step": 25530
+    },
+    {
+      "epoch": 53.08731808731809,
+      "grad_norm": 0.0003157792962156236,
+      "learning_rate": 1.4472917362403704e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19950544,
+      "step": 25535
+    },
+    {
+      "epoch": 53.0977130977131,
+      "grad_norm": 0.00013002964260522276,
+      "learning_rate": 1.4464013506992224e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19954608,
+      "step": 25540
+    },
+    {
+      "epoch": 53.108108108108105,
+      "grad_norm": 3.971412297687493e-05,
+      "learning_rate": 1.4455111276362277e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19958480,
+      "step": 25545
+    },
+    {
+      "epoch": 53.11850311850312,
+      "grad_norm": 0.00032755843130871654,
+      "learning_rate": 1.4446210671886676e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19962448,
+      "step": 25550
+    },
+    {
+      "epoch": 53.12889812889813,
+      "grad_norm": 0.00011922736302949488,
+      "learning_rate": 1.4437311694938015e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19966320,
+      "step": 25555
+    },
+    {
+      "epoch": 53.13929313929314,
+      "grad_norm": 6.7168366513215005e-06,
+      "learning_rate": 1.442841434688864e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19970192,
+      "step": 25560
+    },
+    {
+      "epoch": 53.14968814968815,
+      "grad_norm": 1.3759351531916764e-05,
+      "learning_rate": 1.4419518629110615e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19974096,
+      "step": 25565
+    },
+    {
+      "epoch": 53.16008316008316,
+      "grad_norm": 1.2498493560997304e-05,
+      "learning_rate": 1.4410624542975778e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19978096,
+      "step": 25570
+    },
+    {
+      "epoch": 53.17047817047817,
+      "grad_norm": 1.3091188520775177e-05,
+      "learning_rate": 1.4401732089855724e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19982032,
+      "step": 25575
+    },
+    {
+      "epoch": 53.18087318087318,
+      "grad_norm": 0.00013744120951741934,
+      "learning_rate": 1.4392841271121754e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 19985904,
+      "step": 25580
+    },
+    {
+      "epoch": 53.19126819126819,
+      "grad_norm": 0.14987251162528992,
+      "learning_rate": 1.438395208814497e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 19989808,
+      "step": 25585
+    },
+    {
+      "epoch": 53.2016632016632,
+      "grad_norm": 9.456072439206764e-05,
+      "learning_rate": 1.4375064542296174e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19993808,
+      "step": 25590
+    },
+    {
+      "epoch": 53.21205821205821,
+      "grad_norm": 9.491976015851833e-06,
+      "learning_rate": 1.4366178634945946e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 19997648,
+      "step": 25595
+    },
+    {
+      "epoch": 53.222453222453225,
+      "grad_norm": 0.00015765425632707775,
+      "learning_rate": 1.4357294367464616e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20001552,
+      "step": 25600
+    },
+    {
+      "epoch": 53.222453222453225,
+      "eval_loss": 0.5287191271781921,
+      "eval_runtime": 11.6822,
+      "eval_samples_per_second": 73.274,
+      "eval_steps_per_second": 18.319,
+      "num_input_tokens_seen": 20001552,
+      "step": 25600
+    },
+    {
+      "epoch": 53.232848232848234,
+      "grad_norm": 1.7194986867252737e-05,
+      "learning_rate": 1.434841174122224e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20005392,
+      "step": 25605
+    },
+    {
+      "epoch": 53.24324324324324,
+      "grad_norm": 8.58824478200404e-06,
+      "learning_rate": 1.4339530757588615e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20009328,
+      "step": 25610
+    },
+    {
+      "epoch": 53.25363825363825,
+      "grad_norm": 1.8379927496425807e-05,
+      "learning_rate": 1.433065141793333e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20013392,
+      "step": 25615
+    },
+    {
+      "epoch": 53.264033264033266,
+      "grad_norm": 0.00020260240125935525,
+      "learning_rate": 1.4321773723625665e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20017296,
+      "step": 25620
+    },
+    {
+      "epoch": 53.274428274428274,
+      "grad_norm": 1.2522721590357833e-05,
+      "learning_rate": 1.4312897676034693e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20021200,
+      "step": 25625
+    },
+    {
+      "epoch": 53.28482328482328,
+      "grad_norm": 3.113100092377863e-06,
+      "learning_rate": 1.4304023276529188e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20025072,
+      "step": 25630
+    },
+    {
+      "epoch": 53.2952182952183,
+      "grad_norm": 1.0581230526440777e-05,
+      "learning_rate": 1.4295150526477712e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20028976,
+      "step": 25635
+    },
+    {
+      "epoch": 53.305613305613306,
+      "grad_norm": 5.073535066912882e-05,
+      "learning_rate": 1.4286279427248562e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 20032976,
+      "step": 25640
+    },
+    {
+      "epoch": 53.316008316008315,
+      "grad_norm": 5.013713234802708e-05,
+      "learning_rate": 1.4277409980209747e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20036784,
+      "step": 25645
+    },
+    {
+      "epoch": 53.32640332640332,
+      "grad_norm": 1.3280166058393661e-05,
+      "learning_rate": 1.4268542186729061e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20040752,
+      "step": 25650
+    },
+    {
+      "epoch": 53.33679833679834,
+      "grad_norm": 6.081428182369564e-06,
+      "learning_rate": 1.4259676048174043e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20044496,
+      "step": 25655
+    },
+    {
+      "epoch": 53.34719334719335,
+      "grad_norm": 7.75946828071028e-05,
+      "learning_rate": 1.4250811565911937e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 20048400,
+      "step": 25660
+    },
+    {
+      "epoch": 53.357588357588355,
+      "grad_norm": 1.4391652257472742e-05,
+      "learning_rate": 1.4241948741309782e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20052208,
+      "step": 25665
+    },
+    {
+      "epoch": 53.36798336798337,
+      "grad_norm": 0.0006211532163433731,
+      "learning_rate": 1.4233087575734317e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20056240,
+      "step": 25670
+    },
+    {
+      "epoch": 53.37837837837838,
+      "grad_norm": 0.00010691677744034678,
+      "learning_rate": 1.422422807055206e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20060016,
+      "step": 25675
+    },
+    {
+      "epoch": 53.38877338877339,
+      "grad_norm": 7.938502676552162e-05,
+      "learning_rate": 1.4215370227129243e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20063856,
+      "step": 25680
+    },
+    {
+      "epoch": 53.3991683991684,
+      "grad_norm": 0.0004668447654694319,
+      "learning_rate": 1.4206514046831876e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20067728,
+      "step": 25685
+    },
+    {
+      "epoch": 53.40956340956341,
+      "grad_norm": 5.392667844716925e-06,
+      "learning_rate": 1.419765953102567e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20071696,
+      "step": 25690
+    },
+    {
+      "epoch": 53.41995841995842,
+      "grad_norm": 8.143336162902415e-05,
+      "learning_rate": 1.4188806681076125e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20075632,
+      "step": 25695
+    },
+    {
+      "epoch": 53.43035343035343,
+      "grad_norm": 0.00015649055421818048,
+      "learning_rate": 1.4179955498348443e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20079472,
+      "step": 25700
+    },
+    {
+      "epoch": 53.44074844074844,
+      "grad_norm": 0.0011744080111384392,
+      "learning_rate": 1.4171105984207605e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20083280,
+      "step": 25705
+    },
+    {
+      "epoch": 53.45114345114345,
+      "grad_norm": 1.2007573786831927e-05,
+      "learning_rate": 1.4162258140018304e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20087184,
+      "step": 25710
+    },
+    {
+      "epoch": 53.46153846153846,
+      "grad_norm": 0.0004405644431244582,
+      "learning_rate": 1.4153411967144986e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20091280,
+      "step": 25715
+    },
+    {
+      "epoch": 53.471933471933475,
+      "grad_norm": 0.10980650037527084,
+      "learning_rate": 1.4144567466951864e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 20095024,
+      "step": 25720
+    },
+    {
+      "epoch": 53.482328482328484,
+      "grad_norm": 1.1656554306682665e-05,
+      "learning_rate": 1.4135724640802844e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20098928,
+      "step": 25725
+    },
+    {
+      "epoch": 53.49272349272349,
+      "grad_norm": 1.0550775186857209e-05,
+      "learning_rate": 1.4126883490061615e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20102576,
+      "step": 25730
+    },
+    {
+      "epoch": 53.5031185031185,
+      "grad_norm": 4.6070872485870495e-06,
+      "learning_rate": 1.4118044016091603e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20106480,
+      "step": 25735
+    },
+    {
+      "epoch": 53.513513513513516,
+      "grad_norm": 4.9894690164364874e-05,
+      "learning_rate": 1.410920622025594e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20110384,
+      "step": 25740
+    },
+    {
+      "epoch": 53.523908523908524,
+      "grad_norm": 0.00019873888231813908,
+      "learning_rate": 1.4100370103917554e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20114384,
+      "step": 25745
+    },
+    {
+      "epoch": 53.53430353430353,
+      "grad_norm": 4.113233444513753e-05,
+      "learning_rate": 1.409153566843907e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20118384,
+      "step": 25750
+    },
+    {
+      "epoch": 53.54469854469855,
+      "grad_norm": 7.723829912720248e-05,
+      "learning_rate": 1.408270291518286e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20122192,
+      "step": 25755
+    },
+    {
+      "epoch": 53.555093555093556,
+      "grad_norm": 9.054440852196421e-06,
+      "learning_rate": 1.407387184551107e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 20126064,
+      "step": 25760
+    },
+    {
+      "epoch": 53.565488565488565,
+      "grad_norm": 6.256264896364883e-05,
+      "learning_rate": 1.4065042460785532e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20130064,
+      "step": 25765
+    },
+    {
+      "epoch": 53.57588357588357,
+      "grad_norm": 6.809596925450023e-06,
+      "learning_rate": 1.405621476236787e-05,
+      "loss": 0.0062,
+      "num_input_tokens_seen": 20134000,
+      "step": 25770
+    },
+    {
+      "epoch": 53.58627858627859,
+      "grad_norm": 2.2014031856087968e-05,
+      "learning_rate": 1.4047388751619423e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 20137904,
+      "step": 25775
+    },
+    {
+      "epoch": 53.5966735966736,
+      "grad_norm": 6.095435674069449e-05,
+      "learning_rate": 1.4038564429901264e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 20141936,
+      "step": 25780
+    },
+    {
+      "epoch": 53.607068607068605,
+      "grad_norm": 0.00011970556806772947,
+      "learning_rate": 1.4029741798574227e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20145712,
+      "step": 25785
+    },
+    {
+      "epoch": 53.61746361746362,
+      "grad_norm": 3.8119057990115834e-06,
+      "learning_rate": 1.402092085899886e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 20149616,
+      "step": 25790
+    },
+    {
+      "epoch": 53.62785862785863,
+      "grad_norm": 2.817297172441613e-05,
+      "learning_rate": 1.4012101612535464e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20153488,
+      "step": 25795
+    },
+    {
+      "epoch": 53.63825363825364,
+      "grad_norm": 6.1061537053319626e-06,
+      "learning_rate": 1.4003284060544092e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20157424,
+      "step": 25800
+    },
+    {
+      "epoch": 53.63825363825364,
+      "eval_loss": 0.5364317297935486,
+      "eval_runtime": 11.6878,
+      "eval_samples_per_second": 73.239,
+      "eval_steps_per_second": 18.31,
+      "num_input_tokens_seen": 20157424,
+      "step": 25800
+    },
+    {
+      "epoch": 53.648648648648646,
+      "grad_norm": 0.007422337774187326,
+      "learning_rate": 1.3994468204384504e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20161328,
+      "step": 25805
+    },
+    {
+      "epoch": 53.65904365904366,
+      "grad_norm": 5.201307430979796e-05,
+      "learning_rate": 1.398565404541622e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20165136,
+      "step": 25810
+    },
+    {
+      "epoch": 53.66943866943867,
+      "grad_norm": 3.886279046128038e-06,
+      "learning_rate": 1.3976841584998513e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20169072,
+      "step": 25815
+    },
+    {
+      "epoch": 53.67983367983368,
+      "grad_norm": 0.11910151690244675,
+      "learning_rate": 1.3968030824490352e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 20172976,
+      "step": 25820
+    },
+    {
+      "epoch": 53.69022869022869,
+      "grad_norm": 6.243923417059705e-05,
+      "learning_rate": 1.3959221765250469e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20176784,
+      "step": 25825
+    },
+    {
+      "epoch": 53.7006237006237,
+      "grad_norm": 8.601310400990769e-05,
+      "learning_rate": 1.3950414408637343e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 20180816,
+      "step": 25830
+    },
+    {
+      "epoch": 53.71101871101871,
+      "grad_norm": 1.3048331311438233e-05,
+      "learning_rate": 1.3941608756009166e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20184592,
+      "step": 25835
+    },
+    {
+      "epoch": 53.72141372141372,
+      "grad_norm": 4.014191290480085e-05,
+      "learning_rate": 1.3932804808723898e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20188592,
+      "step": 25840
+    },
+    {
+      "epoch": 53.731808731808734,
+      "grad_norm": 0.0008380291983485222,
+      "learning_rate": 1.3924002568139194e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20192496,
+      "step": 25845
+    },
+    {
+      "epoch": 53.74220374220374,
+      "grad_norm": 3.0937021620047744e-06,
+      "learning_rate": 1.3915202035612485e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20196368,
+      "step": 25850
+    },
+    {
+      "epoch": 53.75259875259875,
+      "grad_norm": 9.262654202757403e-05,
+      "learning_rate": 1.3906403212500935e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20200304,
+      "step": 25855
+    },
+    {
+      "epoch": 53.762993762993766,
+      "grad_norm": 1.885544043034315e-05,
+      "learning_rate": 1.3897606100161409e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20204112,
+      "step": 25860
+    },
+    {
+      "epoch": 53.773388773388774,
+      "grad_norm": 5.065724326414056e-05,
+      "learning_rate": 1.388881069995055e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20208016,
+      "step": 25865
+    },
+    {
+      "epoch": 53.78378378378378,
+      "grad_norm": 0.00010862765338970348,
+      "learning_rate": 1.3880017013224708e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20211984,
+      "step": 25870
+    },
+    {
+      "epoch": 53.79417879417879,
+      "grad_norm": 0.14207011461257935,
+      "learning_rate": 1.3871225041339984e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 20215920,
+      "step": 25875
+    },
+    {
+      "epoch": 53.804573804573806,
+      "grad_norm": 4.3889314838452265e-05,
+      "learning_rate": 1.386243478565222e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20219856,
+      "step": 25880
+    },
+    {
+      "epoch": 53.814968814968815,
+      "grad_norm": 8.067941053013783e-06,
+      "learning_rate": 1.3853646247516966e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20223824,
+      "step": 25885
+    },
+    {
+      "epoch": 53.82536382536382,
+      "grad_norm": 4.535461812338326e-06,
+      "learning_rate": 1.3844859428289545e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20227792,
+      "step": 25890
+    },
+    {
+      "epoch": 53.83575883575884,
+      "grad_norm": 1.719027932267636e-05,
+      "learning_rate": 1.3836074329324984e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20231600,
+      "step": 25895
+    },
+    {
+      "epoch": 53.84615384615385,
+      "grad_norm": 1.0656212907633744e-05,
+      "learning_rate": 1.3827290951978044e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20235568,
+      "step": 25900
+    },
+    {
+      "epoch": 53.856548856548855,
+      "grad_norm": 6.96006027283147e-05,
+      "learning_rate": 1.381850929760326e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20239440,
+      "step": 25905
+    },
+    {
+      "epoch": 53.86694386694387,
+      "grad_norm": 7.241807907121256e-06,
+      "learning_rate": 1.3809729367554842e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20243344,
+      "step": 25910
+    },
+    {
+      "epoch": 53.87733887733888,
+      "grad_norm": 0.1473158895969391,
+      "learning_rate": 1.3800951163186784e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 20247216,
+      "step": 25915
+    },
+    {
+      "epoch": 53.88773388773389,
+      "grad_norm": 6.765192210878013e-06,
+      "learning_rate": 1.3792174685852801e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20251248,
+      "step": 25920
+    },
+    {
+      "epoch": 53.898128898128896,
+      "grad_norm": 0.00012064431939506903,
+      "learning_rate": 1.378339993690632e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20255152,
+      "step": 25925
+    },
+    {
+      "epoch": 53.90852390852391,
+      "grad_norm": 0.00011406522389734164,
+      "learning_rate": 1.3774626917700523e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20258992,
+      "step": 25930
+    },
+    {
+      "epoch": 53.91891891891892,
+      "grad_norm": 0.0004517340858001262,
+      "learning_rate": 1.3765855629588334e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20262768,
+      "step": 25935
+    },
+    {
+      "epoch": 53.92931392931393,
+      "grad_norm": 4.5310744098969735e-06,
+      "learning_rate": 1.3757086073922374e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20266608,
+      "step": 25940
+    },
+    {
+      "epoch": 53.93970893970894,
+      "grad_norm": 0.000331732677295804,
+      "learning_rate": 1.3748318252055038e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20270608,
+      "step": 25945
+    },
+    {
+      "epoch": 53.95010395010395,
+      "grad_norm": 6.329904408630682e-06,
+      "learning_rate": 1.3739552165338416e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20274576,
+      "step": 25950
+    },
+    {
+      "epoch": 53.96049896049896,
+      "grad_norm": 0.00014304964861366898,
+      "learning_rate": 1.3730787815124354e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 20278416,
+      "step": 25955
+    },
+    {
+      "epoch": 53.97089397089397,
+      "grad_norm": 1.6613825209788047e-05,
+      "learning_rate": 1.3722025202764443e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20282416,
+      "step": 25960
+    },
+    {
+      "epoch": 53.981288981288984,
+      "grad_norm": 1.1007047760358546e-05,
+      "learning_rate": 1.371326432960997e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20286448,
+      "step": 25965
+    },
+    {
+      "epoch": 53.99168399168399,
+      "grad_norm": 8.514361979905516e-06,
+      "learning_rate": 1.3704505197011969e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 20290352,
+      "step": 25970
+    },
+    {
+      "epoch": 54.002079002079,
+      "grad_norm": 6.665914406767115e-05,
+      "learning_rate": 1.3695747806321224e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 20294080,
+      "step": 25975
+    },
+    {
+      "epoch": 54.012474012474016,
+      "grad_norm": 3.1774688977748156e-05,
+      "learning_rate": 1.3686992158888212e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20298080,
+      "step": 25980
+    },
+    {
+      "epoch": 54.022869022869024,
+      "grad_norm": 1.6722182408557273e-05,
+      "learning_rate": 1.367823825606319e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20301952,
+      "step": 25985
+    },
+    {
+      "epoch": 54.03326403326403,
+      "grad_norm": 6.211551863088971e-06,
+      "learning_rate": 1.36694860991961e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20305952,
+      "step": 25990
+    },
+    {
+      "epoch": 54.04365904365904,
+      "grad_norm": 0.005859998054802418,
+      "learning_rate": 1.3660735689636636e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20309664,
+      "step": 25995
+    },
+    {
+      "epoch": 54.054054054054056,
+      "grad_norm": 7.087628910085186e-05,
+      "learning_rate": 1.365198702873424e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 20313440,
+      "step": 26000
+    },
+    {
+      "epoch": 54.054054054054056,
+      "eval_loss": 0.5265080332756042,
+      "eval_runtime": 11.7092,
+      "eval_samples_per_second": 73.105,
+      "eval_steps_per_second": 18.276,
+      "num_input_tokens_seen": 20313440,
+      "step": 26000
+    },
+    {
+      "epoch": 54.064449064449065,
+      "grad_norm": 3.643439777079038e-05,
+      "learning_rate": 1.364324011783804e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20317440,
+      "step": 26005
+    },
+    {
+      "epoch": 54.07484407484407,
+      "grad_norm": 2.4152162950485945e-05,
+      "learning_rate": 1.3634494958296934e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20321344,
+      "step": 26010
+    },
+    {
+      "epoch": 54.08523908523909,
+      "grad_norm": 3.5358960303710774e-05,
+      "learning_rate": 1.3625751551459542e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 20325248,
+      "step": 26015
+    },
+    {
+      "epoch": 54.0956340956341,
+      "grad_norm": 7.84049552748911e-06,
+      "learning_rate": 1.3617009898674188e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20329184,
+      "step": 26020
+    },
+    {
+      "epoch": 54.106029106029105,
+      "grad_norm": 1.5260298823704943e-05,
+      "learning_rate": 1.3608270001288967e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20333184,
+      "step": 26025
+    },
+    {
+      "epoch": 54.11642411642411,
+      "grad_norm": 5.9245330703561194e-06,
+      "learning_rate": 1.359953186065166e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20337024,
+      "step": 26030
+    },
+    {
+      "epoch": 54.12681912681913,
+      "grad_norm": 3.0713049454789143e-06,
+      "learning_rate": 1.3590795478109814e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20341024,
+      "step": 26035
+    },
+    {
+      "epoch": 54.13721413721414,
+      "grad_norm": 2.362839222769253e-05,
+      "learning_rate": 1.3582060855010675e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20344992,
+      "step": 26040
+    },
+    {
+      "epoch": 54.147609147609145,
+      "grad_norm": 8.690239337738603e-05,
+      "learning_rate": 1.3573327992701245e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20349024,
+      "step": 26045
+    },
+    {
+      "epoch": 54.15800415800416,
+      "grad_norm": 1.8810324036167003e-06,
+      "learning_rate": 1.356459689252823e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20352832,
+      "step": 26050
+    },
+    {
+      "epoch": 54.16839916839917,
+      "grad_norm": 1.772476935002487e-05,
+      "learning_rate": 1.3555867555838087e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 20356768,
+      "step": 26055
+    },
+    {
+      "epoch": 54.17879417879418,
+      "grad_norm": 1.3075710739940405e-05,
+      "learning_rate": 1.3547139983976975e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 20360832,
+      "step": 26060
+    },
+    {
+      "epoch": 54.189189189189186,
+      "grad_norm": 0.0003133807913400233,
+      "learning_rate": 1.3538414178290815e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20364832,
+      "step": 26065
+    },
+    {
+      "epoch": 54.1995841995842,
+      "grad_norm": 3.6468878533924e-05,
+      "learning_rate": 1.3529690140125209e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20368800,
+      "step": 26070
+    },
+    {
+      "epoch": 54.20997920997921,
+      "grad_norm": 0.1513853669166565,
+      "learning_rate": 1.352096787082553e-05,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 20372608,
+      "step": 26075
+    },
+    {
+      "epoch": 54.22037422037422,
+      "grad_norm": 0.0001695121027296409,
+      "learning_rate": 1.3512247371736871e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20376480,
+      "step": 26080
+    },
+    {
+      "epoch": 54.23076923076923,
+      "grad_norm": 0.00014386487600859255,
+      "learning_rate": 1.3503528644204022e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 20380384,
+      "step": 26085
+    },
+    {
+      "epoch": 54.24116424116424,
+      "grad_norm": 4.355283090262674e-05,
+      "learning_rate": 1.349481168957153e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20384256,
+      "step": 26090
+    },
+    {
+      "epoch": 54.25155925155925,
+      "grad_norm": 0.00012482561578508466,
+      "learning_rate": 1.3486096509183665e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20388032,
+      "step": 26095
+    },
+    {
+      "epoch": 54.26195426195426,
+      "grad_norm": 3.823362203547731e-05,
+      "learning_rate": 1.3477383104384406e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20391904,
+      "step": 26100
+    },
+    {
+      "epoch": 54.272349272349274,
+      "grad_norm": 8.732253263588063e-06,
+      "learning_rate": 1.3468671476517481e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20395648,
+      "step": 26105
+    },
+    {
+      "epoch": 54.28274428274428,
+      "grad_norm": 6.852744991192594e-05,
+      "learning_rate": 1.3459961626926326e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20399616,
+      "step": 26110
+    },
+    {
+      "epoch": 54.29313929313929,
+      "grad_norm": 8.074047218542546e-05,
+      "learning_rate": 1.3451253556954101e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20403712,
+      "step": 26115
+    },
+    {
+      "epoch": 54.303534303534306,
+      "grad_norm": 3.793836003751494e-05,
+      "learning_rate": 1.3442547267943717e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20407584,
+      "step": 26120
+    },
+    {
+      "epoch": 54.313929313929314,
+      "grad_norm": 1.0991791896231007e-05,
+      "learning_rate": 1.3433842761237774e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 20411392,
+      "step": 26125
+    },
+    {
+      "epoch": 54.32432432432432,
+      "grad_norm": 0.0005918701062910259,
+      "learning_rate": 1.3425140038178639e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20415200,
+      "step": 26130
+    },
+    {
+      "epoch": 54.33471933471934,
+      "grad_norm": 0.12926670908927917,
+      "learning_rate": 1.3416439100108358e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 20419168,
+      "step": 26135
+    },
+    {
+      "epoch": 54.34511434511435,
+      "grad_norm": 1.1566669854801148e-05,
+      "learning_rate": 1.3407739948368734e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20423104,
+      "step": 26140
+    },
+    {
+      "epoch": 54.355509355509355,
+      "grad_norm": 3.1289666367229074e-05,
+      "learning_rate": 1.3399042584301298e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 20426944,
+      "step": 26145
+    },
+    {
+      "epoch": 54.36590436590436,
+      "grad_norm": 8.68869319674559e-05,
+      "learning_rate": 1.3390347009247272e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20431008,
+      "step": 26150
+    },
+    {
+      "epoch": 54.37629937629938,
+      "grad_norm": 2.0760578991030343e-05,
+      "learning_rate": 1.3381653224547635e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20434944,
+      "step": 26155
+    },
+    {
+      "epoch": 54.38669438669439,
+      "grad_norm": 2.7942858650931157e-05,
+      "learning_rate": 1.3372961231543086e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 20438880,
+      "step": 26160
+    },
+    {
+      "epoch": 54.397089397089395,
+      "grad_norm": 0.005593139678239822,
+      "learning_rate": 1.3364271031574016e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20442784,
+      "step": 26165
+    },
+    {
+      "epoch": 54.40748440748441,
+      "grad_norm": 8.483843885187525e-06,
+      "learning_rate": 1.335558262598059e-05,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 20446752,
+      "step": 26170
+    },
+    {
+      "epoch": 54.41787941787942,
+      "grad_norm": 3.654574538813904e-05,
+      "learning_rate": 1.3346896016102645e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20450464,
+      "step": 26175
+    },
+    {
+      "epoch": 54.42827442827443,
+      "grad_norm": 8.532582796760835e-06,
+      "learning_rate": 1.3338211203279788e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20454240,
+      "step": 26180
+    },
+    {
+      "epoch": 54.438669438669436,
+      "grad_norm": 5.179619984119199e-05,
+      "learning_rate": 1.3329528188851303e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20458112,
+      "step": 26185
+    },
+    {
+      "epoch": 54.44906444906445,
+      "grad_norm": 0.00010834434942808002,
+      "learning_rate": 1.3320846974156242e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20461920,
+      "step": 26190
+    },
+    {
+      "epoch": 54.45945945945946,
+      "grad_norm": 5.509948096005246e-06,
+      "learning_rate": 1.3312167560533337e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20465888,
+      "step": 26195
+    },
+    {
+      "epoch": 54.46985446985447,
+      "grad_norm": 0.11920446902513504,
+      "learning_rate": 1.3303489949321082e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 20469664,
+      "step": 26200
+    },
+    {
+      "epoch": 54.46985446985447,
+      "eval_loss": 0.5381605625152588,
+      "eval_runtime": 11.7291,
+      "eval_samples_per_second": 72.981,
+      "eval_steps_per_second": 18.245,
+      "num_input_tokens_seen": 20469664,
+      "step": 26200
+    },
+    {
+      "epoch": 54.48024948024948,
+      "grad_norm": 1.1501906556077302e-05,
+      "learning_rate": 1.3294814141857653e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20473536,
+      "step": 26205
+    },
+    {
+      "epoch": 54.49064449064449,
+      "grad_norm": 1.3089221283735242e-05,
+      "learning_rate": 1.3286140139480992e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20477312,
+      "step": 26210
+    },
+    {
+      "epoch": 54.5010395010395,
+      "grad_norm": 2.9605880627059378e-05,
+      "learning_rate": 1.3277467943528719e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20481120,
+      "step": 26215
+    },
+    {
+      "epoch": 54.51143451143451,
+      "grad_norm": 1.2813337889383547e-05,
+      "learning_rate": 1.3268797555338203e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20485248,
+      "step": 26220
+    },
+    {
+      "epoch": 54.521829521829524,
+      "grad_norm": 4.6407654735958204e-05,
+      "learning_rate": 1.3260128976246533e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20489120,
+      "step": 26225
+    },
+    {
+      "epoch": 54.53222453222453,
+      "grad_norm": 6.1059017752995715e-06,
+      "learning_rate": 1.32514622075905e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20493024,
+      "step": 26230
+    },
+    {
+      "epoch": 54.54261954261954,
+      "grad_norm": 2.876486541936174e-05,
+      "learning_rate": 1.3242797250706638e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20496832,
+      "step": 26235
+    },
+    {
+      "epoch": 54.553014553014556,
+      "grad_norm": 7.285965239134384e-06,
+      "learning_rate": 1.3234134106931195e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20500800,
+      "step": 26240
+    },
+    {
+      "epoch": 54.563409563409564,
+      "grad_norm": 6.0757462051697075e-05,
+      "learning_rate": 1.322547277760013e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20504672,
+      "step": 26245
+    },
+    {
+      "epoch": 54.57380457380457,
+      "grad_norm": 1.3247628885437734e-05,
+      "learning_rate": 1.3216813264049132e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20508544,
+      "step": 26250
+    },
+    {
+      "epoch": 54.58419958419958,
+      "grad_norm": 2.987709876833833e-06,
+      "learning_rate": 1.32081555676136e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20512512,
+      "step": 26255
+    },
+    {
+      "epoch": 54.5945945945946,
+      "grad_norm": 0.00010111611481988803,
+      "learning_rate": 1.3199499689628674e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20516288,
+      "step": 26260
+    },
+    {
+      "epoch": 54.604989604989605,
+      "grad_norm": 2.7269368274573935e-06,
+      "learning_rate": 1.3190845631429192e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 20520032,
+      "step": 26265
+    },
+    {
+      "epoch": 54.61538461538461,
+      "grad_norm": 1.9076889657299034e-05,
+      "learning_rate": 1.3182193394349704e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20523968,
+      "step": 26270
+    },
+    {
+      "epoch": 54.62577962577963,
+      "grad_norm": 1.9411369066801853e-05,
+      "learning_rate": 1.3173542979724507e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20527904,
+      "step": 26275
+    },
+    {
+      "epoch": 54.63617463617464,
+      "grad_norm": 4.8620822781231254e-06,
+      "learning_rate": 1.3164894388887617e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20531744,
+      "step": 26280
+    },
+    {
+      "epoch": 54.646569646569645,
+      "grad_norm": 1.6003942846509744e-06,
+      "learning_rate": 1.3156247623172727e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20535552,
+      "step": 26285
+    },
+    {
+      "epoch": 54.656964656964654,
+      "grad_norm": 5.882871391804656e-06,
+      "learning_rate": 1.3147602683913302e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20539488,
+      "step": 26290
+    },
+    {
+      "epoch": 54.66735966735967,
+      "grad_norm": 3.33782663801685e-05,
+      "learning_rate": 1.3138959572442481e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20543360,
+      "step": 26295
+    },
+    {
+      "epoch": 54.67775467775468,
+      "grad_norm": 4.228206307743676e-05,
+      "learning_rate": 1.3130318290093146e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20547328,
+      "step": 26300
+    },
+    {
+      "epoch": 54.688149688149686,
+      "grad_norm": 9.75114744505845e-06,
+      "learning_rate": 1.3121678838197909e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20551488,
+      "step": 26305
+    },
+    {
+      "epoch": 54.6985446985447,
+      "grad_norm": 5.966080789221451e-05,
+      "learning_rate": 1.3113041218089056e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20555136,
+      "step": 26310
+    },
+    {
+      "epoch": 54.70893970893971,
+      "grad_norm": 1.6128694824146805e-06,
+      "learning_rate": 1.3104405431098626e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20558848,
+      "step": 26315
+    },
+    {
+      "epoch": 54.71933471933472,
+      "grad_norm": 1.160515057563316e-05,
+      "learning_rate": 1.3095771478558377e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20563040,
+      "step": 26320
+    },
+    {
+      "epoch": 54.729729729729726,
+      "grad_norm": 4.000321860075928e-05,
+      "learning_rate": 1.3087139361799766e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 20567040,
+      "step": 26325
+    },
+    {
+      "epoch": 54.74012474012474,
+      "grad_norm": 3.5393725283938693e-06,
+      "learning_rate": 1.3078509082153964e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20571040,
+      "step": 26330
+    },
+    {
+      "epoch": 54.75051975051975,
+      "grad_norm": 1.141186567110708e-05,
+      "learning_rate": 1.3069880640951885e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20574848,
+      "step": 26335
+    },
+    {
+      "epoch": 54.76091476091476,
+      "grad_norm": 0.0042166560888290405,
+      "learning_rate": 1.3061254039524123e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20578720,
+      "step": 26340
+    },
+    {
+      "epoch": 54.771309771309774,
+      "grad_norm": 0.00031177871278487146,
+      "learning_rate": 1.3052629279201028e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20582720,
+      "step": 26345
+    },
+    {
+      "epoch": 54.78170478170478,
+      "grad_norm": 0.13462944328784943,
+      "learning_rate": 1.3044006361312633e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 20586688,
+      "step": 26350
+    },
+    {
+      "epoch": 54.79209979209979,
+      "grad_norm": 1.840831828303635e-05,
+      "learning_rate": 1.30353852871887e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 20590560,
+      "step": 26355
+    },
+    {
+      "epoch": 54.802494802494806,
+      "grad_norm": 2.3829783458495513e-05,
+      "learning_rate": 1.302676605815873e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20594400,
+      "step": 26360
+    },
+    {
+      "epoch": 54.812889812889814,
+      "grad_norm": 2.282280729559716e-05,
+      "learning_rate": 1.3018148675551884e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 20598208,
+      "step": 26365
+    },
+    {
+      "epoch": 54.82328482328482,
+      "grad_norm": 1.4019410627952311e-05,
+      "learning_rate": 1.3009533140697094e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20602208,
+      "step": 26370
+    },
+    {
+      "epoch": 54.83367983367983,
+      "grad_norm": 6.492867396445945e-05,
+      "learning_rate": 1.3000919454922966e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20606208,
+      "step": 26375
+    },
+    {
+      "epoch": 54.84407484407485,
+      "grad_norm": 0.0006551023107022047,
+      "learning_rate": 1.299230761955785e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20610112,
+      "step": 26380
+    },
+    {
+      "epoch": 54.854469854469855,
+      "grad_norm": 8.178786083590239e-05,
+      "learning_rate": 1.2983697635929807e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20614176,
+      "step": 26385
+    },
+    {
+      "epoch": 54.86486486486486,
+      "grad_norm": 4.702832939074142e-06,
+      "learning_rate": 1.2975089505366584e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20618144,
+      "step": 26390
+    },
+    {
+      "epoch": 54.87525987525988,
+      "grad_norm": 7.850887777749449e-05,
+      "learning_rate": 1.2966483229195683e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20622144,
+      "step": 26395
+    },
+    {
+      "epoch": 54.88565488565489,
+      "grad_norm": 0.17786775529384613,
+      "learning_rate": 1.2957878808744283e-05,
+      "loss": 0.0058,
+      "num_input_tokens_seen": 20625984,
+      "step": 26400
+    },
+    {
+      "epoch": 54.88565488565489,
+      "eval_loss": 0.5314427018165588,
+      "eval_runtime": 11.6905,
+      "eval_samples_per_second": 73.222,
+      "eval_steps_per_second": 18.305,
+      "num_input_tokens_seen": 20625984,
+      "step": 26400
+    },
+    {
+      "epoch": 54.896049896049895,
+      "grad_norm": 0.00021922255109529942,
+      "learning_rate": 1.294927624533931e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20629952,
+      "step": 26405
+    },
+    {
+      "epoch": 54.906444906444904,
+      "grad_norm": 2.5698847821331583e-05,
+      "learning_rate": 1.2940675540307378e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20633792,
+      "step": 26410
+    },
+    {
+      "epoch": 54.91683991683992,
+      "grad_norm": 7.524596185248811e-06,
+      "learning_rate": 1.2932076694974814e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20637824,
+      "step": 26415
+    },
+    {
+      "epoch": 54.92723492723493,
+      "grad_norm": 0.00010072839359054342,
+      "learning_rate": 1.2923479710667682e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20641824,
+      "step": 26420
+    },
+    {
+      "epoch": 54.937629937629936,
+      "grad_norm": 0.003069140948355198,
+      "learning_rate": 1.2914884588711751e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 20645600,
+      "step": 26425
+    },
+    {
+      "epoch": 54.94802494802495,
+      "grad_norm": 2.339772208870272e-06,
+      "learning_rate": 1.2906291330432475e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20649408,
+      "step": 26430
+    },
+    {
+      "epoch": 54.95841995841996,
+      "grad_norm": 0.0037556691095232964,
+      "learning_rate": 1.2897699937155055e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20653280,
+      "step": 26435
+    },
+    {
+      "epoch": 54.96881496881497,
+      "grad_norm": 6.549831596203148e-06,
+      "learning_rate": 1.2889110410204403e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20657184,
+      "step": 26440
+    },
+    {
+      "epoch": 54.979209979209976,
+      "grad_norm": 1.760614031809382e-05,
+      "learning_rate": 1.2880522750905111e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20661184,
+      "step": 26445
+    },
+    {
+      "epoch": 54.98960498960499,
+      "grad_norm": 1.3056412171863485e-05,
+      "learning_rate": 1.2871936960581523e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20665088,
+      "step": 26450
+    },
+    {
+      "epoch": 55.0,
+      "grad_norm": 1.010305368254194e-05,
+      "learning_rate": 1.2863353040557658e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20668784,
+      "step": 26455
+    },
+    {
+      "epoch": 55.01039501039501,
+      "grad_norm": 0.0009589128894731402,
+      "learning_rate": 1.2854770992157273e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20672848,
+      "step": 26460
+    },
+    {
+      "epoch": 55.020790020790024,
+      "grad_norm": 4.023521341878222e-06,
+      "learning_rate": 1.2846190816703835e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20676688,
+      "step": 26465
+    },
+    {
+      "epoch": 55.03118503118503,
+      "grad_norm": 4.0816677937982604e-05,
+      "learning_rate": 1.2837612515520498e-05,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 20680624,
+      "step": 26470
+    },
+    {
+      "epoch": 55.04158004158004,
+      "grad_norm": 1.5024997992441058e-05,
+      "learning_rate": 1.2829036089930163e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 20684560,
+      "step": 26475
+    },
+    {
+      "epoch": 55.05197505197505,
+      "grad_norm": 0.0001227207394549623,
+      "learning_rate": 1.2820461541255412e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20688528,
+      "step": 26480
+    },
+    {
+      "epoch": 55.062370062370064,
+      "grad_norm": 7.329399522859603e-05,
+      "learning_rate": 1.2811888870818543e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20692368,
+      "step": 26485
+    },
+    {
+      "epoch": 55.07276507276507,
+      "grad_norm": 2.9609229841298657e-06,
+      "learning_rate": 1.2803318079941581e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20696304,
+      "step": 26490
+    },
+    {
+      "epoch": 55.08316008316008,
+      "grad_norm": 1.1242688742640894e-05,
+      "learning_rate": 1.2794749169946235e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20700144,
+      "step": 26495
+    },
+    {
+      "epoch": 55.093555093555096,
+      "grad_norm": 1.4286410078057088e-05,
+      "learning_rate": 1.2786182142153952e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20704176,
+      "step": 26500
+    },
+    {
+      "epoch": 55.103950103950105,
+      "grad_norm": 4.194651046418585e-05,
+      "learning_rate": 1.2777616997885878e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20708048,
+      "step": 26505
+    },
+    {
+      "epoch": 55.11434511434511,
+      "grad_norm": 1.2508235158747993e-06,
+      "learning_rate": 1.2769053738462847e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20711888,
+      "step": 26510
+    },
+    {
+      "epoch": 55.12474012474012,
+      "grad_norm": 2.9549501050496474e-06,
+      "learning_rate": 1.2760492365205434e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20715664,
+      "step": 26515
+    },
+    {
+      "epoch": 55.13513513513514,
+      "grad_norm": 1.2021429029118735e-05,
+      "learning_rate": 1.2751932879433919e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20719344,
+      "step": 26520
+    },
+    {
+      "epoch": 55.145530145530145,
+      "grad_norm": 1.3958157978777308e-05,
+      "learning_rate": 1.2743375282468267e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20723152,
+      "step": 26525
+    },
+    {
+      "epoch": 55.15592515592515,
+      "grad_norm": 7.926594662421849e-06,
+      "learning_rate": 1.2734819575628182e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20727152,
+      "step": 26530
+    },
+    {
+      "epoch": 55.16632016632017,
+      "grad_norm": 0.00019224575953558087,
+      "learning_rate": 1.2726265760233039e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20731024,
+      "step": 26535
+    },
+    {
+      "epoch": 55.17671517671518,
+      "grad_norm": 1.0835070497705601e-05,
+      "learning_rate": 1.271771383760197e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20734864,
+      "step": 26540
+    },
+    {
+      "epoch": 55.187110187110186,
+      "grad_norm": 0.00029224733589217067,
+      "learning_rate": 1.2709163809053764e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 20738896,
+      "step": 26545
+    },
+    {
+      "epoch": 55.197505197505194,
+      "grad_norm": 4.2369971197331324e-05,
+      "learning_rate": 1.2700615675906963e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 20743024,
+      "step": 26550
+    },
+    {
+      "epoch": 55.20790020790021,
+      "grad_norm": 6.462755118263885e-05,
+      "learning_rate": 1.269206943947978e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20747088,
+      "step": 26555
+    },
+    {
+      "epoch": 55.21829521829522,
+      "grad_norm": 0.1278562843799591,
+      "learning_rate": 1.2683525101090177e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 20751120,
+      "step": 26560
+    },
+    {
+      "epoch": 55.228690228690226,
+      "grad_norm": 0.0004452753346413374,
+      "learning_rate": 1.2674982662055765e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20754928,
+      "step": 26565
+    },
+    {
+      "epoch": 55.23908523908524,
+      "grad_norm": 3.5237146221334115e-05,
+      "learning_rate": 1.2666442123693922e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20758640,
+      "step": 26570
+    },
+    {
+      "epoch": 55.24948024948025,
+      "grad_norm": 0.000505782722029835,
+      "learning_rate": 1.265790348732169e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20762480,
+      "step": 26575
+    },
+    {
+      "epoch": 55.25987525987526,
+      "grad_norm": 5.887769020773703e-06,
+      "learning_rate": 1.264936675425584e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20766192,
+      "step": 26580
+    },
+    {
+      "epoch": 55.270270270270274,
+      "grad_norm": 9.487928764428943e-06,
+      "learning_rate": 1.2640831925812852e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20769968,
+      "step": 26585
+    },
+    {
+      "epoch": 55.28066528066528,
+      "grad_norm": 2.2214373530005105e-05,
+      "learning_rate": 1.263229900330889e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20773904,
+      "step": 26590
+    },
+    {
+      "epoch": 55.29106029106029,
+      "grad_norm": 3.237548980905558e-06,
+      "learning_rate": 1.2623767988059843e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20778064,
+      "step": 26595
+    },
+    {
+      "epoch": 55.3014553014553,
+      "grad_norm": 1.5702275049989112e-05,
+      "learning_rate": 1.2615238881381309e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 20781904,
+      "step": 26600
+    },
+    {
+      "epoch": 55.3014553014553,
+      "eval_loss": 0.5448802709579468,
+      "eval_runtime": 11.6843,
+      "eval_samples_per_second": 73.261,
+      "eval_steps_per_second": 18.315,
+      "num_input_tokens_seen": 20781904,
+      "step": 26600
+    },
+    {
+      "epoch": 55.311850311850314,
+      "grad_norm": 0.00010398891754448414,
+      "learning_rate": 1.2606711684588568e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20785936,
+      "step": 26605
+    },
+    {
+      "epoch": 55.32224532224532,
+      "grad_norm": 5.4271285989671014e-06,
+      "learning_rate": 1.2598186398996636e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 20789872,
+      "step": 26610
+    },
+    {
+      "epoch": 55.33264033264033,
+      "grad_norm": 0.00011519777763169259,
+      "learning_rate": 1.2589663025920207e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20793776,
+      "step": 26615
+    },
+    {
+      "epoch": 55.343035343035346,
+      "grad_norm": 3.509876933094347e-06,
+      "learning_rate": 1.2581141566673705e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20797712,
+      "step": 26620
+    },
+    {
+      "epoch": 55.353430353430355,
+      "grad_norm": 8.23427180876024e-05,
+      "learning_rate": 1.257262202257124e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20801552,
+      "step": 26625
+    },
+    {
+      "epoch": 55.36382536382536,
+      "grad_norm": 2.217124892922584e-05,
+      "learning_rate": 1.2564104394926618e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20805520,
+      "step": 26630
+    },
+    {
+      "epoch": 55.37422037422037,
+      "grad_norm": 4.4913587771588936e-05,
+      "learning_rate": 1.2555588685053383e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20809424,
+      "step": 26635
+    },
+    {
+      "epoch": 55.38461538461539,
+      "grad_norm": 1.9516395695973188e-05,
+      "learning_rate": 1.2547074894264762e-05,
+      "loss": 0.0101,
+      "num_input_tokens_seen": 20813328,
+      "step": 26640
+    },
+    {
+      "epoch": 55.395010395010395,
+      "grad_norm": 1.583288621986867e-06,
+      "learning_rate": 1.2538563023873679e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20817200,
+      "step": 26645
+    },
+    {
+      "epoch": 55.4054054054054,
+      "grad_norm": 6.314153142739087e-05,
+      "learning_rate": 1.2530053075192789e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20821232,
+      "step": 26650
+    },
+    {
+      "epoch": 55.41580041580042,
+      "grad_norm": 0.00016895649605430663,
+      "learning_rate": 1.252154504953441e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20825008,
+      "step": 26655
+    },
+    {
+      "epoch": 55.42619542619543,
+      "grad_norm": 6.458801180997398e-06,
+      "learning_rate": 1.25130389482106e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20828848,
+      "step": 26660
+    },
+    {
+      "epoch": 55.436590436590436,
+      "grad_norm": 1.671795871516224e-06,
+      "learning_rate": 1.2504534772533116e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20832816,
+      "step": 26665
+    },
+    {
+      "epoch": 55.446985446985444,
+      "grad_norm": 3.435575490584597e-05,
+      "learning_rate": 1.2496032523813387e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20836592,
+      "step": 26670
+    },
+    {
+      "epoch": 55.45738045738046,
+      "grad_norm": 6.616926839342341e-06,
+      "learning_rate": 1.2487532203362576e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 20840560,
+      "step": 26675
+    },
+    {
+      "epoch": 55.46777546777547,
+      "grad_norm": 8.560236892662942e-06,
+      "learning_rate": 1.247903381249155e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20844624,
+      "step": 26680
+    },
+    {
+      "epoch": 55.478170478170476,
+      "grad_norm": 1.3053943803242873e-05,
+      "learning_rate": 1.2470537352510853e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20848528,
+      "step": 26685
+    },
+    {
+      "epoch": 55.48856548856549,
+      "grad_norm": 0.0001060044378391467,
+      "learning_rate": 1.2462042824730758e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20852432,
+      "step": 26690
+    },
+    {
+      "epoch": 55.4989604989605,
+      "grad_norm": 5.85245834372472e-05,
+      "learning_rate": 1.245355023046122e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20856464,
+      "step": 26695
+    },
+    {
+      "epoch": 55.50935550935551,
+      "grad_norm": 8.954346412792802e-06,
+      "learning_rate": 1.2445059571011896e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20860432,
+      "step": 26700
+    },
+    {
+      "epoch": 55.51975051975052,
+      "grad_norm": 1.3707774087379221e-05,
+      "learning_rate": 1.2436570847692173e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20864432,
+      "step": 26705
+    },
+    {
+      "epoch": 55.53014553014553,
+      "grad_norm": 1.883345248643309e-05,
+      "learning_rate": 1.2428084061811096e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20868400,
+      "step": 26710
+    },
+    {
+      "epoch": 55.54054054054054,
+      "grad_norm": 1.963604699994903e-05,
+      "learning_rate": 1.2419599214677447e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20872336,
+      "step": 26715
+    },
+    {
+      "epoch": 55.55093555093555,
+      "grad_norm": 2.974719973281026e-05,
+      "learning_rate": 1.2411116307599702e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20876208,
+      "step": 26720
+    },
+    {
+      "epoch": 55.561330561330564,
+      "grad_norm": 0.13562330603599548,
+      "learning_rate": 1.2402635341886016e-05,
+      "loss": 0.0064,
+      "num_input_tokens_seen": 20880080,
+      "step": 26725
+    },
+    {
+      "epoch": 55.57172557172557,
+      "grad_norm": 3.246588312322274e-05,
+      "learning_rate": 1.2394156318844278e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20883952,
+      "step": 26730
+    },
+    {
+      "epoch": 55.58212058212058,
+      "grad_norm": 5.406051514000865e-06,
+      "learning_rate": 1.2385679239782039e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20887856,
+      "step": 26735
+    },
+    {
+      "epoch": 55.59251559251559,
+      "grad_norm": 8.817821071716025e-05,
+      "learning_rate": 1.2377204106006585e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20891664,
+      "step": 26740
+    },
+    {
+      "epoch": 55.602910602910605,
+      "grad_norm": 3.42712810379453e-05,
+      "learning_rate": 1.2368730918824891e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20895600,
+      "step": 26745
+    },
+    {
+      "epoch": 55.61330561330561,
+      "grad_norm": 2.8099319024477154e-05,
+      "learning_rate": 1.236025967954362e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 20899504,
+      "step": 26750
+    },
+    {
+      "epoch": 55.62370062370062,
+      "grad_norm": 0.14141511917114258,
+      "learning_rate": 1.2351790389469153e-05,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 20903536,
+      "step": 26755
+    },
+    {
+      "epoch": 55.63409563409564,
+      "grad_norm": 4.854766757489415e-06,
+      "learning_rate": 1.234332304990755e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20907376,
+      "step": 26760
+    },
+    {
+      "epoch": 55.644490644490645,
+      "grad_norm": 0.00019539934874046594,
+      "learning_rate": 1.2334857662164593e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20911184,
+      "step": 26765
+    },
+    {
+      "epoch": 55.65488565488565,
+      "grad_norm": 6.00580642640125e-06,
+      "learning_rate": 1.2326394227545743e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20915216,
+      "step": 26770
+    },
+    {
+      "epoch": 55.66528066528066,
+      "grad_norm": 1.3509580639947671e-05,
+      "learning_rate": 1.2317932747356162e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20919120,
+      "step": 26775
+    },
+    {
+      "epoch": 55.67567567567568,
+      "grad_norm": 0.00035255466355010867,
+      "learning_rate": 1.2309473222900726e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20922896,
+      "step": 26780
+    },
+    {
+      "epoch": 55.686070686070686,
+      "grad_norm": 2.213505831605289e-05,
+      "learning_rate": 1.2301015655484006e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20926768,
+      "step": 26785
+    },
+    {
+      "epoch": 55.696465696465694,
+      "grad_norm": 0.0007698889821767807,
+      "learning_rate": 1.2292560046410245e-05,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 20930672,
+      "step": 26790
+    },
+    {
+      "epoch": 55.70686070686071,
+      "grad_norm": 0.0002042495325440541,
+      "learning_rate": 1.228410639698343e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20934672,
+      "step": 26795
+    },
+    {
+      "epoch": 55.71725571725572,
+      "grad_norm": 0.00010143122199224308,
+      "learning_rate": 1.2275654708507195e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20938512,
+      "step": 26800
+    },
+    {
+      "epoch": 55.71725571725572,
+      "eval_loss": 0.5368707776069641,
+      "eval_runtime": 11.7323,
+      "eval_samples_per_second": 72.961,
+      "eval_steps_per_second": 18.24,
+      "num_input_tokens_seen": 20938512,
+      "step": 26800
+    },
+    {
+      "epoch": 55.727650727650726,
+      "grad_norm": 0.0001256241084774956,
+      "learning_rate": 1.2267204982284908e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20942544,
+      "step": 26805
+    },
+    {
+      "epoch": 55.73804573804574,
+      "grad_norm": 0.00010160940291825682,
+      "learning_rate": 1.2258757219619635e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20946704,
+      "step": 26810
+    },
+    {
+      "epoch": 55.74844074844075,
+      "grad_norm": 8.849706318869721e-06,
+      "learning_rate": 1.2250311421814104e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20950672,
+      "step": 26815
+    },
+    {
+      "epoch": 55.75883575883576,
+      "grad_norm": 5.2014488574059214e-06,
+      "learning_rate": 1.2241867590170772e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20954512,
+      "step": 26820
+    },
+    {
+      "epoch": 55.76923076923077,
+      "grad_norm": 3.8289930671453476e-05,
+      "learning_rate": 1.2233425725991799e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20958480,
+      "step": 26825
+    },
+    {
+      "epoch": 55.77962577962578,
+      "grad_norm": 8.215883281081915e-05,
+      "learning_rate": 1.2224985830579003e-05,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 20962384,
+      "step": 26830
+    },
+    {
+      "epoch": 55.79002079002079,
+      "grad_norm": 6.5076910686912015e-06,
+      "learning_rate": 1.2216547905233944e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20966256,
+      "step": 26835
+    },
+    {
+      "epoch": 55.8004158004158,
+      "grad_norm": 8.478533345623873e-06,
+      "learning_rate": 1.2208111951257842e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 20970288,
+      "step": 26840
+    },
+    {
+      "epoch": 55.810810810810814,
+      "grad_norm": 2.7536016204976477e-05,
+      "learning_rate": 1.2199677969951622e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20974000,
+      "step": 26845
+    },
+    {
+      "epoch": 55.82120582120582,
+      "grad_norm": 0.00023213273379951715,
+      "learning_rate": 1.2191245962615927e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20977808,
+      "step": 26850
+    },
+    {
+      "epoch": 55.83160083160083,
+      "grad_norm": 5.6137734645744786e-05,
+      "learning_rate": 1.218281593055106e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20981680,
+      "step": 26855
+    },
+    {
+      "epoch": 55.84199584199584,
+      "grad_norm": 1.1467636795714498e-05,
+      "learning_rate": 1.217438787505705e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 20985552,
+      "step": 26860
+    },
+    {
+      "epoch": 55.852390852390855,
+      "grad_norm": 7.412298145936802e-06,
+      "learning_rate": 1.2165961797433615e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20989392,
+      "step": 26865
+    },
+    {
+      "epoch": 55.86278586278586,
+      "grad_norm": 3.896246198564768e-05,
+      "learning_rate": 1.215753769898014e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 20993264,
+      "step": 26870
+    },
+    {
+      "epoch": 55.87318087318087,
+      "grad_norm": 6.997279706411064e-05,
+      "learning_rate": 1.2149115580995755e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 20997136,
+      "step": 26875
+    },
+    {
+      "epoch": 55.88357588357589,
+      "grad_norm": 0.000188039819477126,
+      "learning_rate": 1.2140695444779227e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21001168,
+      "step": 26880
+    },
+    {
+      "epoch": 55.893970893970895,
+      "grad_norm": 1.0829337952600326e-05,
+      "learning_rate": 1.2132277291629066e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21004944,
+      "step": 26885
+    },
+    {
+      "epoch": 55.9043659043659,
+      "grad_norm": 7.3744859037105925e-06,
+      "learning_rate": 1.2123861122843458e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21008752,
+      "step": 26890
+    },
+    {
+      "epoch": 55.91476091476091,
+      "grad_norm": 2.2966843971516937e-05,
+      "learning_rate": 1.2115446939720271e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21012784,
+      "step": 26895
+    },
+    {
+      "epoch": 55.92515592515593,
+      "grad_norm": 3.4294542274437845e-05,
+      "learning_rate": 1.210703474355708e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21016752,
+      "step": 26900
+    },
+    {
+      "epoch": 55.935550935550935,
+      "grad_norm": 5.001049430575222e-05,
+      "learning_rate": 1.2098624535651164e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21020496,
+      "step": 26905
+    },
+    {
+      "epoch": 55.945945945945944,
+      "grad_norm": 6.106388809712371e-06,
+      "learning_rate": 1.2090216317299477e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21024336,
+      "step": 26910
+    },
+    {
+      "epoch": 55.95634095634096,
+      "grad_norm": 1.8216989701613784e-05,
+      "learning_rate": 1.2081810089798668e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21028336,
+      "step": 26915
+    },
+    {
+      "epoch": 55.96673596673597,
+      "grad_norm": 6.240437414817279e-06,
+      "learning_rate": 1.2073405854445072e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21032400,
+      "step": 26920
+    },
+    {
+      "epoch": 55.977130977130976,
+      "grad_norm": 3.509196858431096e-06,
+      "learning_rate": 1.206500361253474e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21036272,
+      "step": 26925
+    },
+    {
+      "epoch": 55.987525987525984,
+      "grad_norm": 0.11834053695201874,
+      "learning_rate": 1.2056603365363409e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 21040144,
+      "step": 26930
+    },
+    {
+      "epoch": 55.997920997921,
+      "grad_norm": 2.5653213015175425e-05,
+      "learning_rate": 1.2048205114226487e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21044080,
+      "step": 26935
+    },
+    {
+      "epoch": 56.00831600831601,
+      "grad_norm": 6.947001566004474e-06,
+      "learning_rate": 1.2039808860419102e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21047840,
+      "step": 26940
+    },
+    {
+      "epoch": 56.018711018711016,
+      "grad_norm": 7.702044968027622e-06,
+      "learning_rate": 1.2031414605236066e-05,
+      "loss": 0.0075,
+      "num_input_tokens_seen": 21051808,
+      "step": 26945
+    },
+    {
+      "epoch": 56.02910602910603,
+      "grad_norm": 0.0004226279561407864,
+      "learning_rate": 1.2023022349971862e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21055744,
+      "step": 26950
+    },
+    {
+      "epoch": 56.03950103950104,
+      "grad_norm": 2.440297748762532e-06,
+      "learning_rate": 1.20146320959207e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21059680,
+      "step": 26955
+    },
+    {
+      "epoch": 56.04989604989605,
+      "grad_norm": 3.384592855582014e-05,
+      "learning_rate": 1.2006243844376445e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21063584,
+      "step": 26960
+    },
+    {
+      "epoch": 56.06029106029106,
+      "grad_norm": 2.9528207960538566e-05,
+      "learning_rate": 1.1997857596632678e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21067424,
+      "step": 26965
+    },
+    {
+      "epoch": 56.07068607068607,
+      "grad_norm": 8.53714082040824e-05,
+      "learning_rate": 1.1989473353982672e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21071552,
+      "step": 26970
+    },
+    {
+      "epoch": 56.08108108108108,
+      "grad_norm": 4.886633178102784e-06,
+      "learning_rate": 1.198109111771937e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21075424,
+      "step": 26975
+    },
+    {
+      "epoch": 56.09147609147609,
+      "grad_norm": 2.4533967007300816e-05,
+      "learning_rate": 1.197271088913543e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21079264,
+      "step": 26980
+    },
+    {
+      "epoch": 56.101871101871104,
+      "grad_norm": 2.80129042948829e-05,
+      "learning_rate": 1.1964332669523182e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21083296,
+      "step": 26985
+    },
+    {
+      "epoch": 56.11226611226611,
+      "grad_norm": 8.485299622407183e-05,
+      "learning_rate": 1.1955956460174645e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21087168,
+      "step": 26990
+    },
+    {
+      "epoch": 56.12266112266112,
+      "grad_norm": 1.0459991244715638e-05,
+      "learning_rate": 1.1947582262381552e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21091104,
+      "step": 26995
+    },
+    {
+      "epoch": 56.13305613305613,
+      "grad_norm": 2.0745812435052358e-05,
+      "learning_rate": 1.1939210077435293e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21095008,
+      "step": 27000
+    },
+    {
+      "epoch": 56.13305613305613,
+      "eval_loss": 0.5501160025596619,
+      "eval_runtime": 11.7198,
+      "eval_samples_per_second": 73.039,
+      "eval_steps_per_second": 18.26,
+      "num_input_tokens_seen": 21095008,
+      "step": 27000
+    },
+    {
+      "epoch": 56.143451143451145,
+      "grad_norm": 7.535571057815105e-06,
+      "learning_rate": 1.193083990662697e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21098624,
+      "step": 27005
+    },
+    {
+      "epoch": 56.15384615384615,
+      "grad_norm": 1.65355704666581e-05,
+      "learning_rate": 1.192247175124738e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21102432,
+      "step": 27010
+    },
+    {
+      "epoch": 56.16424116424116,
+      "grad_norm": 1.7666843632468954e-05,
+      "learning_rate": 1.191410561258698e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 21106272,
+      "step": 27015
+    },
+    {
+      "epoch": 56.17463617463618,
+      "grad_norm": 1.2773204616678413e-05,
+      "learning_rate": 1.1905741491935944e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21110048,
+      "step": 27020
+    },
+    {
+      "epoch": 56.185031185031185,
+      "grad_norm": 1.835068906075321e-05,
+      "learning_rate": 1.1897379390584129e-05,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 21113856,
+      "step": 27025
+    },
+    {
+      "epoch": 56.195426195426194,
+      "grad_norm": 8.899865497369319e-05,
+      "learning_rate": 1.1889019309821062e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21117920,
+      "step": 27030
+    },
+    {
+      "epoch": 56.20582120582121,
+      "grad_norm": 3.400586501811631e-05,
+      "learning_rate": 1.188066125093599e-05,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 21121952,
+      "step": 27035
+    },
+    {
+      "epoch": 56.21621621621622,
+      "grad_norm": 6.1982364059076644e-06,
+      "learning_rate": 1.1872305215217811e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21125792,
+      "step": 27040
+    },
+    {
+      "epoch": 56.226611226611226,
+      "grad_norm": 9.585134102962911e-05,
+      "learning_rate": 1.186395120395514e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21129696,
+      "step": 27045
+    },
+    {
+      "epoch": 56.237006237006234,
+      "grad_norm": 7.277079021150712e-06,
+      "learning_rate": 1.1855599218436283e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21133664,
+      "step": 27050
+    },
+    {
+      "epoch": 56.24740124740125,
+      "grad_norm": 7.868999819038436e-05,
+      "learning_rate": 1.1847249259949209e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 21137664,
+      "step": 27055
+    },
+    {
+      "epoch": 56.25779625779626,
+      "grad_norm": 1.4188673958415166e-05,
+      "learning_rate": 1.1838901329781574e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21141632,
+      "step": 27060
+    },
+    {
+      "epoch": 56.268191268191266,
+      "grad_norm": 0.0003470042138360441,
+      "learning_rate": 1.1830555429220758e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 21145536,
+      "step": 27065
+    },
+    {
+      "epoch": 56.27858627858628,
+      "grad_norm": 2.2744652596884407e-06,
+      "learning_rate": 1.1822211559553784e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21149504,
+      "step": 27070
+    },
+    {
+      "epoch": 56.28898128898129,
+      "grad_norm": 3.0303806397569133e-06,
+      "learning_rate": 1.18138697220674e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21153408,
+      "step": 27075
+    },
+    {
+      "epoch": 56.2993762993763,
+      "grad_norm": 1.8525159248383716e-05,
+      "learning_rate": 1.1805529918048e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21157248,
+      "step": 27080
+    },
+    {
+      "epoch": 56.30977130977131,
+      "grad_norm": 2.010169737332035e-05,
+      "learning_rate": 1.1797192148781702e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21161056,
+      "step": 27085
+    },
+    {
+      "epoch": 56.32016632016632,
+      "grad_norm": 3.254523835494183e-05,
+      "learning_rate": 1.1788856415554297e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21164864,
+      "step": 27090
+    },
+    {
+      "epoch": 56.33056133056133,
+      "grad_norm": 2.4104378098854795e-05,
+      "learning_rate": 1.1780522719651249e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21168736,
+      "step": 27095
+    },
+    {
+      "epoch": 56.34095634095634,
+      "grad_norm": 0.00013924931408837438,
+      "learning_rate": 1.1772191062357721e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21172608,
+      "step": 27100
+    },
+    {
+      "epoch": 56.351351351351354,
+      "grad_norm": 5.987561507936334e-06,
+      "learning_rate": 1.1763861444958573e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21176512,
+      "step": 27105
+    },
+    {
+      "epoch": 56.36174636174636,
+      "grad_norm": 4.0699342207517475e-05,
+      "learning_rate": 1.1755533868738317e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21180512,
+      "step": 27110
+    },
+    {
+      "epoch": 56.37214137214137,
+      "grad_norm": 0.00036862344131805,
+      "learning_rate": 1.1747208334981185e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21184480,
+      "step": 27115
+    },
+    {
+      "epoch": 56.38253638253638,
+      "grad_norm": 1.7348143956041895e-05,
+      "learning_rate": 1.1738884844971067e-05,
+      "loss": 0.0043,
+      "num_input_tokens_seen": 21188544,
+      "step": 27120
+    },
+    {
+      "epoch": 56.392931392931395,
+      "grad_norm": 5.4802508202556055e-06,
+      "learning_rate": 1.1730563399991563e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21192320,
+      "step": 27125
+    },
+    {
+      "epoch": 56.4033264033264,
+      "grad_norm": 8.19688466435764e-06,
+      "learning_rate": 1.1722244001325938e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21196352,
+      "step": 27130
+    },
+    {
+      "epoch": 56.41372141372141,
+      "grad_norm": 5.353280357667245e-05,
+      "learning_rate": 1.1713926650257137e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21200224,
+      "step": 27135
+    },
+    {
+      "epoch": 56.42411642411643,
+      "grad_norm": 0.0003728808951564133,
+      "learning_rate": 1.170561134806781e-05,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 21204096,
+      "step": 27140
+    },
+    {
+      "epoch": 56.434511434511435,
+      "grad_norm": 1.859264011727646e-05,
+      "learning_rate": 1.1697298096040287e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 21208096,
+      "step": 27145
+    },
+    {
+      "epoch": 56.444906444906444,
+      "grad_norm": 0.00012893075472675264,
+      "learning_rate": 1.1688986895456567e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21212128,
+      "step": 27150
+    },
+    {
+      "epoch": 56.45530145530145,
+      "grad_norm": 6.2166527641238645e-06,
+      "learning_rate": 1.1680677747598349e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 21215904,
+      "step": 27155
+    },
+    {
+      "epoch": 56.46569646569647,
+      "grad_norm": 1.5746460121590644e-05,
+      "learning_rate": 1.1672370653746995e-05,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 21219744,
+      "step": 27160
+    },
+    {
+      "epoch": 56.476091476091476,
+      "grad_norm": 5.3556639613816515e-05,
+      "learning_rate": 1.166406561518357e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 21223744,
+      "step": 27165
+    },
+    {
+      "epoch": 56.486486486486484,
+      "grad_norm": 6.295858383964514e-06,
+      "learning_rate": 1.1655762633188826e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21227744,
+      "step": 27170
+    },
+    {
+      "epoch": 56.4968814968815,
+      "grad_norm": 1.2129611604905222e-05,
+      "learning_rate": 1.1647461709043172e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21231680,
+      "step": 27175
+    },
+    {
+      "epoch": 56.50727650727651,
+      "grad_norm": 9.257550846086815e-05,
+      "learning_rate": 1.1639162844026722e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21235520,
+      "step": 27180
+    },
+    {
+      "epoch": 56.517671517671516,
+      "grad_norm": 2.2467753296950832e-05,
+      "learning_rate": 1.163086603941927e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21239424,
+      "step": 27185
+    },
+    {
+      "epoch": 56.528066528066525,
+      "grad_norm": 0.00020304491044953465,
+      "learning_rate": 1.1622571296500273e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21243168,
+      "step": 27190
+    },
+    {
+      "epoch": 56.53846153846154,
+      "grad_norm": 6.652022420894355e-05,
+      "learning_rate": 1.1614278616548904e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21247360,
+      "step": 27195
+    },
+    {
+      "epoch": 56.54885654885655,
+      "grad_norm": 9.256851626560092e-05,
+      "learning_rate": 1.1605988000843986e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21251264,
+      "step": 27200
+    },
+    {
+      "epoch": 56.54885654885655,
+      "eval_loss": 0.5637614727020264,
+      "eval_runtime": 11.6866,
+      "eval_samples_per_second": 73.247,
+      "eval_steps_per_second": 18.312,
+      "num_input_tokens_seen": 21251264,
+      "step": 27200
+    },
+    {
+      "epoch": 56.55925155925156,
+      "grad_norm": 9.206669346895069e-06,
+      "learning_rate": 1.1597699450664028e-05,
+      "loss": 0.002,
+      "num_input_tokens_seen": 21255168,
+      "step": 27205
+    },
+    {
+      "epoch": 56.56964656964657,
+      "grad_norm": 0.0018069507787004113,
+      "learning_rate": 1.1589412967287252e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21259040,
+      "step": 27210
+    },
+    {
+      "epoch": 56.58004158004158,
+      "grad_norm": 0.00047293282113969326,
+      "learning_rate": 1.1581128551991514e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21262848,
+      "step": 27215
+    },
+    {
+      "epoch": 56.59043659043659,
+      "grad_norm": 0.0004951197770424187,
+      "learning_rate": 1.1572846206054383e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21266720,
+      "step": 27220
+    },
+    {
+      "epoch": 56.6008316008316,
+      "grad_norm": 1.8455039025866427e-06,
+      "learning_rate": 1.1564565930753113e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21270752,
+      "step": 27225
+    },
+    {
+      "epoch": 56.61122661122661,
+      "grad_norm": 8.904317837732378e-06,
+      "learning_rate": 1.1556287727364606e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21274592,
+      "step": 27230
+    },
+    {
+      "epoch": 56.62162162162162,
+      "grad_norm": 0.00027461297577247024,
+      "learning_rate": 1.1548011597165489e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21278656,
+      "step": 27235
+    },
+    {
+      "epoch": 56.63201663201663,
+      "grad_norm": 6.677795681753196e-06,
+      "learning_rate": 1.1539737541432019e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21282720,
+      "step": 27240
+    },
+    {
+      "epoch": 56.642411642411645,
+      "grad_norm": 0.00015273346798494458,
+      "learning_rate": 1.1531465561440174e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21286688,
+      "step": 27245
+    },
+    {
+      "epoch": 56.65280665280665,
+      "grad_norm": 1.9913679807359586e-06,
+      "learning_rate": 1.1523195658465605e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21290720,
+      "step": 27250
+    },
+    {
+      "epoch": 56.66320166320166,
+      "grad_norm": 5.70058227822301e-06,
+      "learning_rate": 1.1514927833783618e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21294688,
+      "step": 27255
+    },
+    {
+      "epoch": 56.67359667359668,
+      "grad_norm": 1.660834277572576e-05,
+      "learning_rate": 1.150666208866922e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21298528,
+      "step": 27260
+    },
+    {
+      "epoch": 56.683991683991685,
+      "grad_norm": 3.087196773776668e-06,
+      "learning_rate": 1.1498398424397106e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21302368,
+      "step": 27265
+    },
+    {
+      "epoch": 56.694386694386694,
+      "grad_norm": 8.014917693799362e-05,
+      "learning_rate": 1.1490136842241628e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21306208,
+      "step": 27270
+    },
+    {
+      "epoch": 56.7047817047817,
+      "grad_norm": 0.0005204054177738726,
+      "learning_rate": 1.1481877343476813e-05,
+      "loss": 0.0088,
+      "num_input_tokens_seen": 21310144,
+      "step": 27275
+    },
+    {
+      "epoch": 56.71517671517672,
+      "grad_norm": 0.00030476725078187883,
+      "learning_rate": 1.14736199293764e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21314016,
+      "step": 27280
+    },
+    {
+      "epoch": 56.725571725571726,
+      "grad_norm": 0.00034329431946389377,
+      "learning_rate": 1.1465364601213771e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21317792,
+      "step": 27285
+    },
+    {
+      "epoch": 56.735966735966734,
+      "grad_norm": 7.131896381906699e-06,
+      "learning_rate": 1.1457111360262012e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21321728,
+      "step": 27290
+    },
+    {
+      "epoch": 56.74636174636175,
+      "grad_norm": 7.335855480050668e-05,
+      "learning_rate": 1.1448860207793869e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21325696,
+      "step": 27295
+    },
+    {
+      "epoch": 56.75675675675676,
+      "grad_norm": 7.347588962147711e-06,
+      "learning_rate": 1.144061114508177e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21329632,
+      "step": 27300
+    },
+    {
+      "epoch": 56.767151767151766,
+      "grad_norm": 5.632722604786977e-05,
+      "learning_rate": 1.1432364173397842e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21333664,
+      "step": 27305
+    },
+    {
+      "epoch": 56.777546777546775,
+      "grad_norm": 7.869850378483534e-06,
+      "learning_rate": 1.1424119294013852e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21337536,
+      "step": 27310
+    },
+    {
+      "epoch": 56.78794178794179,
+      "grad_norm": 0.00014832538727205247,
+      "learning_rate": 1.1415876508201279e-05,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 21341536,
+      "step": 27315
+    },
+    {
+      "epoch": 56.7983367983368,
+      "grad_norm": 6.0054603636672255e-06,
+      "learning_rate": 1.140763581723125e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21345472,
+      "step": 27320
+    },
+    {
+      "epoch": 56.80873180873181,
+      "grad_norm": 6.713036418659613e-05,
+      "learning_rate": 1.1399397222374588e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21349280,
+      "step": 27325
+    },
+    {
+      "epoch": 56.81912681912682,
+      "grad_norm": 9.59547560341889e-06,
+      "learning_rate": 1.1391160724901804e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21353248,
+      "step": 27330
+    },
+    {
+      "epoch": 56.82952182952183,
+      "grad_norm": 0.0001541633391752839,
+      "learning_rate": 1.138292632608304e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21357088,
+      "step": 27335
+    },
+    {
+      "epoch": 56.83991683991684,
+      "grad_norm": 6.821239367127419e-05,
+      "learning_rate": 1.1374694027188174e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21361120,
+      "step": 27340
+    },
+    {
+      "epoch": 56.85031185031185,
+      "grad_norm": 0.0003034454130101949,
+      "learning_rate": 1.1366463829486711e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21364992,
+      "step": 27345
+    },
+    {
+      "epoch": 56.86070686070686,
+      "grad_norm": 3.8885886169737205e-05,
+      "learning_rate": 1.1358235734247849e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21368928,
+      "step": 27350
+    },
+    {
+      "epoch": 56.87110187110187,
+      "grad_norm": 2.6275598429492675e-05,
+      "learning_rate": 1.1350009742740478e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21372864,
+      "step": 27355
+    },
+    {
+      "epoch": 56.88149688149688,
+      "grad_norm": 1.0720594218582846e-05,
+      "learning_rate": 1.134178585623313e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21376768,
+      "step": 27360
+    },
+    {
+      "epoch": 56.891891891891895,
+      "grad_norm": 6.3592942751711234e-06,
+      "learning_rate": 1.1333564075994047e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21380608,
+      "step": 27365
+    },
+    {
+      "epoch": 56.9022869022869,
+      "grad_norm": 2.0187124391668476e-05,
+      "learning_rate": 1.1325344403291133e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21384480,
+      "step": 27370
+    },
+    {
+      "epoch": 56.91268191268191,
+      "grad_norm": 1.398664517182624e-05,
+      "learning_rate": 1.1317126839391951e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 21388352,
+      "step": 27375
+    },
+    {
+      "epoch": 56.92307692307692,
+      "grad_norm": 1.2204441190988291e-05,
+      "learning_rate": 1.1308911385563766e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21392160,
+      "step": 27380
+    },
+    {
+      "epoch": 56.933471933471935,
+      "grad_norm": 0.14389312267303467,
+      "learning_rate": 1.1300698043073494e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 21396128,
+      "step": 27385
+    },
+    {
+      "epoch": 56.943866943866944,
+      "grad_norm": 0.00029670074582099915,
+      "learning_rate": 1.1292486813187736e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21400064,
+      "step": 27390
+    },
+    {
+      "epoch": 56.95426195426195,
+      "grad_norm": 3.843385638901964e-05,
+      "learning_rate": 1.1284277697172782e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21403936,
+      "step": 27395
+    },
+    {
+      "epoch": 56.96465696465697,
+      "grad_norm": 2.5288196411565877e-05,
+      "learning_rate": 1.127607069629456e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21407744,
+      "step": 27400
+    },
+    {
+      "epoch": 56.96465696465697,
+      "eval_loss": 0.5524911880493164,
+      "eval_runtime": 11.6901,
+      "eval_samples_per_second": 73.224,
+      "eval_steps_per_second": 18.306,
+      "num_input_tokens_seen": 21407744,
+      "step": 27400
+    },
+    {
+      "epoch": 56.975051975051976,
+      "grad_norm": 0.0012285147095099092,
+      "learning_rate": 1.1267865811818701e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21411776,
+      "step": 27405
+    },
+    {
+      "epoch": 56.985446985446984,
+      "grad_norm": 5.729627446271479e-06,
+      "learning_rate": 1.1259663045010513e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21415648,
+      "step": 27410
+    },
+    {
+      "epoch": 56.99584199584199,
+      "grad_norm": 0.00011242563050473109,
+      "learning_rate": 1.1251462397134957e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21419744,
+      "step": 27415
+    },
+    {
+      "epoch": 57.00623700623701,
+      "grad_norm": 6.2019089455134235e-06,
+      "learning_rate": 1.1243263869456664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21423728,
+      "step": 27420
+    },
+    {
+      "epoch": 57.016632016632016,
+      "grad_norm": 1.5191085367405321e-05,
+      "learning_rate": 1.1235067463239967e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 21427696,
+      "step": 27425
+    },
+    {
+      "epoch": 57.027027027027025,
+      "grad_norm": 1.952250613612705e-06,
+      "learning_rate": 1.122687317974884e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21431600,
+      "step": 27430
+    },
+    {
+      "epoch": 57.03742203742204,
+      "grad_norm": 5.597527433565119e-06,
+      "learning_rate": 1.1218681020246963e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21435504,
+      "step": 27435
+    },
+    {
+      "epoch": 57.04781704781705,
+      "grad_norm": 0.00012984176282770932,
+      "learning_rate": 1.1210490985997652e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21439408,
+      "step": 27440
+    },
+    {
+      "epoch": 57.05821205821206,
+      "grad_norm": 2.9330025427043438e-05,
+      "learning_rate": 1.1202303078263917e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21443248,
+      "step": 27445
+    },
+    {
+      "epoch": 57.06860706860707,
+      "grad_norm": 0.1010606437921524,
+      "learning_rate": 1.1194117298308451e-05,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 21447216,
+      "step": 27450
+    },
+    {
+      "epoch": 57.07900207900208,
+      "grad_norm": 9.614248483558185e-06,
+      "learning_rate": 1.1185933647393585e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 21451152,
+      "step": 27455
+    },
+    {
+      "epoch": 57.08939708939709,
+      "grad_norm": 3.999910768470727e-05,
+      "learning_rate": 1.1177752126781354e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21455280,
+      "step": 27460
+    },
+    {
+      "epoch": 57.0997920997921,
+      "grad_norm": 0.0002878968371078372,
+      "learning_rate": 1.1169572737733441e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 21459312,
+      "step": 27465
+    },
+    {
+      "epoch": 57.11018711018711,
+      "grad_norm": 0.0007984951371327043,
+      "learning_rate": 1.1161395481511216e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21463248,
+      "step": 27470
+    },
+    {
+      "epoch": 57.12058212058212,
+      "grad_norm": 0.00041582409176044166,
+      "learning_rate": 1.1153220359375722e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21467120,
+      "step": 27475
+    },
+    {
+      "epoch": 57.13097713097713,
+      "grad_norm": 0.00013764886534772813,
+      "learning_rate": 1.114504737258765e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21470960,
+      "step": 27480
+    },
+    {
+      "epoch": 57.141372141372145,
+      "grad_norm": 7.249226655403618e-06,
+      "learning_rate": 1.1136876522407393e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21474896,
+      "step": 27485
+    },
+    {
+      "epoch": 57.15176715176715,
+      "grad_norm": 1.5100588825589512e-05,
+      "learning_rate": 1.1128707810094985e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21478704,
+      "step": 27490
+    },
+    {
+      "epoch": 57.16216216216216,
+      "grad_norm": 2.4906714315875433e-05,
+      "learning_rate": 1.1120541236910157e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21482608,
+      "step": 27495
+    },
+    {
+      "epoch": 57.17255717255717,
+      "grad_norm": 1.1343206551828189e-06,
+      "learning_rate": 1.111237680411229e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21486640,
+      "step": 27500
+    },
+    {
+      "epoch": 57.182952182952185,
+      "grad_norm": 4.614582940121181e-05,
+      "learning_rate": 1.1104214512960433e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21490608,
+      "step": 27505
+    },
+    {
+      "epoch": 57.19334719334719,
+      "grad_norm": 5.399288420449011e-05,
+      "learning_rate": 1.1096054364713327e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21494384,
+      "step": 27510
+    },
+    {
+      "epoch": 57.2037422037422,
+      "grad_norm": 0.00010859203757718205,
+      "learning_rate": 1.1087896360629371e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21498128,
+      "step": 27515
+    },
+    {
+      "epoch": 57.21413721413722,
+      "grad_norm": 4.172568878857419e-06,
+      "learning_rate": 1.107974050196662e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21502064,
+      "step": 27520
+    },
+    {
+      "epoch": 57.224532224532226,
+      "grad_norm": 6.590229168068618e-05,
+      "learning_rate": 1.1071586789982816e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 21506064,
+      "step": 27525
+    },
+    {
+      "epoch": 57.234927234927234,
+      "grad_norm": 6.96684219292365e-05,
+      "learning_rate": 1.1063435225935373e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21510000,
+      "step": 27530
+    },
+    {
+      "epoch": 57.24532224532224,
+      "grad_norm": 2.1237692635622807e-06,
+      "learning_rate": 1.1055285811081348e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21513936,
+      "step": 27535
+    },
+    {
+      "epoch": 57.25571725571726,
+      "grad_norm": 2.1259565983200446e-05,
+      "learning_rate": 1.1047138546677499e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21517968,
+      "step": 27540
+    },
+    {
+      "epoch": 57.266112266112266,
+      "grad_norm": 8.453285772702657e-06,
+      "learning_rate": 1.1038993433980219e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21521744,
+      "step": 27545
+    },
+    {
+      "epoch": 57.276507276507274,
+      "grad_norm": 0.00035770487738773227,
+      "learning_rate": 1.1030850474245597e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21525648,
+      "step": 27550
+    },
+    {
+      "epoch": 57.28690228690229,
+      "grad_norm": 3.1719573598820716e-05,
+      "learning_rate": 1.102270966872939e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21529552,
+      "step": 27555
+    },
+    {
+      "epoch": 57.2972972972973,
+      "grad_norm": 0.0001359283342026174,
+      "learning_rate": 1.1014571018687e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21533456,
+      "step": 27560
+    },
+    {
+      "epoch": 57.30769230769231,
+      "grad_norm": 7.618964446010068e-06,
+      "learning_rate": 1.1006434525373502e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21537392,
+      "step": 27565
+    },
+    {
+      "epoch": 57.318087318087315,
+      "grad_norm": 4.6715490498172585e-06,
+      "learning_rate": 1.0998300190043664e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21541264,
+      "step": 27570
+    },
+    {
+      "epoch": 57.32848232848233,
+      "grad_norm": 4.955471922585275e-06,
+      "learning_rate": 1.0990168013951882e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21545168,
+      "step": 27575
+    },
+    {
+      "epoch": 57.33887733887734,
+      "grad_norm": 2.21091568164411e-06,
+      "learning_rate": 1.0982037998352263e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21549040,
+      "step": 27580
+    },
+    {
+      "epoch": 57.34927234927235,
+      "grad_norm": 2.951882743218448e-05,
+      "learning_rate": 1.0973910144498534e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21552880,
+      "step": 27585
+    },
+    {
+      "epoch": 57.35966735966736,
+      "grad_norm": 2.7297130145598203e-05,
+      "learning_rate": 1.0965784453644123e-05,
+      "loss": 0.0013,
+      "num_input_tokens_seen": 21556752,
+      "step": 27590
+    },
+    {
+      "epoch": 57.37006237006237,
+      "grad_norm": 0.0007616601069457829,
+      "learning_rate": 1.0957660927042127e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21560784,
+      "step": 27595
+    },
+    {
+      "epoch": 57.38045738045738,
+      "grad_norm": 2.6042065655929036e-05,
+      "learning_rate": 1.094953956594527e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21564560,
+      "step": 27600
+    },
+    {
+      "epoch": 57.38045738045738,
+      "eval_loss": 0.5526473522186279,
+      "eval_runtime": 11.7016,
+      "eval_samples_per_second": 73.152,
+      "eval_steps_per_second": 18.288,
+      "num_input_tokens_seen": 21564560,
+      "step": 27600
+    },
+    {
+      "epoch": 57.39085239085239,
+      "grad_norm": 0.07382059842348099,
+      "learning_rate": 1.0941420371605981e-05,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 21568432,
+      "step": 27605
+    },
+    {
+      "epoch": 57.4012474012474,
+      "grad_norm": 0.2189631313085556,
+      "learning_rate": 1.0933303345276354e-05,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 21572272,
+      "step": 27610
+    },
+    {
+      "epoch": 57.41164241164241,
+      "grad_norm": 2.9779521355521865e-05,
+      "learning_rate": 1.0925188488208112e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21576144,
+      "step": 27615
+    },
+    {
+      "epoch": 57.42203742203742,
+      "grad_norm": 0.19431845843791962,
+      "learning_rate": 1.0917075801652694e-05,
+      "loss": 0.0045,
+      "num_input_tokens_seen": 21580048,
+      "step": 27620
+    },
+    {
+      "epoch": 57.432432432432435,
+      "grad_norm": 6.250629667192698e-05,
+      "learning_rate": 1.0908965286861151e-05,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 21583824,
+      "step": 27625
+    },
+    {
+      "epoch": 57.44282744282744,
+      "grad_norm": 6.803455107728951e-06,
+      "learning_rate": 1.090085694508425e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 21587920,
+      "step": 27630
+    },
+    {
+      "epoch": 57.45322245322245,
+      "grad_norm": 2.0876505004707724e-05,
+      "learning_rate": 1.089275077757238e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21591792,
+      "step": 27635
+    },
+    {
+      "epoch": 57.46361746361746,
+      "grad_norm": 0.00010616792133077979,
+      "learning_rate": 1.0884646785575633e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21595696,
+      "step": 27640
+    },
+    {
+      "epoch": 57.474012474012476,
+      "grad_norm": 6.170541018946096e-05,
+      "learning_rate": 1.0876544970343728e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21599632,
+      "step": 27645
+    },
+    {
+      "epoch": 57.484407484407484,
+      "grad_norm": 9.581928679835983e-06,
+      "learning_rate": 1.0868445333126082e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21603664,
+      "step": 27650
+    },
+    {
+      "epoch": 57.49480249480249,
+      "grad_norm": 3.8254805986071005e-05,
+      "learning_rate": 1.0860347875171745e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21607696,
+      "step": 27655
+    },
+    {
+      "epoch": 57.50519750519751,
+      "grad_norm": 2.9452963644871488e-05,
+      "learning_rate": 1.0852252597729465e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 21611600,
+      "step": 27660
+    },
+    {
+      "epoch": 57.515592515592516,
+      "grad_norm": 7.578309578093467e-06,
+      "learning_rate": 1.0844159502047615e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21615600,
+      "step": 27665
+    },
+    {
+      "epoch": 57.525987525987524,
+      "grad_norm": 1.1475814062578138e-05,
+      "learning_rate": 1.0836068589374265e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21619504,
+      "step": 27670
+    },
+    {
+      "epoch": 57.53638253638254,
+      "grad_norm": 1.392189551552292e-05,
+      "learning_rate": 1.0827979860957144e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21623408,
+      "step": 27675
+    },
+    {
+      "epoch": 57.54677754677755,
+      "grad_norm": 0.13632050156593323,
+      "learning_rate": 1.0819893318043615e-05,
+      "loss": 0.0059,
+      "num_input_tokens_seen": 21627376,
+      "step": 27680
+    },
+    {
+      "epoch": 57.55717255717256,
+      "grad_norm": 7.967545570863876e-06,
+      "learning_rate": 1.0811808961880734e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21631376,
+      "step": 27685
+    },
+    {
+      "epoch": 57.567567567567565,
+      "grad_norm": 0.00016346355550922453,
+      "learning_rate": 1.080372679371522e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21635216,
+      "step": 27690
+    },
+    {
+      "epoch": 57.57796257796258,
+      "grad_norm": 0.00030771212186664343,
+      "learning_rate": 1.0795646814793428e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21638992,
+      "step": 27695
+    },
+    {
+      "epoch": 57.58835758835759,
+      "grad_norm": 5.768333357991651e-05,
+      "learning_rate": 1.078756902636141e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21642832,
+      "step": 27700
+    },
+    {
+      "epoch": 57.5987525987526,
+      "grad_norm": 3.1473435228690505e-05,
+      "learning_rate": 1.077949342966485e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21646832,
+      "step": 27705
+    },
+    {
+      "epoch": 57.60914760914761,
+      "grad_norm": 2.2186754904396366e-06,
+      "learning_rate": 1.0771420025949103e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21650960,
+      "step": 27710
+    },
+    {
+      "epoch": 57.61954261954262,
+      "grad_norm": 1.5487363270949572e-05,
+      "learning_rate": 1.0763348816459204e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21654736,
+      "step": 27715
+    },
+    {
+      "epoch": 57.62993762993763,
+      "grad_norm": 1.179105311166495e-06,
+      "learning_rate": 1.0755279802439816e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21658512,
+      "step": 27720
+    },
+    {
+      "epoch": 57.64033264033264,
+      "grad_norm": 4.5659485294891056e-06,
+      "learning_rate": 1.0747212985135293e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21662320,
+      "step": 27725
+    },
+    {
+      "epoch": 57.65072765072765,
+      "grad_norm": 8.295461157104e-06,
+      "learning_rate": 1.073914836578965e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21666192,
+      "step": 27730
+    },
+    {
+      "epoch": 57.66112266112266,
+      "grad_norm": 0.00013903190847486258,
+      "learning_rate": 1.0731085945646529e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21670192,
+      "step": 27735
+    },
+    {
+      "epoch": 57.67151767151767,
+      "grad_norm": 9.048989340953995e-06,
+      "learning_rate": 1.0723025725949285e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 21674256,
+      "step": 27740
+    },
+    {
+      "epoch": 57.681912681912685,
+      "grad_norm": 1.4548672879755031e-06,
+      "learning_rate": 1.0714967707940875e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21678064,
+      "step": 27745
+    },
+    {
+      "epoch": 57.69230769230769,
+      "grad_norm": 0.00020687883079517633,
+      "learning_rate": 1.0706911892863963e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21681872,
+      "step": 27750
+    },
+    {
+      "epoch": 57.7027027027027,
+      "grad_norm": 1.3252521284812246e-06,
+      "learning_rate": 1.0698858281960866e-05,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 21685872,
+      "step": 27755
+    },
+    {
+      "epoch": 57.71309771309771,
+      "grad_norm": 0.0003490673843771219,
+      "learning_rate": 1.069080687647353e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 21689712,
+      "step": 27760
+    },
+    {
+      "epoch": 57.723492723492726,
+      "grad_norm": 1.685560346231796e-05,
+      "learning_rate": 1.0682757677643596e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21693616,
+      "step": 27765
+    },
+    {
+      "epoch": 57.733887733887734,
+      "grad_norm": 6.6570332819537725e-06,
+      "learning_rate": 1.0674710686712359e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21697488,
+      "step": 27770
+    },
+    {
+      "epoch": 57.74428274428274,
+      "grad_norm": 3.0366509236046113e-06,
+      "learning_rate": 1.0666665904920756e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21701168,
+      "step": 27775
+    },
+    {
+      "epoch": 57.75467775467776,
+      "grad_norm": 0.00021639003534801304,
+      "learning_rate": 1.0658623333509385e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21704976,
+      "step": 27780
+    },
+    {
+      "epoch": 57.765072765072766,
+      "grad_norm": 2.500496520951856e-06,
+      "learning_rate": 1.0650582973718532e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21708880,
+      "step": 27785
+    },
+    {
+      "epoch": 57.775467775467774,
+      "grad_norm": 4.712782538263127e-05,
+      "learning_rate": 1.0642544826788098e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21712784,
+      "step": 27790
+    },
+    {
+      "epoch": 57.78586278586278,
+      "grad_norm": 0.0002645804488565773,
+      "learning_rate": 1.063450889395769e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21716688,
+      "step": 27795
+    },
+    {
+      "epoch": 57.7962577962578,
+      "grad_norm": 5.8120676840189844e-06,
+      "learning_rate": 1.062647517646653e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21720560,
+      "step": 27800
+    },
+    {
+      "epoch": 57.7962577962578,
+      "eval_loss": 0.5543342232704163,
+      "eval_runtime": 11.6931,
+      "eval_samples_per_second": 73.205,
+      "eval_steps_per_second": 18.301,
+      "num_input_tokens_seen": 21720560,
+      "step": 27800
+    },
+    {
+      "epoch": 57.80665280665281,
+      "grad_norm": 1.6471558410557918e-06,
+      "learning_rate": 1.0618443675553527e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21724336,
+      "step": 27805
+    },
+    {
+      "epoch": 57.817047817047815,
+      "grad_norm": 1.5209343473543413e-06,
+      "learning_rate": 1.0610414392457247e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21728240,
+      "step": 27810
+    },
+    {
+      "epoch": 57.82744282744283,
+      "grad_norm": 0.00019291718490421772,
+      "learning_rate": 1.0602387328415888e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21732080,
+      "step": 27815
+    },
+    {
+      "epoch": 57.83783783783784,
+      "grad_norm": 0.10430651903152466,
+      "learning_rate": 1.0594362484667347e-05,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 21736144,
+      "step": 27820
+    },
+    {
+      "epoch": 57.84823284823285,
+      "grad_norm": 0.0004722480080090463,
+      "learning_rate": 1.0586339862449132e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21740272,
+      "step": 27825
+    },
+    {
+      "epoch": 57.858627858627855,
+      "grad_norm": 7.321260454773437e-06,
+      "learning_rate": 1.0578319462998445e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21744144,
+      "step": 27830
+    },
+    {
+      "epoch": 57.86902286902287,
+      "grad_norm": 0.1778549998998642,
+      "learning_rate": 1.057030128755214e-05,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 21748048,
+      "step": 27835
+    },
+    {
+      "epoch": 57.87941787941788,
+      "grad_norm": 0.0022503877989947796,
+      "learning_rate": 1.0562285337346703e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21751984,
+      "step": 27840
+    },
+    {
+      "epoch": 57.88981288981289,
+      "grad_norm": 4.66372694063466e-06,
+      "learning_rate": 1.0554271613618308e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21755920,
+      "step": 27845
+    },
+    {
+      "epoch": 57.9002079002079,
+      "grad_norm": 7.365787314483896e-05,
+      "learning_rate": 1.054626011760276e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21760080,
+      "step": 27850
+    },
+    {
+      "epoch": 57.91060291060291,
+      "grad_norm": 6.412403763533803e-06,
+      "learning_rate": 1.0538250850535549e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21763824,
+      "step": 27855
+    },
+    {
+      "epoch": 57.92099792099792,
+      "grad_norm": 6.971906987018883e-05,
+      "learning_rate": 1.0530243813651794e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21767760,
+      "step": 27860
+    },
+    {
+      "epoch": 57.931392931392935,
+      "grad_norm": 1.2401163985487074e-05,
+      "learning_rate": 1.0522239008186271e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21771664,
+      "step": 27865
+    },
+    {
+      "epoch": 57.94178794178794,
+      "grad_norm": 3.951815870095743e-06,
+      "learning_rate": 1.0514236435373434e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21775440,
+      "step": 27870
+    },
+    {
+      "epoch": 57.95218295218295,
+      "grad_norm": 0.00016997050261124969,
+      "learning_rate": 1.0506236096447386e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21779312,
+      "step": 27875
+    },
+    {
+      "epoch": 57.96257796257796,
+      "grad_norm": 2.145747384929564e-05,
+      "learning_rate": 1.049823799264186e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21783184,
+      "step": 27880
+    },
+    {
+      "epoch": 57.972972972972975,
+      "grad_norm": 6.0313227550068405e-06,
+      "learning_rate": 1.049024212519028e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21787024,
+      "step": 27885
+    },
+    {
+      "epoch": 57.983367983367984,
+      "grad_norm": 0.00010592392936814576,
+      "learning_rate": 1.0482248495325713e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21790928,
+      "step": 27890
+    },
+    {
+      "epoch": 57.99376299376299,
+      "grad_norm": 0.0006148538086563349,
+      "learning_rate": 1.047425710428086e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21794864,
+      "step": 27895
+    },
+    {
+      "epoch": 58.00415800415801,
+      "grad_norm": 1.9865822196152294e-06,
+      "learning_rate": 1.0466267953288114e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21798688,
+      "step": 27900
+    },
+    {
+      "epoch": 58.014553014553016,
+      "grad_norm": 8.156200237863231e-06,
+      "learning_rate": 1.0458281043579482e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21802656,
+      "step": 27905
+    },
+    {
+      "epoch": 58.024948024948024,
+      "grad_norm": 1.8269611246068962e-05,
+      "learning_rate": 1.0450296376386657e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21806592,
+      "step": 27910
+    },
+    {
+      "epoch": 58.03534303534303,
+      "grad_norm": 4.432805781107163e-06,
+      "learning_rate": 1.044231395294098e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21810560,
+      "step": 27915
+    },
+    {
+      "epoch": 58.04573804573805,
+      "grad_norm": 1.1323259059281554e-05,
+      "learning_rate": 1.0434333774473435e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21814400,
+      "step": 27920
+    },
+    {
+      "epoch": 58.056133056133056,
+      "grad_norm": 8.150508801918477e-05,
+      "learning_rate": 1.0426355842214657e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21818464,
+      "step": 27925
+    },
+    {
+      "epoch": 58.066528066528065,
+      "grad_norm": 7.234665827127174e-05,
+      "learning_rate": 1.0418380157394963e-05,
+      "loss": 0.003,
+      "num_input_tokens_seen": 21822304,
+      "step": 27930
+    },
+    {
+      "epoch": 58.07692307692308,
+      "grad_norm": 6.752919034624938e-06,
+      "learning_rate": 1.0410406721244281e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21826240,
+      "step": 27935
+    },
+    {
+      "epoch": 58.08731808731809,
+      "grad_norm": 5.3051062423037365e-05,
+      "learning_rate": 1.0402435534992238e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21830080,
+      "step": 27940
+    },
+    {
+      "epoch": 58.0977130977131,
+      "grad_norm": 6.335513262456516e-06,
+      "learning_rate": 1.0394466599868071e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21834112,
+      "step": 27945
+    },
+    {
+      "epoch": 58.108108108108105,
+      "grad_norm": 1.571741267980542e-05,
+      "learning_rate": 1.0386499917100697e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21837984,
+      "step": 27950
+    },
+    {
+      "epoch": 58.11850311850312,
+      "grad_norm": 3.4302971471333876e-05,
+      "learning_rate": 1.0378535487918692e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 21841888,
+      "step": 27955
+    },
+    {
+      "epoch": 58.12889812889813,
+      "grad_norm": 1.772209543560166e-06,
+      "learning_rate": 1.037057331355025e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21845792,
+      "step": 27960
+    },
+    {
+      "epoch": 58.13929313929314,
+      "grad_norm": 1.2770729881594889e-05,
+      "learning_rate": 1.0362613395223247e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21849600,
+      "step": 27965
+    },
+    {
+      "epoch": 58.14968814968815,
+      "grad_norm": 2.0199604477966204e-06,
+      "learning_rate": 1.0354655734165212e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21853568,
+      "step": 27970
+    },
+    {
+      "epoch": 58.16008316008316,
+      "grad_norm": 6.724601553287357e-05,
+      "learning_rate": 1.03467003316033e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21857472,
+      "step": 27975
+    },
+    {
+      "epoch": 58.17047817047817,
+      "grad_norm": 1.063793547473324e-06,
+      "learning_rate": 1.033874718876435e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21861472,
+      "step": 27980
+    },
+    {
+      "epoch": 58.18087318087318,
+      "grad_norm": 4.666073073167354e-05,
+      "learning_rate": 1.0330796306874818e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21865344,
+      "step": 27985
+    },
+    {
+      "epoch": 58.19126819126819,
+      "grad_norm": 1.004175101115834e-05,
+      "learning_rate": 1.032284768716085e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21869216,
+      "step": 27990
+    },
+    {
+      "epoch": 58.2016632016632,
+      "grad_norm": 6.051576929166913e-05,
+      "learning_rate": 1.0314901330848206e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21873088,
+      "step": 27995
+    },
+    {
+      "epoch": 58.21205821205821,
+      "grad_norm": 0.00010089955321745947,
+      "learning_rate": 1.030695723916233e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21877024,
+      "step": 28000
+    },
+    {
+      "epoch": 58.21205821205821,
+      "eval_loss": 0.5647297501564026,
+      "eval_runtime": 11.6899,
+      "eval_samples_per_second": 73.226,
+      "eval_steps_per_second": 18.306,
+      "num_input_tokens_seen": 21877024,
+      "step": 28000
+    },
+    {
+      "epoch": 58.222453222453225,
+      "grad_norm": 5.481528205564246e-06,
+      "learning_rate": 1.0299015413328289e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21880800,
+      "step": 28005
+    },
+    {
+      "epoch": 58.232848232848234,
+      "grad_norm": 6.983463663345901e-06,
+      "learning_rate": 1.0291075854570809e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21884608,
+      "step": 28010
+    },
+    {
+      "epoch": 58.24324324324324,
+      "grad_norm": 0.14588628709316254,
+      "learning_rate": 1.0283138564114275e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 21888480,
+      "step": 28015
+    },
+    {
+      "epoch": 58.25363825363825,
+      "grad_norm": 3.485222623567097e-05,
+      "learning_rate": 1.027520354318273e-05,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 21892256,
+      "step": 28020
+    },
+    {
+      "epoch": 58.264033264033266,
+      "grad_norm": 0.0005972976214252412,
+      "learning_rate": 1.0267270792999828e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21896288,
+      "step": 28025
+    },
+    {
+      "epoch": 58.274428274428274,
+      "grad_norm": 1.5206101124931592e-05,
+      "learning_rate": 1.0259340314788919e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21900288,
+      "step": 28030
+    },
+    {
+      "epoch": 58.28482328482328,
+      "grad_norm": 1.542847530799918e-05,
+      "learning_rate": 1.0251412109772979e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21904032,
+      "step": 28035
+    },
+    {
+      "epoch": 58.2952182952183,
+      "grad_norm": 1.5923535556794377e-06,
+      "learning_rate": 1.0243486179174627e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21908000,
+      "step": 28040
+    },
+    {
+      "epoch": 58.305613305613306,
+      "grad_norm": 4.1780995161389e-06,
+      "learning_rate": 1.0235562524216158e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21912000,
+      "step": 28045
+    },
+    {
+      "epoch": 58.316008316008315,
+      "grad_norm": 8.585613977629691e-05,
+      "learning_rate": 1.022764114611948e-05,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 21915840,
+      "step": 28050
+    },
+    {
+      "epoch": 58.32640332640332,
+      "grad_norm": 3.288475636509247e-05,
+      "learning_rate": 1.0219722046106178e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21919808,
+      "step": 28055
+    },
+    {
+      "epoch": 58.33679833679834,
+      "grad_norm": 4.2310239223297685e-05,
+      "learning_rate": 1.0211805225397486e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21923712,
+      "step": 28060
+    },
+    {
+      "epoch": 58.34719334719335,
+      "grad_norm": 1.976314933926915e-06,
+      "learning_rate": 1.020389068521426e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21927488,
+      "step": 28065
+    },
+    {
+      "epoch": 58.357588357588355,
+      "grad_norm": 8.859044464770705e-05,
+      "learning_rate": 1.0195978426777039e-05,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 21931360,
+      "step": 28070
+    },
+    {
+      "epoch": 58.36798336798337,
+      "grad_norm": 4.644476211979054e-05,
+      "learning_rate": 1.0188068451305982e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21935296,
+      "step": 28075
+    },
+    {
+      "epoch": 58.37837837837838,
+      "grad_norm": 5.870974564459175e-05,
+      "learning_rate": 1.0180160760020902e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21939328,
+      "step": 28080
+    },
+    {
+      "epoch": 58.38877338877339,
+      "grad_norm": 6.795448825869244e-06,
+      "learning_rate": 1.0172255354141278e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21943200,
+      "step": 28085
+    },
+    {
+      "epoch": 58.3991683991684,
+      "grad_norm": 2.66369097516872e-05,
+      "learning_rate": 1.0164352234886205e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21947072,
+      "step": 28090
+    },
+    {
+      "epoch": 58.40956340956341,
+      "grad_norm": 1.5322822946473025e-05,
+      "learning_rate": 1.0156451403474454e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21950816,
+      "step": 28095
+    },
+    {
+      "epoch": 58.41995841995842,
+      "grad_norm": 5.021153629058972e-05,
+      "learning_rate": 1.0148552861124443e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21954752,
+      "step": 28100
+    },
+    {
+      "epoch": 58.43035343035343,
+      "grad_norm": 1.0402688985777786e-06,
+      "learning_rate": 1.0140656609054205e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21958784,
+      "step": 28105
+    },
+    {
+      "epoch": 58.44074844074844,
+      "grad_norm": 3.2083138648886234e-05,
+      "learning_rate": 1.0132762648481455e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21962720,
+      "step": 28110
+    },
+    {
+      "epoch": 58.45114345114345,
+      "grad_norm": 0.00012195880844956264,
+      "learning_rate": 1.0124870980623543e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21966560,
+      "step": 28115
+    },
+    {
+      "epoch": 58.46153846153846,
+      "grad_norm": 0.1064908504486084,
+      "learning_rate": 1.0116981606697453e-05,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 21970528,
+      "step": 28120
+    },
+    {
+      "epoch": 58.471933471933475,
+      "grad_norm": 5.387404598877765e-05,
+      "learning_rate": 1.0109094527919838e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21974464,
+      "step": 28125
+    },
+    {
+      "epoch": 58.482328482328484,
+      "grad_norm": 3.899784496752545e-05,
+      "learning_rate": 1.010120974550697e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21978496,
+      "step": 28130
+    },
+    {
+      "epoch": 58.49272349272349,
+      "grad_norm": 0.15800607204437256,
+      "learning_rate": 1.0093327260674795e-05,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 21982432,
+      "step": 28135
+    },
+    {
+      "epoch": 58.5031185031185,
+      "grad_norm": 1.6617077562841587e-05,
+      "learning_rate": 1.0085447074638878e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21986368,
+      "step": 28140
+    },
+    {
+      "epoch": 58.513513513513516,
+      "grad_norm": 1.9105566025245935e-05,
+      "learning_rate": 1.0077569188614461e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21990400,
+      "step": 28145
+    },
+    {
+      "epoch": 58.523908523908524,
+      "grad_norm": 0.0002592729579191655,
+      "learning_rate": 1.0069693603816393e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 21994208,
+      "step": 28150
+    },
+    {
+      "epoch": 58.53430353430353,
+      "grad_norm": 1.0432087037770543e-05,
+      "learning_rate": 1.0061820321459204e-05,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 21998048,
+      "step": 28155
+    },
+    {
+      "epoch": 58.54469854469855,
+      "grad_norm": 1.9151195374433883e-05,
+      "learning_rate": 1.0053949342757038e-05,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 22001888,
+      "step": 28160
+    },
+    {
+      "epoch": 58.555093555093556,
+      "grad_norm": 1.1517206530697877e-06,
+      "learning_rate": 1.0046080668923717e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22005824,
+      "step": 28165
+    },
+    {
+      "epoch": 58.565488565488565,
+      "grad_norm": 2.46012382376648e-06,
+      "learning_rate": 1.003821430117267e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22009856,
+      "step": 28170
+    },
+    {
+      "epoch": 58.57588357588357,
+      "grad_norm": 0.0004056073958054185,
+      "learning_rate": 1.0030350240716999e-05,
+      "loss": 0.004,
+      "num_input_tokens_seen": 22013920,
+      "step": 28175
+    },
+    {
+      "epoch": 58.58627858627859,
+      "grad_norm": 0.00010427665984025225,
+      "learning_rate": 1.0022488488769449e-05,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22017696,
+      "step": 28180
+    },
+    {
+      "epoch": 58.5966735966736,
+      "grad_norm": 1.451784646633314e-05,
+      "learning_rate": 1.0014629046542387e-05,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 22021760,
+      "step": 28185
+    },
+    {
+      "epoch": 58.607068607068605,
+      "grad_norm": 2.356100048928056e-05,
+      "learning_rate": 1.0006771915247842e-05,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 22025664,
+      "step": 28190
+    },
+    {
+      "epoch": 58.61746361746362,
+      "grad_norm": 9.7997535704053e-06,
+      "learning_rate": 9.998917096097495e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22029568,
+      "step": 28195
+    },
+    {
+      "epoch": 58.62785862785863,
+      "grad_norm": 1.732523855935142e-06,
+      "learning_rate": 9.991064590302638e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22033344,
+      "step": 28200
+    },
+    {
+      "epoch": 58.62785862785863,
+      "eval_loss": 0.563560426235199,
+      "eval_runtime": 11.7015,
+      "eval_samples_per_second": 73.153,
+      "eval_steps_per_second": 18.288,
+      "num_input_tokens_seen": 22033344,
+      "step": 28200
+    },
+    {
+      "epoch": 58.63825363825364,
+      "grad_norm": 8.033763151615858e-05,
+      "learning_rate": 9.983214399074241e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22037344,
+      "step": 28205
+    },
+    {
+      "epoch": 58.648648648648646,
+      "grad_norm": 7.975526386871934e-06,
+      "learning_rate": 9.975366523622893e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22041312,
+      "step": 28210
+    },
+    {
+      "epoch": 58.65904365904366,
+      "grad_norm": 1.8112798443326028e-06,
+      "learning_rate": 9.967520965158841e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22045120,
+      "step": 28215
+    },
+    {
+      "epoch": 58.66943866943867,
+      "grad_norm": 2.2712973077432252e-05,
+      "learning_rate": 9.95967772489197e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 22048992,
+      "step": 28220
+    },
+    {
+      "epoch": 58.67983367983368,
+      "grad_norm": 2.0859812138951384e-05,
+      "learning_rate": 9.951836804031794e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22052800,
+      "step": 28225
+    },
+    {
+      "epoch": 58.69022869022869,
+      "grad_norm": 0.00040200664079748094,
+      "learning_rate": 9.943998203787489e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22056768,
+      "step": 28230
+    },
+    {
+      "epoch": 58.7006237006237,
+      "grad_norm": 0.0028526331298053265,
+      "learning_rate": 9.936161925367874e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22060640,
+      "step": 28235
+    },
+    {
+      "epoch": 58.71101871101871,
+      "grad_norm": 1.23496729429462e-05,
+      "learning_rate": 9.928327969981386e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22064544,
+      "step": 28240
+    },
+    {
+      "epoch": 58.72141372141372,
+      "grad_norm": 3.000692686327966e-06,
+      "learning_rate": 9.920496338836135e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22068320,
+      "step": 28245
+    },
+    {
+      "epoch": 58.731808731808734,
+      "grad_norm": 3.574452057364397e-05,
+      "learning_rate": 9.912667033139844e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22072192,
+      "step": 28250
+    },
+    {
+      "epoch": 58.74220374220374,
+      "grad_norm": 1.7827729607233778e-05,
+      "learning_rate": 9.904840054099893e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22076032,
+      "step": 28255
+    },
+    {
+      "epoch": 58.75259875259875,
+      "grad_norm": 0.00012190599227324128,
+      "learning_rate": 9.897015402923312e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 22079936,
+      "step": 28260
+    },
+    {
+      "epoch": 58.762993762993766,
+      "grad_norm": 1.3541532098315656e-05,
+      "learning_rate": 9.889193080816744e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22083904,
+      "step": 28265
+    },
+    {
+      "epoch": 58.773388773388774,
+      "grad_norm": 0.00012032757513225079,
+      "learning_rate": 9.881373088986498e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22087776,
+      "step": 28270
+    },
+    {
+      "epoch": 58.78378378378378,
+      "grad_norm": 3.512956027407199e-05,
+      "learning_rate": 9.873555428638523e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22091648,
+      "step": 28275
+    },
+    {
+      "epoch": 58.79417879417879,
+      "grad_norm": 4.435895334609086e-06,
+      "learning_rate": 9.865740100978383e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22095776,
+      "step": 28280
+    },
+    {
+      "epoch": 58.804573804573806,
+      "grad_norm": 8.031865377233771e-07,
+      "learning_rate": 9.857927107211315e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22099776,
+      "step": 28285
+    },
+    {
+      "epoch": 58.814968814968815,
+      "grad_norm": 6.62800157442689e-05,
+      "learning_rate": 9.850116448542177e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22103680,
+      "step": 28290
+    },
+    {
+      "epoch": 58.82536382536382,
+      "grad_norm": 4.300555883673951e-05,
+      "learning_rate": 9.842308126175457e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22107584,
+      "step": 28295
+    },
+    {
+      "epoch": 58.83575883575884,
+      "grad_norm": 2.147639315808192e-05,
+      "learning_rate": 9.834502141315315e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22111488,
+      "step": 28300
+    },
+    {
+      "epoch": 58.84615384615385,
+      "grad_norm": 4.601170076057315e-05,
+      "learning_rate": 9.82669849516552e-06,
+      "loss": 0.0042,
+      "num_input_tokens_seen": 22115392,
+      "step": 28305
+    },
+    {
+      "epoch": 58.856548856548855,
+      "grad_norm": 2.416685310890898e-05,
+      "learning_rate": 9.818897188929493e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22119424,
+      "step": 28310
+    },
+    {
+      "epoch": 58.86694386694387,
+      "grad_norm": 5.749626143369824e-05,
+      "learning_rate": 9.811098223810309e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22123200,
+      "step": 28315
+    },
+    {
+      "epoch": 58.87733887733888,
+      "grad_norm": 2.3337519451160915e-05,
+      "learning_rate": 9.803301601010641e-06,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 22127072,
+      "step": 28320
+    },
+    {
+      "epoch": 58.88773388773389,
+      "grad_norm": 6.652654974459438e-06,
+      "learning_rate": 9.795507321732853e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22131040,
+      "step": 28325
+    },
+    {
+      "epoch": 58.898128898128896,
+      "grad_norm": 7.813757292751689e-06,
+      "learning_rate": 9.787715387178898e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22134976,
+      "step": 28330
+    },
+    {
+      "epoch": 58.90852390852391,
+      "grad_norm": 0.0001531464367872104,
+      "learning_rate": 9.779925798550399e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22138816,
+      "step": 28335
+    },
+    {
+      "epoch": 58.91891891891892,
+      "grad_norm": 0.10585395246744156,
+      "learning_rate": 9.772138557048619e-06,
+      "loss": 0.002,
+      "num_input_tokens_seen": 22142720,
+      "step": 28340
+    },
+    {
+      "epoch": 58.92931392931393,
+      "grad_norm": 0.00024069545906968415,
+      "learning_rate": 9.764353663874426e-06,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 22146656,
+      "step": 28345
+    },
+    {
+      "epoch": 58.93970893970894,
+      "grad_norm": 3.4724705528788036e-06,
+      "learning_rate": 9.756571120228375e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22150464,
+      "step": 28350
+    },
+    {
+      "epoch": 58.95010395010395,
+      "grad_norm": 0.0003206194960512221,
+      "learning_rate": 9.748790927310605e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22154368,
+      "step": 28355
+    },
+    {
+      "epoch": 58.96049896049896,
+      "grad_norm": 1.3725774579143035e-06,
+      "learning_rate": 9.741013086320946e-06,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 22158272,
+      "step": 28360
+    },
+    {
+      "epoch": 58.97089397089397,
+      "grad_norm": 8.592268386564683e-06,
+      "learning_rate": 9.733237598458821e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22162208,
+      "step": 28365
+    },
+    {
+      "epoch": 58.981288981288984,
+      "grad_norm": 0.00012612840509973466,
+      "learning_rate": 9.725464464923308e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22166240,
+      "step": 28370
+    },
+    {
+      "epoch": 58.99168399168399,
+      "grad_norm": 7.843011644581566e-07,
+      "learning_rate": 9.717693686913123e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22170016,
+      "step": 28375
+    },
+    {
+      "epoch": 59.002079002079,
+      "grad_norm": 2.2077414541854523e-05,
+      "learning_rate": 9.709925265626632e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22174064,
+      "step": 28380
+    },
+    {
+      "epoch": 59.012474012474016,
+      "grad_norm": 1.302307850892248e-06,
+      "learning_rate": 9.702159202261801e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22178096,
+      "step": 28385
+    },
+    {
+      "epoch": 59.022869022869024,
+      "grad_norm": 4.0739894757280126e-05,
+      "learning_rate": 9.694395498016268e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22182032,
+      "step": 28390
+    },
+    {
+      "epoch": 59.03326403326403,
+      "grad_norm": 4.124216047785012e-06,
+      "learning_rate": 9.686634154087298e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 22185936,
+      "step": 28395
+    },
+    {
+      "epoch": 59.04365904365904,
+      "grad_norm": 2.3743664314679336e-06,
+      "learning_rate": 9.678875171671776e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22189872,
+      "step": 28400
+    },
+    {
+      "epoch": 59.04365904365904,
+      "eval_loss": 0.5683529376983643,
+      "eval_runtime": 11.6901,
+      "eval_samples_per_second": 73.224,
+      "eval_steps_per_second": 18.306,
+      "num_input_tokens_seen": 22189872,
+      "step": 28400
+    },
+    {
+      "epoch": 59.054054054054056,
+      "grad_norm": 0.13799622654914856,
+      "learning_rate": 9.671118551966246e-06,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 22193936,
+      "step": 28405
+    },
+    {
+      "epoch": 59.064449064449065,
+      "grad_norm": 4.688013359555043e-05,
+      "learning_rate": 9.66336429616686e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22197808,
+      "step": 28410
+    },
+    {
+      "epoch": 59.07484407484407,
+      "grad_norm": 0.0012707271380349994,
+      "learning_rate": 9.655612405469436e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 22201872,
+      "step": 28415
+    },
+    {
+      "epoch": 59.08523908523909,
+      "grad_norm": 6.308913725661114e-06,
+      "learning_rate": 9.647862881069413e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22205776,
+      "step": 28420
+    },
+    {
+      "epoch": 59.0956340956341,
+      "grad_norm": 4.514958618528908e-06,
+      "learning_rate": 9.640115724161855e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22209744,
+      "step": 28425
+    },
+    {
+      "epoch": 59.106029106029105,
+      "grad_norm": 7.820400242053438e-06,
+      "learning_rate": 9.632370935941483e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22213744,
+      "step": 28430
+    },
+    {
+      "epoch": 59.11642411642411,
+      "grad_norm": 8.653808436065447e-06,
+      "learning_rate": 9.624628517602634e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22217584,
+      "step": 28435
+    },
+    {
+      "epoch": 59.12681912681913,
+      "grad_norm": 1.2371361663099378e-05,
+      "learning_rate": 9.61688847033928e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22221424,
+      "step": 28440
+    },
+    {
+      "epoch": 59.13721413721414,
+      "grad_norm": 3.184111847076565e-05,
+      "learning_rate": 9.609150795345051e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22225168,
+      "step": 28445
+    },
+    {
+      "epoch": 59.147609147609145,
+      "grad_norm": 1.578452247485984e-05,
+      "learning_rate": 9.601415493813171e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22228976,
+      "step": 28450
+    },
+    {
+      "epoch": 59.15800415800416,
+      "grad_norm": 6.041274900780991e-06,
+      "learning_rate": 9.593682566936533e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 22232848,
+      "step": 28455
+    },
+    {
+      "epoch": 59.16839916839917,
+      "grad_norm": 1.2390652273097658e-06,
+      "learning_rate": 9.58595201590766e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22236720,
+      "step": 28460
+    },
+    {
+      "epoch": 59.17879417879418,
+      "grad_norm": 1.2423653970472515e-05,
+      "learning_rate": 9.578223841918681e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22240464,
+      "step": 28465
+    },
+    {
+      "epoch": 59.189189189189186,
+      "grad_norm": 4.396133590489626e-06,
+      "learning_rate": 9.570498046161389e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 22244400,
+      "step": 28470
+    },
+    {
+      "epoch": 59.1995841995842,
+      "grad_norm": 9.910919516187278e-07,
+      "learning_rate": 9.562774629827206e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22248208,
+      "step": 28475
+    },
+    {
+      "epoch": 59.20997920997921,
+      "grad_norm": 2.568257968960097e-06,
+      "learning_rate": 9.555053594107163e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22252048,
+      "step": 28480
+    },
+    {
+      "epoch": 59.22037422037422,
+      "grad_norm": 8.657592843519524e-06,
+      "learning_rate": 9.547334940191957e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22256048,
+      "step": 28485
+    },
+    {
+      "epoch": 59.23076923076923,
+      "grad_norm": 3.065765440624091e-06,
+      "learning_rate": 9.539618669271886e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22260080,
+      "step": 28490
+    },
+    {
+      "epoch": 59.24116424116424,
+      "grad_norm": 0.00021692457085009664,
+      "learning_rate": 9.531904782536904e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22263888,
+      "step": 28495
+    },
+    {
+      "epoch": 59.25155925155925,
+      "grad_norm": 7.308773638214916e-05,
+      "learning_rate": 9.524193281176597e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22267856,
+      "step": 28500
+    },
+    {
+      "epoch": 59.26195426195426,
+      "grad_norm": 2.203772737630061e-06,
+      "learning_rate": 9.516484166380165e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22271728,
+      "step": 28505
+    },
+    {
+      "epoch": 59.272349272349274,
+      "grad_norm": 3.118450695183128e-05,
+      "learning_rate": 9.508777439336447e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22275504,
+      "step": 28510
+    },
+    {
+      "epoch": 59.28274428274428,
+      "grad_norm": 6.716713687637821e-05,
+      "learning_rate": 9.50107310123393e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22279344,
+      "step": 28515
+    },
+    {
+      "epoch": 59.29313929313929,
+      "grad_norm": 0.0003934826818294823,
+      "learning_rate": 9.493371153260702e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22283184,
+      "step": 28520
+    },
+    {
+      "epoch": 59.303534303534306,
+      "grad_norm": 1.2935673112224322e-05,
+      "learning_rate": 9.485671596604523e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22287120,
+      "step": 28525
+    },
+    {
+      "epoch": 59.313929313929314,
+      "grad_norm": 2.0830992070841603e-05,
+      "learning_rate": 9.477974432452738e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22290992,
+      "step": 28530
+    },
+    {
+      "epoch": 59.32432432432432,
+      "grad_norm": 7.3987584983115084e-06,
+      "learning_rate": 9.470279661992356e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22294928,
+      "step": 28535
+    },
+    {
+      "epoch": 59.33471933471934,
+      "grad_norm": 3.579013718990609e-05,
+      "learning_rate": 9.462587286410021e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22298864,
+      "step": 28540
+    },
+    {
+      "epoch": 59.34511434511435,
+      "grad_norm": 7.542656385339797e-05,
+      "learning_rate": 9.454897306891972e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22302864,
+      "step": 28545
+    },
+    {
+      "epoch": 59.355509355509355,
+      "grad_norm": 3.669860234367661e-05,
+      "learning_rate": 9.44720972462411e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22306832,
+      "step": 28550
+    },
+    {
+      "epoch": 59.36590436590436,
+      "grad_norm": 1.2996618352190126e-05,
+      "learning_rate": 9.439524540791964e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22310800,
+      "step": 28555
+    },
+    {
+      "epoch": 59.37629937629938,
+      "grad_norm": 0.000276874634437263,
+      "learning_rate": 9.431841756580673e-06,
+      "loss": 0.004,
+      "num_input_tokens_seen": 22314672,
+      "step": 28560
+    },
+    {
+      "epoch": 59.38669438669439,
+      "grad_norm": 2.0350029444671236e-05,
+      "learning_rate": 9.42416137317503e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22318544,
+      "step": 28565
+    },
+    {
+      "epoch": 59.397089397089395,
+      "grad_norm": 3.935947916033911e-06,
+      "learning_rate": 9.416483391759437e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22322352,
+      "step": 28570
+    },
+    {
+      "epoch": 59.40748440748441,
+      "grad_norm": 1.0592590342639596e-06,
+      "learning_rate": 9.408807813517945e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 22326224,
+      "step": 28575
+    },
+    {
+      "epoch": 59.41787941787942,
+      "grad_norm": 0.15057963132858276,
+      "learning_rate": 9.401134639634221e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 22330128,
+      "step": 28580
+    },
+    {
+      "epoch": 59.42827442827443,
+      "grad_norm": 9.104132914217189e-06,
+      "learning_rate": 9.393463871291555e-06,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 22333936,
+      "step": 28585
+    },
+    {
+      "epoch": 59.438669438669436,
+      "grad_norm": 9.393044820171781e-06,
+      "learning_rate": 9.385795509672881e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22337776,
+      "step": 28590
+    },
+    {
+      "epoch": 59.44906444906445,
+      "grad_norm": 8.415070624323562e-05,
+      "learning_rate": 9.378129555960771e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22341840,
+      "step": 28595
+    },
+    {
+      "epoch": 59.45945945945946,
+      "grad_norm": 6.484659388661385e-05,
+      "learning_rate": 9.370466011337392e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22345712,
+      "step": 28600
+    },
+    {
+      "epoch": 59.45945945945946,
+      "eval_loss": 0.5754206776618958,
+      "eval_runtime": 11.6633,
+      "eval_samples_per_second": 73.392,
+      "eval_steps_per_second": 18.348,
+      "num_input_tokens_seen": 22345712,
+      "step": 28600
+    },
+    {
+      "epoch": 59.46985446985447,
+      "grad_norm": 8.485558282700367e-06,
+      "learning_rate": 9.362804876984573e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22349648,
+      "step": 28605
+    },
+    {
+      "epoch": 59.48024948024948,
+      "grad_norm": 3.1231761568051297e-06,
+      "learning_rate": 9.355146154083747e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22353648,
+      "step": 28610
+    },
+    {
+      "epoch": 59.49064449064449,
+      "grad_norm": 2.7355079510016367e-05,
+      "learning_rate": 9.347489843815987e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22357616,
+      "step": 28615
+    },
+    {
+      "epoch": 59.5010395010395,
+      "grad_norm": 7.471584831364453e-05,
+      "learning_rate": 9.339835947362002e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22361680,
+      "step": 28620
+    },
+    {
+      "epoch": 59.51143451143451,
+      "grad_norm": 9.786127520783339e-06,
+      "learning_rate": 9.332184465902105e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22365808,
+      "step": 28625
+    },
+    {
+      "epoch": 59.521829521829524,
+      "grad_norm": 4.8536672693444416e-05,
+      "learning_rate": 9.324535400616266e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22369744,
+      "step": 28630
+    },
+    {
+      "epoch": 59.53222453222453,
+      "grad_norm": 0.00037325642188079655,
+      "learning_rate": 9.31688875268405e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22373680,
+      "step": 28635
+    },
+    {
+      "epoch": 59.54261954261954,
+      "grad_norm": 1.627365577405726e-06,
+      "learning_rate": 9.309244523284674e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22377808,
+      "step": 28640
+    },
+    {
+      "epoch": 59.553014553014556,
+      "grad_norm": 6.253489118535072e-05,
+      "learning_rate": 9.301602713596982e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22381712,
+      "step": 28645
+    },
+    {
+      "epoch": 59.563409563409564,
+      "grad_norm": 1.7965810911846347e-05,
+      "learning_rate": 9.293963324799432e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22385712,
+      "step": 28650
+    },
+    {
+      "epoch": 59.57380457380457,
+      "grad_norm": 0.00011114088556496426,
+      "learning_rate": 9.286326358070104e-06,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 22389680,
+      "step": 28655
+    },
+    {
+      "epoch": 59.58419958419958,
+      "grad_norm": 0.12446413189172745,
+      "learning_rate": 9.278691814586729e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 22393552,
+      "step": 28660
+    },
+    {
+      "epoch": 59.5945945945946,
+      "grad_norm": 2.311266325705219e-05,
+      "learning_rate": 9.271059695526635e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22397424,
+      "step": 28665
+    },
+    {
+      "epoch": 59.604989604989605,
+      "grad_norm": 1.3021637641941197e-05,
+      "learning_rate": 9.263430002066805e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22401360,
+      "step": 28670
+    },
+    {
+      "epoch": 59.61538461538461,
+      "grad_norm": 7.712403021287173e-05,
+      "learning_rate": 9.25580273538382e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22405104,
+      "step": 28675
+    },
+    {
+      "epoch": 59.62577962577963,
+      "grad_norm": 2.523401599319186e-05,
+      "learning_rate": 9.248177896653907e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22408784,
+      "step": 28680
+    },
+    {
+      "epoch": 59.63617463617464,
+      "grad_norm": 0.00025221193209290504,
+      "learning_rate": 9.240555487052918e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22412752,
+      "step": 28685
+    },
+    {
+      "epoch": 59.646569646569645,
+      "grad_norm": 6.048175237083342e-06,
+      "learning_rate": 9.232935507756313e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22416592,
+      "step": 28690
+    },
+    {
+      "epoch": 59.656964656964654,
+      "grad_norm": 1.5235931414281367e-06,
+      "learning_rate": 9.225317959939193e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22420496,
+      "step": 28695
+    },
+    {
+      "epoch": 59.66735966735967,
+      "grad_norm": 6.883007699798327e-06,
+      "learning_rate": 9.217702844776287e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 22424496,
+      "step": 28700
+    },
+    {
+      "epoch": 59.67775467775468,
+      "grad_norm": 1.4697833648824599e-05,
+      "learning_rate": 9.210090163441929e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22428560,
+      "step": 28705
+    },
+    {
+      "epoch": 59.688149688149686,
+      "grad_norm": 0.003943577408790588,
+      "learning_rate": 9.202479917110105e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22432496,
+      "step": 28710
+    },
+    {
+      "epoch": 59.6985446985447,
+      "grad_norm": 0.00010452951391926035,
+      "learning_rate": 9.194872106954392e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22436432,
+      "step": 28715
+    },
+    {
+      "epoch": 59.70893970893971,
+      "grad_norm": 1.0995316870321403e-06,
+      "learning_rate": 9.187266734148029e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22440432,
+      "step": 28720
+    },
+    {
+      "epoch": 59.71933471933472,
+      "grad_norm": 9.189800493913935e-07,
+      "learning_rate": 9.179663799863849e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22444400,
+      "step": 28725
+    },
+    {
+      "epoch": 59.729729729729726,
+      "grad_norm": 2.85159421764547e-06,
+      "learning_rate": 9.172063305274317e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22448304,
+      "step": 28730
+    },
+    {
+      "epoch": 59.74012474012474,
+      "grad_norm": 2.341600520594511e-06,
+      "learning_rate": 9.164465251551527e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22452048,
+      "step": 28735
+    },
+    {
+      "epoch": 59.75051975051975,
+      "grad_norm": 0.00037331721978262067,
+      "learning_rate": 9.156869639867205e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 22455920,
+      "step": 28740
+    },
+    {
+      "epoch": 59.76091476091476,
+      "grad_norm": 6.840093647042522e-06,
+      "learning_rate": 9.149276471392677e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 22459792,
+      "step": 28745
+    },
+    {
+      "epoch": 59.771309771309774,
+      "grad_norm": 1.3881237919122213e-06,
+      "learning_rate": 9.141685747298914e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22463760,
+      "step": 28750
+    },
+    {
+      "epoch": 59.78170478170478,
+      "grad_norm": 1.4377754951055977e-06,
+      "learning_rate": 9.13409746875649e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22467632,
+      "step": 28755
+    },
+    {
+      "epoch": 59.79209979209979,
+      "grad_norm": 1.9834110389638226e-06,
+      "learning_rate": 9.12651163693562e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22471696,
+      "step": 28760
+    },
+    {
+      "epoch": 59.802494802494806,
+      "grad_norm": 0.00016865342331584543,
+      "learning_rate": 9.11892825300614e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22475504,
+      "step": 28765
+    },
+    {
+      "epoch": 59.812889812889814,
+      "grad_norm": 1.3389430932875257e-05,
+      "learning_rate": 9.111347318137491e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22479504,
+      "step": 28770
+    },
+    {
+      "epoch": 59.82328482328482,
+      "grad_norm": 6.686840060865507e-05,
+      "learning_rate": 9.103768833498755e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22483312,
+      "step": 28775
+    },
+    {
+      "epoch": 59.83367983367983,
+      "grad_norm": 5.251452876109397e-06,
+      "learning_rate": 9.096192800258639e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22487184,
+      "step": 28780
+    },
+    {
+      "epoch": 59.84407484407485,
+      "grad_norm": 2.1929939975962043e-05,
+      "learning_rate": 9.088619219585443e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22491120,
+      "step": 28785
+    },
+    {
+      "epoch": 59.854469854469855,
+      "grad_norm": 0.002437376417219639,
+      "learning_rate": 9.081048092647127e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22494832,
+      "step": 28790
+    },
+    {
+      "epoch": 59.86486486486486,
+      "grad_norm": 1.2085027265129611e-05,
+      "learning_rate": 9.073479420611245e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22498640,
+      "step": 28795
+    },
+    {
+      "epoch": 59.87525987525988,
+      "grad_norm": 3.4936031170218484e-06,
+      "learning_rate": 9.065913204644974e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22502352,
+      "step": 28800
+    },
+    {
+      "epoch": 59.87525987525988,
+      "eval_loss": 0.5927456617355347,
+      "eval_runtime": 11.683,
+      "eval_samples_per_second": 73.269,
+      "eval_steps_per_second": 18.317,
+      "num_input_tokens_seen": 22502352,
+      "step": 28800
+    },
+    {
+      "epoch": 59.88565488565489,
+      "grad_norm": 1.0444960025779437e-05,
+      "learning_rate": 9.058349445915135e-06,
+      "loss": 0.0071,
+      "num_input_tokens_seen": 22506192,
+      "step": 28805
+    },
+    {
+      "epoch": 59.896049896049895,
+      "grad_norm": 4.188681850791909e-05,
+      "learning_rate": 9.050788145588138e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22510064,
+      "step": 28810
+    },
+    {
+      "epoch": 59.906444906444904,
+      "grad_norm": 3.01641534861119e-06,
+      "learning_rate": 9.043229304830039e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22513904,
+      "step": 28815
+    },
+    {
+      "epoch": 59.91683991683992,
+      "grad_norm": 9.607417268853169e-06,
+      "learning_rate": 9.035672924806515e-06,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 22517872,
+      "step": 28820
+    },
+    {
+      "epoch": 59.92723492723493,
+      "grad_norm": 1.3317501725396141e-05,
+      "learning_rate": 9.028119006682839e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 22521808,
+      "step": 28825
+    },
+    {
+      "epoch": 59.937629937629936,
+      "grad_norm": 3.9115957406465895e-06,
+      "learning_rate": 9.020567551623935e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22525712,
+      "step": 28830
+    },
+    {
+      "epoch": 59.94802494802495,
+      "grad_norm": 0.0006222022930160165,
+      "learning_rate": 9.013018560794318e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22529680,
+      "step": 28835
+    },
+    {
+      "epoch": 59.95841995841996,
+      "grad_norm": 5.308140316628851e-05,
+      "learning_rate": 9.005472035358139e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22533584,
+      "step": 28840
+    },
+    {
+      "epoch": 59.96881496881497,
+      "grad_norm": 0.0005855459021404386,
+      "learning_rate": 8.997927976479185e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22537424,
+      "step": 28845
+    },
+    {
+      "epoch": 59.979209979209976,
+      "grad_norm": 4.26090628025122e-05,
+      "learning_rate": 8.99038638532082e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22541392,
+      "step": 28850
+    },
+    {
+      "epoch": 59.98960498960499,
+      "grad_norm": 0.000208350844332017,
+      "learning_rate": 8.982847263046065e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22545136,
+      "step": 28855
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 4.007488678325899e-05,
+      "learning_rate": 8.975310610817555e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22548968,
+      "step": 28860
+    },
+    {
+      "epoch": 60.01039501039501,
+      "grad_norm": 1.7788582908906392e-06,
+      "learning_rate": 8.967776429797528e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22552872,
+      "step": 28865
+    },
+    {
+      "epoch": 60.020790020790024,
+      "grad_norm": 3.781547320613754e-06,
+      "learning_rate": 8.960244721147842e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22556552,
+      "step": 28870
+    },
+    {
+      "epoch": 60.03118503118503,
+      "grad_norm": 2.923226566053927e-05,
+      "learning_rate": 8.952715486029995e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22560360,
+      "step": 28875
+    },
+    {
+      "epoch": 60.04158004158004,
+      "grad_norm": 0.10662927478551865,
+      "learning_rate": 8.945188725605075e-06,
+      "loss": 0.002,
+      "num_input_tokens_seen": 22564456,
+      "step": 28880
+    },
+    {
+      "epoch": 60.05197505197505,
+      "grad_norm": 2.3219442937261192e-06,
+      "learning_rate": 8.937664441033817e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22568488,
+      "step": 28885
+    },
+    {
+      "epoch": 60.062370062370064,
+      "grad_norm": 2.989404947584262e-06,
+      "learning_rate": 8.930142633476549e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22572488,
+      "step": 28890
+    },
+    {
+      "epoch": 60.07276507276507,
+      "grad_norm": 5.008429980080109e-06,
+      "learning_rate": 8.92262330409323e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22576360,
+      "step": 28895
+    },
+    {
+      "epoch": 60.08316008316008,
+      "grad_norm": 3.279645397924469e-06,
+      "learning_rate": 8.915106454043448e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22580200,
+      "step": 28900
+    },
+    {
+      "epoch": 60.093555093555096,
+      "grad_norm": 6.150755325506907e-06,
+      "learning_rate": 8.90759208448638e-06,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 22584072,
+      "step": 28905
+    },
+    {
+      "epoch": 60.103950103950105,
+      "grad_norm": 5.919992872804869e-06,
+      "learning_rate": 8.900080196580848e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22588168,
+      "step": 28910
+    },
+    {
+      "epoch": 60.11434511434511,
+      "grad_norm": 8.322930807480589e-05,
+      "learning_rate": 8.892570791485267e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22592040,
+      "step": 28915
+    },
+    {
+      "epoch": 60.12474012474012,
+      "grad_norm": 1.1008835826942232e-05,
+      "learning_rate": 8.885063870357688e-06,
+      "loss": 0.0017,
+      "num_input_tokens_seen": 22595944,
+      "step": 28920
+    },
+    {
+      "epoch": 60.13513513513514,
+      "grad_norm": 0.0007061601500026882,
+      "learning_rate": 8.87755943435578e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22599912,
+      "step": 28925
+    },
+    {
+      "epoch": 60.145530145530145,
+      "grad_norm": 2.040127947111614e-05,
+      "learning_rate": 8.87005748463681e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22603816,
+      "step": 28930
+    },
+    {
+      "epoch": 60.15592515592515,
+      "grad_norm": 4.14623764299904e-06,
+      "learning_rate": 8.862558022357681e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22607496,
+      "step": 28935
+    },
+    {
+      "epoch": 60.16632016632017,
+      "grad_norm": 9.17216584639391e-06,
+      "learning_rate": 8.855061048674903e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22611464,
+      "step": 28940
+    },
+    {
+      "epoch": 60.17671517671518,
+      "grad_norm": 0.00033371636527590454,
+      "learning_rate": 8.847566564744595e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22615368,
+      "step": 28945
+    },
+    {
+      "epoch": 60.187110187110186,
+      "grad_norm": 5.549041452468373e-05,
+      "learning_rate": 8.840074571722512e-06,
+      "loss": 0.0012,
+      "num_input_tokens_seen": 22619240,
+      "step": 28950
+    },
+    {
+      "epoch": 60.197505197505194,
+      "grad_norm": 2.968296394101344e-05,
+      "learning_rate": 8.832585070764002e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22623144,
+      "step": 28955
+    },
+    {
+      "epoch": 60.20790020790021,
+      "grad_norm": 0.001809881068766117,
+      "learning_rate": 8.825098063024045e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22627112,
+      "step": 28960
+    },
+    {
+      "epoch": 60.21829521829522,
+      "grad_norm": 9.24247651710175e-05,
+      "learning_rate": 8.817613549657244e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22631144,
+      "step": 28965
+    },
+    {
+      "epoch": 60.228690228690226,
+      "grad_norm": 4.514949978329241e-06,
+      "learning_rate": 8.810131531817783e-06,
+      "loss": 0.0009,
+      "num_input_tokens_seen": 22634984,
+      "step": 28970
+    },
+    {
+      "epoch": 60.23908523908524,
+      "grad_norm": 8.171280205715448e-05,
+      "learning_rate": 8.802652010659496e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22638952,
+      "step": 28975
+    },
+    {
+      "epoch": 60.24948024948025,
+      "grad_norm": 1.0778341675177217e-05,
+      "learning_rate": 8.795174987335827e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22642856,
+      "step": 28980
+    },
+    {
+      "epoch": 60.25987525987526,
+      "grad_norm": 9.830393537413329e-06,
+      "learning_rate": 8.787700462999807e-06,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 22646824,
+      "step": 28985
+    },
+    {
+      "epoch": 60.270270270270274,
+      "grad_norm": 0.00011290181282674894,
+      "learning_rate": 8.780228438804122e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22650696,
+      "step": 28990
+    },
+    {
+      "epoch": 60.28066528066528,
+      "grad_norm": 6.523622687382158e-06,
+      "learning_rate": 8.772758915901032e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22654536,
+      "step": 28995
+    },
+    {
+      "epoch": 60.29106029106029,
+      "grad_norm": 9.774858881428372e-06,
+      "learning_rate": 8.765291895442443e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22658440,
+      "step": 29000
+    },
+    {
+      "epoch": 60.29106029106029,
+      "eval_loss": 0.5580564737319946,
+      "eval_runtime": 11.6931,
+      "eval_samples_per_second": 73.205,
+      "eval_steps_per_second": 18.301,
+      "num_input_tokens_seen": 22658440,
+      "step": 29000
+    },
+    {
+      "epoch": 60.3014553014553,
+      "grad_norm": 5.759463965659961e-05,
+      "learning_rate": 8.75782737857987e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22662312,
+      "step": 29005
+    },
+    {
+      "epoch": 60.311850311850314,
+      "grad_norm": 2.4006483727134764e-05,
+      "learning_rate": 8.750365366464425e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22666216,
+      "step": 29010
+    },
+    {
+      "epoch": 60.32224532224532,
+      "grad_norm": 3.555860530468635e-05,
+      "learning_rate": 8.742905860246838e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22670120,
+      "step": 29015
+    },
+    {
+      "epoch": 60.33264033264033,
+      "grad_norm": 0.0007573412731289864,
+      "learning_rate": 8.735448861077478e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22673992,
+      "step": 29020
+    },
+    {
+      "epoch": 60.343035343035346,
+      "grad_norm": 5.776680336566642e-05,
+      "learning_rate": 8.727994370106288e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22677640,
+      "step": 29025
+    },
+    {
+      "epoch": 60.353430353430355,
+      "grad_norm": 0.00016782643797341734,
+      "learning_rate": 8.720542388482861e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22681512,
+      "step": 29030
+    },
+    {
+      "epoch": 60.36382536382536,
+      "grad_norm": 1.7747748870533542e-06,
+      "learning_rate": 8.71309291735637e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22685416,
+      "step": 29035
+    },
+    {
+      "epoch": 60.37422037422037,
+      "grad_norm": 0.0005919092218391597,
+      "learning_rate": 8.705645957875621e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22689416,
+      "step": 29040
+    },
+    {
+      "epoch": 60.38461538461539,
+      "grad_norm": 1.0081781738335849e-06,
+      "learning_rate": 8.698201511189048e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22693320,
+      "step": 29045
+    },
+    {
+      "epoch": 60.395010395010395,
+      "grad_norm": 2.972431138914544e-05,
+      "learning_rate": 8.690759578444649e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22697256,
+      "step": 29050
+    },
+    {
+      "epoch": 60.4054054054054,
+      "grad_norm": 0.00017640854639466852,
+      "learning_rate": 8.68332016079008e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22701064,
+      "step": 29055
+    },
+    {
+      "epoch": 60.41580041580042,
+      "grad_norm": 1.1757744005080895e-06,
+      "learning_rate": 8.6758832593726e-06,
+      "loss": 0.0044,
+      "num_input_tokens_seen": 22704904,
+      "step": 29060
+    },
+    {
+      "epoch": 60.42619542619543,
+      "grad_norm": 0.0015927419299259782,
+      "learning_rate": 8.668448875339053e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22708648,
+      "step": 29065
+    },
+    {
+      "epoch": 60.436590436590436,
+      "grad_norm": 1.591596083017066e-05,
+      "learning_rate": 8.661017009835933e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22712584,
+      "step": 29070
+    },
+    {
+      "epoch": 60.446985446985444,
+      "grad_norm": 1.5030213944555726e-05,
+      "learning_rate": 8.653587664009311e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22716456,
+      "step": 29075
+    },
+    {
+      "epoch": 60.45738045738046,
+      "grad_norm": 0.00014545391604769975,
+      "learning_rate": 8.646160839004902e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22720392,
+      "step": 29080
+    },
+    {
+      "epoch": 60.46777546777547,
+      "grad_norm": 9.487565876042936e-06,
+      "learning_rate": 8.638736535967998e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22724264,
+      "step": 29085
+    },
+    {
+      "epoch": 60.478170478170476,
+      "grad_norm": 1.3146261153451633e-05,
+      "learning_rate": 8.631314756043535e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 22728168,
+      "step": 29090
+    },
+    {
+      "epoch": 60.48856548856549,
+      "grad_norm": 2.8706803277600557e-05,
+      "learning_rate": 8.62389550037603e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22732040,
+      "step": 29095
+    },
+    {
+      "epoch": 60.4989604989605,
+      "grad_norm": 5.669633173965849e-05,
+      "learning_rate": 8.616478770109646e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22735912,
+      "step": 29100
+    },
+    {
+      "epoch": 60.50935550935551,
+      "grad_norm": 1.0934272722806782e-06,
+      "learning_rate": 8.609064566388111e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22739784,
+      "step": 29105
+    },
+    {
+      "epoch": 60.51975051975052,
+      "grad_norm": 1.3058405784249771e-05,
+      "learning_rate": 8.601652890354815e-06,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 22743688,
+      "step": 29110
+    },
+    {
+      "epoch": 60.53014553014553,
+      "grad_norm": 4.276564504834823e-05,
+      "learning_rate": 8.594243743152705e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22747592,
+      "step": 29115
+    },
+    {
+      "epoch": 60.54054054054054,
+      "grad_norm": 2.451657928759232e-05,
+      "learning_rate": 8.58683712592438e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22751560,
+      "step": 29120
+    },
+    {
+      "epoch": 60.55093555093555,
+      "grad_norm": 6.407393357221736e-06,
+      "learning_rate": 8.579433039812037e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22755432,
+      "step": 29125
+    },
+    {
+      "epoch": 60.561330561330564,
+      "grad_norm": 1.7931422917172313e-05,
+      "learning_rate": 8.572031485957466e-06,
+      "loss": 0.002,
+      "num_input_tokens_seen": 22759400,
+      "step": 29130
+    },
+    {
+      "epoch": 60.57172557172557,
+      "grad_norm": 0.0013878816971555352,
+      "learning_rate": 8.564632465502084e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22763336,
+      "step": 29135
+    },
+    {
+      "epoch": 60.58212058212058,
+      "grad_norm": 1.8945769625133835e-05,
+      "learning_rate": 8.557235979586928e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22767176,
+      "step": 29140
+    },
+    {
+      "epoch": 60.59251559251559,
+      "grad_norm": 8.513110515195876e-05,
+      "learning_rate": 8.549842029352606e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22771176,
+      "step": 29145
+    },
+    {
+      "epoch": 60.602910602910605,
+      "grad_norm": 1.3823523659084458e-05,
+      "learning_rate": 8.542450615939376e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22775080,
+      "step": 29150
+    },
+    {
+      "epoch": 60.61330561330561,
+      "grad_norm": 1.251869798579719e-05,
+      "learning_rate": 8.535061740487082e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22778952,
+      "step": 29155
+    },
+    {
+      "epoch": 60.62370062370062,
+      "grad_norm": 2.800974925776245e-06,
+      "learning_rate": 8.527675404135168e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22782920,
+      "step": 29160
+    },
+    {
+      "epoch": 60.63409563409564,
+      "grad_norm": 5.645807505061384e-06,
+      "learning_rate": 8.520291608022724e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22786824,
+      "step": 29165
+    },
+    {
+      "epoch": 60.644490644490645,
+      "grad_norm": 7.813795491529163e-06,
+      "learning_rate": 8.512910353288398e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22790632,
+      "step": 29170
+    },
+    {
+      "epoch": 60.65488565488565,
+      "grad_norm": 1.6179850717890076e-05,
+      "learning_rate": 8.505531641070486e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22794440,
+      "step": 29175
+    },
+    {
+      "epoch": 60.66528066528066,
+      "grad_norm": 1.27543198686908e-05,
+      "learning_rate": 8.498155472506885e-06,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 22798280,
+      "step": 29180
+    },
+    {
+      "epoch": 60.67567567567568,
+      "grad_norm": 1.665105446591042e-05,
+      "learning_rate": 8.49078184873508e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22802152,
+      "step": 29185
+    },
+    {
+      "epoch": 60.686070686070686,
+      "grad_norm": 7.366140380327124e-06,
+      "learning_rate": 8.483410770892188e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22806088,
+      "step": 29190
+    },
+    {
+      "epoch": 60.696465696465694,
+      "grad_norm": 5.91809020988876e-06,
+      "learning_rate": 8.476042240114909e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22810056,
+      "step": 29195
+    },
+    {
+      "epoch": 60.70686070686071,
+      "grad_norm": 5.9542438975768164e-05,
+      "learning_rate": 8.468676257539568e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22814056,
+      "step": 29200
+    },
+    {
+      "epoch": 60.70686070686071,
+      "eval_loss": 0.5676396489143372,
+      "eval_runtime": 11.7203,
+      "eval_samples_per_second": 73.036,
+      "eval_steps_per_second": 18.259,
+      "num_input_tokens_seen": 22814056,
+      "step": 29200
+    },
+    {
+      "epoch": 60.71725571725572,
+      "grad_norm": 2.6194920792477205e-06,
+      "learning_rate": 8.4613128243021e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22817928,
+      "step": 29205
+    },
+    {
+      "epoch": 60.727650727650726,
+      "grad_norm": 3.414784032429452e-06,
+      "learning_rate": 8.453951941538028e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22821832,
+      "step": 29210
+    },
+    {
+      "epoch": 60.73804573804574,
+      "grad_norm": 0.0002469606988597661,
+      "learning_rate": 8.446593610382495e-06,
+      "loss": 0.0049,
+      "num_input_tokens_seen": 22825736,
+      "step": 29215
+    },
+    {
+      "epoch": 60.74844074844075,
+      "grad_norm": 3.266129351686686e-05,
+      "learning_rate": 8.439237831970259e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22829480,
+      "step": 29220
+    },
+    {
+      "epoch": 60.75883575883576,
+      "grad_norm": 2.5895826183841564e-05,
+      "learning_rate": 8.431884607435667e-06,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 22833640,
+      "step": 29225
+    },
+    {
+      "epoch": 60.76923076923077,
+      "grad_norm": 5.162347406439949e-06,
+      "learning_rate": 8.424533937912665e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22837576,
+      "step": 29230
+    },
+    {
+      "epoch": 60.77962577962578,
+      "grad_norm": 2.9896355044911616e-06,
+      "learning_rate": 8.41718582453484e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22841544,
+      "step": 29235
+    },
+    {
+      "epoch": 60.79002079002079,
+      "grad_norm": 2.258352424178156e-06,
+      "learning_rate": 8.409840268435346e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22845640,
+      "step": 29240
+    },
+    {
+      "epoch": 60.8004158004158,
+      "grad_norm": 2.2998074200586416e-06,
+      "learning_rate": 8.402497270746976e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22849640,
+      "step": 29245
+    },
+    {
+      "epoch": 60.810810810810814,
+      "grad_norm": 7.801746687619016e-05,
+      "learning_rate": 8.395156832602095e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22853608,
+      "step": 29250
+    },
+    {
+      "epoch": 60.82120582120582,
+      "grad_norm": 1.7156005469587399e-06,
+      "learning_rate": 8.387818955132707e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22857608,
+      "step": 29255
+    },
+    {
+      "epoch": 60.83160083160083,
+      "grad_norm": 8.793768756731879e-06,
+      "learning_rate": 8.38048363947039e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 22861576,
+      "step": 29260
+    },
+    {
+      "epoch": 60.84199584199584,
+      "grad_norm": 4.313632416597102e-06,
+      "learning_rate": 8.373150886746351e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 22865448,
+      "step": 29265
+    },
+    {
+      "epoch": 60.852390852390855,
+      "grad_norm": 2.5020008251885884e-06,
+      "learning_rate": 8.365820698091397e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22869384,
+      "step": 29270
+    },
+    {
+      "epoch": 60.86278586278586,
+      "grad_norm": 1.9852386685670353e-05,
+      "learning_rate": 8.358493074635922e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22873256,
+      "step": 29275
+    },
+    {
+      "epoch": 60.87318087318087,
+      "grad_norm": 2.4289340672112303e-06,
+      "learning_rate": 8.351168017509948e-06,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 22877224,
+      "step": 29280
+    },
+    {
+      "epoch": 60.88357588357589,
+      "grad_norm": 1.7623864550841972e-05,
+      "learning_rate": 8.343845527843094e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22880968,
+      "step": 29285
+    },
+    {
+      "epoch": 60.893970893970895,
+      "grad_norm": 2.147032546417904e-06,
+      "learning_rate": 8.336525606764566e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22884968,
+      "step": 29290
+    },
+    {
+      "epoch": 60.9043659043659,
+      "grad_norm": 1.1217764949833509e-05,
+      "learning_rate": 8.329208255403204e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22889064,
+      "step": 29295
+    },
+    {
+      "epoch": 60.91476091476091,
+      "grad_norm": 1.4246232240111567e-05,
+      "learning_rate": 8.321893474887426e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22892936,
+      "step": 29300
+    },
+    {
+      "epoch": 60.92515592515593,
+      "grad_norm": 8.091843483271077e-05,
+      "learning_rate": 8.31458126634526e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22896712,
+      "step": 29305
+    },
+    {
+      "epoch": 60.935550935550935,
+      "grad_norm": 2.9828366677975282e-05,
+      "learning_rate": 8.30727163090435e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22900520,
+      "step": 29310
+    },
+    {
+      "epoch": 60.945945945945944,
+      "grad_norm": 2.606747148092836e-05,
+      "learning_rate": 8.29996456969192e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22904232,
+      "step": 29315
+    },
+    {
+      "epoch": 60.95634095634096,
+      "grad_norm": 3.616548565332778e-05,
+      "learning_rate": 8.292660083834818e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22908040,
+      "step": 29320
+    },
+    {
+      "epoch": 60.96673596673597,
+      "grad_norm": 0.00012522345059551299,
+      "learning_rate": 8.2853581744595e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22912104,
+      "step": 29325
+    },
+    {
+      "epoch": 60.977130977130976,
+      "grad_norm": 8.211390377255157e-05,
+      "learning_rate": 8.278058842691991e-06,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 22916008,
+      "step": 29330
+    },
+    {
+      "epoch": 60.987525987525984,
+      "grad_norm": 1.5489406450797105e-06,
+      "learning_rate": 8.27076208965796e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 22919944,
+      "step": 29335
+    },
+    {
+      "epoch": 60.997920997921,
+      "grad_norm": 0.1371215432882309,
+      "learning_rate": 8.263467916482637e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 22923816,
+      "step": 29340
+    },
+    {
+      "epoch": 61.00831600831601,
+      "grad_norm": 3.972274043917423e-06,
+      "learning_rate": 8.256176324290885e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22927704,
+      "step": 29345
+    },
+    {
+      "epoch": 61.018711018711016,
+      "grad_norm": 1.538376068310754e-06,
+      "learning_rate": 8.248887314207168e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22931672,
+      "step": 29350
+    },
+    {
+      "epoch": 61.02910602910603,
+      "grad_norm": 3.368487523403019e-05,
+      "learning_rate": 8.24160088735553e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22935544,
+      "step": 29355
+    },
+    {
+      "epoch": 61.03950103950104,
+      "grad_norm": 3.118732638540678e-05,
+      "learning_rate": 8.234317044859629e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22939512,
+      "step": 29360
+    },
+    {
+      "epoch": 61.04989604989605,
+      "grad_norm": 3.912306965503376e-06,
+      "learning_rate": 8.227035787842744e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22943320,
+      "step": 29365
+    },
+    {
+      "epoch": 61.06029106029106,
+      "grad_norm": 0.09985901415348053,
+      "learning_rate": 8.219757117427721e-06,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 22947288,
+      "step": 29370
+    },
+    {
+      "epoch": 61.07068607068607,
+      "grad_norm": 0.15823283791542053,
+      "learning_rate": 8.212481034737014e-06,
+      "loss": 0.007,
+      "num_input_tokens_seen": 22951192,
+      "step": 29375
+    },
+    {
+      "epoch": 61.08108108108108,
+      "grad_norm": 3.416836989345029e-06,
+      "learning_rate": 8.205207540892707e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 22955064,
+      "step": 29380
+    },
+    {
+      "epoch": 61.09147609147609,
+      "grad_norm": 1.8991160686709918e-05,
+      "learning_rate": 8.197936637016442e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22959096,
+      "step": 29385
+    },
+    {
+      "epoch": 61.101871101871104,
+      "grad_norm": 1.4190339243214112e-05,
+      "learning_rate": 8.190668324229508e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22962808,
+      "step": 29390
+    },
+    {
+      "epoch": 61.11226611226611,
+      "grad_norm": 1.4025052905708435e-06,
+      "learning_rate": 8.183402603652749e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 22966744,
+      "step": 29395
+    },
+    {
+      "epoch": 61.12266112266112,
+      "grad_norm": 1.811898619052954e-05,
+      "learning_rate": 8.176139476406635e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22970680,
+      "step": 29400
+    },
+    {
+      "epoch": 61.12266112266112,
+      "eval_loss": 0.5812044143676758,
+      "eval_runtime": 11.6809,
+      "eval_samples_per_second": 73.282,
+      "eval_steps_per_second": 18.32,
+      "num_input_tokens_seen": 22970680,
+      "step": 29400
+    },
+    {
+      "epoch": 61.13305613305613,
+      "grad_norm": 5.90671197642223e-06,
+      "learning_rate": 8.16887894361125e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22974488,
+      "step": 29405
+    },
+    {
+      "epoch": 61.143451143451145,
+      "grad_norm": 3.409777127671987e-05,
+      "learning_rate": 8.161621006386233e-06,
+      "loss": 0.002,
+      "num_input_tokens_seen": 22978360,
+      "step": 29410
+    },
+    {
+      "epoch": 61.15384615384615,
+      "grad_norm": 1.889500163088087e-05,
+      "learning_rate": 8.154365665850869e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22982232,
+      "step": 29415
+    },
+    {
+      "epoch": 61.16424116424116,
+      "grad_norm": 0.000356766366166994,
+      "learning_rate": 8.147112923124005e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22986168,
+      "step": 29420
+    },
+    {
+      "epoch": 61.17463617463618,
+      "grad_norm": 4.121717211091891e-06,
+      "learning_rate": 8.13986277932412e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22990136,
+      "step": 29425
+    },
+    {
+      "epoch": 61.185031185031185,
+      "grad_norm": 9.752279765962157e-06,
+      "learning_rate": 8.132615235569277e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22993944,
+      "step": 29430
+    },
+    {
+      "epoch": 61.195426195426194,
+      "grad_norm": 4.289739081286825e-05,
+      "learning_rate": 8.125370292977124e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 22997784,
+      "step": 29435
+    },
+    {
+      "epoch": 61.20582120582121,
+      "grad_norm": 1.9780953152803704e-05,
+      "learning_rate": 8.118127952664944e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23001656,
+      "step": 29440
+    },
+    {
+      "epoch": 61.21621621621622,
+      "grad_norm": 2.1794843632960692e-05,
+      "learning_rate": 8.110888215749574e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23005560,
+      "step": 29445
+    },
+    {
+      "epoch": 61.226611226611226,
+      "grad_norm": 1.8821867342921905e-05,
+      "learning_rate": 8.10365108334749e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23009400,
+      "step": 29450
+    },
+    {
+      "epoch": 61.237006237006234,
+      "grad_norm": 4.280981556803454e-06,
+      "learning_rate": 8.096416556574743e-06,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 23013368,
+      "step": 29455
+    },
+    {
+      "epoch": 61.24740124740125,
+      "grad_norm": 0.0003386393073014915,
+      "learning_rate": 8.08918463654698e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23017336,
+      "step": 29460
+    },
+    {
+      "epoch": 61.25779625779626,
+      "grad_norm": 2.002370820264332e-06,
+      "learning_rate": 8.081955324379458e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23021240,
+      "step": 29465
+    },
+    {
+      "epoch": 61.268191268191266,
+      "grad_norm": 1.3494715176420868e-06,
+      "learning_rate": 8.074728621187039e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23025144,
+      "step": 29470
+    },
+    {
+      "epoch": 61.27858627858628,
+      "grad_norm": 4.01115212298464e-05,
+      "learning_rate": 8.067504528084158e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 23029208,
+      "step": 29475
+    },
+    {
+      "epoch": 61.28898128898129,
+      "grad_norm": 1.3320298421604093e-05,
+      "learning_rate": 8.060283046184861e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23033144,
+      "step": 29480
+    },
+    {
+      "epoch": 61.2993762993763,
+      "grad_norm": 7.101057417457923e-05,
+      "learning_rate": 8.053064176602806e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23037016,
+      "step": 29485
+    },
+    {
+      "epoch": 61.30977130977131,
+      "grad_norm": 1.0991126146109309e-05,
+      "learning_rate": 8.045847920451216e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23040760,
+      "step": 29490
+    },
+    {
+      "epoch": 61.32016632016632,
+      "grad_norm": 3.3383619211235782e-06,
+      "learning_rate": 8.038634278842944e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23044792,
+      "step": 29495
+    },
+    {
+      "epoch": 61.33056133056133,
+      "grad_norm": 5.97867947362829e-05,
+      "learning_rate": 8.031423252890408e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23048728,
+      "step": 29500
+    },
+    {
+      "epoch": 61.34095634095634,
+      "grad_norm": 6.26034170636558e-06,
+      "learning_rate": 8.024214843705646e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23052536,
+      "step": 29505
+    },
+    {
+      "epoch": 61.351351351351354,
+      "grad_norm": 0.0001059701680787839,
+      "learning_rate": 8.017009052400295e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23056472,
+      "step": 29510
+    },
+    {
+      "epoch": 61.36174636174636,
+      "grad_norm": 8.235268614953384e-05,
+      "learning_rate": 8.00980588008557e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23060376,
+      "step": 29515
+    },
+    {
+      "epoch": 61.37214137214137,
+      "grad_norm": 0.000488348538056016,
+      "learning_rate": 8.002605327872282e-06,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 23064440,
+      "step": 29520
+    },
+    {
+      "epoch": 61.38253638253638,
+      "grad_norm": 3.806592576438561e-05,
+      "learning_rate": 7.995407396870862e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23068248,
+      "step": 29525
+    },
+    {
+      "epoch": 61.392931392931395,
+      "grad_norm": 0.0002646205830387771,
+      "learning_rate": 7.988212088191307e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23072280,
+      "step": 29530
+    },
+    {
+      "epoch": 61.4033264033264,
+      "grad_norm": 2.946229142253287e-05,
+      "learning_rate": 7.98101940294324e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23076056,
+      "step": 29535
+    },
+    {
+      "epoch": 61.41372141372141,
+      "grad_norm": 7.021949568297714e-05,
+      "learning_rate": 7.973829342235847e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23079864,
+      "step": 29540
+    },
+    {
+      "epoch": 61.42411642411643,
+      "grad_norm": 5.789734586869599e-06,
+      "learning_rate": 7.966641907177936e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23083832,
+      "step": 29545
+    },
+    {
+      "epoch": 61.434511434511435,
+      "grad_norm": 8.602752927799884e-07,
+      "learning_rate": 7.959457098877901e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23087640,
+      "step": 29550
+    },
+    {
+      "epoch": 61.444906444906444,
+      "grad_norm": 7.965162512846291e-05,
+      "learning_rate": 7.952274918443719e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23091608,
+      "step": 29555
+    },
+    {
+      "epoch": 61.45530145530145,
+      "grad_norm": 3.058876973227598e-05,
+      "learning_rate": 7.945095366982983e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23095512,
+      "step": 29560
+    },
+    {
+      "epoch": 61.46569646569647,
+      "grad_norm": 3.953209215978859e-06,
+      "learning_rate": 7.937918445602871e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 23099448,
+      "step": 29565
+    },
+    {
+      "epoch": 61.476091476091476,
+      "grad_norm": 4.238609562889906e-06,
+      "learning_rate": 7.930744155410145e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23103320,
+      "step": 29570
+    },
+    {
+      "epoch": 61.486486486486484,
+      "grad_norm": 9.311416215496138e-05,
+      "learning_rate": 7.923572497511181e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23107160,
+      "step": 29575
+    },
+    {
+      "epoch": 61.4968814968815,
+      "grad_norm": 2.1856805688003078e-05,
+      "learning_rate": 7.916403473011927e-06,
+      "loss": 0.0037,
+      "num_input_tokens_seen": 23111192,
+      "step": 29580
+    },
+    {
+      "epoch": 61.50727650727651,
+      "grad_norm": 0.00044447006075643003,
+      "learning_rate": 7.909237083017953e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23115064,
+      "step": 29585
+    },
+    {
+      "epoch": 61.517671517671516,
+      "grad_norm": 1.0206653087152517e-06,
+      "learning_rate": 7.902073328634389e-06,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 23119000,
+      "step": 29590
+    },
+    {
+      "epoch": 61.528066528066525,
+      "grad_norm": 0.00042006056173704565,
+      "learning_rate": 7.894912210965987e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23123000,
+      "step": 29595
+    },
+    {
+      "epoch": 61.53846153846154,
+      "grad_norm": 1.5497893173233024e-06,
+      "learning_rate": 7.887753731117075e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23126776,
+      "step": 29600
+    },
+    {
+      "epoch": 61.53846153846154,
+      "eval_loss": 0.5788075923919678,
+      "eval_runtime": 11.6858,
+      "eval_samples_per_second": 73.251,
+      "eval_steps_per_second": 18.313,
+      "num_input_tokens_seen": 23126776,
+      "step": 29600
+    },
+    {
+      "epoch": 61.54885654885655,
+      "grad_norm": 2.9415616609185236e-06,
+      "learning_rate": 7.880597890191587e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23130520,
+      "step": 29605
+    },
+    {
+      "epoch": 61.55925155925156,
+      "grad_norm": 4.90494494442828e-05,
+      "learning_rate": 7.873444689293036e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23134456,
+      "step": 29610
+    },
+    {
+      "epoch": 61.56964656964657,
+      "grad_norm": 0.00016418313316535205,
+      "learning_rate": 7.866294129524548e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23138296,
+      "step": 29615
+    },
+    {
+      "epoch": 61.58004158004158,
+      "grad_norm": 3.1594121537636966e-05,
+      "learning_rate": 7.859146211988811e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23142232,
+      "step": 29620
+    },
+    {
+      "epoch": 61.59043659043659,
+      "grad_norm": 1.9370903828530572e-05,
+      "learning_rate": 7.852000937788134e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 23146104,
+      "step": 29625
+    },
+    {
+      "epoch": 61.6008316008316,
+      "grad_norm": 1.997405661313678e-06,
+      "learning_rate": 7.844858308024416e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23150072,
+      "step": 29630
+    },
+    {
+      "epoch": 61.61122661122661,
+      "grad_norm": 4.600573083735071e-05,
+      "learning_rate": 7.837718323799122e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 23154008,
+      "step": 29635
+    },
+    {
+      "epoch": 61.62162162162162,
+      "grad_norm": 5.549586421693675e-05,
+      "learning_rate": 7.83058098621334e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23157944,
+      "step": 29640
+    },
+    {
+      "epoch": 61.63201663201663,
+      "grad_norm": 5.024104757467285e-05,
+      "learning_rate": 7.823446296367739e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23161784,
+      "step": 29645
+    },
+    {
+      "epoch": 61.642411642411645,
+      "grad_norm": 5.388169029174605e-06,
+      "learning_rate": 7.81631425536257e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23165656,
+      "step": 29650
+    },
+    {
+      "epoch": 61.65280665280665,
+      "grad_norm": 6.303517238848144e-06,
+      "learning_rate": 7.809184864297689e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23169656,
+      "step": 29655
+    },
+    {
+      "epoch": 61.66320166320166,
+      "grad_norm": 5.889438034500927e-05,
+      "learning_rate": 7.802058124272532e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23173496,
+      "step": 29660
+    },
+    {
+      "epoch": 61.67359667359668,
+      "grad_norm": 1.8969931261381134e-05,
+      "learning_rate": 7.79493403638614e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23177432,
+      "step": 29665
+    },
+    {
+      "epoch": 61.683991683991685,
+      "grad_norm": 1.874080226116348e-06,
+      "learning_rate": 7.787812601737132e-06,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 23181272,
+      "step": 29670
+    },
+    {
+      "epoch": 61.694386694386694,
+      "grad_norm": 7.096611625456717e-06,
+      "learning_rate": 7.780693821423715e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23185176,
+      "step": 29675
+    },
+    {
+      "epoch": 61.7047817047817,
+      "grad_norm": 0.00010901577479671687,
+      "learning_rate": 7.773577696543705e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23189144,
+      "step": 29680
+    },
+    {
+      "epoch": 61.71517671517672,
+      "grad_norm": 4.983558483218076e-06,
+      "learning_rate": 7.7664642281945e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23193080,
+      "step": 29685
+    },
+    {
+      "epoch": 61.725571725571726,
+      "grad_norm": 5.3580888561555184e-06,
+      "learning_rate": 7.759353417473072e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23196856,
+      "step": 29690
+    },
+    {
+      "epoch": 61.735966735966734,
+      "grad_norm": 0.0002582994638942182,
+      "learning_rate": 7.752245265476016e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 23200696,
+      "step": 29695
+    },
+    {
+      "epoch": 61.74636174636175,
+      "grad_norm": 5.294995935400948e-05,
+      "learning_rate": 7.745139773299481e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23204696,
+      "step": 29700
+    },
+    {
+      "epoch": 61.75675675675676,
+      "grad_norm": 2.5159522920148447e-05,
+      "learning_rate": 7.738036942039232e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23208664,
+      "step": 29705
+    },
+    {
+      "epoch": 61.767151767151766,
+      "grad_norm": 2.2272966816672124e-05,
+      "learning_rate": 7.73093677279062e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23212600,
+      "step": 29710
+    },
+    {
+      "epoch": 61.777546777546775,
+      "grad_norm": 1.644731491978746e-05,
+      "learning_rate": 7.72383926664857e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 23216376,
+      "step": 29715
+    },
+    {
+      "epoch": 61.78794178794179,
+      "grad_norm": 0.1434052735567093,
+      "learning_rate": 7.716744424707606e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 23220344,
+      "step": 29720
+    },
+    {
+      "epoch": 61.7983367983368,
+      "grad_norm": 5.617029728455236e-06,
+      "learning_rate": 7.709652248061858e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23224312,
+      "step": 29725
+    },
+    {
+      "epoch": 61.80873180873181,
+      "grad_norm": 1.9758135749725625e-05,
+      "learning_rate": 7.702562737805017e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23228408,
+      "step": 29730
+    },
+    {
+      "epoch": 61.81912681912682,
+      "grad_norm": 8.632569915789645e-06,
+      "learning_rate": 7.695475895030365e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23232216,
+      "step": 29735
+    },
+    {
+      "epoch": 61.82952182952183,
+      "grad_norm": 1.0115449185832404e-05,
+      "learning_rate": 7.6883917208308e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23236152,
+      "step": 29740
+    },
+    {
+      "epoch": 61.83991683991684,
+      "grad_norm": 2.5011124762386316e-06,
+      "learning_rate": 7.681310216298778e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23239992,
+      "step": 29745
+    },
+    {
+      "epoch": 61.85031185031185,
+      "grad_norm": 6.574324652319774e-05,
+      "learning_rate": 7.674231382526367e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23244024,
+      "step": 29750
+    },
+    {
+      "epoch": 61.86070686070686,
+      "grad_norm": 4.446233106136788e-06,
+      "learning_rate": 7.667155220605198e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 23247832,
+      "step": 29755
+    },
+    {
+      "epoch": 61.87110187110187,
+      "grad_norm": 2.161528027500026e-05,
+      "learning_rate": 7.660081731626515e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23251800,
+      "step": 29760
+    },
+    {
+      "epoch": 61.88149688149688,
+      "grad_norm": 3.010184173035668e-06,
+      "learning_rate": 7.653010916681141e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 23255736,
+      "step": 29765
+    },
+    {
+      "epoch": 61.891891891891895,
+      "grad_norm": 1.3943924841441913e-06,
+      "learning_rate": 7.645942776859472e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23259480,
+      "step": 29770
+    },
+    {
+      "epoch": 61.9022869022869,
+      "grad_norm": 6.404900432244176e-06,
+      "learning_rate": 7.63887731325152e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23263288,
+      "step": 29775
+    },
+    {
+      "epoch": 61.91268191268191,
+      "grad_norm": 0.00015020201681181788,
+      "learning_rate": 7.63181452694685e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23267224,
+      "step": 29780
+    },
+    {
+      "epoch": 61.92307692307692,
+      "grad_norm": 1.369483811686223e-06,
+      "learning_rate": 7.624754419034644e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23271064,
+      "step": 29785
+    },
+    {
+      "epoch": 61.933471933471935,
+      "grad_norm": 1.1044435268559027e-05,
+      "learning_rate": 7.6176969906036645e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23275032,
+      "step": 29790
+    },
+    {
+      "epoch": 61.943866943866944,
+      "grad_norm": 9.428580597159453e-06,
+      "learning_rate": 7.610642242742242e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23279000,
+      "step": 29795
+    },
+    {
+      "epoch": 61.95426195426195,
+      "grad_norm": 1.4826824553892948e-05,
+      "learning_rate": 7.603590176538322e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23283064,
+      "step": 29800
+    },
+    {
+      "epoch": 61.95426195426195,
+      "eval_loss": 0.5783374309539795,
+      "eval_runtime": 11.6708,
+      "eval_samples_per_second": 73.346,
+      "eval_steps_per_second": 18.336,
+      "num_input_tokens_seen": 23283064,
+      "step": 29800
+    },
+    {
+      "epoch": 61.96465696465697,
+      "grad_norm": 1.303837780142203e-05,
+      "learning_rate": 7.596540793079404e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23286936,
+      "step": 29805
+    },
+    {
+      "epoch": 61.975051975051976,
+      "grad_norm": 1.8731043383013457e-06,
+      "learning_rate": 7.5894940934526125e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23290968,
+      "step": 29810
+    },
+    {
+      "epoch": 61.985446985446984,
+      "grad_norm": 9.787034650798887e-05,
+      "learning_rate": 7.582450078744621e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23294840,
+      "step": 29815
+    },
+    {
+      "epoch": 61.99584199584199,
+      "grad_norm": 2.8803256100218277e-06,
+      "learning_rate": 7.575408750041707e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23298808,
+      "step": 29820
+    },
+    {
+      "epoch": 62.00623700623701,
+      "grad_norm": 2.1769463273813017e-05,
+      "learning_rate": 7.568370108429732e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23302816,
+      "step": 29825
+    },
+    {
+      "epoch": 62.016632016632016,
+      "grad_norm": 4.530157457338646e-05,
+      "learning_rate": 7.561334154994154e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23306816,
+      "step": 29830
+    },
+    {
+      "epoch": 62.027027027027025,
+      "grad_norm": 1.156839425675571e-05,
+      "learning_rate": 7.55430089081999e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23310688,
+      "step": 29835
+    },
+    {
+      "epoch": 62.03742203742204,
+      "grad_norm": 6.472925451816991e-05,
+      "learning_rate": 7.547270316991864e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23314752,
+      "step": 29840
+    },
+    {
+      "epoch": 62.04781704781705,
+      "grad_norm": 7.231147174024954e-05,
+      "learning_rate": 7.5402424345939884e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 23318720,
+      "step": 29845
+    },
+    {
+      "epoch": 62.05821205821206,
+      "grad_norm": 1.9836149931506952e-06,
+      "learning_rate": 7.533217244710133e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23322656,
+      "step": 29850
+    },
+    {
+      "epoch": 62.06860706860707,
+      "grad_norm": 1.2310938473092392e-05,
+      "learning_rate": 7.52619474842369e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23326560,
+      "step": 29855
+    },
+    {
+      "epoch": 62.07900207900208,
+      "grad_norm": 0.054654382169246674,
+      "learning_rate": 7.519174946817597e-06,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 23330464,
+      "step": 29860
+    },
+    {
+      "epoch": 62.08939708939709,
+      "grad_norm": 1.1628616221059929e-06,
+      "learning_rate": 7.512157840974407e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23334368,
+      "step": 29865
+    },
+    {
+      "epoch": 62.0997920997921,
+      "grad_norm": 1.3498513908416498e-05,
+      "learning_rate": 7.5051434319762496e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23338208,
+      "step": 29870
+    },
+    {
+      "epoch": 62.11018711018711,
+      "grad_norm": 1.1963273209403269e-05,
+      "learning_rate": 7.498131720904822e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23342144,
+      "step": 29875
+    },
+    {
+      "epoch": 62.12058212058212,
+      "grad_norm": 4.7361801989609376e-05,
+      "learning_rate": 7.491122708841433e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23346080,
+      "step": 29880
+    },
+    {
+      "epoch": 62.13097713097713,
+      "grad_norm": 9.323935228167102e-05,
+      "learning_rate": 7.4841163968669524e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23349920,
+      "step": 29885
+    },
+    {
+      "epoch": 62.141372141372145,
+      "grad_norm": 3.8327685615513474e-05,
+      "learning_rate": 7.4771127860618355e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23353920,
+      "step": 29890
+    },
+    {
+      "epoch": 62.15176715176715,
+      "grad_norm": 3.213826175851864e-06,
+      "learning_rate": 7.470111877506139e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23357856,
+      "step": 29895
+    },
+    {
+      "epoch": 62.16216216216216,
+      "grad_norm": 3.870716227538651e-06,
+      "learning_rate": 7.463113672279479e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23361696,
+      "step": 29900
+    },
+    {
+      "epoch": 62.17255717255717,
+      "grad_norm": 3.0036997031857027e-06,
+      "learning_rate": 7.456118171461071e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 23365696,
+      "step": 29905
+    },
+    {
+      "epoch": 62.182952182952185,
+      "grad_norm": 3.4165304896305315e-06,
+      "learning_rate": 7.449125376129721e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23369632,
+      "step": 29910
+    },
+    {
+      "epoch": 62.19334719334719,
+      "grad_norm": 6.405556632671505e-05,
+      "learning_rate": 7.442135287363788e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23373536,
+      "step": 29915
+    },
+    {
+      "epoch": 62.2037422037422,
+      "grad_norm": 1.605070792720653e-05,
+      "learning_rate": 7.435147906241247e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23377536,
+      "step": 29920
+    },
+    {
+      "epoch": 62.21413721413722,
+      "grad_norm": 8.915890816751926e-07,
+      "learning_rate": 7.428163233839624e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23381568,
+      "step": 29925
+    },
+    {
+      "epoch": 62.224532224532226,
+      "grad_norm": 3.607055305110407e-06,
+      "learning_rate": 7.4211812712360525e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23385216,
+      "step": 29930
+    },
+    {
+      "epoch": 62.234927234927234,
+      "grad_norm": 1.1826662557723466e-05,
+      "learning_rate": 7.4142020195072464e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23389024,
+      "step": 29935
+    },
+    {
+      "epoch": 62.24532224532224,
+      "grad_norm": 4.4544573029270396e-05,
+      "learning_rate": 7.407225479729479e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23393024,
+      "step": 29940
+    },
+    {
+      "epoch": 62.25571725571726,
+      "grad_norm": 1.1948260180361103e-05,
+      "learning_rate": 7.400251652978632e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23396736,
+      "step": 29945
+    },
+    {
+      "epoch": 62.266112266112266,
+      "grad_norm": 8.049923962971661e-06,
+      "learning_rate": 7.393280540330147e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23400704,
+      "step": 29950
+    },
+    {
+      "epoch": 62.276507276507274,
+      "grad_norm": 5.051261541666463e-05,
+      "learning_rate": 7.386312142859069e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23404768,
+      "step": 29955
+    },
+    {
+      "epoch": 62.28690228690229,
+      "grad_norm": 1.9525090465322137e-05,
+      "learning_rate": 7.379346461640008e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23408736,
+      "step": 29960
+    },
+    {
+      "epoch": 62.2972972972973,
+      "grad_norm": 0.0004994393093511462,
+      "learning_rate": 7.372383497747149e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23412608,
+      "step": 29965
+    },
+    {
+      "epoch": 62.30769230769231,
+      "grad_norm": 9.420147080163588e-07,
+      "learning_rate": 7.3654232522542775e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23416448,
+      "step": 29970
+    },
+    {
+      "epoch": 62.318087318087315,
+      "grad_norm": 1.4056789041205775e-05,
+      "learning_rate": 7.358465726234756e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23420384,
+      "step": 29975
+    },
+    {
+      "epoch": 62.32848232848233,
+      "grad_norm": 1.8419061234453693e-05,
+      "learning_rate": 7.351510920761512e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23424224,
+      "step": 29980
+    },
+    {
+      "epoch": 62.33887733887734,
+      "grad_norm": 1.1927303603442851e-05,
+      "learning_rate": 7.344558836907067e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23428160,
+      "step": 29985
+    },
+    {
+      "epoch": 62.34927234927235,
+      "grad_norm": 9.81261928245658e-06,
+      "learning_rate": 7.3376094757435285e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23432160,
+      "step": 29990
+    },
+    {
+      "epoch": 62.35966735966736,
+      "grad_norm": 4.993944457964972e-06,
+      "learning_rate": 7.330662838342561e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23436128,
+      "step": 29995
+    },
+    {
+      "epoch": 62.37006237006237,
+      "grad_norm": 6.16006218479015e-05,
+      "learning_rate": 7.323718925775438e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23440000,
+      "step": 30000
+    },
+    {
+      "epoch": 62.37006237006237,
+      "eval_loss": 0.596027135848999,
+      "eval_runtime": 11.6894,
+      "eval_samples_per_second": 73.229,
+      "eval_steps_per_second": 18.307,
+      "num_input_tokens_seen": 23440000,
+      "step": 30000
+    },
+    {
+      "epoch": 62.38045738045738,
+      "grad_norm": 1.4988857401476707e-05,
+      "learning_rate": 7.316777739112985e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23443744,
+      "step": 30005
+    },
+    {
+      "epoch": 62.39085239085239,
+      "grad_norm": 7.3088590397674125e-06,
+      "learning_rate": 7.309839279425626e-06,
+      "loss": 0.0038,
+      "num_input_tokens_seen": 23447680,
+      "step": 30010
+    },
+    {
+      "epoch": 62.4012474012474,
+      "grad_norm": 0.15468865633010864,
+      "learning_rate": 7.302903547783366e-06,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 23451648,
+      "step": 30015
+    },
+    {
+      "epoch": 62.41164241164241,
+      "grad_norm": 4.351005372882355e-06,
+      "learning_rate": 7.2959705452557644e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23455680,
+      "step": 30020
+    },
+    {
+      "epoch": 62.42203742203742,
+      "grad_norm": 1.0443382052471861e-05,
+      "learning_rate": 7.289040272911996e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23459520,
+      "step": 30025
+    },
+    {
+      "epoch": 62.432432432432435,
+      "grad_norm": 6.0886755818501115e-05,
+      "learning_rate": 7.282112731820789e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23463392,
+      "step": 30030
+    },
+    {
+      "epoch": 62.44282744282744,
+      "grad_norm": 1.0417775229143444e-06,
+      "learning_rate": 7.275187923050447e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23467168,
+      "step": 30035
+    },
+    {
+      "epoch": 62.45322245322245,
+      "grad_norm": 2.6086976504302584e-05,
+      "learning_rate": 7.268265847668879e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23471232,
+      "step": 30040
+    },
+    {
+      "epoch": 62.46361746361746,
+      "grad_norm": 1.7964221115107648e-05,
+      "learning_rate": 7.261346506743538e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23474976,
+      "step": 30045
+    },
+    {
+      "epoch": 62.474012474012476,
+      "grad_norm": 1.6597349485891755e-06,
+      "learning_rate": 7.254429901341486e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23478848,
+      "step": 30050
+    },
+    {
+      "epoch": 62.484407484407484,
+      "grad_norm": 8.955615157901775e-06,
+      "learning_rate": 7.247516032529356e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23482752,
+      "step": 30055
+    },
+    {
+      "epoch": 62.49480249480249,
+      "grad_norm": 3.6454621294979006e-05,
+      "learning_rate": 7.240604901373338e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 23486784,
+      "step": 30060
+    },
+    {
+      "epoch": 62.50519750519751,
+      "grad_norm": 0.00012871927174273878,
+      "learning_rate": 7.233696508939223e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23490592,
+      "step": 30065
+    },
+    {
+      "epoch": 62.515592515592516,
+      "grad_norm": 1.5889860378592857e-06,
+      "learning_rate": 7.226790856292376e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23494432,
+      "step": 30070
+    },
+    {
+      "epoch": 62.525987525987524,
+      "grad_norm": 2.0054433207405964e-06,
+      "learning_rate": 7.219887944497727e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23498208,
+      "step": 30075
+    },
+    {
+      "epoch": 62.53638253638254,
+      "grad_norm": 0.0001318963768426329,
+      "learning_rate": 7.2129877746198e-06,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 23502176,
+      "step": 30080
+    },
+    {
+      "epoch": 62.54677754677755,
+      "grad_norm": 0.10487383604049683,
+      "learning_rate": 7.20609034772268e-06,
+      "loss": 0.0039,
+      "num_input_tokens_seen": 23505824,
+      "step": 30085
+    },
+    {
+      "epoch": 62.55717255717256,
+      "grad_norm": 5.986337782815099e-06,
+      "learning_rate": 7.19919566487004e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23509696,
+      "step": 30090
+    },
+    {
+      "epoch": 62.567567567567565,
+      "grad_norm": 1.982106186915189e-05,
+      "learning_rate": 7.192303727125132e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23513504,
+      "step": 30095
+    },
+    {
+      "epoch": 62.57796257796258,
+      "grad_norm": 4.281833753339015e-06,
+      "learning_rate": 7.185414535550777e-06,
+      "loss": 0.008,
+      "num_input_tokens_seen": 23517440,
+      "step": 30100
+    },
+    {
+      "epoch": 62.58835758835759,
+      "grad_norm": 0.11358395963907242,
+      "learning_rate": 7.178528091209363e-06,
+      "loss": 0.0061,
+      "num_input_tokens_seen": 23521408,
+      "step": 30105
+    },
+    {
+      "epoch": 62.5987525987526,
+      "grad_norm": 2.8019574529025704e-05,
+      "learning_rate": 7.171644395162888e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23525408,
+      "step": 30110
+    },
+    {
+      "epoch": 62.60914760914761,
+      "grad_norm": 1.2338975466263946e-05,
+      "learning_rate": 7.164763448472881e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23529344,
+      "step": 30115
+    },
+    {
+      "epoch": 62.61954261954262,
+      "grad_norm": 6.438281707232818e-05,
+      "learning_rate": 7.157885252200491e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23533152,
+      "step": 30120
+    },
+    {
+      "epoch": 62.62993762993763,
+      "grad_norm": 9.900316945277154e-05,
+      "learning_rate": 7.151009807406403e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23536928,
+      "step": 30125
+    },
+    {
+      "epoch": 62.64033264033264,
+      "grad_norm": 4.305300990381511e-06,
+      "learning_rate": 7.144137115150909e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23540768,
+      "step": 30130
+    },
+    {
+      "epoch": 62.65072765072765,
+      "grad_norm": 1.8374940964349662e-06,
+      "learning_rate": 7.1372671764938725e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23544736,
+      "step": 30135
+    },
+    {
+      "epoch": 62.66112266112266,
+      "grad_norm": 1.1286704648227897e-05,
+      "learning_rate": 7.130399992494705e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23548640,
+      "step": 30140
+    },
+    {
+      "epoch": 62.67151767151767,
+      "grad_norm": 6.793460670451168e-06,
+      "learning_rate": 7.123535564212419e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23552576,
+      "step": 30145
+    },
+    {
+      "epoch": 62.681912681912685,
+      "grad_norm": 0.00023925327695906162,
+      "learning_rate": 7.116673892705611e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23556416,
+      "step": 30150
+    },
+    {
+      "epoch": 62.69230769230769,
+      "grad_norm": 0.00025814169202931225,
+      "learning_rate": 7.109814979032415e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23560288,
+      "step": 30155
+    },
+    {
+      "epoch": 62.7027027027027,
+      "grad_norm": 1.6727330148569308e-05,
+      "learning_rate": 7.102958824250577e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23564160,
+      "step": 30160
+    },
+    {
+      "epoch": 62.71309771309771,
+      "grad_norm": 2.6566931410343386e-06,
+      "learning_rate": 7.096105429417393e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23568096,
+      "step": 30165
+    },
+    {
+      "epoch": 62.723492723492726,
+      "grad_norm": 3.488137008389458e-05,
+      "learning_rate": 7.0892547955897506e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23572160,
+      "step": 30170
+    },
+    {
+      "epoch": 62.733887733887734,
+      "grad_norm": 9.040745499078184e-05,
+      "learning_rate": 7.0824069238241e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23576128,
+      "step": 30175
+    },
+    {
+      "epoch": 62.74428274428274,
+      "grad_norm": 7.528301921411185e-06,
+      "learning_rate": 7.075561815176462e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23580096,
+      "step": 30180
+    },
+    {
+      "epoch": 62.75467775467776,
+      "grad_norm": 4.0587992771179415e-06,
+      "learning_rate": 7.068719470702445e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23584128,
+      "step": 30185
+    },
+    {
+      "epoch": 62.765072765072766,
+      "grad_norm": 1.8323021322430577e-06,
+      "learning_rate": 7.061879891457229e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23588256,
+      "step": 30190
+    },
+    {
+      "epoch": 62.775467775467774,
+      "grad_norm": 8.37563638924621e-05,
+      "learning_rate": 7.0550430784955515e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 23592192,
+      "step": 30195
+    },
+    {
+      "epoch": 62.78586278586278,
+      "grad_norm": 0.1536676585674286,
+      "learning_rate": 7.048209032871752e-06,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 23596224,
+      "step": 30200
+    },
+    {
+      "epoch": 62.78586278586278,
+      "eval_loss": 0.5770869255065918,
+      "eval_runtime": 11.6933,
+      "eval_samples_per_second": 73.204,
+      "eval_steps_per_second": 18.301,
+      "num_input_tokens_seen": 23596224,
+      "step": 30200
+    },
+    {
+      "epoch": 62.7962577962578,
+      "grad_norm": 4.2046167436637916e-06,
+      "learning_rate": 7.0413777556397055e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23600256,
+      "step": 30205
+    },
+    {
+      "epoch": 62.80665280665281,
+      "grad_norm": 1.8027947589871474e-05,
+      "learning_rate": 7.0345492478528925e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23604288,
+      "step": 30210
+    },
+    {
+      "epoch": 62.817047817047815,
+      "grad_norm": 0.13812023401260376,
+      "learning_rate": 7.02772351056436e-06,
+      "loss": 0.0063,
+      "num_input_tokens_seen": 23608128,
+      "step": 30215
+    },
+    {
+      "epoch": 62.82744282744283,
+      "grad_norm": 4.23971459895256e-06,
+      "learning_rate": 7.020900544826709e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23612096,
+      "step": 30220
+    },
+    {
+      "epoch": 62.83783783783784,
+      "grad_norm": 2.203791336796712e-05,
+      "learning_rate": 7.014080351692134e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23616096,
+      "step": 30225
+    },
+    {
+      "epoch": 62.84823284823285,
+      "grad_norm": 8.659779268782586e-06,
+      "learning_rate": 7.0072629322124024e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 23619872,
+      "step": 30230
+    },
+    {
+      "epoch": 62.858627858627855,
+      "grad_norm": 3.267851752752904e-06,
+      "learning_rate": 7.000448287438827e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23623808,
+      "step": 30235
+    },
+    {
+      "epoch": 62.86902286902287,
+      "grad_norm": 0.0002127560874214396,
+      "learning_rate": 6.993636418422331e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23627776,
+      "step": 30240
+    },
+    {
+      "epoch": 62.87941787941788,
+      "grad_norm": 0.0033943040762096643,
+      "learning_rate": 6.986827326213383e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23631616,
+      "step": 30245
+    },
+    {
+      "epoch": 62.88981288981289,
+      "grad_norm": 0.00014110915071796626,
+      "learning_rate": 6.9800210118620205e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23635584,
+      "step": 30250
+    },
+    {
+      "epoch": 62.9002079002079,
+      "grad_norm": 1.1809273701146594e-06,
+      "learning_rate": 6.973217476417876e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23639296,
+      "step": 30255
+    },
+    {
+      "epoch": 62.91060291060291,
+      "grad_norm": 1.8973400074173696e-05,
+      "learning_rate": 6.96641672093013e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23643200,
+      "step": 30260
+    },
+    {
+      "epoch": 62.92099792099792,
+      "grad_norm": 2.7066891561844386e-05,
+      "learning_rate": 6.95961874644755e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23647040,
+      "step": 30265
+    },
+    {
+      "epoch": 62.931392931392935,
+      "grad_norm": 4.3476106839079875e-06,
+      "learning_rate": 6.952823554018476e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 23650880,
+      "step": 30270
+    },
+    {
+      "epoch": 62.94178794178794,
+      "grad_norm": 4.230303238728084e-05,
+      "learning_rate": 6.946031144690798e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23654752,
+      "step": 30275
+    },
+    {
+      "epoch": 62.95218295218295,
+      "grad_norm": 8.639498446427751e-06,
+      "learning_rate": 6.939241519512005e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23658592,
+      "step": 30280
+    },
+    {
+      "epoch": 62.96257796257796,
+      "grad_norm": 4.568458825815469e-05,
+      "learning_rate": 6.932454679529129e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23662560,
+      "step": 30285
+    },
+    {
+      "epoch": 62.972972972972975,
+      "grad_norm": 2.996169087055023e-06,
+      "learning_rate": 6.925670625788791e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23666464,
+      "step": 30290
+    },
+    {
+      "epoch": 62.983367983367984,
+      "grad_norm": 9.583167411619797e-06,
+      "learning_rate": 6.918889359337186e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23670304,
+      "step": 30295
+    },
+    {
+      "epoch": 62.99376299376299,
+      "grad_norm": 0.0001499650243204087,
+      "learning_rate": 6.912110881220058e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23674272,
+      "step": 30300
+    },
+    {
+      "epoch": 63.00415800415801,
+      "grad_norm": 5.572974714596057e-06,
+      "learning_rate": 6.905335192482735e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23678248,
+      "step": 30305
+    },
+    {
+      "epoch": 63.014553014553016,
+      "grad_norm": 0.00011027855362044647,
+      "learning_rate": 6.8985622941701275e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23682024,
+      "step": 30310
+    },
+    {
+      "epoch": 63.024948024948024,
+      "grad_norm": 3.84700542781502e-05,
+      "learning_rate": 6.89179218732669e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23686024,
+      "step": 30315
+    },
+    {
+      "epoch": 63.03534303534303,
+      "grad_norm": 8.496391092194244e-05,
+      "learning_rate": 6.8850248729964595e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23689768,
+      "step": 30320
+    },
+    {
+      "epoch": 63.04573804573805,
+      "grad_norm": 1.8579308743937872e-05,
+      "learning_rate": 6.8782603522230314e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23693640,
+      "step": 30325
+    },
+    {
+      "epoch": 63.056133056133056,
+      "grad_norm": 0.00012184533261461183,
+      "learning_rate": 6.871498626049591e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23697512,
+      "step": 30330
+    },
+    {
+      "epoch": 63.066528066528065,
+      "grad_norm": 5.248837624094449e-06,
+      "learning_rate": 6.8647396955188875e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23701416,
+      "step": 30335
+    },
+    {
+      "epoch": 63.07692307692308,
+      "grad_norm": 1.7327110981568694e-06,
+      "learning_rate": 6.857983561673218e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23705256,
+      "step": 30340
+    },
+    {
+      "epoch": 63.08731808731809,
+      "grad_norm": 1.4535941772919614e-05,
+      "learning_rate": 6.851230225554467e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 23709192,
+      "step": 30345
+    },
+    {
+      "epoch": 63.0977130977131,
+      "grad_norm": 0.00015600073675159365,
+      "learning_rate": 6.8444796882040946e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23713032,
+      "step": 30350
+    },
+    {
+      "epoch": 63.108108108108105,
+      "grad_norm": 2.5562371774867643e-06,
+      "learning_rate": 6.837731950663106e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23716936,
+      "step": 30355
+    },
+    {
+      "epoch": 63.11850311850312,
+      "grad_norm": 1.0089744137076195e-05,
+      "learning_rate": 6.830987013972098e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23720872,
+      "step": 30360
+    },
+    {
+      "epoch": 63.12889812889813,
+      "grad_norm": 4.008174073533155e-06,
+      "learning_rate": 6.82424487917121e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23724712,
+      "step": 30365
+    },
+    {
+      "epoch": 63.13929313929314,
+      "grad_norm": 9.446345757169183e-06,
+      "learning_rate": 6.8175055473001735e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23728616,
+      "step": 30370
+    },
+    {
+      "epoch": 63.14968814968815,
+      "grad_norm": 2.376587326580193e-05,
+      "learning_rate": 6.8107690193982855e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23732488,
+      "step": 30375
+    },
+    {
+      "epoch": 63.16008316008316,
+      "grad_norm": 2.644259984663222e-05,
+      "learning_rate": 6.804035296504385e-06,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 23736392,
+      "step": 30380
+    },
+    {
+      "epoch": 63.17047817047817,
+      "grad_norm": 3.681895759655163e-05,
+      "learning_rate": 6.797304379656916e-06,
+      "loss": 0.002,
+      "num_input_tokens_seen": 23740232,
+      "step": 30385
+    },
+    {
+      "epoch": 63.18087318087318,
+      "grad_norm": 0.0005880056414753199,
+      "learning_rate": 6.790576269893861e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23744072,
+      "step": 30390
+    },
+    {
+      "epoch": 63.19126819126819,
+      "grad_norm": 0.0003157756582368165,
+      "learning_rate": 6.783850968252772e-06,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 23748008,
+      "step": 30395
+    },
+    {
+      "epoch": 63.2016632016632,
+      "grad_norm": 2.8930116968695074e-05,
+      "learning_rate": 6.777128475770789e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23751880,
+      "step": 30400
+    },
+    {
+      "epoch": 63.2016632016632,
+      "eval_loss": 0.5874329805374146,
+      "eval_runtime": 11.7187,
+      "eval_samples_per_second": 73.046,
+      "eval_steps_per_second": 18.261,
+      "num_input_tokens_seen": 23751880,
+      "step": 30400
+    },
+    {
+      "epoch": 63.21205821205821,
+      "grad_norm": 0.10658536851406097,
+      "learning_rate": 6.77040879348459e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 23755880,
+      "step": 30405
+    },
+    {
+      "epoch": 63.222453222453225,
+      "grad_norm": 1.609554828974069e-06,
+      "learning_rate": 6.763691922430443e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23759720,
+      "step": 30410
+    },
+    {
+      "epoch": 63.232848232848234,
+      "grad_norm": 8.744869410293177e-05,
+      "learning_rate": 6.756977863644178e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23763688,
+      "step": 30415
+    },
+    {
+      "epoch": 63.24324324324324,
+      "grad_norm": 3.4115673770429567e-06,
+      "learning_rate": 6.7502666181611804e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23767784,
+      "step": 30420
+    },
+    {
+      "epoch": 63.25363825363825,
+      "grad_norm": 4.548783635982545e-06,
+      "learning_rate": 6.743558187016405e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 23771656,
+      "step": 30425
+    },
+    {
+      "epoch": 63.264033264033266,
+      "grad_norm": 4.087205525138415e-06,
+      "learning_rate": 6.7368525712443925e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23775496,
+      "step": 30430
+    },
+    {
+      "epoch": 63.274428274428274,
+      "grad_norm": 3.610890780691989e-05,
+      "learning_rate": 6.7301497718792155e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23779656,
+      "step": 30435
+    },
+    {
+      "epoch": 63.28482328482328,
+      "grad_norm": 1.4738549907633569e-05,
+      "learning_rate": 6.723449789954544e-06,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 23783624,
+      "step": 30440
+    },
+    {
+      "epoch": 63.2952182952183,
+      "grad_norm": 7.993349754542578e-06,
+      "learning_rate": 6.716752626503586e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23787496,
+      "step": 30445
+    },
+    {
+      "epoch": 63.305613305613306,
+      "grad_norm": 4.430190983839566e-06,
+      "learning_rate": 6.710058282559131e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 23791336,
+      "step": 30450
+    },
+    {
+      "epoch": 63.316008316008315,
+      "grad_norm": 5.928745849814732e-06,
+      "learning_rate": 6.703366759153545e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23795304,
+      "step": 30455
+    },
+    {
+      "epoch": 63.32640332640332,
+      "grad_norm": 3.269732360422495e-06,
+      "learning_rate": 6.6966780573187335e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23799112,
+      "step": 30460
+    },
+    {
+      "epoch": 63.33679833679834,
+      "grad_norm": 4.300531873013824e-05,
+      "learning_rate": 6.689992178086174e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23803016,
+      "step": 30465
+    },
+    {
+      "epoch": 63.34719334719335,
+      "grad_norm": 6.4306404965464026e-06,
+      "learning_rate": 6.683309122486925e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23806856,
+      "step": 30470
+    },
+    {
+      "epoch": 63.357588357588355,
+      "grad_norm": 9.711528036859818e-06,
+      "learning_rate": 6.676628891551584e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23810760,
+      "step": 30475
+    },
+    {
+      "epoch": 63.36798336798337,
+      "grad_norm": 0.15057380497455597,
+      "learning_rate": 6.6699514863103385e-06,
+      "loss": 0.0033,
+      "num_input_tokens_seen": 23814632,
+      "step": 30480
+    },
+    {
+      "epoch": 63.37837837837838,
+      "grad_norm": 1.2574548236443661e-05,
+      "learning_rate": 6.663276907792921e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 23818504,
+      "step": 30485
+    },
+    {
+      "epoch": 63.38877338877339,
+      "grad_norm": 1.1383893252059352e-05,
+      "learning_rate": 6.656605157028634e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23822280,
+      "step": 30490
+    },
+    {
+      "epoch": 63.3991683991684,
+      "grad_norm": 2.411856439721305e-05,
+      "learning_rate": 6.649936235046358e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23826184,
+      "step": 30495
+    },
+    {
+      "epoch": 63.40956340956341,
+      "grad_norm": 9.249216418538708e-06,
+      "learning_rate": 6.643270142874508e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23830120,
+      "step": 30500
+    },
+    {
+      "epoch": 63.41995841995842,
+      "grad_norm": 7.924921192170586e-06,
+      "learning_rate": 6.636606881541094e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23833992,
+      "step": 30505
+    },
+    {
+      "epoch": 63.43035343035343,
+      "grad_norm": 2.4360584575333633e-05,
+      "learning_rate": 6.629946452073662e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23837992,
+      "step": 30510
+    },
+    {
+      "epoch": 63.44074844074844,
+      "grad_norm": 1.0985923836415168e-05,
+      "learning_rate": 6.6232888554993375e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23841800,
+      "step": 30515
+    },
+    {
+      "epoch": 63.45114345114345,
+      "grad_norm": 5.839113782712957e-06,
+      "learning_rate": 6.616634092844817e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 23845768,
+      "step": 30520
+    },
+    {
+      "epoch": 63.46153846153846,
+      "grad_norm": 1.164020250143949e-05,
+      "learning_rate": 6.609982165136331e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23849512,
+      "step": 30525
+    },
+    {
+      "epoch": 63.471933471933475,
+      "grad_norm": 4.82806317450013e-05,
+      "learning_rate": 6.603333073399706e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23853416,
+      "step": 30530
+    },
+    {
+      "epoch": 63.482328482328484,
+      "grad_norm": 0.0018969493685290217,
+      "learning_rate": 6.596686818660308e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23857480,
+      "step": 30535
+    },
+    {
+      "epoch": 63.49272349272349,
+      "grad_norm": 7.065432873787358e-05,
+      "learning_rate": 6.590043401943066e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 23861416,
+      "step": 30540
+    },
+    {
+      "epoch": 63.5031185031185,
+      "grad_norm": 8.657410944579169e-05,
+      "learning_rate": 6.583402824272494e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23865224,
+      "step": 30545
+    },
+    {
+      "epoch": 63.513513513513516,
+      "grad_norm": 9.116742148762569e-06,
+      "learning_rate": 6.576765086672634e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23869128,
+      "step": 30550
+    },
+    {
+      "epoch": 63.523908523908524,
+      "grad_norm": 1.27582188724773e-05,
+      "learning_rate": 6.57013019016712e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23873032,
+      "step": 30555
+    },
+    {
+      "epoch": 63.53430353430353,
+      "grad_norm": 4.794592678081244e-05,
+      "learning_rate": 6.563498135779142e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23876936,
+      "step": 30560
+    },
+    {
+      "epoch": 63.54469854469855,
+      "grad_norm": 0.0009808127069845796,
+      "learning_rate": 6.556868924531431e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 23880712,
+      "step": 30565
+    },
+    {
+      "epoch": 63.555093555093556,
+      "grad_norm": 9.115913599089254e-06,
+      "learning_rate": 6.550242557446304e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23884584,
+      "step": 30570
+    },
+    {
+      "epoch": 63.565488565488565,
+      "grad_norm": 1.8464668301021447e-06,
+      "learning_rate": 6.543619035545634e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23888456,
+      "step": 30575
+    },
+    {
+      "epoch": 63.57588357588357,
+      "grad_norm": 2.0323054741311353e-06,
+      "learning_rate": 6.53699835985084e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23892328,
+      "step": 30580
+    },
+    {
+      "epoch": 63.58627858627859,
+      "grad_norm": 3.6734985769726336e-05,
+      "learning_rate": 6.530380531382927e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23896104,
+      "step": 30585
+    },
+    {
+      "epoch": 63.5966735966736,
+      "grad_norm": 8.958414241533319e-07,
+      "learning_rate": 6.523765551162433e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23899944,
+      "step": 30590
+    },
+    {
+      "epoch": 63.607068607068605,
+      "grad_norm": 0.000916124670766294,
+      "learning_rate": 6.517153420209476e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 23903752,
+      "step": 30595
+    },
+    {
+      "epoch": 63.61746361746362,
+      "grad_norm": 2.7128326109959744e-05,
+      "learning_rate": 6.510544139543739e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23907624,
+      "step": 30600
+    },
+    {
+      "epoch": 63.61746361746362,
+      "eval_loss": 0.5837193131446838,
+      "eval_runtime": 11.7033,
+      "eval_samples_per_second": 73.142,
+      "eval_steps_per_second": 18.285,
+      "num_input_tokens_seen": 23907624,
+      "step": 30600
+    },
+    {
+      "epoch": 63.62785862785863,
+      "grad_norm": 3.514997297315858e-05,
+      "learning_rate": 6.503937710184452e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23911528,
+      "step": 30605
+    },
+    {
+      "epoch": 63.63825363825364,
+      "grad_norm": 9.304919331043493e-06,
+      "learning_rate": 6.4973341331503954e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23915400,
+      "step": 30610
+    },
+    {
+      "epoch": 63.648648648648646,
+      "grad_norm": 0.00014560746785718948,
+      "learning_rate": 6.490733409459942e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23919304,
+      "step": 30615
+    },
+    {
+      "epoch": 63.65904365904366,
+      "grad_norm": 0.12596996128559113,
+      "learning_rate": 6.484135540130995e-06,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 23923144,
+      "step": 30620
+    },
+    {
+      "epoch": 63.66943866943867,
+      "grad_norm": 1.905808130686637e-05,
+      "learning_rate": 6.4775405261810364e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23927048,
+      "step": 30625
+    },
+    {
+      "epoch": 63.67983367983368,
+      "grad_norm": 1.5399837138829753e-06,
+      "learning_rate": 6.470948368627092e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23930920,
+      "step": 30630
+    },
+    {
+      "epoch": 63.69022869022869,
+      "grad_norm": 0.00024184626818168908,
+      "learning_rate": 6.464359068485756e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23934824,
+      "step": 30635
+    },
+    {
+      "epoch": 63.7006237006237,
+      "grad_norm": 0.0007961113005876541,
+      "learning_rate": 6.457772626773195e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23938728,
+      "step": 30640
+    },
+    {
+      "epoch": 63.71101871101871,
+      "grad_norm": 4.690695277531631e-06,
+      "learning_rate": 6.451189044505104e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23942536,
+      "step": 30645
+    },
+    {
+      "epoch": 63.72141372141372,
+      "grad_norm": 1.74071919900598e-05,
+      "learning_rate": 6.44460832269676e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23946472,
+      "step": 30650
+    },
+    {
+      "epoch": 63.731808731808734,
+      "grad_norm": 0.00037397159030660987,
+      "learning_rate": 6.438030462363001e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23950408,
+      "step": 30655
+    },
+    {
+      "epoch": 63.74220374220374,
+      "grad_norm": 0.00011505904694786295,
+      "learning_rate": 6.431455464518205e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23954376,
+      "step": 30660
+    },
+    {
+      "epoch": 63.75259875259875,
+      "grad_norm": 0.00029707630164921284,
+      "learning_rate": 6.424883330176326e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23958184,
+      "step": 30665
+    },
+    {
+      "epoch": 63.762993762993766,
+      "grad_norm": 3.752880365937017e-05,
+      "learning_rate": 6.418314060350864e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23962056,
+      "step": 30670
+    },
+    {
+      "epoch": 63.773388773388774,
+      "grad_norm": 6.487079372163862e-05,
+      "learning_rate": 6.4117476560548895e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23965800,
+      "step": 30675
+    },
+    {
+      "epoch": 63.78378378378378,
+      "grad_norm": 1.0186712415816146e-06,
+      "learning_rate": 6.405184118301016e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23969832,
+      "step": 30680
+    },
+    {
+      "epoch": 63.79417879417879,
+      "grad_norm": 5.945029261056334e-05,
+      "learning_rate": 6.398623448101434e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23973928,
+      "step": 30685
+    },
+    {
+      "epoch": 63.804573804573806,
+      "grad_norm": 8.241846080636606e-05,
+      "learning_rate": 6.392065646467871e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23977896,
+      "step": 30690
+    },
+    {
+      "epoch": 63.814968814968815,
+      "grad_norm": 0.00019303400767967105,
+      "learning_rate": 6.385510714411632e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23981800,
+      "step": 30695
+    },
+    {
+      "epoch": 63.82536382536382,
+      "grad_norm": 1.7235439372598194e-05,
+      "learning_rate": 6.378958652943559e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23985736,
+      "step": 30700
+    },
+    {
+      "epoch": 63.83575883575884,
+      "grad_norm": 1.232449085364351e-05,
+      "learning_rate": 6.3724094630740776e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23989896,
+      "step": 30705
+    },
+    {
+      "epoch": 63.84615384615385,
+      "grad_norm": 0.00027765275444835424,
+      "learning_rate": 6.365863145813136e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 23993864,
+      "step": 30710
+    },
+    {
+      "epoch": 63.856548856548855,
+      "grad_norm": 0.17371265590190887,
+      "learning_rate": 6.359319702170269e-06,
+      "loss": 0.0041,
+      "num_input_tokens_seen": 23997736,
+      "step": 30715
+    },
+    {
+      "epoch": 63.86694386694387,
+      "grad_norm": 7.042263860057574e-06,
+      "learning_rate": 6.352779133154566e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24001736,
+      "step": 30720
+    },
+    {
+      "epoch": 63.87733887733888,
+      "grad_norm": 0.00015231822908390313,
+      "learning_rate": 6.346241439774648e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24005576,
+      "step": 30725
+    },
+    {
+      "epoch": 63.88773388773389,
+      "grad_norm": 3.2435928005725145e-06,
+      "learning_rate": 6.339706623038716e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 24009480,
+      "step": 30730
+    },
+    {
+      "epoch": 63.898128898128896,
+      "grad_norm": 1.6822190445964225e-05,
+      "learning_rate": 6.333174683954532e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24013416,
+      "step": 30735
+    },
+    {
+      "epoch": 63.90852390852391,
+      "grad_norm": 0.0002265327057102695,
+      "learning_rate": 6.326645623529387e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24017224,
+      "step": 30740
+    },
+    {
+      "epoch": 63.91891891891892,
+      "grad_norm": 2.3316317765420536e-06,
+      "learning_rate": 6.320119442770156e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24021224,
+      "step": 30745
+    },
+    {
+      "epoch": 63.92931392931393,
+      "grad_norm": 0.00043116844608448446,
+      "learning_rate": 6.313596142683254e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24025032,
+      "step": 30750
+    },
+    {
+      "epoch": 63.93970893970894,
+      "grad_norm": 0.12097246944904327,
+      "learning_rate": 6.307075724274647e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 24028872,
+      "step": 30755
+    },
+    {
+      "epoch": 63.95010395010395,
+      "grad_norm": 9.577967148288735e-07,
+      "learning_rate": 6.300558188549882e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24032680,
+      "step": 30760
+    },
+    {
+      "epoch": 63.96049896049896,
+      "grad_norm": 1.3395200767263304e-05,
+      "learning_rate": 6.29404353651403e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24036424,
+      "step": 30765
+    },
+    {
+      "epoch": 63.97089397089397,
+      "grad_norm": 1.765196770975308e-06,
+      "learning_rate": 6.287531769171737e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24040232,
+      "step": 30770
+    },
+    {
+      "epoch": 63.981288981288984,
+      "grad_norm": 2.8874806048406754e-06,
+      "learning_rate": 6.2810228875272045e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24044200,
+      "step": 30775
+    },
+    {
+      "epoch": 63.99168399168399,
+      "grad_norm": 6.246028078749077e-06,
+      "learning_rate": 6.274516892584179e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24048136,
+      "step": 30780
+    },
+    {
+      "epoch": 64.002079002079,
+      "grad_norm": 9.97140432446031e-07,
+      "learning_rate": 6.268013785345969e-06,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 24052152,
+      "step": 30785
+    },
+    {
+      "epoch": 64.01247401247402,
+      "grad_norm": 7.519641076214612e-05,
+      "learning_rate": 6.26151356681543e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24056152,
+      "step": 30790
+    },
+    {
+      "epoch": 64.02286902286902,
+      "grad_norm": 1.335685283265775e-06,
+      "learning_rate": 6.255016237994981e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24060024,
+      "step": 30795
+    },
+    {
+      "epoch": 64.03326403326403,
+      "grad_norm": 4.856049599766266e-06,
+      "learning_rate": 6.248521799886603e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24063864,
+      "step": 30800
+    },
+    {
+      "epoch": 64.03326403326403,
+      "eval_loss": 0.582709789276123,
+      "eval_runtime": 11.7134,
+      "eval_samples_per_second": 73.079,
+      "eval_steps_per_second": 18.27,
+      "num_input_tokens_seen": 24063864,
+      "step": 30800
+    },
+    {
+      "epoch": 64.04365904365905,
+      "grad_norm": 0.00010171454778173938,
+      "learning_rate": 6.242030253491798e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24067704,
+      "step": 30805
+    },
+    {
+      "epoch": 64.05405405405405,
+      "grad_norm": 6.909145304234698e-05,
+      "learning_rate": 6.235541599811656e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24071672,
+      "step": 30810
+    },
+    {
+      "epoch": 64.06444906444906,
+      "grad_norm": 1.1429352525738068e-05,
+      "learning_rate": 6.229055839846814e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24075608,
+      "step": 30815
+    },
+    {
+      "epoch": 64.07484407484408,
+      "grad_norm": 5.113497900310904e-05,
+      "learning_rate": 6.222572974597455e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24079352,
+      "step": 30820
+    },
+    {
+      "epoch": 64.08523908523908,
+      "grad_norm": 1.951649892362184e-06,
+      "learning_rate": 6.216093005063306e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24083128,
+      "step": 30825
+    },
+    {
+      "epoch": 64.0956340956341,
+      "grad_norm": 3.830623973044567e-05,
+      "learning_rate": 6.209615932243678e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24087032,
+      "step": 30830
+    },
+    {
+      "epoch": 64.10602910602911,
+      "grad_norm": 4.797531346412143e-06,
+      "learning_rate": 6.203141757137399e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24091032,
+      "step": 30835
+    },
+    {
+      "epoch": 64.11642411642411,
+      "grad_norm": 7.379437079180207e-07,
+      "learning_rate": 6.196670480742886e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24095032,
+      "step": 30840
+    },
+    {
+      "epoch": 64.12681912681913,
+      "grad_norm": 4.039702616864815e-06,
+      "learning_rate": 6.190202104058074e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 24098744,
+      "step": 30845
+    },
+    {
+      "epoch": 64.13721413721414,
+      "grad_norm": 1.238152526639169e-05,
+      "learning_rate": 6.183736628080475e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24102648,
+      "step": 30850
+    },
+    {
+      "epoch": 64.14760914760915,
+      "grad_norm": 2.8667420338024385e-05,
+      "learning_rate": 6.177274053807155e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24106520,
+      "step": 30855
+    },
+    {
+      "epoch": 64.15800415800416,
+      "grad_norm": 2.2685057956550736e-06,
+      "learning_rate": 6.170814382234713e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24110360,
+      "step": 30860
+    },
+    {
+      "epoch": 64.16839916839916,
+      "grad_norm": 1.6798798242234625e-05,
+      "learning_rate": 6.16435761435932e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24114456,
+      "step": 30865
+    },
+    {
+      "epoch": 64.17879417879418,
+      "grad_norm": 1.0831286090251524e-06,
+      "learning_rate": 6.157903751176681e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 24118424,
+      "step": 30870
+    },
+    {
+      "epoch": 64.1891891891892,
+      "grad_norm": 8.725719453650527e-06,
+      "learning_rate": 6.151452793682066e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24122264,
+      "step": 30875
+    },
+    {
+      "epoch": 64.1995841995842,
+      "grad_norm": 1.969276308955159e-05,
+      "learning_rate": 6.145004742870305e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24126072,
+      "step": 30880
+    },
+    {
+      "epoch": 64.20997920997921,
+      "grad_norm": 2.445184463795158e-06,
+      "learning_rate": 6.138559599735752e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 24129912,
+      "step": 30885
+    },
+    {
+      "epoch": 64.22037422037423,
+      "grad_norm": 9.822421134231263e-07,
+      "learning_rate": 6.132117365272344e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24133848,
+      "step": 30890
+    },
+    {
+      "epoch": 64.23076923076923,
+      "grad_norm": 2.8181117158965208e-05,
+      "learning_rate": 6.125678040473545e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24137720,
+      "step": 30895
+    },
+    {
+      "epoch": 64.24116424116424,
+      "grad_norm": 1.105163937609177e-05,
+      "learning_rate": 6.1192416263323755e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24141624,
+      "step": 30900
+    },
+    {
+      "epoch": 64.25155925155926,
+      "grad_norm": 0.000199042129679583,
+      "learning_rate": 6.112808123841424e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24145432,
+      "step": 30905
+    },
+    {
+      "epoch": 64.26195426195426,
+      "grad_norm": 1.7030981325660832e-05,
+      "learning_rate": 6.106377533992805e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24149304,
+      "step": 30910
+    },
+    {
+      "epoch": 64.27234927234927,
+      "grad_norm": 4.560704473988153e-05,
+      "learning_rate": 6.099949857778204e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24153016,
+      "step": 30915
+    },
+    {
+      "epoch": 64.28274428274429,
+      "grad_norm": 7.5500793172977865e-06,
+      "learning_rate": 6.093525096188852e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24156760,
+      "step": 30920
+    },
+    {
+      "epoch": 64.29313929313929,
+      "grad_norm": 0.14440639317035675,
+      "learning_rate": 6.087103250215518e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 24160760,
+      "step": 30925
+    },
+    {
+      "epoch": 64.3035343035343,
+      "grad_norm": 0.05841416120529175,
+      "learning_rate": 6.080684320848537e-06,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 24164760,
+      "step": 30930
+    },
+    {
+      "epoch": 64.31392931392931,
+      "grad_norm": 3.129672768409364e-05,
+      "learning_rate": 6.074268309077794e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24168696,
+      "step": 30935
+    },
+    {
+      "epoch": 64.32432432432432,
+      "grad_norm": 3.417518382775597e-05,
+      "learning_rate": 6.067855215892709e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24172472,
+      "step": 30940
+    },
+    {
+      "epoch": 64.33471933471934,
+      "grad_norm": 0.1311100572347641,
+      "learning_rate": 6.061445042282271e-06,
+      "loss": 0.0078,
+      "num_input_tokens_seen": 24176312,
+      "step": 30945
+    },
+    {
+      "epoch": 64.34511434511434,
+      "grad_norm": 0.0001383092167088762,
+      "learning_rate": 6.055037789234999e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24180344,
+      "step": 30950
+    },
+    {
+      "epoch": 64.35550935550935,
+      "grad_norm": 4.708332198788412e-06,
+      "learning_rate": 6.048633457738975e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24184312,
+      "step": 30955
+    },
+    {
+      "epoch": 64.36590436590437,
+      "grad_norm": 0.14447589218616486,
+      "learning_rate": 6.042232048781837e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 24188312,
+      "step": 30960
+    },
+    {
+      "epoch": 64.37629937629937,
+      "grad_norm": 3.219889322281233e-06,
+      "learning_rate": 6.035833563350757e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24192184,
+      "step": 30965
+    },
+    {
+      "epoch": 64.38669438669439,
+      "grad_norm": 1.0027822554548038e-06,
+      "learning_rate": 6.0294380024324525e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24196024,
+      "step": 30970
+    },
+    {
+      "epoch": 64.3970893970894,
+      "grad_norm": 0.13949961960315704,
+      "learning_rate": 6.023045367013213e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 24200120,
+      "step": 30975
+    },
+    {
+      "epoch": 64.4074844074844,
+      "grad_norm": 0.0001434675941709429,
+      "learning_rate": 6.016655658078851e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24204024,
+      "step": 30980
+    },
+    {
+      "epoch": 64.41787941787942,
+      "grad_norm": 1.7261203538510017e-05,
+      "learning_rate": 6.010268876614753e-06,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 24207960,
+      "step": 30985
+    },
+    {
+      "epoch": 64.42827442827443,
+      "grad_norm": 9.645678801462054e-06,
+      "learning_rate": 6.0038850236058266e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24211832,
+      "step": 30990
+    },
+    {
+      "epoch": 64.43866943866944,
+      "grad_norm": 2.405224449830712e-06,
+      "learning_rate": 5.997504100036549e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24215800,
+      "step": 30995
+    },
+    {
+      "epoch": 64.44906444906445,
+      "grad_norm": 0.00014471961185336113,
+      "learning_rate": 5.991126106890949e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24219608,
+      "step": 31000
+    },
+    {
+      "epoch": 64.44906444906445,
+      "eval_loss": 0.5825332403182983,
+      "eval_runtime": 11.6992,
+      "eval_samples_per_second": 73.168,
+      "eval_steps_per_second": 18.292,
+      "num_input_tokens_seen": 24219608,
+      "step": 31000
+    },
+    {
+      "epoch": 64.45945945945945,
+      "grad_norm": 0.00018954541883431375,
+      "learning_rate": 5.984751045152576e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24223640,
+      "step": 31005
+    },
+    {
+      "epoch": 64.46985446985447,
+      "grad_norm": 4.850691766478121e-06,
+      "learning_rate": 5.978378915804553e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24227576,
+      "step": 31010
+    },
+    {
+      "epoch": 64.48024948024948,
+      "grad_norm": 3.0910105124348775e-05,
+      "learning_rate": 5.972009719829547e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24231416,
+      "step": 31015
+    },
+    {
+      "epoch": 64.49064449064448,
+      "grad_norm": 5.543435236177174e-06,
+      "learning_rate": 5.965643458209755e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24235256,
+      "step": 31020
+    },
+    {
+      "epoch": 64.5010395010395,
+      "grad_norm": 2.3110926122171804e-05,
+      "learning_rate": 5.95928013192695e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24239320,
+      "step": 31025
+    },
+    {
+      "epoch": 64.51143451143452,
+      "grad_norm": 1.1385382094886154e-06,
+      "learning_rate": 5.952919741962423e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24243160,
+      "step": 31030
+    },
+    {
+      "epoch": 64.52182952182952,
+      "grad_norm": 1.8467771951691248e-06,
+      "learning_rate": 5.946562289297042e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24247096,
+      "step": 31035
+    },
+    {
+      "epoch": 64.53222453222453,
+      "grad_norm": 1.5164510841714218e-05,
+      "learning_rate": 5.9402077749111855e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24251128,
+      "step": 31040
+    },
+    {
+      "epoch": 64.54261954261955,
+      "grad_norm": 1.1295214790152386e-06,
+      "learning_rate": 5.933856199784821e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24255064,
+      "step": 31045
+    },
+    {
+      "epoch": 64.55301455301455,
+      "grad_norm": 7.485194510081783e-05,
+      "learning_rate": 5.927507564897419e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24259128,
+      "step": 31050
+    },
+    {
+      "epoch": 64.56340956340956,
+      "grad_norm": 9.241379643754044e-07,
+      "learning_rate": 5.9211618712280395e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24263064,
+      "step": 31055
+    },
+    {
+      "epoch": 64.57380457380458,
+      "grad_norm": 3.201506842742674e-05,
+      "learning_rate": 5.914819119755255e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24266904,
+      "step": 31060
+    },
+    {
+      "epoch": 64.58419958419958,
+      "grad_norm": 1.4724276297783945e-05,
+      "learning_rate": 5.908479311457205e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24270904,
+      "step": 31065
+    },
+    {
+      "epoch": 64.5945945945946,
+      "grad_norm": 1.0078788363898639e-05,
+      "learning_rate": 5.902142447311559e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24274776,
+      "step": 31070
+    },
+    {
+      "epoch": 64.60498960498961,
+      "grad_norm": 6.266951913858065e-06,
+      "learning_rate": 5.895808528295546e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 24278744,
+      "step": 31075
+    },
+    {
+      "epoch": 64.61538461538461,
+      "grad_norm": 6.911760692673852e-07,
+      "learning_rate": 5.889477555385941e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24282776,
+      "step": 31080
+    },
+    {
+      "epoch": 64.62577962577963,
+      "grad_norm": 5.117149157740641e-06,
+      "learning_rate": 5.883149529559051e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24286744,
+      "step": 31085
+    },
+    {
+      "epoch": 64.63617463617463,
+      "grad_norm": 2.6247604182572104e-05,
+      "learning_rate": 5.876824451790738e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24290584,
+      "step": 31090
+    },
+    {
+      "epoch": 64.64656964656965,
+      "grad_norm": 7.963184907566756e-05,
+      "learning_rate": 5.87050232305642e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24294424,
+      "step": 31095
+    },
+    {
+      "epoch": 64.65696465696466,
+      "grad_norm": 1.2258786227903329e-05,
+      "learning_rate": 5.864183144331034e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24298456,
+      "step": 31100
+    },
+    {
+      "epoch": 64.66735966735966,
+      "grad_norm": 3.6113938222115394e-06,
+      "learning_rate": 5.857866916589089e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24302328,
+      "step": 31105
+    },
+    {
+      "epoch": 64.67775467775468,
+      "grad_norm": 3.856508919852786e-05,
+      "learning_rate": 5.8515536408046216e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24306296,
+      "step": 31110
+    },
+    {
+      "epoch": 64.6881496881497,
+      "grad_norm": 1.0626499715726823e-06,
+      "learning_rate": 5.845243317951208e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 24310328,
+      "step": 31115
+    },
+    {
+      "epoch": 64.6985446985447,
+      "grad_norm": 0.11612988263368607,
+      "learning_rate": 5.838935949001997e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 24314168,
+      "step": 31120
+    },
+    {
+      "epoch": 64.70893970893971,
+      "grad_norm": 6.028707502991892e-06,
+      "learning_rate": 5.8326315349296476e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24318200,
+      "step": 31125
+    },
+    {
+      "epoch": 64.71933471933473,
+      "grad_norm": 4.080422968399944e-06,
+      "learning_rate": 5.826330076706396e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 24322200,
+      "step": 31130
+    },
+    {
+      "epoch": 64.72972972972973,
+      "grad_norm": 2.858098469005199e-06,
+      "learning_rate": 5.820031575303988e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24326168,
+      "step": 31135
+    },
+    {
+      "epoch": 64.74012474012474,
+      "grad_norm": 0.00022922684729564935,
+      "learning_rate": 5.813736031693745e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 24330008,
+      "step": 31140
+    },
+    {
+      "epoch": 64.75051975051976,
+      "grad_norm": 0.13254927098751068,
+      "learning_rate": 5.807443446846522e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 24333848,
+      "step": 31145
+    },
+    {
+      "epoch": 64.76091476091476,
+      "grad_norm": 3.4717354537860956e-06,
+      "learning_rate": 5.801153821732699e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24337656,
+      "step": 31150
+    },
+    {
+      "epoch": 64.77130977130977,
+      "grad_norm": 8.009662451513577e-06,
+      "learning_rate": 5.794867157322229e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24341528,
+      "step": 31155
+    },
+    {
+      "epoch": 64.78170478170478,
+      "grad_norm": 8.264402822533157e-06,
+      "learning_rate": 5.788583454584593e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24345560,
+      "step": 31160
+    },
+    {
+      "epoch": 64.79209979209979,
+      "grad_norm": 5.183203938941006e-06,
+      "learning_rate": 5.7823027144888075e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 24349400,
+      "step": 31165
+    },
+    {
+      "epoch": 64.8024948024948,
+      "grad_norm": 1.597879054315854e-05,
+      "learning_rate": 5.776024938003455e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24353176,
+      "step": 31170
+    },
+    {
+      "epoch": 64.81288981288981,
+      "grad_norm": 2.4777789803920314e-05,
+      "learning_rate": 5.7697501260966345e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24357080,
+      "step": 31175
+    },
+    {
+      "epoch": 64.82328482328482,
+      "grad_norm": 3.0160256301314803e-06,
+      "learning_rate": 5.7634782797360145e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24361048,
+      "step": 31180
+    },
+    {
+      "epoch": 64.83367983367984,
+      "grad_norm": 3.077289875363931e-05,
+      "learning_rate": 5.757209399888777e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24364984,
+      "step": 31185
+    },
+    {
+      "epoch": 64.84407484407484,
+      "grad_norm": 7.94328298070468e-06,
+      "learning_rate": 5.750943487521679e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24368920,
+      "step": 31190
+    },
+    {
+      "epoch": 64.85446985446985,
+      "grad_norm": 9.565066648065113e-06,
+      "learning_rate": 5.744680543600986e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24372856,
+      "step": 31195
+    },
+    {
+      "epoch": 64.86486486486487,
+      "grad_norm": 3.913967884727754e-05,
+      "learning_rate": 5.738420569092537e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24376856,
+      "step": 31200
+    },
+    {
+      "epoch": 64.86486486486487,
+      "eval_loss": 0.5768327713012695,
+      "eval_runtime": 11.7072,
+      "eval_samples_per_second": 73.117,
+      "eval_steps_per_second": 18.279,
+      "num_input_tokens_seen": 24376856,
+      "step": 31200
+    },
+    {
+      "epoch": 64.87525987525987,
+      "grad_norm": 6.617599410674302e-06,
+      "learning_rate": 5.732163564961684e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24380792,
+      "step": 31205
+    },
+    {
+      "epoch": 64.88565488565489,
+      "grad_norm": 7.56062800064683e-05,
+      "learning_rate": 5.725909532173354e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24384664,
+      "step": 31210
+    },
+    {
+      "epoch": 64.8960498960499,
+      "grad_norm": 2.189466977142729e-05,
+      "learning_rate": 5.719658471691977e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24388664,
+      "step": 31215
+    },
+    {
+      "epoch": 64.9064449064449,
+      "grad_norm": 9.4311440079764e-07,
+      "learning_rate": 5.71341038448156e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24392632,
+      "step": 31220
+    },
+    {
+      "epoch": 64.91683991683992,
+      "grad_norm": 8.630751835880801e-06,
+      "learning_rate": 5.707165271505635e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24396440,
+      "step": 31225
+    },
+    {
+      "epoch": 64.92723492723492,
+      "grad_norm": 0.0002267190720885992,
+      "learning_rate": 5.700923133727271e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24400248,
+      "step": 31230
+    },
+    {
+      "epoch": 64.93762993762994,
+      "grad_norm": 2.429306960038957e-06,
+      "learning_rate": 5.694683972109083e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24404248,
+      "step": 31235
+    },
+    {
+      "epoch": 64.94802494802495,
+      "grad_norm": 5.686976692231838e-06,
+      "learning_rate": 5.688447787613241e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24408120,
+      "step": 31240
+    },
+    {
+      "epoch": 64.95841995841995,
+      "grad_norm": 0.0001661209826124832,
+      "learning_rate": 5.6822145812014285e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24411896,
+      "step": 31245
+    },
+    {
+      "epoch": 64.96881496881497,
+      "grad_norm": 3.3624084608163685e-05,
+      "learning_rate": 5.675984353834896e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24415768,
+      "step": 31250
+    },
+    {
+      "epoch": 64.97920997920998,
+      "grad_norm": 4.1156949919241015e-06,
+      "learning_rate": 5.66975710647441e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24419768,
+      "step": 31255
+    },
+    {
+      "epoch": 64.98960498960498,
+      "grad_norm": 1.2354919817880727e-05,
+      "learning_rate": 5.663532840080304e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24423768,
+      "step": 31260
+    },
+    {
+      "epoch": 65.0,
+      "grad_norm": 4.186566002317704e-06,
+      "learning_rate": 5.6573115556124325e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24427656,
+      "step": 31265
+    },
+    {
+      "epoch": 65.01039501039502,
+      "grad_norm": 2.6940957468468696e-05,
+      "learning_rate": 5.651093254030185e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24431720,
+      "step": 31270
+    },
+    {
+      "epoch": 65.02079002079002,
+      "grad_norm": 0.12981854379177094,
+      "learning_rate": 5.644877936292514e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 24435656,
+      "step": 31275
+    },
+    {
+      "epoch": 65.03118503118503,
+      "grad_norm": 7.637820090167224e-06,
+      "learning_rate": 5.638665603357901e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24439624,
+      "step": 31280
+    },
+    {
+      "epoch": 65.04158004158005,
+      "grad_norm": 3.719209871633211e-06,
+      "learning_rate": 5.632456256184357e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24443528,
+      "step": 31285
+    },
+    {
+      "epoch": 65.05197505197505,
+      "grad_norm": 1.2314434570726007e-05,
+      "learning_rate": 5.626249895729452e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24447400,
+      "step": 31290
+    },
+    {
+      "epoch": 65.06237006237006,
+      "grad_norm": 0.00018452816584613174,
+      "learning_rate": 5.620046522950273e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24451272,
+      "step": 31295
+    },
+    {
+      "epoch": 65.07276507276508,
+      "grad_norm": 1.0551651712376042e-06,
+      "learning_rate": 5.613846138803464e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24455048,
+      "step": 31300
+    },
+    {
+      "epoch": 65.08316008316008,
+      "grad_norm": 1.216405780724017e-05,
+      "learning_rate": 5.607648744245206e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24458984,
+      "step": 31305
+    },
+    {
+      "epoch": 65.0935550935551,
+      "grad_norm": 3.5828245017910376e-05,
+      "learning_rate": 5.601454340231207e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24462888,
+      "step": 31310
+    },
+    {
+      "epoch": 65.1039501039501,
+      "grad_norm": 0.0002973927475977689,
+      "learning_rate": 5.595262927716724e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24466920,
+      "step": 31315
+    },
+    {
+      "epoch": 65.11434511434511,
+      "grad_norm": 0.15407633781433105,
+      "learning_rate": 5.589074507656561e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 24470888,
+      "step": 31320
+    },
+    {
+      "epoch": 65.12474012474013,
+      "grad_norm": 1.5461766452062875e-05,
+      "learning_rate": 5.582889081005044e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24474920,
+      "step": 31325
+    },
+    {
+      "epoch": 65.13513513513513,
+      "grad_norm": 8.36864419397898e-05,
+      "learning_rate": 5.5767066487160316e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24478760,
+      "step": 31330
+    },
+    {
+      "epoch": 65.14553014553015,
+      "grad_norm": 1.8023034726866172e-06,
+      "learning_rate": 5.570527211742949e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24482632,
+      "step": 31335
+    },
+    {
+      "epoch": 65.15592515592516,
+      "grad_norm": 1.0238204595225397e-06,
+      "learning_rate": 5.564350771038731e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24486504,
+      "step": 31340
+    },
+    {
+      "epoch": 65.16632016632016,
+      "grad_norm": 3.106670419583679e-06,
+      "learning_rate": 5.558177327555875e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24490472,
+      "step": 31345
+    },
+    {
+      "epoch": 65.17671517671518,
+      "grad_norm": 4.411116606206633e-06,
+      "learning_rate": 5.552006882246388e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24494472,
+      "step": 31350
+    },
+    {
+      "epoch": 65.18711018711019,
+      "grad_norm": 0.00018531776731833816,
+      "learning_rate": 5.545839436061839e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24498440,
+      "step": 31355
+    },
+    {
+      "epoch": 65.1975051975052,
+      "grad_norm": 0.0001100105801015161,
+      "learning_rate": 5.539674989953331e-06,
+      "loss": 0.0057,
+      "num_input_tokens_seen": 24502312,
+      "step": 31360
+    },
+    {
+      "epoch": 65.20790020790021,
+      "grad_norm": 3.815427589870524e-06,
+      "learning_rate": 5.533513544871488e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24506248,
+      "step": 31365
+    },
+    {
+      "epoch": 65.21829521829522,
+      "grad_norm": 1.0196107723459136e-05,
+      "learning_rate": 5.527355101766493e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24510120,
+      "step": 31370
+    },
+    {
+      "epoch": 65.22869022869023,
+      "grad_norm": 0.00010258357360726222,
+      "learning_rate": 5.521199661588044e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24514056,
+      "step": 31375
+    },
+    {
+      "epoch": 65.23908523908524,
+      "grad_norm": 3.042109938178328e-06,
+      "learning_rate": 5.5150472252853944e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24517800,
+      "step": 31380
+    },
+    {
+      "epoch": 65.24948024948024,
+      "grad_norm": 6.257212226046249e-05,
+      "learning_rate": 5.50889779380733e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24521608,
+      "step": 31385
+    },
+    {
+      "epoch": 65.25987525987526,
+      "grad_norm": 1.2776503126588068e-06,
+      "learning_rate": 5.5027513681021605e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24525448,
+      "step": 31390
+    },
+    {
+      "epoch": 65.27027027027027,
+      "grad_norm": 0.00021610093244817108,
+      "learning_rate": 5.4966079491177545e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24529384,
+      "step": 31395
+    },
+    {
+      "epoch": 65.28066528066527,
+      "grad_norm": 1.0501783435756806e-05,
+      "learning_rate": 5.490467537801491e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24533352,
+      "step": 31400
+    },
+    {
+      "epoch": 65.28066528066527,
+      "eval_loss": 0.576035737991333,
+      "eval_runtime": 11.7117,
+      "eval_samples_per_second": 73.089,
+      "eval_steps_per_second": 18.272,
+      "num_input_tokens_seen": 24533352,
+      "step": 31400
+    },
+    {
+      "epoch": 65.29106029106029,
+      "grad_norm": 3.5591936011769576e-06,
+      "learning_rate": 5.484330135100313e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24537224,
+      "step": 31405
+    },
+    {
+      "epoch": 65.3014553014553,
+      "grad_norm": 6.974882580834674e-06,
+      "learning_rate": 5.4781957419606785e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 24541032,
+      "step": 31410
+    },
+    {
+      "epoch": 65.3118503118503,
+      "grad_norm": 0.003225161461159587,
+      "learning_rate": 5.472064359328577e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24544840,
+      "step": 31415
+    },
+    {
+      "epoch": 65.32224532224532,
+      "grad_norm": 1.513460483693052e-05,
+      "learning_rate": 5.4659359881495565e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24548616,
+      "step": 31420
+    },
+    {
+      "epoch": 65.33264033264034,
+      "grad_norm": 1.6632330925858696e-06,
+      "learning_rate": 5.4598106293686916e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24552328,
+      "step": 31425
+    },
+    {
+      "epoch": 65.34303534303534,
+      "grad_norm": 1.0216986083833035e-05,
+      "learning_rate": 5.45368828393058e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24556232,
+      "step": 31430
+    },
+    {
+      "epoch": 65.35343035343035,
+      "grad_norm": 8.739265467738733e-06,
+      "learning_rate": 5.44756895277937e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 24560168,
+      "step": 31435
+    },
+    {
+      "epoch": 65.36382536382537,
+      "grad_norm": 1.574273301230278e-05,
+      "learning_rate": 5.441452636858746e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 24563944,
+      "step": 31440
+    },
+    {
+      "epoch": 65.37422037422037,
+      "grad_norm": 2.1223220755928196e-05,
+      "learning_rate": 5.435339337111905e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 24567848,
+      "step": 31445
+    },
+    {
+      "epoch": 65.38461538461539,
+      "grad_norm": 4.675492618844146e-06,
+      "learning_rate": 5.42922905448161e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24571688,
+      "step": 31450
+    },
+    {
+      "epoch": 65.39501039501039,
+      "grad_norm": 0.00011736318992916495,
+      "learning_rate": 5.423121789910129e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24575528,
+      "step": 31455
+    },
+    {
+      "epoch": 65.4054054054054,
+      "grad_norm": 5.36227562406566e-05,
+      "learning_rate": 5.417017544339287e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24579304,
+      "step": 31460
+    },
+    {
+      "epoch": 65.41580041580042,
+      "grad_norm": 2.6806403639056953e-06,
+      "learning_rate": 5.410916318710443e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24583208,
+      "step": 31465
+    },
+    {
+      "epoch": 65.42619542619542,
+      "grad_norm": 2.8659380859608063e-06,
+      "learning_rate": 5.404818113964466e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 24587016,
+      "step": 31470
+    },
+    {
+      "epoch": 65.43659043659044,
+      "grad_norm": 2.6597244868753478e-05,
+      "learning_rate": 5.398722931041792e-06,
+      "loss": 0.0061,
+      "num_input_tokens_seen": 24590792,
+      "step": 31475
+    },
+    {
+      "epoch": 65.44698544698545,
+      "grad_norm": 1.344968313787831e-05,
+      "learning_rate": 5.392630770882367e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24594760,
+      "step": 31480
+    },
+    {
+      "epoch": 65.45738045738045,
+      "grad_norm": 4.342289230407914e-06,
+      "learning_rate": 5.3865416344256705e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24598760,
+      "step": 31485
+    },
+    {
+      "epoch": 65.46777546777547,
+      "grad_norm": 4.069660008099163e-06,
+      "learning_rate": 5.380455522610742e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24602760,
+      "step": 31490
+    },
+    {
+      "epoch": 65.47817047817048,
+      "grad_norm": 0.0005612883833236992,
+      "learning_rate": 5.374372436376116e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24606664,
+      "step": 31495
+    },
+    {
+      "epoch": 65.48856548856548,
+      "grad_norm": 8.279694156954065e-05,
+      "learning_rate": 5.368292376659895e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24610792,
+      "step": 31500
+    },
+    {
+      "epoch": 65.4989604989605,
+      "grad_norm": 0.0005552992806769907,
+      "learning_rate": 5.362215344399701e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24614600,
+      "step": 31505
+    },
+    {
+      "epoch": 65.50935550935552,
+      "grad_norm": 8.332059951499104e-05,
+      "learning_rate": 5.356141340532678e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24618504,
+      "step": 31510
+    },
+    {
+      "epoch": 65.51975051975052,
+      "grad_norm": 1.427464849257376e-05,
+      "learning_rate": 5.350070365995522e-06,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 24622280,
+      "step": 31515
+    },
+    {
+      "epoch": 65.53014553014553,
+      "grad_norm": 2.9230093787191436e-05,
+      "learning_rate": 5.344002421724459e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24626216,
+      "step": 31520
+    },
+    {
+      "epoch": 65.54054054054055,
+      "grad_norm": 1.1785150491050445e-05,
+      "learning_rate": 5.337937508655228e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24630152,
+      "step": 31525
+    },
+    {
+      "epoch": 65.55093555093555,
+      "grad_norm": 1.1083277968282346e-05,
+      "learning_rate": 5.331875627723126e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24634088,
+      "step": 31530
+    },
+    {
+      "epoch": 65.56133056133056,
+      "grad_norm": 3.951205144403502e-06,
+      "learning_rate": 5.325816779862963e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24638088,
+      "step": 31535
+    },
+    {
+      "epoch": 65.57172557172557,
+      "grad_norm": 3.0260685889516026e-05,
+      "learning_rate": 5.319760966009102e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24641864,
+      "step": 31540
+    },
+    {
+      "epoch": 65.58212058212058,
+      "grad_norm": 2.3484210032620467e-05,
+      "learning_rate": 5.3137081870954096e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24645768,
+      "step": 31545
+    },
+    {
+      "epoch": 65.5925155925156,
+      "grad_norm": 0.13204526901245117,
+      "learning_rate": 5.307658444055313e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 24649800,
+      "step": 31550
+    },
+    {
+      "epoch": 65.6029106029106,
+      "grad_norm": 0.15280582010746002,
+      "learning_rate": 5.301611737821749e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 24653672,
+      "step": 31555
+    },
+    {
+      "epoch": 65.61330561330561,
+      "grad_norm": 2.5094695956795476e-06,
+      "learning_rate": 5.295568069327206e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24657576,
+      "step": 31560
+    },
+    {
+      "epoch": 65.62370062370063,
+      "grad_norm": 3.0860319384373724e-06,
+      "learning_rate": 5.289527439503683e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24661384,
+      "step": 31565
+    },
+    {
+      "epoch": 65.63409563409563,
+      "grad_norm": 4.534053914539982e-06,
+      "learning_rate": 5.28348984928273e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24665224,
+      "step": 31570
+    },
+    {
+      "epoch": 65.64449064449065,
+      "grad_norm": 0.00070799799868837,
+      "learning_rate": 5.27745529959541e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24669096,
+      "step": 31575
+    },
+    {
+      "epoch": 65.65488565488566,
+      "grad_norm": 8.708270229362824e-07,
+      "learning_rate": 5.271423791372335e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24672872,
+      "step": 31580
+    },
+    {
+      "epoch": 65.66528066528066,
+      "grad_norm": 3.8261685403995216e-05,
+      "learning_rate": 5.26539532554364e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24676872,
+      "step": 31585
+    },
+    {
+      "epoch": 65.67567567567568,
+      "grad_norm": 6.273001054069027e-05,
+      "learning_rate": 5.25936990303898e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24680904,
+      "step": 31590
+    },
+    {
+      "epoch": 65.68607068607069,
+      "grad_norm": 0.00029044513939879835,
+      "learning_rate": 5.253347524787555e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24684904,
+      "step": 31595
+    },
+    {
+      "epoch": 65.6964656964657,
+      "grad_norm": 5.1656745199579746e-05,
+      "learning_rate": 5.2473281917181035e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 24688616,
+      "step": 31600
+    },
+    {
+      "epoch": 65.6964656964657,
+      "eval_loss": 0.5757218599319458,
+      "eval_runtime": 11.7066,
+      "eval_samples_per_second": 73.121,
+      "eval_steps_per_second": 18.28,
+      "num_input_tokens_seen": 24688616,
+      "step": 31600
+    },
+    {
+      "epoch": 65.70686070686071,
+      "grad_norm": 2.602905260573607e-05,
+      "learning_rate": 5.241311904758864e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24692680,
+      "step": 31605
+    },
+    {
+      "epoch": 65.71725571725571,
+      "grad_norm": 6.383168511092663e-05,
+      "learning_rate": 5.23529866483764e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24696712,
+      "step": 31610
+    },
+    {
+      "epoch": 65.72765072765073,
+      "grad_norm": 9.972583029593807e-06,
+      "learning_rate": 5.229288472881732e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24700680,
+      "step": 31615
+    },
+    {
+      "epoch": 65.73804573804574,
+      "grad_norm": 5.078427989246848e-07,
+      "learning_rate": 5.2232813298180025e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24704552,
+      "step": 31620
+    },
+    {
+      "epoch": 65.74844074844074,
+      "grad_norm": 1.1604330211412162e-05,
+      "learning_rate": 5.217277236572824e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24708424,
+      "step": 31625
+    },
+    {
+      "epoch": 65.75883575883576,
+      "grad_norm": 1.6485440710312105e-06,
+      "learning_rate": 5.211276194072093e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24712232,
+      "step": 31630
+    },
+    {
+      "epoch": 65.76923076923077,
+      "grad_norm": 2.607677060950664e-06,
+      "learning_rate": 5.205278203241254e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24716232,
+      "step": 31635
+    },
+    {
+      "epoch": 65.77962577962577,
+      "grad_norm": 1.742325366649311e-05,
+      "learning_rate": 5.199283265005278e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24720264,
+      "step": 31640
+    },
+    {
+      "epoch": 65.79002079002079,
+      "grad_norm": 3.766037480090745e-06,
+      "learning_rate": 5.193291380288648e-06,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 24724168,
+      "step": 31645
+    },
+    {
+      "epoch": 65.8004158004158,
+      "grad_norm": 1.0427480447106063e-05,
+      "learning_rate": 5.1873025500153995e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24728136,
+      "step": 31650
+    },
+    {
+      "epoch": 65.8108108108108,
+      "grad_norm": 5.827236691402504e-06,
+      "learning_rate": 5.181316775109071e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24732040,
+      "step": 31655
+    },
+    {
+      "epoch": 65.82120582120582,
+      "grad_norm": 4.631625415640883e-06,
+      "learning_rate": 5.1753340564927564e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24735848,
+      "step": 31660
+    },
+    {
+      "epoch": 65.83160083160084,
+      "grad_norm": 3.6499994166661054e-05,
+      "learning_rate": 5.169354395089068e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24739656,
+      "step": 31665
+    },
+    {
+      "epoch": 65.84199584199584,
+      "grad_norm": 1.8714024918153882e-05,
+      "learning_rate": 5.1633777918201346e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24743528,
+      "step": 31670
+    },
+    {
+      "epoch": 65.85239085239085,
+      "grad_norm": 2.896259138651658e-05,
+      "learning_rate": 5.157404247607625e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24747432,
+      "step": 31675
+    },
+    {
+      "epoch": 65.86278586278586,
+      "grad_norm": 1.5331237591453828e-05,
+      "learning_rate": 5.1514337633727454e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24751176,
+      "step": 31680
+    },
+    {
+      "epoch": 65.87318087318087,
+      "grad_norm": 1.0576673048490193e-05,
+      "learning_rate": 5.145466340036206e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24755048,
+      "step": 31685
+    },
+    {
+      "epoch": 65.88357588357589,
+      "grad_norm": 0.00010538085916778073,
+      "learning_rate": 5.139501978518274e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 24759112,
+      "step": 31690
+    },
+    {
+      "epoch": 65.89397089397089,
+      "grad_norm": 0.0006465492770075798,
+      "learning_rate": 5.133540679738716e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24763016,
+      "step": 31695
+    },
+    {
+      "epoch": 65.9043659043659,
+      "grad_norm": 0.12396880984306335,
+      "learning_rate": 5.127582444616838e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 24766952,
+      "step": 31700
+    },
+    {
+      "epoch": 65.91476091476092,
+      "grad_norm": 5.12362748850137e-06,
+      "learning_rate": 5.121627274071486e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24770760,
+      "step": 31705
+    },
+    {
+      "epoch": 65.92515592515592,
+      "grad_norm": 1.143999816122232e-05,
+      "learning_rate": 5.115675169021009e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24774696,
+      "step": 31710
+    },
+    {
+      "epoch": 65.93555093555094,
+      "grad_norm": 3.057191497646272e-05,
+      "learning_rate": 5.1097261303832994e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24778600,
+      "step": 31715
+    },
+    {
+      "epoch": 65.94594594594595,
+      "grad_norm": 1.3229027899797074e-05,
+      "learning_rate": 5.103780159075788e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24782632,
+      "step": 31720
+    },
+    {
+      "epoch": 65.95634095634095,
+      "grad_norm": 4.877037440564891e-07,
+      "learning_rate": 5.0978372560154e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 24786472,
+      "step": 31725
+    },
+    {
+      "epoch": 65.96673596673597,
+      "grad_norm": 1.4074691534915473e-05,
+      "learning_rate": 5.091897422118619e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24790408,
+      "step": 31730
+    },
+    {
+      "epoch": 65.97713097713098,
+      "grad_norm": 5.84632687150588e-07,
+      "learning_rate": 5.0859606583014305e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24794280,
+      "step": 31735
+    },
+    {
+      "epoch": 65.98752598752598,
+      "grad_norm": 1.0248490070807748e-05,
+      "learning_rate": 5.080026965479365e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24798120,
+      "step": 31740
+    },
+    {
+      "epoch": 65.997920997921,
+      "grad_norm": 1.9996393803012324e-06,
+      "learning_rate": 5.074096344567475e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 24802088,
+      "step": 31745
+    },
+    {
+      "epoch": 66.00831600831602,
+      "grad_norm": 0.136662095785141,
+      "learning_rate": 5.0681687964803294e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 24805920,
+      "step": 31750
+    },
+    {
+      "epoch": 66.01871101871102,
+      "grad_norm": 0.0002377223427174613,
+      "learning_rate": 5.06224432213204e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 24809952,
+      "step": 31755
+    },
+    {
+      "epoch": 66.02910602910603,
+      "grad_norm": 1.8175261402575416e-06,
+      "learning_rate": 5.056322922436224e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24813920,
+      "step": 31760
+    },
+    {
+      "epoch": 66.03950103950103,
+      "grad_norm": 7.5579655458568595e-06,
+      "learning_rate": 5.0504045983060465e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24817824,
+      "step": 31765
+    },
+    {
+      "epoch": 66.04989604989605,
+      "grad_norm": 0.00013089407002553344,
+      "learning_rate": 5.044489350654183e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24821632,
+      "step": 31770
+    },
+    {
+      "epoch": 66.06029106029106,
+      "grad_norm": 1.0328589041819214e-06,
+      "learning_rate": 5.038577180392831e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24825472,
+      "step": 31775
+    },
+    {
+      "epoch": 66.07068607068607,
+      "grad_norm": 0.001863394514657557,
+      "learning_rate": 5.032668088433729e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24829312,
+      "step": 31780
+    },
+    {
+      "epoch": 66.08108108108108,
+      "grad_norm": 7.747241284050688e-07,
+      "learning_rate": 5.02676207568814e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 24833152,
+      "step": 31785
+    },
+    {
+      "epoch": 66.0914760914761,
+      "grad_norm": 2.8275480872252956e-06,
+      "learning_rate": 5.02085914306683e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24837056,
+      "step": 31790
+    },
+    {
+      "epoch": 66.1018711018711,
+      "grad_norm": 4.8720394261181355e-06,
+      "learning_rate": 5.014959291480123e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24840928,
+      "step": 31795
+    },
+    {
+      "epoch": 66.11226611226611,
+      "grad_norm": 2.220504484284902e-06,
+      "learning_rate": 5.009062521837835e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 24844832,
+      "step": 31800
+    },
+    {
+      "epoch": 66.11226611226611,
+      "eval_loss": 0.5854148268699646,
+      "eval_runtime": 11.7046,
+      "eval_samples_per_second": 73.134,
+      "eval_steps_per_second": 18.283,
+      "num_input_tokens_seen": 24844832,
+      "step": 31800
+    },
+    {
+      "epoch": 66.12266112266113,
+      "grad_norm": 6.610517175431596e-06,
+      "learning_rate": 5.003168835049324e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24848576,
+      "step": 31805
+    },
+    {
+      "epoch": 66.13305613305613,
+      "grad_norm": 3.5868406484951265e-06,
+      "learning_rate": 4.997278232023483e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24852512,
+      "step": 31810
+    },
+    {
+      "epoch": 66.14345114345114,
+      "grad_norm": 1.4809093954681884e-05,
+      "learning_rate": 4.9913907136687036e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24856384,
+      "step": 31815
+    },
+    {
+      "epoch": 66.15384615384616,
+      "grad_norm": 0.14557743072509766,
+      "learning_rate": 4.985506280892918e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 24860352,
+      "step": 31820
+    },
+    {
+      "epoch": 66.16424116424116,
+      "grad_norm": 2.1575258415396092e-06,
+      "learning_rate": 4.979624934603589e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24864256,
+      "step": 31825
+    },
+    {
+      "epoch": 66.17463617463618,
+      "grad_norm": 4.537373115454102e-06,
+      "learning_rate": 4.97374667570768e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24868192,
+      "step": 31830
+    },
+    {
+      "epoch": 66.18503118503118,
+      "grad_norm": 1.98644147531013e-06,
+      "learning_rate": 4.967871505111704e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24872128,
+      "step": 31835
+    },
+    {
+      "epoch": 66.1954261954262,
+      "grad_norm": 3.4830811728170374e-06,
+      "learning_rate": 4.961999423721686e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24876000,
+      "step": 31840
+    },
+    {
+      "epoch": 66.20582120582121,
+      "grad_norm": 5.7558850130590145e-06,
+      "learning_rate": 4.956130432443159e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24880096,
+      "step": 31845
+    },
+    {
+      "epoch": 66.21621621621621,
+      "grad_norm": 0.0005532302311621606,
+      "learning_rate": 4.950264532181215e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24884064,
+      "step": 31850
+    },
+    {
+      "epoch": 66.22661122661123,
+      "grad_norm": 0.00024515012046322227,
+      "learning_rate": 4.944401723840433e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 24888064,
+      "step": 31855
+    },
+    {
+      "epoch": 66.23700623700624,
+      "grad_norm": 2.619003453219193e-06,
+      "learning_rate": 4.938542008324942e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24892000,
+      "step": 31860
+    },
+    {
+      "epoch": 66.24740124740124,
+      "grad_norm": 0.00011473505583126098,
+      "learning_rate": 4.9326853865383855e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24896064,
+      "step": 31865
+    },
+    {
+      "epoch": 66.25779625779626,
+      "grad_norm": 1.1687066034937743e-06,
+      "learning_rate": 4.926831859383918e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24900224,
+      "step": 31870
+    },
+    {
+      "epoch": 66.26819126819127,
+      "grad_norm": 0.00016171195602510124,
+      "learning_rate": 4.92098142776424e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24904128,
+      "step": 31875
+    },
+    {
+      "epoch": 66.27858627858627,
+      "grad_norm": 1.8772652765619569e-06,
+      "learning_rate": 4.91513409258155e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 24907904,
+      "step": 31880
+    },
+    {
+      "epoch": 66.28898128898129,
+      "grad_norm": 8.4004057043785e-07,
+      "learning_rate": 4.909289854737581e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 24911744,
+      "step": 31885
+    },
+    {
+      "epoch": 66.2993762993763,
+      "grad_norm": 8.889811056178587e-07,
+      "learning_rate": 4.903448715133602e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24915648,
+      "step": 31890
+    },
+    {
+      "epoch": 66.3097713097713,
+      "grad_norm": 1.6677782923579798e-06,
+      "learning_rate": 4.897610674670372e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24919552,
+      "step": 31895
+    },
+    {
+      "epoch": 66.32016632016632,
+      "grad_norm": 1.1939395335502923e-05,
+      "learning_rate": 4.8917757342482e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24923360,
+      "step": 31900
+    },
+    {
+      "epoch": 66.33056133056132,
+      "grad_norm": 5.334814431989798e-06,
+      "learning_rate": 4.885943894766909e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24927424,
+      "step": 31905
+    },
+    {
+      "epoch": 66.34095634095634,
+      "grad_norm": 5.080389018985443e-05,
+      "learning_rate": 4.880115157125842e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 24931520,
+      "step": 31910
+    },
+    {
+      "epoch": 66.35135135135135,
+      "grad_norm": 7.247660391840327e-07,
+      "learning_rate": 4.874289522223857e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24935424,
+      "step": 31915
+    },
+    {
+      "epoch": 66.36174636174636,
+      "grad_norm": 2.2270376575761475e-05,
+      "learning_rate": 4.868466990959339e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24939296,
+      "step": 31920
+    },
+    {
+      "epoch": 66.37214137214137,
+      "grad_norm": 2.8446747819543816e-05,
+      "learning_rate": 4.8626475642301964e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24943264,
+      "step": 31925
+    },
+    {
+      "epoch": 66.38253638253639,
+      "grad_norm": 9.141262125922367e-06,
+      "learning_rate": 4.856831242933871e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24947200,
+      "step": 31930
+    },
+    {
+      "epoch": 66.39293139293139,
+      "grad_norm": 1.7104529206335428e-06,
+      "learning_rate": 4.851018027967294e-06,
+      "loss": 0.0021,
+      "num_input_tokens_seen": 24951200,
+      "step": 31935
+    },
+    {
+      "epoch": 66.4033264033264,
+      "grad_norm": 7.847598681109957e-06,
+      "learning_rate": 4.845207920226946e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24955040,
+      "step": 31940
+    },
+    {
+      "epoch": 66.41372141372142,
+      "grad_norm": 1.1133201041957363e-05,
+      "learning_rate": 4.839400920608825e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24958976,
+      "step": 31945
+    },
+    {
+      "epoch": 66.42411642411642,
+      "grad_norm": 3.0871442504576407e-06,
+      "learning_rate": 4.83359703000843e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24962976,
+      "step": 31950
+    },
+    {
+      "epoch": 66.43451143451144,
+      "grad_norm": 0.00016358774155378342,
+      "learning_rate": 4.827796249320804e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24966912,
+      "step": 31955
+    },
+    {
+      "epoch": 66.44490644490645,
+      "grad_norm": 6.445702638302464e-06,
+      "learning_rate": 4.82199857944049e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24970880,
+      "step": 31960
+    },
+    {
+      "epoch": 66.45530145530145,
+      "grad_norm": 2.3987729946384206e-05,
+      "learning_rate": 4.8162040212615695e-06,
+      "loss": 0.0019,
+      "num_input_tokens_seen": 24974720,
+      "step": 31965
+    },
+    {
+      "epoch": 66.46569646569647,
+      "grad_norm": 4.649281436286401e-06,
+      "learning_rate": 4.810412575677639e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24978720,
+      "step": 31970
+    },
+    {
+      "epoch": 66.47609147609148,
+      "grad_norm": 4.292501762392931e-05,
+      "learning_rate": 4.804624243581801e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24982784,
+      "step": 31975
+    },
+    {
+      "epoch": 66.48648648648648,
+      "grad_norm": 2.0792745090147946e-06,
+      "learning_rate": 4.798839025866703e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24986688,
+      "step": 31980
+    },
+    {
+      "epoch": 66.4968814968815,
+      "grad_norm": 4.864602306042798e-05,
+      "learning_rate": 4.793056923424491e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24990592,
+      "step": 31985
+    },
+    {
+      "epoch": 66.5072765072765,
+      "grad_norm": 1.171565327240387e-05,
+      "learning_rate": 4.78727793714683e-06,
+      "loss": 0.0018,
+      "num_input_tokens_seen": 24994528,
+      "step": 31990
+    },
+    {
+      "epoch": 66.51767151767152,
+      "grad_norm": 1.3804999525746098e-06,
+      "learning_rate": 4.7815020679249285e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 24998400,
+      "step": 31995
+    },
+    {
+      "epoch": 66.52806652806653,
+      "grad_norm": 1.528370557934977e-05,
+      "learning_rate": 4.775729316649483e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25002240,
+      "step": 32000
+    },
+    {
+      "epoch": 66.52806652806653,
+      "eval_loss": 0.5968259572982788,
+      "eval_runtime": 11.7223,
+      "eval_samples_per_second": 73.023,
+      "eval_steps_per_second": 18.256,
+      "num_input_tokens_seen": 25002240,
+      "step": 32000
+    },
+    {
+      "epoch": 66.53846153846153,
+      "grad_norm": 5.9422404774522875e-06,
+      "learning_rate": 4.769959684210728e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25006272,
+      "step": 32005
+    },
+    {
+      "epoch": 66.54885654885655,
+      "grad_norm": 5.517757017514668e-06,
+      "learning_rate": 4.764193171498426e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25010080,
+      "step": 32010
+    },
+    {
+      "epoch": 66.55925155925156,
+      "grad_norm": 6.619633495574817e-05,
+      "learning_rate": 4.75842977940183e-06,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 25013792,
+      "step": 32015
+    },
+    {
+      "epoch": 66.56964656964657,
+      "grad_norm": 1.0284732525178697e-05,
+      "learning_rate": 4.752669508809729e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25017792,
+      "step": 32020
+    },
+    {
+      "epoch": 66.58004158004158,
+      "grad_norm": 4.711326710093999e-06,
+      "learning_rate": 4.746912360610445e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25021728,
+      "step": 32025
+    },
+    {
+      "epoch": 66.5904365904366,
+      "grad_norm": 1.882653805296286e-06,
+      "learning_rate": 4.741158335691781e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25025440,
+      "step": 32030
+    },
+    {
+      "epoch": 66.6008316008316,
+      "grad_norm": 9.626057726563886e-05,
+      "learning_rate": 4.7354074349410994e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25029376,
+      "step": 32035
+    },
+    {
+      "epoch": 66.61122661122661,
+      "grad_norm": 8.224907105613966e-06,
+      "learning_rate": 4.729659659245245e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25033376,
+      "step": 32040
+    },
+    {
+      "epoch": 66.62162162162163,
+      "grad_norm": 7.713655577390455e-06,
+      "learning_rate": 4.723915009490601e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25037216,
+      "step": 32045
+    },
+    {
+      "epoch": 66.63201663201663,
+      "grad_norm": 0.00020110831246711314,
+      "learning_rate": 4.718173486563077e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25041440,
+      "step": 32050
+    },
+    {
+      "epoch": 66.64241164241164,
+      "grad_norm": 9.863605328064295e-07,
+      "learning_rate": 4.71243509134808e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25045312,
+      "step": 32055
+    },
+    {
+      "epoch": 66.65280665280665,
+      "grad_norm": 1.2399058277878794e-06,
+      "learning_rate": 4.706699824730532e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25049280,
+      "step": 32060
+    },
+    {
+      "epoch": 66.66320166320166,
+      "grad_norm": 0.0003365107986610383,
+      "learning_rate": 4.700967687594901e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25053152,
+      "step": 32065
+    },
+    {
+      "epoch": 66.67359667359668,
+      "grad_norm": 4.704303137259558e-05,
+      "learning_rate": 4.69523868082514e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25056992,
+      "step": 32070
+    },
+    {
+      "epoch": 66.68399168399168,
+      "grad_norm": 5.172835244593443e-06,
+      "learning_rate": 4.689512805304747e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25060928,
+      "step": 32075
+    },
+    {
+      "epoch": 66.6943866943867,
+      "grad_norm": 6.448878593801055e-06,
+      "learning_rate": 4.683790061916707e-06,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 25064864,
+      "step": 32080
+    },
+    {
+      "epoch": 66.70478170478171,
+      "grad_norm": 1.8789243085848284e-06,
+      "learning_rate": 4.678070451543551e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25068736,
+      "step": 32085
+    },
+    {
+      "epoch": 66.71517671517671,
+      "grad_norm": 1.2462112408684334e-06,
+      "learning_rate": 4.6723539750673204e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25072544,
+      "step": 32090
+    },
+    {
+      "epoch": 66.72557172557173,
+      "grad_norm": 1.6464358623125008e-06,
+      "learning_rate": 4.666640633369551e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25076384,
+      "step": 32095
+    },
+    {
+      "epoch": 66.73596673596674,
+      "grad_norm": 3.477445261523826e-06,
+      "learning_rate": 4.660930427331323e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25080224,
+      "step": 32100
+    },
+    {
+      "epoch": 66.74636174636174,
+      "grad_norm": 2.647525616339408e-05,
+      "learning_rate": 4.6552233578332244e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25084096,
+      "step": 32105
+    },
+    {
+      "epoch": 66.75675675675676,
+      "grad_norm": 5.777507681159477e-07,
+      "learning_rate": 4.649519425755347e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25088160,
+      "step": 32110
+    },
+    {
+      "epoch": 66.76715176715177,
+      "grad_norm": 1.7455677152611315e-05,
+      "learning_rate": 4.64381863197732e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25092160,
+      "step": 32115
+    },
+    {
+      "epoch": 66.77754677754677,
+      "grad_norm": 8.635063750261907e-06,
+      "learning_rate": 4.638120977378269e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25096032,
+      "step": 32120
+    },
+    {
+      "epoch": 66.78794178794179,
+      "grad_norm": 4.5709050027653575e-05,
+      "learning_rate": 4.632426462836848e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25099904,
+      "step": 32125
+    },
+    {
+      "epoch": 66.7983367983368,
+      "grad_norm": 6.323739216895774e-05,
+      "learning_rate": 4.626735089231224e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25103712,
+      "step": 32130
+    },
+    {
+      "epoch": 66.8087318087318,
+      "grad_norm": 1.3713815860683098e-06,
+      "learning_rate": 4.621046857439068e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 25107584,
+      "step": 32135
+    },
+    {
+      "epoch": 66.81912681912682,
+      "grad_norm": 1.228042492584791e-06,
+      "learning_rate": 4.615361768337587e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25111456,
+      "step": 32140
+    },
+    {
+      "epoch": 66.82952182952182,
+      "grad_norm": 8.466170697829511e-07,
+      "learning_rate": 4.6096798228034946e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25115360,
+      "step": 32145
+    },
+    {
+      "epoch": 66.83991683991684,
+      "grad_norm": 1.5531573808402754e-05,
+      "learning_rate": 4.604001021713008e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 25119232,
+      "step": 32150
+    },
+    {
+      "epoch": 66.85031185031185,
+      "grad_norm": 4.095723852515221e-06,
+      "learning_rate": 4.598325365941883e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25123072,
+      "step": 32155
+    },
+    {
+      "epoch": 66.86070686070686,
+      "grad_norm": 3.367743192939088e-05,
+      "learning_rate": 4.5926528563653645e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25126944,
+      "step": 32160
+    },
+    {
+      "epoch": 66.87110187110187,
+      "grad_norm": 1.1111554158560466e-05,
+      "learning_rate": 4.5869834938582295e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25130784,
+      "step": 32165
+    },
+    {
+      "epoch": 66.88149688149689,
+      "grad_norm": 1.8760538296191953e-06,
+      "learning_rate": 4.581317279294772e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25134720,
+      "step": 32170
+    },
+    {
+      "epoch": 66.89189189189189,
+      "grad_norm": 9.346565093437675e-06,
+      "learning_rate": 4.57565421354878e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 25138592,
+      "step": 32175
+    },
+    {
+      "epoch": 66.9022869022869,
+      "grad_norm": 9.37428922043182e-06,
+      "learning_rate": 4.569994297493579e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25142560,
+      "step": 32180
+    },
+    {
+      "epoch": 66.91268191268192,
+      "grad_norm": 3.344899232615717e-05,
+      "learning_rate": 4.564337532002002e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 25146528,
+      "step": 32185
+    },
+    {
+      "epoch": 66.92307692307692,
+      "grad_norm": 3.964806182921166e-06,
+      "learning_rate": 4.55868391794638e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25150304,
+      "step": 32190
+    },
+    {
+      "epoch": 66.93347193347194,
+      "grad_norm": 1.8615235603647307e-05,
+      "learning_rate": 4.553033456198588e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25154240,
+      "step": 32195
+    },
+    {
+      "epoch": 66.94386694386695,
+      "grad_norm": 1.4211239431460854e-05,
+      "learning_rate": 4.54738614762999e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 25158144,
+      "step": 32200
+    },
+    {
+      "epoch": 66.94386694386695,
+      "eval_loss": 0.5835604071617126,
+      "eval_runtime": 11.709,
+      "eval_samples_per_second": 73.106,
+      "eval_steps_per_second": 18.277,
+      "num_input_tokens_seen": 25158144,
+      "step": 32200
+    },
+    {
+      "epoch": 66.95426195426195,
+      "grad_norm": 0.0001385412469971925,
+      "learning_rate": 4.541741993111465e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 25161952,
+      "step": 32205
+    },
+    {
+      "epoch": 66.96465696465697,
+      "grad_norm": 1.1186824849573895e-05,
+      "learning_rate": 4.536100993513423e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25165792,
+      "step": 32210
+    },
+    {
+      "epoch": 66.97505197505197,
+      "grad_norm": 2.366246462770505e-06,
+      "learning_rate": 4.530463149705768e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25169824,
+      "step": 32215
+    },
+    {
+      "epoch": 66.98544698544698,
+      "grad_norm": 6.581702564290026e-06,
+      "learning_rate": 4.524828462557934e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25173664,
+      "step": 32220
+    },
+    {
+      "epoch": 66.995841995842,
+      "grad_norm": 2.389136716374196e-05,
+      "learning_rate": 4.5191969329388625e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25177600,
+      "step": 32225
+    },
+    {
+      "epoch": 67.006237006237,
+      "grad_norm": 8.879872552824963e-07,
+      "learning_rate": 4.5135685617169965e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25181488,
+      "step": 32230
+    },
+    {
+      "epoch": 67.01663201663202,
+      "grad_norm": 1.6757461708039045e-05,
+      "learning_rate": 4.507943349760313e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25185392,
+      "step": 32235
+    },
+    {
+      "epoch": 67.02702702702703,
+      "grad_norm": 8.656988939037547e-06,
+      "learning_rate": 4.502321297936277e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25189200,
+      "step": 32240
+    },
+    {
+      "epoch": 67.03742203742203,
+      "grad_norm": 2.0292134195187828e-06,
+      "learning_rate": 4.496702407111888e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25193200,
+      "step": 32245
+    },
+    {
+      "epoch": 67.04781704781705,
+      "grad_norm": 0.000908319721929729,
+      "learning_rate": 4.491086678153653e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 25197232,
+      "step": 32250
+    },
+    {
+      "epoch": 67.05821205821206,
+      "grad_norm": 1.2728187357424758e-06,
+      "learning_rate": 4.485474111927579e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25201296,
+      "step": 32255
+    },
+    {
+      "epoch": 67.06860706860707,
+      "grad_norm": 2.902484766309499e-06,
+      "learning_rate": 4.479864709299197e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25205168,
+      "step": 32260
+    },
+    {
+      "epoch": 67.07900207900208,
+      "grad_norm": 1.2198677723063156e-05,
+      "learning_rate": 4.474258471133555e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25209072,
+      "step": 32265
+    },
+    {
+      "epoch": 67.0893970893971,
+      "grad_norm": 9.289577064919285e-06,
+      "learning_rate": 4.4686553982952014e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25213072,
+      "step": 32270
+    },
+    {
+      "epoch": 67.0997920997921,
+      "grad_norm": 6.183955974847777e-06,
+      "learning_rate": 4.463055491648191e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 25217104,
+      "step": 32275
+    },
+    {
+      "epoch": 67.11018711018711,
+      "grad_norm": 5.762820819654735e-06,
+      "learning_rate": 4.457458752056112e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25221072,
+      "step": 32280
+    },
+    {
+      "epoch": 67.12058212058211,
+      "grad_norm": 0.14333458244800568,
+      "learning_rate": 4.451865180382042e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 25224912,
+      "step": 32285
+    },
+    {
+      "epoch": 67.13097713097713,
+      "grad_norm": 1.9780929505941458e-05,
+      "learning_rate": 4.4462747774885936e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25228912,
+      "step": 32290
+    },
+    {
+      "epoch": 67.14137214137214,
+      "grad_norm": 2.2833675757283345e-05,
+      "learning_rate": 4.440687544237859e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25232944,
+      "step": 32295
+    },
+    {
+      "epoch": 67.15176715176715,
+      "grad_norm": 8.674664968566503e-06,
+      "learning_rate": 4.435103481491471e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25236784,
+      "step": 32300
+    },
+    {
+      "epoch": 67.16216216216216,
+      "grad_norm": 0.00041360571049153805,
+      "learning_rate": 4.429522590110569e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25240688,
+      "step": 32305
+    },
+    {
+      "epoch": 67.17255717255718,
+      "grad_norm": 6.704784141220443e-07,
+      "learning_rate": 4.423944870955779e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25244592,
+      "step": 32310
+    },
+    {
+      "epoch": 67.18295218295218,
+      "grad_norm": 1.4449345144385006e-05,
+      "learning_rate": 4.418370324887272e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25248464,
+      "step": 32315
+    },
+    {
+      "epoch": 67.1933471933472,
+      "grad_norm": 8.704470928933006e-06,
+      "learning_rate": 4.412798952764699e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25252432,
+      "step": 32320
+    },
+    {
+      "epoch": 67.20374220374221,
+      "grad_norm": 8.224931661970913e-06,
+      "learning_rate": 4.407230755447245e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25256272,
+      "step": 32325
+    },
+    {
+      "epoch": 67.21413721413721,
+      "grad_norm": 1.3917200703872368e-05,
+      "learning_rate": 4.401665733793598e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25260112,
+      "step": 32330
+    },
+    {
+      "epoch": 67.22453222453223,
+      "grad_norm": 0.1418980211019516,
+      "learning_rate": 4.3961038886619425e-06,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 25264048,
+      "step": 32335
+    },
+    {
+      "epoch": 67.23492723492724,
+      "grad_norm": 7.690147322136909e-05,
+      "learning_rate": 4.39054522091e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 25267984,
+      "step": 32340
+    },
+    {
+      "epoch": 67.24532224532224,
+      "grad_norm": 8.129670277412515e-06,
+      "learning_rate": 4.384989731394979e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 25271632,
+      "step": 32345
+    },
+    {
+      "epoch": 67.25571725571726,
+      "grad_norm": 9.194549420499243e-06,
+      "learning_rate": 4.379437420973598e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25275600,
+      "step": 32350
+    },
+    {
+      "epoch": 67.26611226611226,
+      "grad_norm": 3.6396918403625023e-06,
+      "learning_rate": 4.373888290502107e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25279600,
+      "step": 32355
+    },
+    {
+      "epoch": 67.27650727650727,
+      "grad_norm": 7.072638709360035e-06,
+      "learning_rate": 4.36834234083624e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25283440,
+      "step": 32360
+    },
+    {
+      "epoch": 67.28690228690229,
+      "grad_norm": 1.4649101558461552e-06,
+      "learning_rate": 4.362799572831258e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25287312,
+      "step": 32365
+    },
+    {
+      "epoch": 67.29729729729729,
+      "grad_norm": 7.5633770393324085e-06,
+      "learning_rate": 4.35725998734193e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25291120,
+      "step": 32370
+    },
+    {
+      "epoch": 67.3076923076923,
+      "grad_norm": 1.1451118098193547e-06,
+      "learning_rate": 4.3517235852225195e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25294992,
+      "step": 32375
+    },
+    {
+      "epoch": 67.31808731808732,
+      "grad_norm": 3.25534820149187e-05,
+      "learning_rate": 4.346190367326822e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25298768,
+      "step": 32380
+    },
+    {
+      "epoch": 67.32848232848232,
+      "grad_norm": 3.309769454062916e-05,
+      "learning_rate": 4.340660334508115e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25302640,
+      "step": 32385
+    },
+    {
+      "epoch": 67.33887733887734,
+      "grad_norm": 4.38424967796891e-06,
+      "learning_rate": 4.335133487619206e-06,
+      "loss": 0.0034,
+      "num_input_tokens_seen": 25306544,
+      "step": 32390
+    },
+    {
+      "epoch": 67.34927234927235,
+      "grad_norm": 3.960438789363252e-06,
+      "learning_rate": 4.329609827512409e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25310416,
+      "step": 32395
+    },
+    {
+      "epoch": 67.35966735966736,
+      "grad_norm": 1.3543035493057687e-06,
+      "learning_rate": 4.324089355039531e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25314384,
+      "step": 32400
+    },
+    {
+      "epoch": 67.35966735966736,
+      "eval_loss": 0.5943719148635864,
+      "eval_runtime": 11.7426,
+      "eval_samples_per_second": 72.897,
+      "eval_steps_per_second": 18.224,
+      "num_input_tokens_seen": 25314384,
+      "step": 32400
+    },
+    {
+      "epoch": 67.37006237006237,
+      "grad_norm": 1.0722674232965801e-05,
+      "learning_rate": 4.3185720710519075e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25318320,
+      "step": 32405
+    },
+    {
+      "epoch": 67.38045738045739,
+      "grad_norm": 5.27079919265816e-06,
+      "learning_rate": 4.3130579764003724e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25322192,
+      "step": 32410
+    },
+    {
+      "epoch": 67.39085239085239,
+      "grad_norm": 3.2633404316584347e-06,
+      "learning_rate": 4.307547071935267e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25326096,
+      "step": 32415
+    },
+    {
+      "epoch": 67.4012474012474,
+      "grad_norm": 1.6513173477505916e-06,
+      "learning_rate": 4.302039358506435e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25329936,
+      "step": 32420
+    },
+    {
+      "epoch": 67.41164241164242,
+      "grad_norm": 2.4489843326591654e-06,
+      "learning_rate": 4.296534836963245e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25333648,
+      "step": 32425
+    },
+    {
+      "epoch": 67.42203742203742,
+      "grad_norm": 2.567072442616336e-05,
+      "learning_rate": 4.291033508154555e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 25337456,
+      "step": 32430
+    },
+    {
+      "epoch": 67.43243243243244,
+      "grad_norm": 1.4650568118668161e-05,
+      "learning_rate": 4.285535372928748e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25341456,
+      "step": 32435
+    },
+    {
+      "epoch": 67.44282744282744,
+      "grad_norm": 1.342518771707546e-06,
+      "learning_rate": 4.280040432133695e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25345456,
+      "step": 32440
+    },
+    {
+      "epoch": 67.45322245322245,
+      "grad_norm": 1.3183847613618127e-06,
+      "learning_rate": 4.274548686616789e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25349392,
+      "step": 32445
+    },
+    {
+      "epoch": 67.46361746361747,
+      "grad_norm": 3.6927875044057146e-05,
+      "learning_rate": 4.2690601372249364e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 25353552,
+      "step": 32450
+    },
+    {
+      "epoch": 67.47401247401247,
+      "grad_norm": 0.0004922813386656344,
+      "learning_rate": 4.263574784804525e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25357392,
+      "step": 32455
+    },
+    {
+      "epoch": 67.48440748440748,
+      "grad_norm": 2.3604552552569658e-05,
+      "learning_rate": 4.258092630201479e-06,
+      "loss": 0.0036,
+      "num_input_tokens_seen": 25361168,
+      "step": 32460
+    },
+    {
+      "epoch": 67.4948024948025,
+      "grad_norm": 3.300210300949402e-05,
+      "learning_rate": 4.252613674261202e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25365072,
+      "step": 32465
+    },
+    {
+      "epoch": 67.5051975051975,
+      "grad_norm": 2.5202965844073333e-05,
+      "learning_rate": 4.2471379178286224e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25369040,
+      "step": 32470
+    },
+    {
+      "epoch": 67.51559251559252,
+      "grad_norm": 7.662098505534232e-05,
+      "learning_rate": 4.241665361748181e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25372976,
+      "step": 32475
+    },
+    {
+      "epoch": 67.52598752598753,
+      "grad_norm": 8.056331353145652e-06,
+      "learning_rate": 4.2361960068637994e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 25377008,
+      "step": 32480
+    },
+    {
+      "epoch": 67.53638253638253,
+      "grad_norm": 4.538008852250641e-06,
+      "learning_rate": 4.230729854018933e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25380912,
+      "step": 32485
+    },
+    {
+      "epoch": 67.54677754677755,
+      "grad_norm": 3.703022957779467e-05,
+      "learning_rate": 4.225266904056521e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25384880,
+      "step": 32490
+    },
+    {
+      "epoch": 67.55717255717256,
+      "grad_norm": 5.006138962926343e-06,
+      "learning_rate": 4.21980715781903e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25388752,
+      "step": 32495
+    },
+    {
+      "epoch": 67.56756756756756,
+      "grad_norm": 1.8149379457099712e-06,
+      "learning_rate": 4.214350616148416e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25392688,
+      "step": 32500
+    },
+    {
+      "epoch": 67.57796257796258,
+      "grad_norm": 1.3159240552340634e-05,
+      "learning_rate": 4.20889727988614e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25396592,
+      "step": 32505
+    },
+    {
+      "epoch": 67.58835758835758,
+      "grad_norm": 5.58895044378005e-06,
+      "learning_rate": 4.20344714987318e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 25400464,
+      "step": 32510
+    },
+    {
+      "epoch": 67.5987525987526,
+      "grad_norm": 5.0071943405782804e-05,
+      "learning_rate": 4.198000226950022e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25404432,
+      "step": 32515
+    },
+    {
+      "epoch": 67.60914760914761,
+      "grad_norm": 8.017790241865441e-05,
+      "learning_rate": 4.192556511956635e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25408496,
+      "step": 32520
+    },
+    {
+      "epoch": 67.61954261954261,
+      "grad_norm": 1.8465617586116423e-06,
+      "learning_rate": 4.18711600573252e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 25412496,
+      "step": 32525
+    },
+    {
+      "epoch": 67.62993762993763,
+      "grad_norm": 5.1970622735098004e-05,
+      "learning_rate": 4.181678709116671e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25416272,
+      "step": 32530
+    },
+    {
+      "epoch": 67.64033264033264,
+      "grad_norm": 1.7067171711460105e-06,
+      "learning_rate": 4.1762446229475785e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25420176,
+      "step": 32535
+    },
+    {
+      "epoch": 67.65072765072765,
+      "grad_norm": 3.0688540846313117e-06,
+      "learning_rate": 4.17081374806326e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25423984,
+      "step": 32540
+    },
+    {
+      "epoch": 67.66112266112266,
+      "grad_norm": 2.8662281692959368e-05,
+      "learning_rate": 4.165386085301212e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25427792,
+      "step": 32545
+    },
+    {
+      "epoch": 67.67151767151768,
+      "grad_norm": 0.0013591740280389786,
+      "learning_rate": 4.1599616354984525e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25431664,
+      "step": 32550
+    },
+    {
+      "epoch": 67.68191268191268,
+      "grad_norm": 1.1036420346499654e-06,
+      "learning_rate": 4.154540399491508e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25435504,
+      "step": 32555
+    },
+    {
+      "epoch": 67.6923076923077,
+      "grad_norm": 0.0002571245713625103,
+      "learning_rate": 4.149122378116394e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25439376,
+      "step": 32560
+    },
+    {
+      "epoch": 67.70270270270271,
+      "grad_norm": 8.112965588225052e-06,
+      "learning_rate": 4.14370757220863e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25443184,
+      "step": 32565
+    },
+    {
+      "epoch": 67.71309771309771,
+      "grad_norm": 8.090904157143086e-05,
+      "learning_rate": 4.138295982603263e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25447216,
+      "step": 32570
+    },
+    {
+      "epoch": 67.72349272349273,
+      "grad_norm": 0.00016385763592552394,
+      "learning_rate": 4.132887610134814e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25451120,
+      "step": 32575
+    },
+    {
+      "epoch": 67.73388773388774,
+      "grad_norm": 3.875670699926559e-06,
+      "learning_rate": 4.127482455637335e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25454928,
+      "step": 32580
+    },
+    {
+      "epoch": 67.74428274428274,
+      "grad_norm": 0.12346284091472626,
+      "learning_rate": 4.1220805199443545e-06,
+      "loss": 0.0052,
+      "num_input_tokens_seen": 25458928,
+      "step": 32585
+    },
+    {
+      "epoch": 67.75467775467776,
+      "grad_norm": 6.848708267170878e-07,
+      "learning_rate": 4.116681803888925e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25462736,
+      "step": 32590
+    },
+    {
+      "epoch": 67.76507276507276,
+      "grad_norm": 3.8470938307000324e-05,
+      "learning_rate": 4.111286308303605e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25466672,
+      "step": 32595
+    },
+    {
+      "epoch": 67.77546777546777,
+      "grad_norm": 1.3075252809358062e-06,
+      "learning_rate": 4.105894034020433e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25470704,
+      "step": 32600
+    },
+    {
+      "epoch": 67.77546777546777,
+      "eval_loss": 0.5913785696029663,
+      "eval_runtime": 11.7189,
+      "eval_samples_per_second": 73.044,
+      "eval_steps_per_second": 18.261,
+      "num_input_tokens_seen": 25470704,
+      "step": 32600
+    },
+    {
+      "epoch": 67.78586278586279,
+      "grad_norm": 0.00015433398948516697,
+      "learning_rate": 4.100504981870975e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25474672,
+      "step": 32605
+    },
+    {
+      "epoch": 67.79625779625779,
+      "grad_norm": 3.5931468573835446e-06,
+      "learning_rate": 4.0951191526862915e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25478576,
+      "step": 32610
+    },
+    {
+      "epoch": 67.8066528066528,
+      "grad_norm": 3.871577973768581e-06,
+      "learning_rate": 4.089736547296938e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25482384,
+      "step": 32615
+    },
+    {
+      "epoch": 67.81704781704782,
+      "grad_norm": 0.0006709403241984546,
+      "learning_rate": 4.08435716653299e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25486416,
+      "step": 32620
+    },
+    {
+      "epoch": 67.82744282744282,
+      "grad_norm": 7.1034141910786275e-06,
+      "learning_rate": 4.0789810112240005e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25490256,
+      "step": 32625
+    },
+    {
+      "epoch": 67.83783783783784,
+      "grad_norm": 8.041750334086828e-06,
+      "learning_rate": 4.073608082199057e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25494288,
+      "step": 32630
+    },
+    {
+      "epoch": 67.84823284823285,
+      "grad_norm": 1.924689286170178e-06,
+      "learning_rate": 4.068238380286718e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25498256,
+      "step": 32635
+    },
+    {
+      "epoch": 67.85862785862786,
+      "grad_norm": 5.170352324057603e-06,
+      "learning_rate": 4.062871906315072e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25502320,
+      "step": 32640
+    },
+    {
+      "epoch": 67.86902286902287,
+      "grad_norm": 0.0026495938654989004,
+      "learning_rate": 4.057508661111686e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25506128,
+      "step": 32645
+    },
+    {
+      "epoch": 67.87941787941789,
+      "grad_norm": 2.0857689833064796e-06,
+      "learning_rate": 4.052148645503648e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25510064,
+      "step": 32650
+    },
+    {
+      "epoch": 67.88981288981289,
+      "grad_norm": 4.288062427804107e-06,
+      "learning_rate": 4.046791860317531e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 25514032,
+      "step": 32655
+    },
+    {
+      "epoch": 67.9002079002079,
+      "grad_norm": 1.7331831259070896e-05,
+      "learning_rate": 4.041438306379431e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25517872,
+      "step": 32660
+    },
+    {
+      "epoch": 67.9106029106029,
+      "grad_norm": 3.505286076688208e-05,
+      "learning_rate": 4.036087984514916e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 25521872,
+      "step": 32665
+    },
+    {
+      "epoch": 67.92099792099792,
+      "grad_norm": 3.812530394498026e-06,
+      "learning_rate": 4.030740895549084e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25525776,
+      "step": 32670
+    },
+    {
+      "epoch": 67.93139293139293,
+      "grad_norm": 6.577989097422687e-06,
+      "learning_rate": 4.025397040306531e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25529680,
+      "step": 32675
+    },
+    {
+      "epoch": 67.94178794178794,
+      "grad_norm": 8.341300417669117e-05,
+      "learning_rate": 4.0200564196113285e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 25533648,
+      "step": 32680
+    },
+    {
+      "epoch": 67.95218295218295,
+      "grad_norm": 1.9791639260802185e-06,
+      "learning_rate": 4.014719034287079e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 25537360,
+      "step": 32685
+    },
+    {
+      "epoch": 67.96257796257797,
+      "grad_norm": 6.811199909861898e-06,
+      "learning_rate": 4.0093848851568775e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25541296,
+      "step": 32690
+    },
+    {
+      "epoch": 67.97297297297297,
+      "grad_norm": 9.807798960537184e-06,
+      "learning_rate": 4.004053973043304e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25545168,
+      "step": 32695
+    },
+    {
+      "epoch": 67.98336798336798,
+      "grad_norm": 5.901354961679317e-05,
+      "learning_rate": 3.998726298768465e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25549040,
+      "step": 32700
+    },
+    {
+      "epoch": 67.993762993763,
+      "grad_norm": 1.4775903764530085e-05,
+      "learning_rate": 3.99340186315395e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25552880,
+      "step": 32705
+    },
+    {
+      "epoch": 68.004158004158,
+      "grad_norm": 4.670803264161805e-06,
+      "learning_rate": 3.988080667020849e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25556864,
+      "step": 32710
+    },
+    {
+      "epoch": 68.01455301455302,
+      "grad_norm": 3.6722015011037e-06,
+      "learning_rate": 3.982762711189766e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25560736,
+      "step": 32715
+    },
+    {
+      "epoch": 68.02494802494803,
+      "grad_norm": 0.001565297832712531,
+      "learning_rate": 3.977447996480785e-06,
+      "loss": 0.002,
+      "num_input_tokens_seen": 25564704,
+      "step": 32720
+    },
+    {
+      "epoch": 68.03534303534303,
+      "grad_norm": 1.8102866306435317e-05,
+      "learning_rate": 3.97213652371351e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25568576,
+      "step": 32725
+    },
+    {
+      "epoch": 68.04573804573805,
+      "grad_norm": 1.0223486697213957e-06,
+      "learning_rate": 3.966828293707042e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25572480,
+      "step": 32730
+    },
+    {
+      "epoch": 68.05613305613305,
+      "grad_norm": 0.00023555295774713159,
+      "learning_rate": 3.961523307279963e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25576352,
+      "step": 32735
+    },
+    {
+      "epoch": 68.06652806652806,
+      "grad_norm": 4.905569221591577e-05,
+      "learning_rate": 3.956221565250382e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25580128,
+      "step": 32740
+    },
+    {
+      "epoch": 68.07692307692308,
+      "grad_norm": 7.970704700710485e-07,
+      "learning_rate": 3.950923068435883e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25584096,
+      "step": 32745
+    },
+    {
+      "epoch": 68.08731808731808,
+      "grad_norm": 3.2453549465571996e-06,
+      "learning_rate": 3.945627817653566e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25588032,
+      "step": 32750
+    },
+    {
+      "epoch": 68.0977130977131,
+      "grad_norm": 1.95779603018309e-06,
+      "learning_rate": 3.9403358137200335e-06,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 25591968,
+      "step": 32755
+    },
+    {
+      "epoch": 68.10810810810811,
+      "grad_norm": 1.1590396979954676e-06,
+      "learning_rate": 3.9350470574513605e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25595968,
+      "step": 32760
+    },
+    {
+      "epoch": 68.11850311850311,
+      "grad_norm": 0.00015430926578119397,
+      "learning_rate": 3.9297615496631525e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25599808,
+      "step": 32765
+    },
+    {
+      "epoch": 68.12889812889813,
+      "grad_norm": 2.8009983452648157e-06,
+      "learning_rate": 3.924479291170505e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25603648,
+      "step": 32770
+    },
+    {
+      "epoch": 68.13929313929314,
+      "grad_norm": 0.12475251406431198,
+      "learning_rate": 3.919200282788002e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 25607520,
+      "step": 32775
+    },
+    {
+      "epoch": 68.14968814968815,
+      "grad_norm": 4.873909347224981e-06,
+      "learning_rate": 3.913924525329726e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25611392,
+      "step": 32780
+    },
+    {
+      "epoch": 68.16008316008316,
+      "grad_norm": 1.00829618077114e-06,
+      "learning_rate": 3.908652019609279e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25615264,
+      "step": 32785
+    },
+    {
+      "epoch": 68.17047817047818,
+      "grad_norm": 1.5252331877491088e-06,
+      "learning_rate": 3.9033827664397364e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25619200,
+      "step": 32790
+    },
+    {
+      "epoch": 68.18087318087318,
+      "grad_norm": 7.952392479637638e-06,
+      "learning_rate": 3.898116766633694e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25623136,
+      "step": 32795
+    },
+    {
+      "epoch": 68.1912681912682,
+      "grad_norm": 2.1265843770379433e-06,
+      "learning_rate": 3.8928540210032225e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25627200,
+      "step": 32800
+    },
+    {
+      "epoch": 68.1912681912682,
+      "eval_loss": 0.5901188850402832,
+      "eval_runtime": 11.724,
+      "eval_samples_per_second": 73.013,
+      "eval_steps_per_second": 18.253,
+      "num_input_tokens_seen": 25627200,
+      "step": 32800
+    },
+    {
+      "epoch": 68.20166320166321,
+      "grad_norm": 5.834729677189898e-07,
+      "learning_rate": 3.887594530359909e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25631104,
+      "step": 32805
+    },
+    {
+      "epoch": 68.21205821205821,
+      "grad_norm": 3.1911026781017426e-06,
+      "learning_rate": 3.88233829551484e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25635264,
+      "step": 32810
+    },
+    {
+      "epoch": 68.22245322245323,
+      "grad_norm": 6.7228770603833254e-06,
+      "learning_rate": 3.877085317278581e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25639136,
+      "step": 32815
+    },
+    {
+      "epoch": 68.23284823284823,
+      "grad_norm": 0.00022962111688684672,
+      "learning_rate": 3.87183559646122e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25642912,
+      "step": 32820
+    },
+    {
+      "epoch": 68.24324324324324,
+      "grad_norm": 2.1062617179268273e-06,
+      "learning_rate": 3.866589133872317e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25646752,
+      "step": 32825
+    },
+    {
+      "epoch": 68.25363825363826,
+      "grad_norm": 2.820444387907628e-05,
+      "learning_rate": 3.861345930320948e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25650720,
+      "step": 32830
+    },
+    {
+      "epoch": 68.26403326403326,
+      "grad_norm": 2.378518729528878e-06,
+      "learning_rate": 3.856105986615688e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25654784,
+      "step": 32835
+    },
+    {
+      "epoch": 68.27442827442827,
+      "grad_norm": 0.0009368057362735271,
+      "learning_rate": 3.850869303564589e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25658656,
+      "step": 32840
+    },
+    {
+      "epoch": 68.28482328482329,
+      "grad_norm": 3.469826333457604e-05,
+      "learning_rate": 3.845635881975226e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25662624,
+      "step": 32845
+    },
+    {
+      "epoch": 68.29521829521829,
+      "grad_norm": 1.6099362255772576e-05,
+      "learning_rate": 3.840405722654647e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25666528,
+      "step": 32850
+    },
+    {
+      "epoch": 68.3056133056133,
+      "grad_norm": 5.4836273193359375e-05,
+      "learning_rate": 3.835178826409419e-06,
+      "loss": 0.0035,
+      "num_input_tokens_seen": 25670336,
+      "step": 32855
+    },
+    {
+      "epoch": 68.31600831600832,
+      "grad_norm": 8.614324542577378e-06,
+      "learning_rate": 3.8299551940455895e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 25674144,
+      "step": 32860
+    },
+    {
+      "epoch": 68.32640332640332,
+      "grad_norm": 5.1521285058697686e-05,
+      "learning_rate": 3.824734826368703e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25678176,
+      "step": 32865
+    },
+    {
+      "epoch": 68.33679833679834,
+      "grad_norm": 1.2025701835227665e-05,
+      "learning_rate": 3.819517724183813e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25682176,
+      "step": 32870
+    },
+    {
+      "epoch": 68.34719334719335,
+      "grad_norm": 4.356612407718785e-05,
+      "learning_rate": 3.8143038882954648e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 25686144,
+      "step": 32875
+    },
+    {
+      "epoch": 68.35758835758836,
+      "grad_norm": 7.380162514891708e-06,
+      "learning_rate": 3.8090933195076867e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25690016,
+      "step": 32880
+    },
+    {
+      "epoch": 68.36798336798337,
+      "grad_norm": 9.376531670568511e-05,
+      "learning_rate": 3.8038860186240198e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25693824,
+      "step": 32885
+    },
+    {
+      "epoch": 68.37837837837837,
+      "grad_norm": 1.8161107391279074e-06,
+      "learning_rate": 3.7986819864475026e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25697728,
+      "step": 32890
+    },
+    {
+      "epoch": 68.38877338877339,
+      "grad_norm": 0.0001698373380349949,
+      "learning_rate": 3.793481223780651e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25701632,
+      "step": 32895
+    },
+    {
+      "epoch": 68.3991683991684,
+      "grad_norm": 1.058812176779611e-05,
+      "learning_rate": 3.788283731425496e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25705504,
+      "step": 32900
+    },
+    {
+      "epoch": 68.4095634095634,
+      "grad_norm": 2.352861520193983e-05,
+      "learning_rate": 3.7830895101835488e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25709504,
+      "step": 32905
+    },
+    {
+      "epoch": 68.41995841995842,
+      "grad_norm": 1.554212644805375e-06,
+      "learning_rate": 3.7778985608558274e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25713440,
+      "step": 32910
+    },
+    {
+      "epoch": 68.43035343035343,
+      "grad_norm": 9.532726039651607e-07,
+      "learning_rate": 3.7727108842428443e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25717312,
+      "step": 32915
+    },
+    {
+      "epoch": 68.44074844074844,
+      "grad_norm": 2.864121825041366e-06,
+      "learning_rate": 3.7675264811446065e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 25721248,
+      "step": 32920
+    },
+    {
+      "epoch": 68.45114345114345,
+      "grad_norm": 2.3524586140410975e-05,
+      "learning_rate": 3.7623453523605994e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25725312,
+      "step": 32925
+    },
+    {
+      "epoch": 68.46153846153847,
+      "grad_norm": 1.701234759821091e-05,
+      "learning_rate": 3.757167498689834e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 25729280,
+      "step": 32930
+    },
+    {
+      "epoch": 68.47193347193347,
+      "grad_norm": 0.00012517474533524364,
+      "learning_rate": 3.7519929209307914e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25733088,
+      "step": 32935
+    },
+    {
+      "epoch": 68.48232848232848,
+      "grad_norm": 0.00010829915845533833,
+      "learning_rate": 3.746821619881463e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25737024,
+      "step": 32940
+    },
+    {
+      "epoch": 68.4927234927235,
+      "grad_norm": 0.00014602280862163752,
+      "learning_rate": 3.74165359633932e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25740960,
+      "step": 32945
+    },
+    {
+      "epoch": 68.5031185031185,
+      "grad_norm": 1.5766655678817187e-06,
+      "learning_rate": 3.736488851101341e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25744800,
+      "step": 32950
+    },
+    {
+      "epoch": 68.51351351351352,
+      "grad_norm": 0.0001235568488482386,
+      "learning_rate": 3.7313273849640035e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25748768,
+      "step": 32955
+    },
+    {
+      "epoch": 68.52390852390852,
+      "grad_norm": 1.1096884009020869e-05,
+      "learning_rate": 3.7261691987232533e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 25752704,
+      "step": 32960
+    },
+    {
+      "epoch": 68.53430353430353,
+      "grad_norm": 2.664585736056324e-06,
+      "learning_rate": 3.7210142931745575e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 25756576,
+      "step": 32965
+    },
+    {
+      "epoch": 68.54469854469855,
+      "grad_norm": 6.856319032522151e-06,
+      "learning_rate": 3.7158626691128712e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25760512,
+      "step": 32970
+    },
+    {
+      "epoch": 68.55509355509355,
+      "grad_norm": 4.554670454126608e-07,
+      "learning_rate": 3.710714327332629e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 25764128,
+      "step": 32975
+    },
+    {
+      "epoch": 68.56548856548856,
+      "grad_norm": 0.13925693929195404,
+      "learning_rate": 3.7055692686277815e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 25768000,
+      "step": 32980
+    },
+    {
+      "epoch": 68.57588357588358,
+      "grad_norm": 4.281182918930426e-05,
+      "learning_rate": 3.70042749379175e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 25771840,
+      "step": 32985
+    },
+    {
+      "epoch": 68.58627858627858,
+      "grad_norm": 2.923273314081598e-05,
+      "learning_rate": 3.6952890036174693e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25775680,
+      "step": 32990
+    },
+    {
+      "epoch": 68.5966735966736,
+      "grad_norm": 1.8368422161074704e-06,
+      "learning_rate": 3.690153798897353e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25779520,
+      "step": 32995
+    },
+    {
+      "epoch": 68.60706860706861,
+      "grad_norm": 7.52680421101104e-07,
+      "learning_rate": 3.6850218804233225e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25783456,
+      "step": 33000
+    },
+    {
+      "epoch": 68.60706860706861,
+      "eval_loss": 0.595913827419281,
+      "eval_runtime": 11.713,
+      "eval_samples_per_second": 73.081,
+      "eval_steps_per_second": 18.27,
+      "num_input_tokens_seen": 25783456,
+      "step": 33000
+    },
+    {
+      "epoch": 68.61746361746361,
+      "grad_norm": 6.2254562180896755e-06,
+      "learning_rate": 3.679893248986779e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25787328,
+      "step": 33005
+    },
+    {
+      "epoch": 68.62785862785863,
+      "grad_norm": 5.581254299613647e-06,
+      "learning_rate": 3.6747679053786147e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25791232,
+      "step": 33010
+    },
+    {
+      "epoch": 68.63825363825364,
+      "grad_norm": 2.1175430447328836e-05,
+      "learning_rate": 3.669645850389228e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25795072,
+      "step": 33015
+    },
+    {
+      "epoch": 68.64864864864865,
+      "grad_norm": 0.0002134252863470465,
+      "learning_rate": 3.664527084808514e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 25799008,
+      "step": 33020
+    },
+    {
+      "epoch": 68.65904365904366,
+      "grad_norm": 4.457516479305923e-05,
+      "learning_rate": 3.6594116094258337e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25802944,
+      "step": 33025
+    },
+    {
+      "epoch": 68.66943866943868,
+      "grad_norm": 5.0540611482574604e-06,
+      "learning_rate": 3.6542994250300665e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25806784,
+      "step": 33030
+    },
+    {
+      "epoch": 68.67983367983368,
+      "grad_norm": 5.640792369376868e-05,
+      "learning_rate": 3.6491905324095825e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 25810752,
+      "step": 33035
+    },
+    {
+      "epoch": 68.6902286902287,
+      "grad_norm": 0.0001016157038975507,
+      "learning_rate": 3.644084932352221e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25814752,
+      "step": 33040
+    },
+    {
+      "epoch": 68.7006237006237,
+      "grad_norm": 1.4135208630250418e-06,
+      "learning_rate": 3.6389826256453457e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25818624,
+      "step": 33045
+    },
+    {
+      "epoch": 68.71101871101871,
+      "grad_norm": 5.6130156735889614e-05,
+      "learning_rate": 3.633883613075781e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25822496,
+      "step": 33050
+    },
+    {
+      "epoch": 68.72141372141373,
+      "grad_norm": 9.334395940641116e-07,
+      "learning_rate": 3.6287878954298693e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25826304,
+      "step": 33055
+    },
+    {
+      "epoch": 68.73180873180873,
+      "grad_norm": 4.469519808480982e-06,
+      "learning_rate": 3.6236954734934354e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25830144,
+      "step": 33060
+    },
+    {
+      "epoch": 68.74220374220374,
+      "grad_norm": 1.1552385331015103e-05,
+      "learning_rate": 3.618606348051784e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25834336,
+      "step": 33065
+    },
+    {
+      "epoch": 68.75259875259876,
+      "grad_norm": 0.13858139514923096,
+      "learning_rate": 3.6135205198897376e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 25838272,
+      "step": 33070
+    },
+    {
+      "epoch": 68.76299376299376,
+      "grad_norm": 7.228946924442425e-05,
+      "learning_rate": 3.6084379897915854e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25842176,
+      "step": 33075
+    },
+    {
+      "epoch": 68.77338877338877,
+      "grad_norm": 0.0002241785841761157,
+      "learning_rate": 3.6033587585411115e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25846144,
+      "step": 33080
+    },
+    {
+      "epoch": 68.78378378378379,
+      "grad_norm": 9.105042408918962e-07,
+      "learning_rate": 3.5982828269216117e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25849920,
+      "step": 33085
+    },
+    {
+      "epoch": 68.79417879417879,
+      "grad_norm": 0.13623416423797607,
+      "learning_rate": 3.593210195715843e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 25854016,
+      "step": 33090
+    },
+    {
+      "epoch": 68.8045738045738,
+      "grad_norm": 1.4237688446883112e-05,
+      "learning_rate": 3.5881408657060773e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25857952,
+      "step": 33095
+    },
+    {
+      "epoch": 68.81496881496882,
+      "grad_norm": 0.0001109786971937865,
+      "learning_rate": 3.583074837674075e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25861952,
+      "step": 33100
+    },
+    {
+      "epoch": 68.82536382536382,
+      "grad_norm": 0.0001505849650129676,
+      "learning_rate": 3.578012112401069e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25865856,
+      "step": 33105
+    },
+    {
+      "epoch": 68.83575883575884,
+      "grad_norm": 6.847103941254318e-05,
+      "learning_rate": 3.5729526906677996e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25869824,
+      "step": 33110
+    },
+    {
+      "epoch": 68.84615384615384,
+      "grad_norm": 5.014860107621644e-06,
+      "learning_rate": 3.5678965732545007e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25873568,
+      "step": 33115
+    },
+    {
+      "epoch": 68.85654885654886,
+      "grad_norm": 1.2742378885377548e-06,
+      "learning_rate": 3.562843760940876e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25877536,
+      "step": 33120
+    },
+    {
+      "epoch": 68.86694386694387,
+      "grad_norm": 1.94664630726038e-06,
+      "learning_rate": 3.5577942545061473e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25881504,
+      "step": 33125
+    },
+    {
+      "epoch": 68.87733887733887,
+      "grad_norm": 6.247585406526923e-05,
+      "learning_rate": 3.5527480547289967e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25885408,
+      "step": 33130
+    },
+    {
+      "epoch": 68.88773388773389,
+      "grad_norm": 2.208136265835492e-06,
+      "learning_rate": 3.547705162387624e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25889312,
+      "step": 33135
+    },
+    {
+      "epoch": 68.8981288981289,
+      "grad_norm": 6.079427294025663e-06,
+      "learning_rate": 3.542665578259699e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25893216,
+      "step": 33140
+    },
+    {
+      "epoch": 68.9085239085239,
+      "grad_norm": 2.6641771455615526e-06,
+      "learning_rate": 3.5376293031223945e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 25897120,
+      "step": 33145
+    },
+    {
+      "epoch": 68.91891891891892,
+      "grad_norm": 1.9423090634518303e-05,
+      "learning_rate": 3.5325963377523614e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25900960,
+      "step": 33150
+    },
+    {
+      "epoch": 68.92931392931393,
+      "grad_norm": 4.610044197761454e-05,
+      "learning_rate": 3.5275666829257536e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25904896,
+      "step": 33155
+    },
+    {
+      "epoch": 68.93970893970894,
+      "grad_norm": 4.178850304015214e-06,
+      "learning_rate": 3.5225403394181955e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25908864,
+      "step": 33160
+    },
+    {
+      "epoch": 68.95010395010395,
+      "grad_norm": 1.0427514098410029e-05,
+      "learning_rate": 3.517517308004828e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25912640,
+      "step": 33165
+    },
+    {
+      "epoch": 68.96049896049897,
+      "grad_norm": 9.030951332533732e-06,
+      "learning_rate": 3.512497589460251e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25916448,
+      "step": 33170
+    },
+    {
+      "epoch": 68.97089397089397,
+      "grad_norm": 8.414569492742885e-06,
+      "learning_rate": 3.5074811845585727e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 25920192,
+      "step": 33175
+    },
+    {
+      "epoch": 68.98128898128898,
+      "grad_norm": 9.66881771091721e-07,
+      "learning_rate": 3.5024680940733937e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 25924096,
+      "step": 33180
+    },
+    {
+      "epoch": 68.99168399168398,
+      "grad_norm": 7.433603741446859e-07,
+      "learning_rate": 3.4974583187777852e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25928160,
+      "step": 33185
+    },
+    {
+      "epoch": 69.002079002079,
+      "grad_norm": 7.967774763528723e-06,
+      "learning_rate": 3.4924518594443204e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25932176,
+      "step": 33190
+    },
+    {
+      "epoch": 69.01247401247402,
+      "grad_norm": 0.00010369920346420258,
+      "learning_rate": 3.4874487168450682e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 25936272,
+      "step": 33195
+    },
+    {
+      "epoch": 69.02286902286902,
+      "grad_norm": 1.708037302705634e-06,
+      "learning_rate": 3.482448891751558e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25940304,
+      "step": 33200
+    },
+    {
+      "epoch": 69.02286902286902,
+      "eval_loss": 0.5941133499145508,
+      "eval_runtime": 11.7191,
+      "eval_samples_per_second": 73.043,
+      "eval_steps_per_second": 18.261,
+      "num_input_tokens_seen": 25940304,
+      "step": 33200
+    },
+    {
+      "epoch": 69.03326403326403,
+      "grad_norm": 1.3825449514115462e-06,
+      "learning_rate": 3.477452384934843e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25944240,
+      "step": 33205
+    },
+    {
+      "epoch": 69.04365904365905,
+      "grad_norm": 1.216466557707463e-06,
+      "learning_rate": 3.472459197165434e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25947920,
+      "step": 33210
+    },
+    {
+      "epoch": 69.05405405405405,
+      "grad_norm": 1.8280092035638518e-06,
+      "learning_rate": 3.4674693292133518e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25951792,
+      "step": 33215
+    },
+    {
+      "epoch": 69.06444906444906,
+      "grad_norm": 6.840633886895375e-06,
+      "learning_rate": 3.4624827818480977e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25955760,
+      "step": 33220
+    },
+    {
+      "epoch": 69.07484407484408,
+      "grad_norm": 3.894581459462643e-05,
+      "learning_rate": 3.4574995558386474e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25959824,
+      "step": 33225
+    },
+    {
+      "epoch": 69.08523908523908,
+      "grad_norm": 2.852151283150306e-06,
+      "learning_rate": 3.452519651953487e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25963792,
+      "step": 33230
+    },
+    {
+      "epoch": 69.0956340956341,
+      "grad_norm": 7.117651875887532e-06,
+      "learning_rate": 3.447543070960585e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25967728,
+      "step": 33235
+    },
+    {
+      "epoch": 69.10602910602911,
+      "grad_norm": 1.7760030459612608e-06,
+      "learning_rate": 3.4425698136273778e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25971344,
+      "step": 33240
+    },
+    {
+      "epoch": 69.11642411642411,
+      "grad_norm": 7.093094609444961e-07,
+      "learning_rate": 3.437599880720821e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25975216,
+      "step": 33245
+    },
+    {
+      "epoch": 69.12681912681913,
+      "grad_norm": 2.331984433112666e-06,
+      "learning_rate": 3.4326332730073267e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25979024,
+      "step": 33250
+    },
+    {
+      "epoch": 69.13721413721414,
+      "grad_norm": 6.618569750571623e-05,
+      "learning_rate": 3.427669991252813e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25982896,
+      "step": 33255
+    },
+    {
+      "epoch": 69.14760914760915,
+      "grad_norm": 1.3322394352144329e-06,
+      "learning_rate": 3.42271003622269e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25986608,
+      "step": 33260
+    },
+    {
+      "epoch": 69.15800415800416,
+      "grad_norm": 0.1317053586244583,
+      "learning_rate": 3.4177534086818286e-06,
+      "loss": 0.0077,
+      "num_input_tokens_seen": 25990768,
+      "step": 33265
+    },
+    {
+      "epoch": 69.16839916839916,
+      "grad_norm": 0.0005964825977571309,
+      "learning_rate": 3.412800109394612e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25994576,
+      "step": 33270
+    },
+    {
+      "epoch": 69.17879417879418,
+      "grad_norm": 2.9494956834241748e-05,
+      "learning_rate": 3.4078501391249044e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 25998512,
+      "step": 33275
+    },
+    {
+      "epoch": 69.1891891891892,
+      "grad_norm": 0.00019018394232261926,
+      "learning_rate": 3.4029034986360453e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26002544,
+      "step": 33280
+    },
+    {
+      "epoch": 69.1995841995842,
+      "grad_norm": 1.151709852820204e-06,
+      "learning_rate": 3.397960188690877e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26006544,
+      "step": 33285
+    },
+    {
+      "epoch": 69.20997920997921,
+      "grad_norm": 4.000279204774415e-06,
+      "learning_rate": 3.393020210051717e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26010512,
+      "step": 33290
+    },
+    {
+      "epoch": 69.22037422037423,
+      "grad_norm": 4.298985913919751e-06,
+      "learning_rate": 3.3880835634803655e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26014416,
+      "step": 33295
+    },
+    {
+      "epoch": 69.23076923076923,
+      "grad_norm": 7.82572606112808e-05,
+      "learning_rate": 3.383150249738126e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26018320,
+      "step": 33300
+    },
+    {
+      "epoch": 69.24116424116424,
+      "grad_norm": 5.147911451786058e-06,
+      "learning_rate": 3.3782202695857663e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 26022224,
+      "step": 33305
+    },
+    {
+      "epoch": 69.25155925155926,
+      "grad_norm": 2.2627536964137107e-06,
+      "learning_rate": 3.373293623783558e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26026128,
+      "step": 33310
+    },
+    {
+      "epoch": 69.26195426195426,
+      "grad_norm": 2.7382368443795713e-06,
+      "learning_rate": 3.368370313091257e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26030000,
+      "step": 33315
+    },
+    {
+      "epoch": 69.27234927234927,
+      "grad_norm": 0.00012953949044458568,
+      "learning_rate": 3.363450338268087e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 26033904,
+      "step": 33320
+    },
+    {
+      "epoch": 69.28274428274429,
+      "grad_norm": 5.43791611562483e-06,
+      "learning_rate": 3.358533700072783e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26037808,
+      "step": 33325
+    },
+    {
+      "epoch": 69.29313929313929,
+      "grad_norm": 3.848133474093629e-06,
+      "learning_rate": 3.3536203992635377e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26041744,
+      "step": 33330
+    },
+    {
+      "epoch": 69.3035343035343,
+      "grad_norm": 1.919114538395661e-06,
+      "learning_rate": 3.348710436598057e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26045648,
+      "step": 33335
+    },
+    {
+      "epoch": 69.31392931392931,
+      "grad_norm": 4.3380361603340134e-05,
+      "learning_rate": 3.3438038128335155e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26049584,
+      "step": 33340
+    },
+    {
+      "epoch": 69.32432432432432,
+      "grad_norm": 4.069819624419324e-06,
+      "learning_rate": 3.338900528726571e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26053456,
+      "step": 33345
+    },
+    {
+      "epoch": 69.33471933471934,
+      "grad_norm": 0.00039053044747561216,
+      "learning_rate": 3.3340005850333812e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26057488,
+      "step": 33350
+    },
+    {
+      "epoch": 69.34511434511434,
+      "grad_norm": 1.2930577213410288e-06,
+      "learning_rate": 3.329103982509568e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26061296,
+      "step": 33355
+    },
+    {
+      "epoch": 69.35550935550935,
+      "grad_norm": 1.7567042505106656e-06,
+      "learning_rate": 3.324210721910259e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26065072,
+      "step": 33360
+    },
+    {
+      "epoch": 69.36590436590437,
+      "grad_norm": 1.721509215713013e-05,
+      "learning_rate": 3.319320803990053e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26068912,
+      "step": 33365
+    },
+    {
+      "epoch": 69.37629937629937,
+      "grad_norm": 5.73795887248707e-06,
+      "learning_rate": 3.3144342295030274e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26072880,
+      "step": 33370
+    },
+    {
+      "epoch": 69.38669438669439,
+      "grad_norm": 1.9868279196089134e-05,
+      "learning_rate": 3.309550999202765e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 26076784,
+      "step": 33375
+    },
+    {
+      "epoch": 69.3970893970894,
+      "grad_norm": 9.079850315174554e-06,
+      "learning_rate": 3.3046711138423197e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26080656,
+      "step": 33380
+    },
+    {
+      "epoch": 69.4074844074844,
+      "grad_norm": 4.811981398233911e-06,
+      "learning_rate": 3.2997945741742255e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26084432,
+      "step": 33385
+    },
+    {
+      "epoch": 69.41787941787942,
+      "grad_norm": 5.639399205392692e-06,
+      "learning_rate": 3.2949213809505082e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26088560,
+      "step": 33390
+    },
+    {
+      "epoch": 69.42827442827443,
+      "grad_norm": 0.0006859556888230145,
+      "learning_rate": 3.2900515349226834e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26092464,
+      "step": 33395
+    },
+    {
+      "epoch": 69.43866943866944,
+      "grad_norm": 0.00020767647947650403,
+      "learning_rate": 3.285185036841731e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 26096432,
+      "step": 33400
+    },
+    {
+      "epoch": 69.43866943866944,
+      "eval_loss": 0.5907233357429504,
+      "eval_runtime": 11.7253,
+      "eval_samples_per_second": 73.005,
+      "eval_steps_per_second": 18.251,
+      "num_input_tokens_seen": 26096432,
+      "step": 33400
+    },
+    {
+      "epoch": 69.44906444906445,
+      "grad_norm": 2.5031033601408126e-06,
+      "learning_rate": 3.2803218874581377e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26100368,
+      "step": 33405
+    },
+    {
+      "epoch": 69.45945945945945,
+      "grad_norm": 5.7356774050276726e-05,
+      "learning_rate": 3.2754620875218494e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26104304,
+      "step": 33410
+    },
+    {
+      "epoch": 69.46985446985447,
+      "grad_norm": 2.13178691410576e-06,
+      "learning_rate": 3.2706056377823146e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 26108208,
+      "step": 33415
+    },
+    {
+      "epoch": 69.48024948024948,
+      "grad_norm": 6.3773582041903865e-06,
+      "learning_rate": 3.2657525389884647e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26112112,
+      "step": 33420
+    },
+    {
+      "epoch": 69.49064449064448,
+      "grad_norm": 1.5233424164762255e-05,
+      "learning_rate": 3.260902791888698e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26115952,
+      "step": 33425
+    },
+    {
+      "epoch": 69.5010395010395,
+      "grad_norm": 3.6224200812284835e-06,
+      "learning_rate": 3.2560563972309166e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26119728,
+      "step": 33430
+    },
+    {
+      "epoch": 69.51143451143452,
+      "grad_norm": 1.0895541890931781e-05,
+      "learning_rate": 3.251213355762489e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26123696,
+      "step": 33435
+    },
+    {
+      "epoch": 69.52182952182952,
+      "grad_norm": 1.7230391677003354e-05,
+      "learning_rate": 3.2463736682302707e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26127632,
+      "step": 33440
+    },
+    {
+      "epoch": 69.53222453222453,
+      "grad_norm": 1.6410650687248562e-06,
+      "learning_rate": 3.2415373353806124e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26131504,
+      "step": 33445
+    },
+    {
+      "epoch": 69.54261954261955,
+      "grad_norm": 2.722611498029437e-06,
+      "learning_rate": 3.236704357959322e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26135408,
+      "step": 33450
+    },
+    {
+      "epoch": 69.55301455301455,
+      "grad_norm": 6.9991920099710114e-06,
+      "learning_rate": 3.2318747367117154e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 26139312,
+      "step": 33455
+    },
+    {
+      "epoch": 69.56340956340956,
+      "grad_norm": 3.3919397992576705e-06,
+      "learning_rate": 3.227048472382585e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26143248,
+      "step": 33460
+    },
+    {
+      "epoch": 69.57380457380458,
+      "grad_norm": 1.2252634178366861e-06,
+      "learning_rate": 3.2222255657161915e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26147248,
+      "step": 33465
+    },
+    {
+      "epoch": 69.58419958419958,
+      "grad_norm": 2.814805156958755e-06,
+      "learning_rate": 3.2174060174562924e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 26150960,
+      "step": 33470
+    },
+    {
+      "epoch": 69.5945945945946,
+      "grad_norm": 4.8826968850335106e-05,
+      "learning_rate": 3.2125898283461298e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26154960,
+      "step": 33475
+    },
+    {
+      "epoch": 69.60498960498961,
+      "grad_norm": 1.0004924661188852e-05,
+      "learning_rate": 3.207776999128406e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26158896,
+      "step": 33480
+    },
+    {
+      "epoch": 69.61538461538461,
+      "grad_norm": 2.4781088541203644e-06,
+      "learning_rate": 3.202967530545331e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26162800,
+      "step": 33485
+    },
+    {
+      "epoch": 69.62577962577963,
+      "grad_norm": 5.753408004238736e-06,
+      "learning_rate": 3.1981614233385778e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 26166864,
+      "step": 33490
+    },
+    {
+      "epoch": 69.63617463617463,
+      "grad_norm": 1.1194670150871389e-06,
+      "learning_rate": 3.1933586782493115e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26170704,
+      "step": 33495
+    },
+    {
+      "epoch": 69.64656964656965,
+      "grad_norm": 1.0782606523207505e-06,
+      "learning_rate": 3.188559296018184e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26174704,
+      "step": 33500
+    },
+    {
+      "epoch": 69.65696465696466,
+      "grad_norm": 0.0006050765514373779,
+      "learning_rate": 3.1837632773853098e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26178704,
+      "step": 33505
+    },
+    {
+      "epoch": 69.66735966735966,
+      "grad_norm": 5.822292905577342e-07,
+      "learning_rate": 3.178970623090294e-06,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 26182480,
+      "step": 33510
+    },
+    {
+      "epoch": 69.67775467775468,
+      "grad_norm": 3.2969826861517504e-05,
+      "learning_rate": 3.174181333872234e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26186416,
+      "step": 33515
+    },
+    {
+      "epoch": 69.6881496881497,
+      "grad_norm": 7.235627890622709e-06,
+      "learning_rate": 3.169395410469686e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26190480,
+      "step": 33520
+    },
+    {
+      "epoch": 69.6985446985447,
+      "grad_norm": 7.547889708803268e-06,
+      "learning_rate": 3.164612853620713e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26194416,
+      "step": 33525
+    },
+    {
+      "epoch": 69.70893970893971,
+      "grad_norm": 3.935169297619723e-05,
+      "learning_rate": 3.1598336640628333e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26198384,
+      "step": 33530
+    },
+    {
+      "epoch": 69.71933471933473,
+      "grad_norm": 0.00010943298548227176,
+      "learning_rate": 3.155057842533063e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 26202256,
+      "step": 33535
+    },
+    {
+      "epoch": 69.72972972972973,
+      "grad_norm": 5.58291674224165e-07,
+      "learning_rate": 3.1502853897678984e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26206256,
+      "step": 33540
+    },
+    {
+      "epoch": 69.74012474012474,
+      "grad_norm": 1.8359752402830054e-06,
+      "learning_rate": 3.1455163065033017e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26210032,
+      "step": 33545
+    },
+    {
+      "epoch": 69.75051975051976,
+      "grad_norm": 2.0462410248001106e-05,
+      "learning_rate": 3.140750593474734e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26214192,
+      "step": 33550
+    },
+    {
+      "epoch": 69.76091476091476,
+      "grad_norm": 1.2478098142310046e-05,
+      "learning_rate": 3.1359882514171294e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26218032,
+      "step": 33555
+    },
+    {
+      "epoch": 69.77130977130977,
+      "grad_norm": 6.672630115645006e-05,
+      "learning_rate": 3.1312292810648903e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26222032,
+      "step": 33560
+    },
+    {
+      "epoch": 69.78170478170478,
+      "grad_norm": 1.3027851082370034e-06,
+      "learning_rate": 3.1264736831519204e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26226000,
+      "step": 33565
+    },
+    {
+      "epoch": 69.79209979209979,
+      "grad_norm": 3.976995503762737e-05,
+      "learning_rate": 3.1217214584115863e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26230064,
+      "step": 33570
+    },
+    {
+      "epoch": 69.8024948024948,
+      "grad_norm": 8.571734610995918e-07,
+      "learning_rate": 3.116972607576746e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26233936,
+      "step": 33575
+    },
+    {
+      "epoch": 69.81288981288981,
+      "grad_norm": 9.203379818245594e-07,
+      "learning_rate": 3.1122271313797303e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26237872,
+      "step": 33580
+    },
+    {
+      "epoch": 69.82328482328482,
+      "grad_norm": 7.714883395237848e-06,
+      "learning_rate": 3.107485030552343e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26241680,
+      "step": 33585
+    },
+    {
+      "epoch": 69.83367983367984,
+      "grad_norm": 0.00022764215827919543,
+      "learning_rate": 3.1027463058258848e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26245552,
+      "step": 33590
+    },
+    {
+      "epoch": 69.84407484407484,
+      "grad_norm": 5.476734713738551e-06,
+      "learning_rate": 3.0980109579311273e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26249488,
+      "step": 33595
+    },
+    {
+      "epoch": 69.85446985446985,
+      "grad_norm": 4.031622040656657e-07,
+      "learning_rate": 3.093278987598314e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26253360,
+      "step": 33600
+    },
+    {
+      "epoch": 69.85446985446985,
+      "eval_loss": 0.5958399772644043,
+      "eval_runtime": 11.7159,
+      "eval_samples_per_second": 73.063,
+      "eval_steps_per_second": 18.266,
+      "num_input_tokens_seen": 26253360,
+      "step": 33600
+    },
+    {
+      "epoch": 69.86486486486487,
+      "grad_norm": 1.5003533007984515e-05,
+      "learning_rate": 3.0885503955571826e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26257168,
+      "step": 33605
+    },
+    {
+      "epoch": 69.87525987525987,
+      "grad_norm": 4.225147677061614e-06,
+      "learning_rate": 3.0838251825369313e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26260848,
+      "step": 33610
+    },
+    {
+      "epoch": 69.88565488565489,
+      "grad_norm": 0.00015897360572125763,
+      "learning_rate": 3.0791033492662517e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26264752,
+      "step": 33615
+    },
+    {
+      "epoch": 69.8960498960499,
+      "grad_norm": 2.039525134023279e-05,
+      "learning_rate": 3.0743848964733203e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26268912,
+      "step": 33620
+    },
+    {
+      "epoch": 69.9064449064449,
+      "grad_norm": 0.0018307652790099382,
+      "learning_rate": 3.0696698248857625e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 26272656,
+      "step": 33625
+    },
+    {
+      "epoch": 69.91683991683992,
+      "grad_norm": 1.2469662578951102e-05,
+      "learning_rate": 3.0649581352307192e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26276592,
+      "step": 33630
+    },
+    {
+      "epoch": 69.92723492723492,
+      "grad_norm": 4.0311060729436576e-05,
+      "learning_rate": 3.060249828234776e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26280432,
+      "step": 33635
+    },
+    {
+      "epoch": 69.93762993762994,
+      "grad_norm": 9.244735679203586e-07,
+      "learning_rate": 3.055544904624025e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26284304,
+      "step": 33640
+    },
+    {
+      "epoch": 69.94802494802495,
+      "grad_norm": 2.4597325136710424e-06,
+      "learning_rate": 3.050843365124026e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26288176,
+      "step": 33645
+    },
+    {
+      "epoch": 69.95841995841995,
+      "grad_norm": 7.673733489355072e-05,
+      "learning_rate": 3.0461452104598083e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26292048,
+      "step": 33650
+    },
+    {
+      "epoch": 69.96881496881497,
+      "grad_norm": 4.098429144505644e-06,
+      "learning_rate": 3.0414504413558836e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26296048,
+      "step": 33655
+    },
+    {
+      "epoch": 69.97920997920998,
+      "grad_norm": 4.291643108444987e-06,
+      "learning_rate": 3.0367590585362564e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26299952,
+      "step": 33660
+    },
+    {
+      "epoch": 69.98960498960498,
+      "grad_norm": 7.867114391046925e-07,
+      "learning_rate": 3.0320710627243813e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26303728,
+      "step": 33665
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 0.0011325186351314187,
+      "learning_rate": 3.027386454643222e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26307456,
+      "step": 33670
+    },
+    {
+      "epoch": 70.01039501039502,
+      "grad_norm": 0.0001460480416426435,
+      "learning_rate": 3.0227052350151914e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26311264,
+      "step": 33675
+    },
+    {
+      "epoch": 70.02079002079002,
+      "grad_norm": 3.3050439469661796e-06,
+      "learning_rate": 3.0180274045621957e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26315104,
+      "step": 33680
+    },
+    {
+      "epoch": 70.03118503118503,
+      "grad_norm": 1.1129865924885962e-05,
+      "learning_rate": 3.013352964005625e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26319168,
+      "step": 33685
+    },
+    {
+      "epoch": 70.04158004158005,
+      "grad_norm": 1.116291150538018e-05,
+      "learning_rate": 3.0086819140663218e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26323136,
+      "step": 33690
+    },
+    {
+      "epoch": 70.05197505197505,
+      "grad_norm": 1.0621608907968039e-06,
+      "learning_rate": 3.0040142554646265e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26326912,
+      "step": 33695
+    },
+    {
+      "epoch": 70.06237006237006,
+      "grad_norm": 3.646146069513634e-06,
+      "learning_rate": 2.999349988920361e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26330720,
+      "step": 33700
+    },
+    {
+      "epoch": 70.07276507276508,
+      "grad_norm": 7.293680027942173e-06,
+      "learning_rate": 2.994689115152796e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26334752,
+      "step": 33705
+    },
+    {
+      "epoch": 70.08316008316008,
+      "grad_norm": 5.463436173158698e-05,
+      "learning_rate": 2.9900316348807105e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26338496,
+      "step": 33710
+    },
+    {
+      "epoch": 70.0935550935551,
+      "grad_norm": 1.231785745403613e-06,
+      "learning_rate": 2.985377548822338e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26342496,
+      "step": 33715
+    },
+    {
+      "epoch": 70.1039501039501,
+      "grad_norm": 1.1039678611268755e-05,
+      "learning_rate": 2.980726857695404e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26346400,
+      "step": 33720
+    },
+    {
+      "epoch": 70.11434511434511,
+      "grad_norm": 0.00022025316138751805,
+      "learning_rate": 2.9760795622171017e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26350464,
+      "step": 33725
+    },
+    {
+      "epoch": 70.12474012474013,
+      "grad_norm": 7.479047781089321e-06,
+      "learning_rate": 2.971435663104094e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26354304,
+      "step": 33730
+    },
+    {
+      "epoch": 70.13513513513513,
+      "grad_norm": 9.096138819586486e-05,
+      "learning_rate": 2.9667951610725385e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26358080,
+      "step": 33735
+    },
+    {
+      "epoch": 70.14553014553015,
+      "grad_norm": 1.78638038050849e-06,
+      "learning_rate": 2.9621580568380575e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26361920,
+      "step": 33740
+    },
+    {
+      "epoch": 70.15592515592516,
+      "grad_norm": 0.13611924648284912,
+      "learning_rate": 2.9575243511157453e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26365664,
+      "step": 33745
+    },
+    {
+      "epoch": 70.16632016632016,
+      "grad_norm": 1.8992524246641551e-06,
+      "learning_rate": 2.952894044620186e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26369504,
+      "step": 33750
+    },
+    {
+      "epoch": 70.17671517671518,
+      "grad_norm": 5.956401946605183e-05,
+      "learning_rate": 2.948267138065419e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26373504,
+      "step": 33755
+    },
+    {
+      "epoch": 70.18711018711019,
+      "grad_norm": 0.00010234364162897691,
+      "learning_rate": 2.943643632164983e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 26377536,
+      "step": 33760
+    },
+    {
+      "epoch": 70.1975051975052,
+      "grad_norm": 0.0010959001956507564,
+      "learning_rate": 2.939023527631879e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26381376,
+      "step": 33765
+    },
+    {
+      "epoch": 70.20790020790021,
+      "grad_norm": 4.463891855266411e-06,
+      "learning_rate": 2.934406825178576e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26385280,
+      "step": 33770
+    },
+    {
+      "epoch": 70.21829521829522,
+      "grad_norm": 5.538709501706762e-06,
+      "learning_rate": 2.9297935255170357e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26389056,
+      "step": 33775
+    },
+    {
+      "epoch": 70.22869022869023,
+      "grad_norm": 9.207589073412237e-07,
+      "learning_rate": 2.925183629358691e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26393152,
+      "step": 33780
+    },
+    {
+      "epoch": 70.23908523908524,
+      "grad_norm": 6.037289949745173e-06,
+      "learning_rate": 2.9205771374144346e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26397088,
+      "step": 33785
+    },
+    {
+      "epoch": 70.24948024948024,
+      "grad_norm": 2.9919519874965772e-05,
+      "learning_rate": 2.915974050394657e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 26401024,
+      "step": 33790
+    },
+    {
+      "epoch": 70.25987525987526,
+      "grad_norm": 2.7944714020122774e-05,
+      "learning_rate": 2.9113743690092067e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 26404960,
+      "step": 33795
+    },
+    {
+      "epoch": 70.27027027027027,
+      "grad_norm": 3.186504545737989e-05,
+      "learning_rate": 2.906778093967402e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 26408736,
+      "step": 33800
+    },
+    {
+      "epoch": 70.27027027027027,
+      "eval_loss": 0.5947487354278564,
+      "eval_runtime": 11.7192,
+      "eval_samples_per_second": 73.042,
+      "eval_steps_per_second": 18.261,
+      "num_input_tokens_seen": 26408736,
+      "step": 33800
+    },
+    {
+      "epoch": 70.28066528066527,
+      "grad_norm": 1.1237541002628859e-05,
+      "learning_rate": 2.9021852259780656e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26412448,
+      "step": 33805
+    },
+    {
+      "epoch": 70.29106029106029,
+      "grad_norm": 1.100747067539487e-05,
+      "learning_rate": 2.8975957657494583e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26416352,
+      "step": 33810
+    },
+    {
+      "epoch": 70.3014553014553,
+      "grad_norm": 0.00011725974036380649,
+      "learning_rate": 2.8930097139893417e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26420352,
+      "step": 33815
+    },
+    {
+      "epoch": 70.3118503118503,
+      "grad_norm": 1.64428220159607e-05,
+      "learning_rate": 2.888427071404945e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26424480,
+      "step": 33820
+    },
+    {
+      "epoch": 70.32224532224532,
+      "grad_norm": 3.893434950441588e-06,
+      "learning_rate": 2.8838478387029606e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26428416,
+      "step": 33825
+    },
+    {
+      "epoch": 70.33264033264034,
+      "grad_norm": 1.5294912145691342e-06,
+      "learning_rate": 2.8792720165895737e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 26432256,
+      "step": 33830
+    },
+    {
+      "epoch": 70.34303534303534,
+      "grad_norm": 1.2529213563539088e-05,
+      "learning_rate": 2.874699605770423e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26436192,
+      "step": 33835
+    },
+    {
+      "epoch": 70.35343035343035,
+      "grad_norm": 6.3577081164112315e-06,
+      "learning_rate": 2.8701306069506383e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26440128,
+      "step": 33840
+    },
+    {
+      "epoch": 70.36382536382537,
+      "grad_norm": 2.7663720175041817e-05,
+      "learning_rate": 2.8655650208348178e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26444064,
+      "step": 33845
+    },
+    {
+      "epoch": 70.37422037422037,
+      "grad_norm": 3.3424639696022496e-05,
+      "learning_rate": 2.8610028481270257e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 26447936,
+      "step": 33850
+    },
+    {
+      "epoch": 70.38461538461539,
+      "grad_norm": 6.14042278357374e-07,
+      "learning_rate": 2.856444089530813e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26451904,
+      "step": 33855
+    },
+    {
+      "epoch": 70.39501039501039,
+      "grad_norm": 1.9684443941514473e-06,
+      "learning_rate": 2.8518887457491955e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 26455840,
+      "step": 33860
+    },
+    {
+      "epoch": 70.4054054054054,
+      "grad_norm": 5.774659075541422e-05,
+      "learning_rate": 2.8473368174846666e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26459552,
+      "step": 33865
+    },
+    {
+      "epoch": 70.41580041580042,
+      "grad_norm": 1.041109157995379e-06,
+      "learning_rate": 2.842788305439184e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26463328,
+      "step": 33870
+    },
+    {
+      "epoch": 70.42619542619542,
+      "grad_norm": 7.544911568402313e-06,
+      "learning_rate": 2.8382432103141925e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 26467232,
+      "step": 33875
+    },
+    {
+      "epoch": 70.43659043659044,
+      "grad_norm": 8.013236219994724e-05,
+      "learning_rate": 2.833701532810598e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26471200,
+      "step": 33880
+    },
+    {
+      "epoch": 70.44698544698545,
+      "grad_norm": 8.418841753154993e-05,
+      "learning_rate": 2.8291632736287877e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26475168,
+      "step": 33885
+    },
+    {
+      "epoch": 70.45738045738045,
+      "grad_norm": 6.125882805463334e-07,
+      "learning_rate": 2.824628433468615e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26479072,
+      "step": 33890
+    },
+    {
+      "epoch": 70.46777546777547,
+      "grad_norm": 3.800981403401238e-06,
+      "learning_rate": 2.8200970130294073e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26482944,
+      "step": 33895
+    },
+    {
+      "epoch": 70.47817047817048,
+      "grad_norm": 5.879008767806226e-06,
+      "learning_rate": 2.8155690130099775e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26486816,
+      "step": 33900
+    },
+    {
+      "epoch": 70.48856548856548,
+      "grad_norm": 4.4535758547681326e-07,
+      "learning_rate": 2.8110444341085895e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26490592,
+      "step": 33905
+    },
+    {
+      "epoch": 70.4989604989605,
+      "grad_norm": 1.7579686755198054e-05,
+      "learning_rate": 2.806523277022996e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26494496,
+      "step": 33910
+    },
+    {
+      "epoch": 70.50935550935552,
+      "grad_norm": 0.00015756304492242634,
+      "learning_rate": 2.802005542450409e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26498336,
+      "step": 33915
+    },
+    {
+      "epoch": 70.51975051975052,
+      "grad_norm": 4.270020144758746e-05,
+      "learning_rate": 2.797491231087526e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26502144,
+      "step": 33920
+    },
+    {
+      "epoch": 70.53014553014553,
+      "grad_norm": 0.0003723642439581454,
+      "learning_rate": 2.7929803436305137e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 26506048,
+      "step": 33925
+    },
+    {
+      "epoch": 70.54054054054055,
+      "grad_norm": 1.024580342345871e-05,
+      "learning_rate": 2.788472880774998e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 26510208,
+      "step": 33930
+    },
+    {
+      "epoch": 70.55093555093555,
+      "grad_norm": 0.00012203981168568134,
+      "learning_rate": 2.7839688432160977e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26514176,
+      "step": 33935
+    },
+    {
+      "epoch": 70.56133056133056,
+      "grad_norm": 2.5106184693868272e-05,
+      "learning_rate": 2.779468231648383e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 26517984,
+      "step": 33940
+    },
+    {
+      "epoch": 70.57172557172557,
+      "grad_norm": 2.6180778149864636e-06,
+      "learning_rate": 2.774971046765906e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26522016,
+      "step": 33945
+    },
+    {
+      "epoch": 70.58212058212058,
+      "grad_norm": 2.5739220291143283e-05,
+      "learning_rate": 2.770477289262194e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26525952,
+      "step": 33950
+    },
+    {
+      "epoch": 70.5925155925156,
+      "grad_norm": 3.2643340546201216e-06,
+      "learning_rate": 2.765986959830233e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26529952,
+      "step": 33955
+    },
+    {
+      "epoch": 70.6029106029106,
+      "grad_norm": 3.3527858249726705e-06,
+      "learning_rate": 2.761500059162492e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26533888,
+      "step": 33960
+    },
+    {
+      "epoch": 70.61330561330561,
+      "grad_norm": 2.541057710914174e-06,
+      "learning_rate": 2.757016587950914e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26537728,
+      "step": 33965
+    },
+    {
+      "epoch": 70.62370062370063,
+      "grad_norm": 0.0006374967633746564,
+      "learning_rate": 2.752536546886897e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26541728,
+      "step": 33970
+    },
+    {
+      "epoch": 70.63409563409563,
+      "grad_norm": 2.941665570688201e-06,
+      "learning_rate": 2.7480599366613234e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26545504,
+      "step": 33975
+    },
+    {
+      "epoch": 70.64449064449065,
+      "grad_norm": 5.095427331980318e-05,
+      "learning_rate": 2.7435867579645473e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26549376,
+      "step": 33980
+    },
+    {
+      "epoch": 70.65488565488566,
+      "grad_norm": 6.864409897389123e-06,
+      "learning_rate": 2.739117011486378e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26553312,
+      "step": 33985
+    },
+    {
+      "epoch": 70.66528066528066,
+      "grad_norm": 8.846793662087293e-07,
+      "learning_rate": 2.7346506979161216e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26557312,
+      "step": 33990
+    },
+    {
+      "epoch": 70.67567567567568,
+      "grad_norm": 8.088144340945291e-07,
+      "learning_rate": 2.7301878179425227e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26561152,
+      "step": 33995
+    },
+    {
+      "epoch": 70.68607068607069,
+      "grad_norm": 1.7763090909284074e-06,
+      "learning_rate": 2.7257283722538244e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26565056,
+      "step": 34000
+    },
+    {
+      "epoch": 70.68607068607069,
+      "eval_loss": 0.599452793598175,
+      "eval_runtime": 11.6991,
+      "eval_samples_per_second": 73.168,
+      "eval_steps_per_second": 18.292,
+      "num_input_tokens_seen": 26565056,
+      "step": 34000
+    },
+    {
+      "epoch": 70.6964656964657,
+      "grad_norm": 6.826692242611898e-07,
+      "learning_rate": 2.7212723615377326e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26569056,
+      "step": 34005
+    },
+    {
+      "epoch": 70.70686070686071,
+      "grad_norm": 5.0257431212230586e-06,
+      "learning_rate": 2.7168197864814145e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26572864,
+      "step": 34010
+    },
+    {
+      "epoch": 70.71725571725571,
+      "grad_norm": 1.8481856386642903e-05,
+      "learning_rate": 2.712370647771509e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26576672,
+      "step": 34015
+    },
+    {
+      "epoch": 70.72765072765073,
+      "grad_norm": 2.2677281776850577e-06,
+      "learning_rate": 2.707924946094137e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26580480,
+      "step": 34020
+    },
+    {
+      "epoch": 70.73804573804574,
+      "grad_norm": 1.78077862074133e-05,
+      "learning_rate": 2.7034826821348723e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26584512,
+      "step": 34025
+    },
+    {
+      "epoch": 70.74844074844074,
+      "grad_norm": 1.6504767700098455e-05,
+      "learning_rate": 2.6990438565787786e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 26588480,
+      "step": 34030
+    },
+    {
+      "epoch": 70.75883575883576,
+      "grad_norm": 2.726818593146163e-06,
+      "learning_rate": 2.6946084701103714e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26592352,
+      "step": 34035
+    },
+    {
+      "epoch": 70.76923076923077,
+      "grad_norm": 3.20157669193577e-05,
+      "learning_rate": 2.6901765234136428e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26596256,
+      "step": 34040
+    },
+    {
+      "epoch": 70.77962577962577,
+      "grad_norm": 4.4657399485004134e-06,
+      "learning_rate": 2.685748017172063e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26600288,
+      "step": 34045
+    },
+    {
+      "epoch": 70.79002079002079,
+      "grad_norm": 2.548387783463113e-05,
+      "learning_rate": 2.681322952068549e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26604224,
+      "step": 34050
+    },
+    {
+      "epoch": 70.8004158004158,
+      "grad_norm": 3.2493285289092455e-06,
+      "learning_rate": 2.6769013287855137e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26608000,
+      "step": 34055
+    },
+    {
+      "epoch": 70.8108108108108,
+      "grad_norm": 6.3487041188636795e-06,
+      "learning_rate": 2.6724831480048286e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26611808,
+      "step": 34060
+    },
+    {
+      "epoch": 70.82120582120582,
+      "grad_norm": 9.510478662377864e-07,
+      "learning_rate": 2.66806841040782e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26615712,
+      "step": 34065
+    },
+    {
+      "epoch": 70.83160083160084,
+      "grad_norm": 1.6297782394758542e-06,
+      "learning_rate": 2.6636571166753083e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26619616,
+      "step": 34070
+    },
+    {
+      "epoch": 70.84199584199584,
+      "grad_norm": 6.355268851621076e-05,
+      "learning_rate": 2.6592492674875598e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26623488,
+      "step": 34075
+    },
+    {
+      "epoch": 70.85239085239085,
+      "grad_norm": 1.6027615856728517e-06,
+      "learning_rate": 2.6548448635243305e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26627424,
+      "step": 34080
+    },
+    {
+      "epoch": 70.86278586278586,
+      "grad_norm": 7.09134724274918e-07,
+      "learning_rate": 2.650443905464828e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26631392,
+      "step": 34085
+    },
+    {
+      "epoch": 70.87318087318087,
+      "grad_norm": 4.281660494598327e-06,
+      "learning_rate": 2.646046393987739e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26635328,
+      "step": 34090
+    },
+    {
+      "epoch": 70.88357588357589,
+      "grad_norm": 7.67347046348732e-06,
+      "learning_rate": 2.64165232977121e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26639232,
+      "step": 34095
+    },
+    {
+      "epoch": 70.89397089397089,
+      "grad_norm": 0.15195953845977783,
+      "learning_rate": 2.6372617134928695e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 26643168,
+      "step": 34100
+    },
+    {
+      "epoch": 70.9043659043659,
+      "grad_norm": 0.001620941562578082,
+      "learning_rate": 2.6328745458297943e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26647232,
+      "step": 34105
+    },
+    {
+      "epoch": 70.91476091476092,
+      "grad_norm": 5.0806843319151085e-06,
+      "learning_rate": 2.6284908274585546e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26651264,
+      "step": 34110
+    },
+    {
+      "epoch": 70.92515592515592,
+      "grad_norm": 1.1396958825571346e-06,
+      "learning_rate": 2.6241105590551595e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26655264,
+      "step": 34115
+    },
+    {
+      "epoch": 70.93555093555094,
+      "grad_norm": 8.446039601039956e-07,
+      "learning_rate": 2.6197337412951105e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26659136,
+      "step": 34120
+    },
+    {
+      "epoch": 70.94594594594595,
+      "grad_norm": 1.4906358046573587e-05,
+      "learning_rate": 2.6153603748533705e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26663008,
+      "step": 34125
+    },
+    {
+      "epoch": 70.95634095634095,
+      "grad_norm": 8.013425031094812e-06,
+      "learning_rate": 2.6109904604043585e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26666880,
+      "step": 34130
+    },
+    {
+      "epoch": 70.96673596673597,
+      "grad_norm": 4.877811079495586e-07,
+      "learning_rate": 2.6066239986219765e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26670784,
+      "step": 34135
+    },
+    {
+      "epoch": 70.97713097713098,
+      "grad_norm": 1.9676135707413778e-06,
+      "learning_rate": 2.602260990179592e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26674592,
+      "step": 34140
+    },
+    {
+      "epoch": 70.98752598752598,
+      "grad_norm": 0.00032600463600829244,
+      "learning_rate": 2.5979014357500248e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 26678432,
+      "step": 34145
+    },
+    {
+      "epoch": 70.997920997921,
+      "grad_norm": 1.923937634273898e-05,
+      "learning_rate": 2.5935453360055844e-06,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 26682304,
+      "step": 34150
+    },
+    {
+      "epoch": 71.00831600831602,
+      "grad_norm": 0.00042291433783248067,
+      "learning_rate": 2.5891926916180283e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26686136,
+      "step": 34155
+    },
+    {
+      "epoch": 71.01871101871102,
+      "grad_norm": 0.12221365422010422,
+      "learning_rate": 2.5848435032585883e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 26689976,
+      "step": 34160
+    },
+    {
+      "epoch": 71.02910602910603,
+      "grad_norm": 1.2139430509705562e-06,
+      "learning_rate": 2.58049777159797e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 26693848,
+      "step": 34165
+    },
+    {
+      "epoch": 71.03950103950103,
+      "grad_norm": 3.3807264117058367e-05,
+      "learning_rate": 2.576155497306332e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26697720,
+      "step": 34170
+    },
+    {
+      "epoch": 71.04989604989605,
+      "grad_norm": 1.339218215434812e-05,
+      "learning_rate": 2.57181668105331e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 26701528,
+      "step": 34175
+    },
+    {
+      "epoch": 71.06029106029106,
+      "grad_norm": 9.994885658670682e-06,
+      "learning_rate": 2.567481323508014e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26705496,
+      "step": 34180
+    },
+    {
+      "epoch": 71.07068607068607,
+      "grad_norm": 5.384164865063212e-07,
+      "learning_rate": 2.5631494253389954e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26709336,
+      "step": 34185
+    },
+    {
+      "epoch": 71.08108108108108,
+      "grad_norm": 5.211734242038801e-05,
+      "learning_rate": 2.5588209872142997e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 26713208,
+      "step": 34190
+    },
+    {
+      "epoch": 71.0914760914761,
+      "grad_norm": 1.2715668162854854e-06,
+      "learning_rate": 2.5544960098014186e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26717240,
+      "step": 34195
+    },
+    {
+      "epoch": 71.1018711018711,
+      "grad_norm": 2.1154000933165662e-05,
+      "learning_rate": 2.550174493767318e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26721176,
+      "step": 34200
+    },
+    {
+      "epoch": 71.1018711018711,
+      "eval_loss": 0.59966641664505,
+      "eval_runtime": 11.6896,
+      "eval_samples_per_second": 73.228,
+      "eval_steps_per_second": 18.307,
+      "num_input_tokens_seen": 26721176,
+      "step": 34200
+    },
+    {
+      "epoch": 71.11226611226611,
+      "grad_norm": 1.3722216181122349e-06,
+      "learning_rate": 2.545856439778438e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26725016,
+      "step": 34205
+    },
+    {
+      "epoch": 71.12266112266113,
+      "grad_norm": 2.509261503291782e-05,
+      "learning_rate": 2.541541848500667e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26728824,
+      "step": 34210
+    },
+    {
+      "epoch": 71.13305613305613,
+      "grad_norm": 1.4907347576809116e-05,
+      "learning_rate": 2.5372307205993733e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26732536,
+      "step": 34215
+    },
+    {
+      "epoch": 71.14345114345114,
+      "grad_norm": 1.0053477126348298e-06,
+      "learning_rate": 2.5329230567393917e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26736472,
+      "step": 34220
+    },
+    {
+      "epoch": 71.15384615384616,
+      "grad_norm": 0.00014038941299077123,
+      "learning_rate": 2.5286188575850164e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26740440,
+      "step": 34225
+    },
+    {
+      "epoch": 71.16424116424116,
+      "grad_norm": 4.497819099924527e-05,
+      "learning_rate": 2.5243181237999984e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26744280,
+      "step": 34230
+    },
+    {
+      "epoch": 71.17463617463618,
+      "grad_norm": 1.6274532299576094e-06,
+      "learning_rate": 2.520020856047578e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 26748120,
+      "step": 34235
+    },
+    {
+      "epoch": 71.18503118503118,
+      "grad_norm": 2.4024535377975553e-05,
+      "learning_rate": 2.515727054990438e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26751896,
+      "step": 34240
+    },
+    {
+      "epoch": 71.1954261954262,
+      "grad_norm": 0.0005199215956963599,
+      "learning_rate": 2.511436721290747e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 26755768,
+      "step": 34245
+    },
+    {
+      "epoch": 71.20582120582121,
+      "grad_norm": 4.185702437098371e-06,
+      "learning_rate": 2.5071498556101164e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 26759672,
+      "step": 34250
+    },
+    {
+      "epoch": 71.21621621621621,
+      "grad_norm": 7.311681997634878e-07,
+      "learning_rate": 2.5028664586096485e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26763416,
+      "step": 34255
+    },
+    {
+      "epoch": 71.22661122661123,
+      "grad_norm": 1.6582503121753689e-06,
+      "learning_rate": 2.498586530949881e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26767256,
+      "step": 34260
+    },
+    {
+      "epoch": 71.23700623700624,
+      "grad_norm": 4.794423603016185e-06,
+      "learning_rate": 2.4943100732908427e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26771256,
+      "step": 34265
+    },
+    {
+      "epoch": 71.24740124740124,
+      "grad_norm": 4.874601927440381e-06,
+      "learning_rate": 2.4900370862920188e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26775128,
+      "step": 34270
+    },
+    {
+      "epoch": 71.25779625779626,
+      "grad_norm": 9.099974704440683e-06,
+      "learning_rate": 2.4857675706123518e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26779096,
+      "step": 34275
+    },
+    {
+      "epoch": 71.26819126819127,
+      "grad_norm": 6.746388407918857e-06,
+      "learning_rate": 2.4815015269102543e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26783032,
+      "step": 34280
+    },
+    {
+      "epoch": 71.27858627858627,
+      "grad_norm": 7.144227538447012e-07,
+      "learning_rate": 2.477238955843611e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 26787000,
+      "step": 34285
+    },
+    {
+      "epoch": 71.28898128898129,
+      "grad_norm": 0.0002837304200511426,
+      "learning_rate": 2.4729798580697573e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26790968,
+      "step": 34290
+    },
+    {
+      "epoch": 71.2993762993763,
+      "grad_norm": 1.4460995316767367e-06,
+      "learning_rate": 2.4687242342455034e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26794872,
+      "step": 34295
+    },
+    {
+      "epoch": 71.3097713097713,
+      "grad_norm": 4.353291296865791e-05,
+      "learning_rate": 2.4644720850271196e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26798744,
+      "step": 34300
+    },
+    {
+      "epoch": 71.32016632016632,
+      "grad_norm": 5.982238917567884e-07,
+      "learning_rate": 2.4602234110703364e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26802744,
+      "step": 34305
+    },
+    {
+      "epoch": 71.33056133056132,
+      "grad_norm": 3.028855871889391e-06,
+      "learning_rate": 2.4559782130303576e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26806680,
+      "step": 34310
+    },
+    {
+      "epoch": 71.34095634095634,
+      "grad_norm": 8.525146881765977e-07,
+      "learning_rate": 2.451736491561843e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26810616,
+      "step": 34315
+    },
+    {
+      "epoch": 71.35135135135135,
+      "grad_norm": 2.7332578156347154e-06,
+      "learning_rate": 2.4474982473189163e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 26814616,
+      "step": 34320
+    },
+    {
+      "epoch": 71.36174636174636,
+      "grad_norm": 0.00016965129179880023,
+      "learning_rate": 2.4432634809551796e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26818712,
+      "step": 34325
+    },
+    {
+      "epoch": 71.37214137214137,
+      "grad_norm": 4.593130142893642e-06,
+      "learning_rate": 2.439032193123675e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26822616,
+      "step": 34330
+    },
+    {
+      "epoch": 71.38253638253639,
+      "grad_norm": 1.3398534974840004e-06,
+      "learning_rate": 2.4348043844769297e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26826520,
+      "step": 34335
+    },
+    {
+      "epoch": 71.39293139293139,
+      "grad_norm": 4.6545305849576835e-06,
+      "learning_rate": 2.4305800556669146e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26830488,
+      "step": 34340
+    },
+    {
+      "epoch": 71.4033264033264,
+      "grad_norm": 1.3319358913577162e-06,
+      "learning_rate": 2.426359207345083e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 26834456,
+      "step": 34345
+    },
+    {
+      "epoch": 71.41372141372142,
+      "grad_norm": 2.6363381039118394e-05,
+      "learning_rate": 2.4221418401623396e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26838264,
+      "step": 34350
+    },
+    {
+      "epoch": 71.42411642411642,
+      "grad_norm": 1.1472714504634496e-05,
+      "learning_rate": 2.4179279547690557e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26842072,
+      "step": 34355
+    },
+    {
+      "epoch": 71.43451143451144,
+      "grad_norm": 2.1382641079981113e-06,
+      "learning_rate": 2.413717551815062e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26846104,
+      "step": 34360
+    },
+    {
+      "epoch": 71.44490644490645,
+      "grad_norm": 8.405842777392536e-07,
+      "learning_rate": 2.409510631949666e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26850104,
+      "step": 34365
+    },
+    {
+      "epoch": 71.45530145530145,
+      "grad_norm": 2.307179329363862e-06,
+      "learning_rate": 2.405307195821618e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26854168,
+      "step": 34370
+    },
+    {
+      "epoch": 71.46569646569647,
+      "grad_norm": 4.206515586702153e-05,
+      "learning_rate": 2.4011072440791372e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26858008,
+      "step": 34375
+    },
+    {
+      "epoch": 71.47609147609148,
+      "grad_norm": 8.110079761536326e-06,
+      "learning_rate": 2.3969107773699233e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26861944,
+      "step": 34380
+    },
+    {
+      "epoch": 71.48648648648648,
+      "grad_norm": 3.994296639575623e-05,
+      "learning_rate": 2.3927177963411096e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 26865816,
+      "step": 34385
+    },
+    {
+      "epoch": 71.4968814968815,
+      "grad_norm": 9.350740583613515e-06,
+      "learning_rate": 2.3885283016393144e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26869720,
+      "step": 34390
+    },
+    {
+      "epoch": 71.5072765072765,
+      "grad_norm": 1.2289501682971604e-06,
+      "learning_rate": 2.3843422939106076e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26873528,
+      "step": 34395
+    },
+    {
+      "epoch": 71.51767151767152,
+      "grad_norm": 7.152313173719449e-06,
+      "learning_rate": 2.380159773800525e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26877368,
+      "step": 34400
+    },
+    {
+      "epoch": 71.51767151767152,
+      "eval_loss": 0.5993248820304871,
+      "eval_runtime": 11.713,
+      "eval_samples_per_second": 73.081,
+      "eval_steps_per_second": 18.27,
+      "num_input_tokens_seen": 26877368,
+      "step": 34400
+    },
+    {
+      "epoch": 71.52806652806653,
+      "grad_norm": 2.676252051969641e-06,
+      "learning_rate": 2.3759807419540675e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26881336,
+      "step": 34405
+    },
+    {
+      "epoch": 71.53846153846153,
+      "grad_norm": 0.13141360878944397,
+      "learning_rate": 2.3718051990156835e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 26885240,
+      "step": 34410
+    },
+    {
+      "epoch": 71.54885654885655,
+      "grad_norm": 2.1164353256608592e-06,
+      "learning_rate": 2.367633145629311e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26889240,
+      "step": 34415
+    },
+    {
+      "epoch": 71.55925155925156,
+      "grad_norm": 8.9633740572026e-06,
+      "learning_rate": 2.363464582438316e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26892984,
+      "step": 34420
+    },
+    {
+      "epoch": 71.56964656964657,
+      "grad_norm": 5.172231340111466e-06,
+      "learning_rate": 2.3592995100855526e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26896824,
+      "step": 34425
+    },
+    {
+      "epoch": 71.58004158004158,
+      "grad_norm": 6.941776518942788e-06,
+      "learning_rate": 2.3551379292133273e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26900824,
+      "step": 34430
+    },
+    {
+      "epoch": 71.5904365904366,
+      "grad_norm": 2.889603138100938e-06,
+      "learning_rate": 2.3509798404634047e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 26904728,
+      "step": 34435
+    },
+    {
+      "epoch": 71.6008316008316,
+      "grad_norm": 4.19814205088187e-06,
+      "learning_rate": 2.346825244477019e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26908536,
+      "step": 34440
+    },
+    {
+      "epoch": 71.61122661122661,
+      "grad_norm": 1.9831925328617217e-06,
+      "learning_rate": 2.3426741418948545e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 26912536,
+      "step": 34445
+    },
+    {
+      "epoch": 71.62162162162163,
+      "grad_norm": 0.0013164922129362822,
+      "learning_rate": 2.3385265333570715e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26916408,
+      "step": 34450
+    },
+    {
+      "epoch": 71.63201663201663,
+      "grad_norm": 3.206542942280066e-06,
+      "learning_rate": 2.334382419503278e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26920280,
+      "step": 34455
+    },
+    {
+      "epoch": 71.64241164241164,
+      "grad_norm": 4.313921635912266e-06,
+      "learning_rate": 2.3302418009725465e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26924056,
+      "step": 34460
+    },
+    {
+      "epoch": 71.65280665280665,
+      "grad_norm": 0.00010934123565675691,
+      "learning_rate": 2.326104678403415e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26928056,
+      "step": 34465
+    },
+    {
+      "epoch": 71.66320166320166,
+      "grad_norm": 3.5727478007174795e-06,
+      "learning_rate": 2.321971052433883e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26932056,
+      "step": 34470
+    },
+    {
+      "epoch": 71.67359667359668,
+      "grad_norm": 7.475659367628396e-05,
+      "learning_rate": 2.3178409237014004e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26935960,
+      "step": 34475
+    },
+    {
+      "epoch": 71.68399168399168,
+      "grad_norm": 6.286681468736788e-07,
+      "learning_rate": 2.313714292842889e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 26939832,
+      "step": 34480
+    },
+    {
+      "epoch": 71.6943866943867,
+      "grad_norm": 1.2072207937308121e-05,
+      "learning_rate": 2.309591160494734e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 26943800,
+      "step": 34485
+    },
+    {
+      "epoch": 71.70478170478171,
+      "grad_norm": 7.861108315410092e-06,
+      "learning_rate": 2.305471527292763e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26947672,
+      "step": 34490
+    },
+    {
+      "epoch": 71.71517671517671,
+      "grad_norm": 1.0289572855981532e-05,
+      "learning_rate": 2.3013553938722817e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26951640,
+      "step": 34495
+    },
+    {
+      "epoch": 71.72557172557173,
+      "grad_norm": 1.1548624570423272e-05,
+      "learning_rate": 2.297242760868043e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26955800,
+      "step": 34500
+    },
+    {
+      "epoch": 71.73596673596674,
+      "grad_norm": 4.104051913600415e-05,
+      "learning_rate": 2.2931336289142735e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26959704,
+      "step": 34505
+    },
+    {
+      "epoch": 71.74636174636174,
+      "grad_norm": 9.031322406372055e-06,
+      "learning_rate": 2.289027998644655e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 26963544,
+      "step": 34510
+    },
+    {
+      "epoch": 71.75675675675676,
+      "grad_norm": 2.41529096456361e-06,
+      "learning_rate": 2.2849258706923228e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26967640,
+      "step": 34515
+    },
+    {
+      "epoch": 71.76715176715177,
+      "grad_norm": 5.322509878169512e-06,
+      "learning_rate": 2.2808272456898705e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26971512,
+      "step": 34520
+    },
+    {
+      "epoch": 71.77754677754677,
+      "grad_norm": 6.23915548203513e-06,
+      "learning_rate": 2.2767321242693707e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26975416,
+      "step": 34525
+    },
+    {
+      "epoch": 71.78794178794179,
+      "grad_norm": 5.40504231594241e-07,
+      "learning_rate": 2.272640507062329e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26979384,
+      "step": 34530
+    },
+    {
+      "epoch": 71.7983367983368,
+      "grad_norm": 4.117750904697459e-06,
+      "learning_rate": 2.2685523946997382e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26983384,
+      "step": 34535
+    },
+    {
+      "epoch": 71.8087318087318,
+      "grad_norm": 1.1685389836202376e-05,
+      "learning_rate": 2.2644677878120245e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26987320,
+      "step": 34540
+    },
+    {
+      "epoch": 71.81912681912682,
+      "grad_norm": 4.42240070697153e-06,
+      "learning_rate": 2.2603866870290897e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 26991256,
+      "step": 34545
+    },
+    {
+      "epoch": 71.82952182952182,
+      "grad_norm": 6.101736289565451e-07,
+      "learning_rate": 2.256309092980294e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26995224,
+      "step": 34550
+    },
+    {
+      "epoch": 71.83991683991684,
+      "grad_norm": 5.3714578825747594e-05,
+      "learning_rate": 2.252235006294448e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 26999160,
+      "step": 34555
+    },
+    {
+      "epoch": 71.85031185031185,
+      "grad_norm": 1.4749822412341018e-06,
+      "learning_rate": 2.2481644275998333e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 27002904,
+      "step": 34560
+    },
+    {
+      "epoch": 71.86070686070686,
+      "grad_norm": 7.507481495849788e-05,
+      "learning_rate": 2.2440973575241832e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27006840,
+      "step": 34565
+    },
+    {
+      "epoch": 71.87110187110187,
+      "grad_norm": 2.2979963887337362e-06,
+      "learning_rate": 2.240033796694685e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27010584,
+      "step": 34570
+    },
+    {
+      "epoch": 71.88149688149689,
+      "grad_norm": 4.745233673020266e-05,
+      "learning_rate": 2.235973745737999e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27014424,
+      "step": 34575
+    },
+    {
+      "epoch": 71.89189189189189,
+      "grad_norm": 0.00010971736628562212,
+      "learning_rate": 2.2319172052802263e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27018296,
+      "step": 34580
+    },
+    {
+      "epoch": 71.9022869022869,
+      "grad_norm": 4.742832970805466e-06,
+      "learning_rate": 2.2278641759469477e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27022136,
+      "step": 34585
+    },
+    {
+      "epoch": 71.91268191268192,
+      "grad_norm": 1.0577108696452342e-05,
+      "learning_rate": 2.2238146583631825e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27026040,
+      "step": 34590
+    },
+    {
+      "epoch": 71.92307692307692,
+      "grad_norm": 6.492549800896086e-06,
+      "learning_rate": 2.2197686531534256e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27030104,
+      "step": 34595
+    },
+    {
+      "epoch": 71.93347193347194,
+      "grad_norm": 3.3467047614976764e-05,
+      "learning_rate": 2.2157261609416087e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27033912,
+      "step": 34600
+    },
+    {
+      "epoch": 71.93347193347194,
+      "eval_loss": 0.5984647274017334,
+      "eval_runtime": 11.7193,
+      "eval_samples_per_second": 73.042,
+      "eval_steps_per_second": 18.26,
+      "num_input_tokens_seen": 27033912,
+      "step": 34600
+    },
+    {
+      "epoch": 71.94386694386695,
+      "grad_norm": 7.778358849463984e-06,
+      "learning_rate": 2.211687182351149e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 27037752,
+      "step": 34605
+    },
+    {
+      "epoch": 71.95426195426195,
+      "grad_norm": 1.0151228480026475e-06,
+      "learning_rate": 2.2076517180048993e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27041592,
+      "step": 34610
+    },
+    {
+      "epoch": 71.96465696465697,
+      "grad_norm": 3.7155765312490985e-05,
+      "learning_rate": 2.2036197685251834e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27045432,
+      "step": 34615
+    },
+    {
+      "epoch": 71.97505197505197,
+      "grad_norm": 6.56778581742401e-07,
+      "learning_rate": 2.199591334533771e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27049304,
+      "step": 34620
+    },
+    {
+      "epoch": 71.98544698544698,
+      "grad_norm": 9.662520460551605e-05,
+      "learning_rate": 2.1955664166519036e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27053240,
+      "step": 34625
+    },
+    {
+      "epoch": 71.995841995842,
+      "grad_norm": 0.00016199327365029603,
+      "learning_rate": 2.1915450155002793e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27057208,
+      "step": 34630
+    },
+    {
+      "epoch": 72.006237006237,
+      "grad_norm": 9.214114470523782e-06,
+      "learning_rate": 2.187527131699038e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27061032,
+      "step": 34635
+    },
+    {
+      "epoch": 72.01663201663202,
+      "grad_norm": 1.7203899460582761e-06,
+      "learning_rate": 2.18351276586779e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 27064904,
+      "step": 34640
+    },
+    {
+      "epoch": 72.02702702702703,
+      "grad_norm": 2.3019015316094737e-06,
+      "learning_rate": 2.1795019186256092e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27068968,
+      "step": 34645
+    },
+    {
+      "epoch": 72.03742203742203,
+      "grad_norm": 2.0117074654990574e-06,
+      "learning_rate": 2.1754945905910094e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27073096,
+      "step": 34650
+    },
+    {
+      "epoch": 72.04781704781705,
+      "grad_norm": 1.298480128752999e-05,
+      "learning_rate": 2.171490782381977e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27076904,
+      "step": 34655
+    },
+    {
+      "epoch": 72.05821205821206,
+      "grad_norm": 0.05883011594414711,
+      "learning_rate": 2.1674904946159425e-06,
+      "loss": 0.005,
+      "num_input_tokens_seen": 27080776,
+      "step": 34660
+    },
+    {
+      "epoch": 72.06860706860707,
+      "grad_norm": 1.1503072528284974e-05,
+      "learning_rate": 2.16349372790981e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27084680,
+      "step": 34665
+    },
+    {
+      "epoch": 72.07900207900208,
+      "grad_norm": 1.8279847608937416e-06,
+      "learning_rate": 2.159500482879928e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27088584,
+      "step": 34670
+    },
+    {
+      "epoch": 72.0893970893971,
+      "grad_norm": 9.216806233780517e-07,
+      "learning_rate": 2.155510760142096e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 27092456,
+      "step": 34675
+    },
+    {
+      "epoch": 72.0997920997921,
+      "grad_norm": 4.890066065854626e-06,
+      "learning_rate": 2.151524560311588e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27096488,
+      "step": 34680
+    },
+    {
+      "epoch": 72.11018711018711,
+      "grad_norm": 2.684035735001089e-06,
+      "learning_rate": 2.147541884003129e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27100520,
+      "step": 34685
+    },
+    {
+      "epoch": 72.12058212058211,
+      "grad_norm": 1.9487158624542644e-06,
+      "learning_rate": 2.1435627318308895e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27104520,
+      "step": 34690
+    },
+    {
+      "epoch": 72.13097713097713,
+      "grad_norm": 1.8318844468012685e-06,
+      "learning_rate": 2.139587104408511e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27108360,
+      "step": 34695
+    },
+    {
+      "epoch": 72.14137214137214,
+      "grad_norm": 0.0008450868772342801,
+      "learning_rate": 2.1356150023490783e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 27112328,
+      "step": 34700
+    },
+    {
+      "epoch": 72.15176715176715,
+      "grad_norm": 1.337514845545229e-06,
+      "learning_rate": 2.1316464262651464e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27116104,
+      "step": 34705
+    },
+    {
+      "epoch": 72.16216216216216,
+      "grad_norm": 0.00011517745588207617,
+      "learning_rate": 2.1276813767687224e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 27119912,
+      "step": 34710
+    },
+    {
+      "epoch": 72.17255717255718,
+      "grad_norm": 9.194783160637598e-06,
+      "learning_rate": 2.123719854471254e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27123880,
+      "step": 34715
+    },
+    {
+      "epoch": 72.18295218295218,
+      "grad_norm": 6.108213028710452e-07,
+      "learning_rate": 2.119761859983668e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 27127720,
+      "step": 34720
+    },
+    {
+      "epoch": 72.1933471933472,
+      "grad_norm": 1.8668304619495757e-05,
+      "learning_rate": 2.1158073939163386e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27131528,
+      "step": 34725
+    },
+    {
+      "epoch": 72.20374220374221,
+      "grad_norm": 2.1376923541538417e-05,
+      "learning_rate": 2.111856456879088e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 27135528,
+      "step": 34730
+    },
+    {
+      "epoch": 72.21413721413721,
+      "grad_norm": 1.67161294939433e-06,
+      "learning_rate": 2.1079090494811993e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27139560,
+      "step": 34735
+    },
+    {
+      "epoch": 72.22453222453223,
+      "grad_norm": 0.051378652453422546,
+      "learning_rate": 2.103965172331418e-06,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 27143368,
+      "step": 34740
+    },
+    {
+      "epoch": 72.23492723492724,
+      "grad_norm": 6.981779279158218e-06,
+      "learning_rate": 2.100024826037933e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27147432,
+      "step": 34745
+    },
+    {
+      "epoch": 72.24532224532224,
+      "grad_norm": 0.0005552049842663109,
+      "learning_rate": 2.0960880112084027e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27151432,
+      "step": 34750
+    },
+    {
+      "epoch": 72.25571725571726,
+      "grad_norm": 1.1374082760085003e-06,
+      "learning_rate": 2.092154728449927e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27155176,
+      "step": 34755
+    },
+    {
+      "epoch": 72.26611226611226,
+      "grad_norm": 8.376459845749196e-06,
+      "learning_rate": 2.0882249783690687e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27158856,
+      "step": 34760
+    },
+    {
+      "epoch": 72.27650727650727,
+      "grad_norm": 1.3363397783905384e-06,
+      "learning_rate": 2.084298761571851e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27162952,
+      "step": 34765
+    },
+    {
+      "epoch": 72.28690228690229,
+      "grad_norm": 1.6286386426145327e-06,
+      "learning_rate": 2.080376078663737e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27166856,
+      "step": 34770
+    },
+    {
+      "epoch": 72.29729729729729,
+      "grad_norm": 6.427235348382965e-06,
+      "learning_rate": 2.0764569302496593e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27170856,
+      "step": 34775
+    },
+    {
+      "epoch": 72.3076923076923,
+      "grad_norm": 0.0002537247200962156,
+      "learning_rate": 2.0725413169339957e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27174664,
+      "step": 34780
+    },
+    {
+      "epoch": 72.31808731808732,
+      "grad_norm": 9.279046935262159e-06,
+      "learning_rate": 2.068629239320588e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27178600,
+      "step": 34785
+    },
+    {
+      "epoch": 72.32848232848232,
+      "grad_norm": 1.0036291314463597e-05,
+      "learning_rate": 2.064720698012726e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27182504,
+      "step": 34790
+    },
+    {
+      "epoch": 72.33887733887734,
+      "grad_norm": 2.3073412194207776e-06,
+      "learning_rate": 2.0608156936131522e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27186312,
+      "step": 34795
+    },
+    {
+      "epoch": 72.34927234927235,
+      "grad_norm": 3.393776978555252e-06,
+      "learning_rate": 2.056914226724074e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27190376,
+      "step": 34800
+    },
+    {
+      "epoch": 72.34927234927235,
+      "eval_loss": 0.6059231162071228,
+      "eval_runtime": 11.6824,
+      "eval_samples_per_second": 73.273,
+      "eval_steps_per_second": 18.318,
+      "num_input_tokens_seen": 27190376,
+      "step": 34800
+    },
+    {
+      "epoch": 72.35966735966736,
+      "grad_norm": 4.3155273488082457e-07,
+      "learning_rate": 2.0530162979471385e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27194120,
+      "step": 34805
+    },
+    {
+      "epoch": 72.37006237006237,
+      "grad_norm": 2.3315589714911766e-05,
+      "learning_rate": 2.0491219078834667e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27198152,
+      "step": 34810
+    },
+    {
+      "epoch": 72.38045738045739,
+      "grad_norm": 2.601944515845389e-06,
+      "learning_rate": 2.045231057133612e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27201960,
+      "step": 34815
+    },
+    {
+      "epoch": 72.39085239085239,
+      "grad_norm": 1.2529436389741022e-05,
+      "learning_rate": 2.0413437462975944e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27205896,
+      "step": 34820
+    },
+    {
+      "epoch": 72.4012474012474,
+      "grad_norm": 1.268406595045235e-05,
+      "learning_rate": 2.0374599759748843e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27209800,
+      "step": 34825
+    },
+    {
+      "epoch": 72.41164241164242,
+      "grad_norm": 0.00011196268314961344,
+      "learning_rate": 2.033579746764419e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27213640,
+      "step": 34830
+    },
+    {
+      "epoch": 72.42203742203742,
+      "grad_norm": 0.0003093570994678885,
+      "learning_rate": 2.029703059264565e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27217544,
+      "step": 34835
+    },
+    {
+      "epoch": 72.43243243243244,
+      "grad_norm": 0.0008452896727249026,
+      "learning_rate": 2.02582991407316e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27221384,
+      "step": 34840
+    },
+    {
+      "epoch": 72.44282744282744,
+      "grad_norm": 0.0002534343511797488,
+      "learning_rate": 2.0219603117874992e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27225384,
+      "step": 34845
+    },
+    {
+      "epoch": 72.45322245322245,
+      "grad_norm": 6.955231583560817e-06,
+      "learning_rate": 2.0180942530043156e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27229448,
+      "step": 34850
+    },
+    {
+      "epoch": 72.46361746361747,
+      "grad_norm": 1.0284438758390024e-06,
+      "learning_rate": 2.0142317383198107e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27233448,
+      "step": 34855
+    },
+    {
+      "epoch": 72.47401247401247,
+      "grad_norm": 1.453430286346702e-06,
+      "learning_rate": 2.0103727683296243e-06,
+      "loss": 0.005,
+      "num_input_tokens_seen": 27237384,
+      "step": 34860
+    },
+    {
+      "epoch": 72.48440748440748,
+      "grad_norm": 6.596599178010365e-07,
+      "learning_rate": 2.0065173436288636e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27241320,
+      "step": 34865
+    },
+    {
+      "epoch": 72.4948024948025,
+      "grad_norm": 3.6931680369889364e-05,
+      "learning_rate": 2.002665464812087e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27245288,
+      "step": 34870
+    },
+    {
+      "epoch": 72.5051975051975,
+      "grad_norm": 4.343422915553674e-06,
+      "learning_rate": 1.998817132473291e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27249224,
+      "step": 34875
+    },
+    {
+      "epoch": 72.51559251559252,
+      "grad_norm": 9.31748581933789e-06,
+      "learning_rate": 1.9949723472059507e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27253032,
+      "step": 34880
+    },
+    {
+      "epoch": 72.52598752598753,
+      "grad_norm": 8.162568701663986e-05,
+      "learning_rate": 1.9911311096029726e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27256936,
+      "step": 34885
+    },
+    {
+      "epoch": 72.53638253638253,
+      "grad_norm": 1.5279045328497887e-05,
+      "learning_rate": 1.9872934202567224e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27260872,
+      "step": 34890
+    },
+    {
+      "epoch": 72.54677754677755,
+      "grad_norm": 9.706949640531093e-06,
+      "learning_rate": 1.9834592797590257e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27265000,
+      "step": 34895
+    },
+    {
+      "epoch": 72.55717255717256,
+      "grad_norm": 4.266621829174255e-07,
+      "learning_rate": 1.979628688701149e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27268872,
+      "step": 34900
+    },
+    {
+      "epoch": 72.56756756756756,
+      "grad_norm": 1.2904257346235681e-05,
+      "learning_rate": 1.9758016476738193e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27272616,
+      "step": 34905
+    },
+    {
+      "epoch": 72.57796257796258,
+      "grad_norm": 2.2262845504883444e-06,
+      "learning_rate": 1.971978157267221e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27276520,
+      "step": 34910
+    },
+    {
+      "epoch": 72.58835758835758,
+      "grad_norm": 1.7050302858478972e-06,
+      "learning_rate": 1.968158218070973e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27280520,
+      "step": 34915
+    },
+    {
+      "epoch": 72.5987525987526,
+      "grad_norm": 7.418652216983901e-07,
+      "learning_rate": 1.9643418306741682e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27284392,
+      "step": 34920
+    },
+    {
+      "epoch": 72.60914760914761,
+      "grad_norm": 9.836894605541602e-05,
+      "learning_rate": 1.9605289956653337e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27288456,
+      "step": 34925
+    },
+    {
+      "epoch": 72.61954261954261,
+      "grad_norm": 6.040081643732265e-07,
+      "learning_rate": 1.9567197136324626e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 27292456,
+      "step": 34930
+    },
+    {
+      "epoch": 72.62993762993763,
+      "grad_norm": 4.018890467705205e-05,
+      "learning_rate": 1.9529139851629935e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27296328,
+      "step": 34935
+    },
+    {
+      "epoch": 72.64033264033264,
+      "grad_norm": 6.993508350205957e-07,
+      "learning_rate": 1.949111810843812e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27300104,
+      "step": 34940
+    },
+    {
+      "epoch": 72.65072765072765,
+      "grad_norm": 1.5028912230263813e-06,
+      "learning_rate": 1.9453131912612694e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27304040,
+      "step": 34945
+    },
+    {
+      "epoch": 72.66112266112266,
+      "grad_norm": 0.13688279688358307,
+      "learning_rate": 1.941518127001149e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 27307848,
+      "step": 34950
+    },
+    {
+      "epoch": 72.67151767151768,
+      "grad_norm": 7.175692735472694e-05,
+      "learning_rate": 1.9377266186487107e-06,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 27311560,
+      "step": 34955
+    },
+    {
+      "epoch": 72.68191268191268,
+      "grad_norm": 6.154900620458648e-05,
+      "learning_rate": 1.9339386667886483e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27315528,
+      "step": 34960
+    },
+    {
+      "epoch": 72.6923076923077,
+      "grad_norm": 2.586180244179559e-06,
+      "learning_rate": 1.9301542720051024e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27319496,
+      "step": 34965
+    },
+    {
+      "epoch": 72.70270270270271,
+      "grad_norm": 1.6498541981491144e-06,
+      "learning_rate": 1.926373434881684e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27323592,
+      "step": 34970
+    },
+    {
+      "epoch": 72.71309771309771,
+      "grad_norm": 5.476083970279433e-05,
+      "learning_rate": 1.9225961560014468e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27327528,
+      "step": 34975
+    },
+    {
+      "epoch": 72.72349272349273,
+      "grad_norm": 1.7078639302781085e-06,
+      "learning_rate": 1.918822435946885e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27331464,
+      "step": 34980
+    },
+    {
+      "epoch": 72.73388773388774,
+      "grad_norm": 1.0076769285660703e-06,
+      "learning_rate": 1.915052275299961e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27335336,
+      "step": 34985
+    },
+    {
+      "epoch": 72.74428274428274,
+      "grad_norm": 1.2571659908644506e-06,
+      "learning_rate": 1.9112856746420854e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27339272,
+      "step": 34990
+    },
+    {
+      "epoch": 72.75467775467776,
+      "grad_norm": 1.6795904230093583e-05,
+      "learning_rate": 1.907522634554104e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27343304,
+      "step": 34995
+    },
+    {
+      "epoch": 72.76507276507276,
+      "grad_norm": 5.048193543188972e-06,
+      "learning_rate": 1.9037631556163337e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27347112,
+      "step": 35000
+    },
+    {
+      "epoch": 72.76507276507276,
+      "eval_loss": 0.5962649583816528,
+      "eval_runtime": 11.6909,
+      "eval_samples_per_second": 73.219,
+      "eval_steps_per_second": 18.305,
+      "num_input_tokens_seen": 27347112,
+      "step": 35000
+    },
+    {
+      "epoch": 72.77546777546777,
+      "grad_norm": 2.4027169274631888e-05,
+      "learning_rate": 1.9000072384085272e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27350984,
+      "step": 35005
+    },
+    {
+      "epoch": 72.78586278586279,
+      "grad_norm": 2.566635430412134e-06,
+      "learning_rate": 1.8962548835098987e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27354888,
+      "step": 35010
+    },
+    {
+      "epoch": 72.79625779625779,
+      "grad_norm": 3.362172719789669e-05,
+      "learning_rate": 1.8925060914991077e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27358760,
+      "step": 35015
+    },
+    {
+      "epoch": 72.8066528066528,
+      "grad_norm": 4.219047332298942e-05,
+      "learning_rate": 1.888760862954264e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27362728,
+      "step": 35020
+    },
+    {
+      "epoch": 72.81704781704782,
+      "grad_norm": 8.654912562633399e-06,
+      "learning_rate": 1.8850191984529309e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27366504,
+      "step": 35025
+    },
+    {
+      "epoch": 72.82744282744282,
+      "grad_norm": 4.2158902147093613e-07,
+      "learning_rate": 1.8812810985721186e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27370440,
+      "step": 35030
+    },
+    {
+      "epoch": 72.83783783783784,
+      "grad_norm": 1.1704658390954137e-05,
+      "learning_rate": 1.8775465638882856e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27374312,
+      "step": 35035
+    },
+    {
+      "epoch": 72.84823284823285,
+      "grad_norm": 6.404540044968599e-07,
+      "learning_rate": 1.8738155949773517e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27378152,
+      "step": 35040
+    },
+    {
+      "epoch": 72.85862785862786,
+      "grad_norm": 7.438933971570805e-05,
+      "learning_rate": 1.8700881924146707e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27381992,
+      "step": 35045
+    },
+    {
+      "epoch": 72.86902286902287,
+      "grad_norm": 7.845400250516832e-05,
+      "learning_rate": 1.8663643567750577e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27385864,
+      "step": 35050
+    },
+    {
+      "epoch": 72.87941787941789,
+      "grad_norm": 3.066471981583163e-05,
+      "learning_rate": 1.8626440886327813e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 27389736,
+      "step": 35055
+    },
+    {
+      "epoch": 72.88981288981289,
+      "grad_norm": 3.153884972562082e-05,
+      "learning_rate": 1.8589273885615432e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27393608,
+      "step": 35060
+    },
+    {
+      "epoch": 72.9002079002079,
+      "grad_norm": 4.535193056653952e-06,
+      "learning_rate": 1.8552142571345133e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27397416,
+      "step": 35065
+    },
+    {
+      "epoch": 72.9106029106029,
+      "grad_norm": 6.958478479646146e-05,
+      "learning_rate": 1.8515046949243025e-06,
+      "loss": 0.0056,
+      "num_input_tokens_seen": 27401320,
+      "step": 35070
+    },
+    {
+      "epoch": 72.92099792099792,
+      "grad_norm": 8.996224642032757e-05,
+      "learning_rate": 1.8477987025029674e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27405288,
+      "step": 35075
+    },
+    {
+      "epoch": 72.93139293139293,
+      "grad_norm": 2.8634613045142032e-05,
+      "learning_rate": 1.8440962804420232e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27409288,
+      "step": 35080
+    },
+    {
+      "epoch": 72.94178794178794,
+      "grad_norm": 5.790951945527922e-06,
+      "learning_rate": 1.8403974293124265e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27413224,
+      "step": 35085
+    },
+    {
+      "epoch": 72.95218295218295,
+      "grad_norm": 3.655887485365383e-05,
+      "learning_rate": 1.8367021496845854e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27417128,
+      "step": 35090
+    },
+    {
+      "epoch": 72.96257796257797,
+      "grad_norm": 6.208939794305479e-06,
+      "learning_rate": 1.8330104421283662e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27420968,
+      "step": 35095
+    },
+    {
+      "epoch": 72.97297297297297,
+      "grad_norm": 0.15497036278247833,
+      "learning_rate": 1.8293223072130717e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 27424904,
+      "step": 35100
+    },
+    {
+      "epoch": 72.98336798336798,
+      "grad_norm": 7.615188678755658e-06,
+      "learning_rate": 1.8256377455074525e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27429000,
+      "step": 35105
+    },
+    {
+      "epoch": 72.993762993763,
+      "grad_norm": 6.726424999214942e-06,
+      "learning_rate": 1.8219567575797263e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27432840,
+      "step": 35110
+    },
+    {
+      "epoch": 73.004158004158,
+      "grad_norm": 5.9867983509320766e-05,
+      "learning_rate": 1.8182793439975365e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27436632,
+      "step": 35115
+    },
+    {
+      "epoch": 73.01455301455302,
+      "grad_norm": 7.519246310039307e-07,
+      "learning_rate": 1.8146055053279958e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27440536,
+      "step": 35120
+    },
+    {
+      "epoch": 73.02494802494803,
+      "grad_norm": 6.568872777279466e-05,
+      "learning_rate": 1.8109352421376486e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27444472,
+      "step": 35125
+    },
+    {
+      "epoch": 73.03534303534303,
+      "grad_norm": 1.0742159247456584e-05,
+      "learning_rate": 1.8072685549924972e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27448344,
+      "step": 35130
+    },
+    {
+      "epoch": 73.04573804573805,
+      "grad_norm": 0.13706724345684052,
+      "learning_rate": 1.8036054444579982e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 27452408,
+      "step": 35135
+    },
+    {
+      "epoch": 73.05613305613305,
+      "grad_norm": 1.290489308303222e-05,
+      "learning_rate": 1.7999459110990407e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27456440,
+      "step": 35140
+    },
+    {
+      "epoch": 73.06652806652806,
+      "grad_norm": 0.00032439129427075386,
+      "learning_rate": 1.7962899554799712e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27460312,
+      "step": 35145
+    },
+    {
+      "epoch": 73.07692307692308,
+      "grad_norm": 3.1673769171902677e-06,
+      "learning_rate": 1.7926375781645937e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27464280,
+      "step": 35150
+    },
+    {
+      "epoch": 73.08731808731808,
+      "grad_norm": 1.8674776356419898e-06,
+      "learning_rate": 1.7889887797161359e-06,
+      "loss": 0.0032,
+      "num_input_tokens_seen": 27468216,
+      "step": 35155
+    },
+    {
+      "epoch": 73.0977130977131,
+      "grad_norm": 3.3644168979662936e-06,
+      "learning_rate": 1.7853435606973028e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27472216,
+      "step": 35160
+    },
+    {
+      "epoch": 73.10810810810811,
+      "grad_norm": 4.631041520042345e-05,
+      "learning_rate": 1.781701921670223e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27476344,
+      "step": 35165
+    },
+    {
+      "epoch": 73.11850311850311,
+      "grad_norm": 2.079460500681307e-05,
+      "learning_rate": 1.7780638631964886e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27480344,
+      "step": 35170
+    },
+    {
+      "epoch": 73.12889812889813,
+      "grad_norm": 0.00023303784837480634,
+      "learning_rate": 1.7744293858371314e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27484216,
+      "step": 35175
+    },
+    {
+      "epoch": 73.13929313929314,
+      "grad_norm": 9.243142994819209e-05,
+      "learning_rate": 1.770798490152631e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27488088,
+      "step": 35180
+    },
+    {
+      "epoch": 73.14968814968815,
+      "grad_norm": 2.442687673465116e-06,
+      "learning_rate": 1.767171176702917e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27491960,
+      "step": 35185
+    },
+    {
+      "epoch": 73.16008316008316,
+      "grad_norm": 3.405121788091492e-06,
+      "learning_rate": 1.7635474460473755e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27495864,
+      "step": 35190
+    },
+    {
+      "epoch": 73.17047817047818,
+      "grad_norm": 0.0007555529591627419,
+      "learning_rate": 1.7599272987448206e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27499640,
+      "step": 35195
+    },
+    {
+      "epoch": 73.18087318087318,
+      "grad_norm": 3.654122565421858e-06,
+      "learning_rate": 1.7563107353535362e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 27503480,
+      "step": 35200
+    },
+    {
+      "epoch": 73.18087318087318,
+      "eval_loss": 0.599449098110199,
+      "eval_runtime": 11.698,
+      "eval_samples_per_second": 73.175,
+      "eval_steps_per_second": 18.294,
+      "num_input_tokens_seen": 27503480,
+      "step": 35200
+    },
+    {
+      "epoch": 73.1912681912682,
+      "grad_norm": 5.417241936811479e-06,
+      "learning_rate": 1.7526977564312263e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27507416,
+      "step": 35205
+    },
+    {
+      "epoch": 73.20166320166321,
+      "grad_norm": 7.050128624541685e-05,
+      "learning_rate": 1.7490883625350701e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27511320,
+      "step": 35210
+    },
+    {
+      "epoch": 73.21205821205821,
+      "grad_norm": 2.9244113193271914e-06,
+      "learning_rate": 1.7454825542216807e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27515416,
+      "step": 35215
+    },
+    {
+      "epoch": 73.22245322245323,
+      "grad_norm": 2.6433856419316726e-06,
+      "learning_rate": 1.7418803320471105e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27519192,
+      "step": 35220
+    },
+    {
+      "epoch": 73.23284823284823,
+      "grad_norm": 1.069706013367977e-05,
+      "learning_rate": 1.7382816965668737e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27522904,
+      "step": 35225
+    },
+    {
+      "epoch": 73.24324324324324,
+      "grad_norm": 7.082783781697799e-07,
+      "learning_rate": 1.7346866483359285e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27526808,
+      "step": 35230
+    },
+    {
+      "epoch": 73.25363825363826,
+      "grad_norm": 0.1496582329273224,
+      "learning_rate": 1.7310951879086657e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 27530712,
+      "step": 35235
+    },
+    {
+      "epoch": 73.26403326403326,
+      "grad_norm": 1.951844978975714e-06,
+      "learning_rate": 1.7275073158389471e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 27534808,
+      "step": 35240
+    },
+    {
+      "epoch": 73.27442827442827,
+      "grad_norm": 0.00017960178956855088,
+      "learning_rate": 1.723923032680061e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27538776,
+      "step": 35245
+    },
+    {
+      "epoch": 73.28482328482329,
+      "grad_norm": 2.52215195359895e-05,
+      "learning_rate": 1.7203423389847428e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27542680,
+      "step": 35250
+    },
+    {
+      "epoch": 73.29521829521829,
+      "grad_norm": 1.8785776774166152e-06,
+      "learning_rate": 1.7167652353051928e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27546616,
+      "step": 35255
+    },
+    {
+      "epoch": 73.3056133056133,
+      "grad_norm": 1.4905296666256618e-05,
+      "learning_rate": 1.7131917221930333e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27550456,
+      "step": 35260
+    },
+    {
+      "epoch": 73.31600831600832,
+      "grad_norm": 1.542813492960704e-06,
+      "learning_rate": 1.7096218001993513e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27554392,
+      "step": 35265
+    },
+    {
+      "epoch": 73.32640332640332,
+      "grad_norm": 1.0594190825941041e-05,
+      "learning_rate": 1.706055469874676e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27558328,
+      "step": 35270
+    },
+    {
+      "epoch": 73.33679833679834,
+      "grad_norm": 3.4025995319098e-06,
+      "learning_rate": 1.702492731768976e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27562296,
+      "step": 35275
+    },
+    {
+      "epoch": 73.34719334719335,
+      "grad_norm": 4.807231107406551e-06,
+      "learning_rate": 1.6989335864316724e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27566232,
+      "step": 35280
+    },
+    {
+      "epoch": 73.35758835758836,
+      "grad_norm": 2.5188473955495283e-05,
+      "learning_rate": 1.6953780344116265e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27570264,
+      "step": 35285
+    },
+    {
+      "epoch": 73.36798336798337,
+      "grad_norm": 9.8641339718597e-07,
+      "learning_rate": 1.6918260762571497e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27574232,
+      "step": 35290
+    },
+    {
+      "epoch": 73.37837837837837,
+      "grad_norm": 7.566212616438861e-07,
+      "learning_rate": 1.6882777125160093e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27578264,
+      "step": 35295
+    },
+    {
+      "epoch": 73.38877338877339,
+      "grad_norm": 8.157462048075104e-07,
+      "learning_rate": 1.6847329437353899e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27581976,
+      "step": 35300
+    },
+    {
+      "epoch": 73.3991683991684,
+      "grad_norm": 1.9266335584688932e-05,
+      "learning_rate": 1.6811917704619511e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27585976,
+      "step": 35305
+    },
+    {
+      "epoch": 73.4095634095634,
+      "grad_norm": 2.5902612833306193e-05,
+      "learning_rate": 1.67765419324179e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27589944,
+      "step": 35310
+    },
+    {
+      "epoch": 73.41995841995842,
+      "grad_norm": 5.6702054280322045e-05,
+      "learning_rate": 1.6741202126204364e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 27593848,
+      "step": 35315
+    },
+    {
+      "epoch": 73.43035343035343,
+      "grad_norm": 1.3715850855078315e-06,
+      "learning_rate": 1.6705898291428767e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 27597720,
+      "step": 35320
+    },
+    {
+      "epoch": 73.44074844074844,
+      "grad_norm": 5.836180207552388e-06,
+      "learning_rate": 1.6670630433535395e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 27601656,
+      "step": 35325
+    },
+    {
+      "epoch": 73.45114345114345,
+      "grad_norm": 8.704852007213049e-06,
+      "learning_rate": 1.6635398557962979e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27605688,
+      "step": 35330
+    },
+    {
+      "epoch": 73.46153846153847,
+      "grad_norm": 9.347139894089196e-06,
+      "learning_rate": 1.660020267014481e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27609528,
+      "step": 35335
+    },
+    {
+      "epoch": 73.47193347193347,
+      "grad_norm": 7.976858000802167e-07,
+      "learning_rate": 1.6565042775508438e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27613496,
+      "step": 35340
+    },
+    {
+      "epoch": 73.48232848232848,
+      "grad_norm": 5.744353757108911e-07,
+      "learning_rate": 1.6529918879475997e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27617400,
+      "step": 35345
+    },
+    {
+      "epoch": 73.4927234927235,
+      "grad_norm": 1.758369876370125e-06,
+      "learning_rate": 1.6494830987464043e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27621272,
+      "step": 35350
+    },
+    {
+      "epoch": 73.5031185031185,
+      "grad_norm": 1.377831995341694e-05,
+      "learning_rate": 1.6459779104883555e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27625176,
+      "step": 35355
+    },
+    {
+      "epoch": 73.51351351351352,
+      "grad_norm": 1.0476050192664843e-05,
+      "learning_rate": 1.6424763237140013e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27629112,
+      "step": 35360
+    },
+    {
+      "epoch": 73.52390852390852,
+      "grad_norm": 1.618240276002325e-05,
+      "learning_rate": 1.6389783389633207e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27632952,
+      "step": 35365
+    },
+    {
+      "epoch": 73.53430353430353,
+      "grad_norm": 6.835940666860552e-07,
+      "learning_rate": 1.6354839567757546e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27636664,
+      "step": 35370
+    },
+    {
+      "epoch": 73.54469854469855,
+      "grad_norm": 1.592941589478869e-05,
+      "learning_rate": 1.6319931776901831e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 27640760,
+      "step": 35375
+    },
+    {
+      "epoch": 73.55509355509355,
+      "grad_norm": 2.410822844467475e-06,
+      "learning_rate": 1.6285060022449229e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27644632,
+      "step": 35380
+    },
+    {
+      "epoch": 73.56548856548856,
+      "grad_norm": 7.475267921108752e-05,
+      "learning_rate": 1.6250224309777434e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27648440,
+      "step": 35385
+    },
+    {
+      "epoch": 73.57588357588358,
+      "grad_norm": 1.7730608306010254e-05,
+      "learning_rate": 1.6215424644258515e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27652248,
+      "step": 35390
+    },
+    {
+      "epoch": 73.58627858627858,
+      "grad_norm": 1.7228727529072785e-06,
+      "learning_rate": 1.6180661031259036e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27656280,
+      "step": 35395
+    },
+    {
+      "epoch": 73.5966735966736,
+      "grad_norm": 1.5938658179948106e-05,
+      "learning_rate": 1.614593347613999e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27660280,
+      "step": 35400
+    },
+    {
+      "epoch": 73.5966735966736,
+      "eval_loss": 0.5971519351005554,
+      "eval_runtime": 11.699,
+      "eval_samples_per_second": 73.168,
+      "eval_steps_per_second": 18.292,
+      "num_input_tokens_seen": 27660280,
+      "step": 35400
+    },
+    {
+      "epoch": 73.60706860706861,
+      "grad_norm": 1.9902941858163103e-05,
+      "learning_rate": 1.6111241984256758e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27664152,
+      "step": 35405
+    },
+    {
+      "epoch": 73.61746361746361,
+      "grad_norm": 7.535884947174054e-07,
+      "learning_rate": 1.6076586560959257e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 27667960,
+      "step": 35410
+    },
+    {
+      "epoch": 73.62785862785863,
+      "grad_norm": 1.4041750091564609e-06,
+      "learning_rate": 1.604196721159182e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27671736,
+      "step": 35415
+    },
+    {
+      "epoch": 73.63825363825364,
+      "grad_norm": 7.425597914334503e-07,
+      "learning_rate": 1.6007383941493092e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 27675544,
+      "step": 35420
+    },
+    {
+      "epoch": 73.64864864864865,
+      "grad_norm": 3.824407031061128e-06,
+      "learning_rate": 1.5972836755996285e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27679384,
+      "step": 35425
+    },
+    {
+      "epoch": 73.65904365904366,
+      "grad_norm": 0.1375211626291275,
+      "learning_rate": 1.5938325660429076e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 27683160,
+      "step": 35430
+    },
+    {
+      "epoch": 73.66943866943868,
+      "grad_norm": 3.915700744983042e-06,
+      "learning_rate": 1.5903850660113378e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27687064,
+      "step": 35435
+    },
+    {
+      "epoch": 73.67983367983368,
+      "grad_norm": 3.5524876693671104e-06,
+      "learning_rate": 1.5869411760365826e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27690968,
+      "step": 35440
+    },
+    {
+      "epoch": 73.6902286902287,
+      "grad_norm": 1.0796788956213277e-05,
+      "learning_rate": 1.58350089664972e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27694904,
+      "step": 35445
+    },
+    {
+      "epoch": 73.7006237006237,
+      "grad_norm": 8.87940768734552e-05,
+      "learning_rate": 1.5800642283812865e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27698776,
+      "step": 35450
+    },
+    {
+      "epoch": 73.71101871101871,
+      "grad_norm": 1.4683167819384835e-06,
+      "learning_rate": 1.5766311717612698e-06,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 27702680,
+      "step": 35455
+    },
+    {
+      "epoch": 73.72141372141373,
+      "grad_norm": 4.9296057113679126e-05,
+      "learning_rate": 1.5732017273190818e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27706584,
+      "step": 35460
+    },
+    {
+      "epoch": 73.73180873180873,
+      "grad_norm": 3.9454709622077644e-05,
+      "learning_rate": 1.5697758955835806e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27710712,
+      "step": 35465
+    },
+    {
+      "epoch": 73.74220374220374,
+      "grad_norm": 0.0002870492171496153,
+      "learning_rate": 1.566353677083085e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27714616,
+      "step": 35470
+    },
+    {
+      "epoch": 73.75259875259876,
+      "grad_norm": 3.935029781132471e-06,
+      "learning_rate": 1.562935072345334e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27718488,
+      "step": 35475
+    },
+    {
+      "epoch": 73.76299376299376,
+      "grad_norm": 0.1383458822965622,
+      "learning_rate": 1.5595200818975281e-06,
+      "loss": 0.0047,
+      "num_input_tokens_seen": 27722296,
+      "step": 35480
+    },
+    {
+      "epoch": 73.77338877338877,
+      "grad_norm": 8.106426321319304e-06,
+      "learning_rate": 1.5561087062662905e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27726168,
+      "step": 35485
+    },
+    {
+      "epoch": 73.78378378378379,
+      "grad_norm": 2.1650162125297356e-06,
+      "learning_rate": 1.5527009459777087e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 27730008,
+      "step": 35490
+    },
+    {
+      "epoch": 73.79417879417879,
+      "grad_norm": 7.481553438992705e-07,
+      "learning_rate": 1.5492968015572984e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27733912,
+      "step": 35495
+    },
+    {
+      "epoch": 73.8045738045738,
+      "grad_norm": 4.7943321987986565e-05,
+      "learning_rate": 1.5458962735300203e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27737784,
+      "step": 35500
+    },
+    {
+      "epoch": 73.81496881496882,
+      "grad_norm": 2.088298742819461e-06,
+      "learning_rate": 1.54249936242028e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27741656,
+      "step": 35505
+    },
+    {
+      "epoch": 73.82536382536382,
+      "grad_norm": 1.7329120964859612e-06,
+      "learning_rate": 1.5391060687519222e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27745624,
+      "step": 35510
+    },
+    {
+      "epoch": 73.83575883575884,
+      "grad_norm": 3.490484959911555e-05,
+      "learning_rate": 1.5357163930482367e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27749560,
+      "step": 35515
+    },
+    {
+      "epoch": 73.84615384615384,
+      "grad_norm": 1.7390036646247609e-06,
+      "learning_rate": 1.532330335831955e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27753432,
+      "step": 35520
+    },
+    {
+      "epoch": 73.85654885654886,
+      "grad_norm": 4.3327182197572256e-07,
+      "learning_rate": 1.5289478976252491e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27757304,
+      "step": 35525
+    },
+    {
+      "epoch": 73.86694386694387,
+      "grad_norm": 6.756848597433418e-05,
+      "learning_rate": 1.5255690789497345e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27761080,
+      "step": 35530
+    },
+    {
+      "epoch": 73.87733887733887,
+      "grad_norm": 5.184791120882437e-07,
+      "learning_rate": 1.5221938803264641e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27765176,
+      "step": 35535
+    },
+    {
+      "epoch": 73.88773388773389,
+      "grad_norm": 2.0482041236391524e-06,
+      "learning_rate": 1.518822302275938e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 27769112,
+      "step": 35540
+    },
+    {
+      "epoch": 73.8981288981289,
+      "grad_norm": 0.0001476290199207142,
+      "learning_rate": 1.5154543453180958e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27773048,
+      "step": 35545
+    },
+    {
+      "epoch": 73.9085239085239,
+      "grad_norm": 1.439976927031239e-06,
+      "learning_rate": 1.5120900099723167e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27776824,
+      "step": 35550
+    },
+    {
+      "epoch": 73.91891891891892,
+      "grad_norm": 4.821960101253353e-05,
+      "learning_rate": 1.5087292967574273e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27780664,
+      "step": 35555
+    },
+    {
+      "epoch": 73.92931392931393,
+      "grad_norm": 5.544422037928598e-06,
+      "learning_rate": 1.5053722061916908e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27784536,
+      "step": 35560
+    },
+    {
+      "epoch": 73.93970893970894,
+      "grad_norm": 4.2211755499010906e-05,
+      "learning_rate": 1.5020187387928124e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 27788472,
+      "step": 35565
+    },
+    {
+      "epoch": 73.95010395010395,
+      "grad_norm": 1.045907902152976e-05,
+      "learning_rate": 1.4986688950779343e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27792408,
+      "step": 35570
+    },
+    {
+      "epoch": 73.96049896049897,
+      "grad_norm": 2.3692969080002513e-06,
+      "learning_rate": 1.495322675563654e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27796504,
+      "step": 35575
+    },
+    {
+      "epoch": 73.97089397089397,
+      "grad_norm": 0.00010289761848980561,
+      "learning_rate": 1.4919800807659922e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 27800344,
+      "step": 35580
+    },
+    {
+      "epoch": 73.98128898128898,
+      "grad_norm": 5.1695988076971844e-05,
+      "learning_rate": 1.4886411112004255e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27804120,
+      "step": 35585
+    },
+    {
+      "epoch": 73.99168399168398,
+      "grad_norm": 0.14661259949207306,
+      "learning_rate": 1.4853057673818588e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 27807960,
+      "step": 35590
+    },
+    {
+      "epoch": 74.002079002079,
+      "grad_norm": 6.327976507236599e-07,
+      "learning_rate": 1.481974049824647e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27811664,
+      "step": 35595
+    },
+    {
+      "epoch": 74.01247401247402,
+      "grad_norm": 0.00035367917735129595,
+      "learning_rate": 1.4786459590425849e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27815536,
+      "step": 35600
+    },
+    {
+      "epoch": 74.01247401247402,
+      "eval_loss": 0.6042081117630005,
+      "eval_runtime": 11.7185,
+      "eval_samples_per_second": 73.047,
+      "eval_steps_per_second": 18.262,
+      "num_input_tokens_seen": 27815536,
+      "step": 35600
+    },
+    {
+      "epoch": 74.02286902286902,
+      "grad_norm": 1.07871176169283e-06,
+      "learning_rate": 1.4753214955489036e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27819632,
+      "step": 35605
+    },
+    {
+      "epoch": 74.03326403326403,
+      "grad_norm": 6.18148487774306e-06,
+      "learning_rate": 1.4720006598562737e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27823600,
+      "step": 35610
+    },
+    {
+      "epoch": 74.04365904365905,
+      "grad_norm": 5.1270842959638685e-05,
+      "learning_rate": 1.4686834524768185e-06,
+      "loss": 0.003,
+      "num_input_tokens_seen": 27827344,
+      "step": 35615
+    },
+    {
+      "epoch": 74.05405405405405,
+      "grad_norm": 6.207497790455818e-05,
+      "learning_rate": 1.4653698739220844e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27831216,
+      "step": 35620
+    },
+    {
+      "epoch": 74.06444906444906,
+      "grad_norm": 8.05395529823727e-07,
+      "learning_rate": 1.4620599247030715e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27835088,
+      "step": 35625
+    },
+    {
+      "epoch": 74.07484407484408,
+      "grad_norm": 4.039237865072209e-06,
+      "learning_rate": 1.4587536053302125e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27839056,
+      "step": 35630
+    },
+    {
+      "epoch": 74.08523908523908,
+      "grad_norm": 2.0773061351064825e-06,
+      "learning_rate": 1.4554509163133862e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27843120,
+      "step": 35635
+    },
+    {
+      "epoch": 74.0956340956341,
+      "grad_norm": 4.1418975342821795e-06,
+      "learning_rate": 1.4521518581619098e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27846960,
+      "step": 35640
+    },
+    {
+      "epoch": 74.10602910602911,
+      "grad_norm": 1.0819751423696289e-06,
+      "learning_rate": 1.4488564313845348e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27850672,
+      "step": 35645
+    },
+    {
+      "epoch": 74.11642411642411,
+      "grad_norm": 7.965421673361561e-07,
+      "learning_rate": 1.4455646364894603e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27854672,
+      "step": 35650
+    },
+    {
+      "epoch": 74.12681912681913,
+      "grad_norm": 6.0887305153300986e-06,
+      "learning_rate": 1.4422764739843247e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27858544,
+      "step": 35655
+    },
+    {
+      "epoch": 74.13721413721414,
+      "grad_norm": 1.1938022907997947e-05,
+      "learning_rate": 1.4389919443762e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27862544,
+      "step": 35660
+    },
+    {
+      "epoch": 74.14760914760915,
+      "grad_norm": 8.427990678683273e-07,
+      "learning_rate": 1.4357110481716063e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27866576,
+      "step": 35665
+    },
+    {
+      "epoch": 74.15800415800416,
+      "grad_norm": 2.9841978630429367e-06,
+      "learning_rate": 1.4324337858764941e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 27870352,
+      "step": 35670
+    },
+    {
+      "epoch": 74.16839916839916,
+      "grad_norm": 3.6651938444265397e-06,
+      "learning_rate": 1.4291601579962622e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27874320,
+      "step": 35675
+    },
+    {
+      "epoch": 74.17879417879418,
+      "grad_norm": 1.3438037740343134e-06,
+      "learning_rate": 1.42589016503574e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27878224,
+      "step": 35680
+    },
+    {
+      "epoch": 74.1891891891892,
+      "grad_norm": 0.00010859673784580082,
+      "learning_rate": 1.4226238074992099e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27882000,
+      "step": 35685
+    },
+    {
+      "epoch": 74.1995841995842,
+      "grad_norm": 3.452319560892647e-06,
+      "learning_rate": 1.4193610858903778e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27886000,
+      "step": 35690
+    },
+    {
+      "epoch": 74.20997920997921,
+      "grad_norm": 2.1442026991280727e-05,
+      "learning_rate": 1.416102000712402e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27889936,
+      "step": 35695
+    },
+    {
+      "epoch": 74.22037422037423,
+      "grad_norm": 1.7079368262784556e-05,
+      "learning_rate": 1.4128465524678668e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27893744,
+      "step": 35700
+    },
+    {
+      "epoch": 74.23076923076923,
+      "grad_norm": 3.205327629984822e-06,
+      "learning_rate": 1.4095947416588124e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 27897616,
+      "step": 35705
+    },
+    {
+      "epoch": 74.24116424116424,
+      "grad_norm": 0.00010346889030188322,
+      "learning_rate": 1.4063465687866983e-06,
+      "loss": 0.0031,
+      "num_input_tokens_seen": 27901584,
+      "step": 35710
+    },
+    {
+      "epoch": 74.25155925155926,
+      "grad_norm": 5.187995952837809e-07,
+      "learning_rate": 1.4031020343524438e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27905552,
+      "step": 35715
+    },
+    {
+      "epoch": 74.26195426195426,
+      "grad_norm": 2.1700143406633288e-06,
+      "learning_rate": 1.3998611388563926e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27909456,
+      "step": 35720
+    },
+    {
+      "epoch": 74.27234927234927,
+      "grad_norm": 1.309353956457926e-05,
+      "learning_rate": 1.3966238827983314e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27913392,
+      "step": 35725
+    },
+    {
+      "epoch": 74.28274428274429,
+      "grad_norm": 9.141041118709836e-06,
+      "learning_rate": 1.393390266677483e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27917200,
+      "step": 35730
+    },
+    {
+      "epoch": 74.29313929313929,
+      "grad_norm": 9.196910832542926e-05,
+      "learning_rate": 1.3901602909925204e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 27921104,
+      "step": 35735
+    },
+    {
+      "epoch": 74.3035343035343,
+      "grad_norm": 0.15178963541984558,
+      "learning_rate": 1.3869339562415373e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 27924944,
+      "step": 35740
+    },
+    {
+      "epoch": 74.31392931392931,
+      "grad_norm": 5.399145265982952e-06,
+      "learning_rate": 1.38371126292208e-06,
+      "loss": 0.0059,
+      "num_input_tokens_seen": 27928944,
+      "step": 35745
+    },
+    {
+      "epoch": 74.32432432432432,
+      "grad_norm": 1.0002576118495199e-06,
+      "learning_rate": 1.3804922115311286e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27932912,
+      "step": 35750
+    },
+    {
+      "epoch": 74.33471933471934,
+      "grad_norm": 1.555890662530146e-06,
+      "learning_rate": 1.3772768025650945e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 27936624,
+      "step": 35755
+    },
+    {
+      "epoch": 74.34511434511434,
+      "grad_norm": 4.654484655475244e-06,
+      "learning_rate": 1.3740650365198448e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27940496,
+      "step": 35760
+    },
+    {
+      "epoch": 74.35550935550935,
+      "grad_norm": 0.1550140678882599,
+      "learning_rate": 1.3708569138906612e-06,
+      "loss": 0.0077,
+      "num_input_tokens_seen": 27944336,
+      "step": 35765
+    },
+    {
+      "epoch": 74.36590436590437,
+      "grad_norm": 6.416226483452192e-07,
+      "learning_rate": 1.367652435172287e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27948304,
+      "step": 35770
+    },
+    {
+      "epoch": 74.37629937629937,
+      "grad_norm": 2.870566049750778e-06,
+      "learning_rate": 1.364451600858893e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27952208,
+      "step": 35775
+    },
+    {
+      "epoch": 74.38669438669439,
+      "grad_norm": 9.079341225515236e-07,
+      "learning_rate": 1.3612544114440823e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27956112,
+      "step": 35780
+    },
+    {
+      "epoch": 74.3970893970894,
+      "grad_norm": 3.8884491004864685e-06,
+      "learning_rate": 1.3580608674209072e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27960048,
+      "step": 35785
+    },
+    {
+      "epoch": 74.4074844074844,
+      "grad_norm": 9.248751302948222e-05,
+      "learning_rate": 1.3548709692818434e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27963920,
+      "step": 35790
+    },
+    {
+      "epoch": 74.41787941787942,
+      "grad_norm": 1.0072077202494256e-05,
+      "learning_rate": 1.3516847175188223e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27967760,
+      "step": 35795
+    },
+    {
+      "epoch": 74.42827442827443,
+      "grad_norm": 8.713461738807382e-07,
+      "learning_rate": 1.348502112623204e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 27971600,
+      "step": 35800
+    },
+    {
+      "epoch": 74.42827442827443,
+      "eval_loss": 0.6031776666641235,
+      "eval_runtime": 11.7244,
+      "eval_samples_per_second": 73.01,
+      "eval_steps_per_second": 18.253,
+      "num_input_tokens_seen": 27971600,
+      "step": 35800
+    },
+    {
+      "epoch": 74.43866943866944,
+      "grad_norm": 2.1268367618176853e-06,
+      "learning_rate": 1.3453231550857787e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27975376,
+      "step": 35805
+    },
+    {
+      "epoch": 74.44906444906445,
+      "grad_norm": 2.307994509465061e-05,
+      "learning_rate": 1.3421478453967878e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27979280,
+      "step": 35810
+    },
+    {
+      "epoch": 74.45945945945945,
+      "grad_norm": 0.0003120009205304086,
+      "learning_rate": 1.3389761840459065e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27983408,
+      "step": 35815
+    },
+    {
+      "epoch": 74.46985446985447,
+      "grad_norm": 9.848034778769943e-07,
+      "learning_rate": 1.3358081715222376e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27987184,
+      "step": 35820
+    },
+    {
+      "epoch": 74.48024948024948,
+      "grad_norm": 7.251144779729657e-06,
+      "learning_rate": 1.3326438083143295e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27991056,
+      "step": 35825
+    },
+    {
+      "epoch": 74.49064449064448,
+      "grad_norm": 5.67336292078835e-06,
+      "learning_rate": 1.3294830949101723e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27995152,
+      "step": 35830
+    },
+    {
+      "epoch": 74.5010395010395,
+      "grad_norm": 3.996504528913647e-05,
+      "learning_rate": 1.3263260317971815e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 27998928,
+      "step": 35835
+    },
+    {
+      "epoch": 74.51143451143452,
+      "grad_norm": 6.597420451726066e-06,
+      "learning_rate": 1.3231726194622208e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28002768,
+      "step": 35840
+    },
+    {
+      "epoch": 74.52182952182952,
+      "grad_norm": 1.5623530771335936e-06,
+      "learning_rate": 1.3200228583915814e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28006672,
+      "step": 35845
+    },
+    {
+      "epoch": 74.53222453222453,
+      "grad_norm": 6.305799615802243e-07,
+      "learning_rate": 1.3168767490709971e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28010576,
+      "step": 35850
+    },
+    {
+      "epoch": 74.54261954261955,
+      "grad_norm": 5.542286817217246e-05,
+      "learning_rate": 1.3137342919856437e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28014448,
+      "step": 35855
+    },
+    {
+      "epoch": 74.55301455301455,
+      "grad_norm": 4.179697498329915e-06,
+      "learning_rate": 1.310595487620117e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28018096,
+      "step": 35860
+    },
+    {
+      "epoch": 74.56340956340956,
+      "grad_norm": 2.5387680580024607e-05,
+      "learning_rate": 1.3074603364584715e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28022000,
+      "step": 35865
+    },
+    {
+      "epoch": 74.57380457380458,
+      "grad_norm": 7.536301382060628e-06,
+      "learning_rate": 1.3043288389841758e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28025840,
+      "step": 35870
+    },
+    {
+      "epoch": 74.58419958419958,
+      "grad_norm": 4.144466947764158e-05,
+      "learning_rate": 1.3012009956801546e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28029840,
+      "step": 35875
+    },
+    {
+      "epoch": 74.5945945945946,
+      "grad_norm": 3.1370170745503856e-06,
+      "learning_rate": 1.2980768070287586e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28033712,
+      "step": 35880
+    },
+    {
+      "epoch": 74.60498960498961,
+      "grad_norm": 9.46012823987985e-06,
+      "learning_rate": 1.2949562735117716e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28037616,
+      "step": 35885
+    },
+    {
+      "epoch": 74.61538461538461,
+      "grad_norm": 6.138524781817978e-07,
+      "learning_rate": 1.291839395610428e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28041328,
+      "step": 35890
+    },
+    {
+      "epoch": 74.62577962577963,
+      "grad_norm": 8.848883226164617e-07,
+      "learning_rate": 1.2887261738053852e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28045328,
+      "step": 35895
+    },
+    {
+      "epoch": 74.63617463617463,
+      "grad_norm": 2.6280547444912372e-06,
+      "learning_rate": 1.2856166085767396e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28049264,
+      "step": 35900
+    },
+    {
+      "epoch": 74.64656964656965,
+      "grad_norm": 9.23966183563607e-07,
+      "learning_rate": 1.2825107004040272e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28053008,
+      "step": 35905
+    },
+    {
+      "epoch": 74.65696465696466,
+      "grad_norm": 5.0916787586174905e-06,
+      "learning_rate": 1.2794084497662146e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28056848,
+      "step": 35910
+    },
+    {
+      "epoch": 74.66735966735966,
+      "grad_norm": 7.967851161083672e-07,
+      "learning_rate": 1.276309857141711e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28060784,
+      "step": 35915
+    },
+    {
+      "epoch": 74.67775467775468,
+      "grad_norm": 1.3058559488854371e-05,
+      "learning_rate": 1.273214923008359e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28064752,
+      "step": 35920
+    },
+    {
+      "epoch": 74.6881496881497,
+      "grad_norm": 7.875459232309368e-06,
+      "learning_rate": 1.2701236478434352e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28068784,
+      "step": 35925
+    },
+    {
+      "epoch": 74.6985446985447,
+      "grad_norm": 6.491897693194915e-06,
+      "learning_rate": 1.2670360321236502e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28072880,
+      "step": 35930
+    },
+    {
+      "epoch": 74.70893970893971,
+      "grad_norm": 3.132753408863209e-05,
+      "learning_rate": 1.2639520763251617e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28076880,
+      "step": 35935
+    },
+    {
+      "epoch": 74.71933471933473,
+      "grad_norm": 2.8232363547431305e-05,
+      "learning_rate": 1.2608717809235448e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28080816,
+      "step": 35940
+    },
+    {
+      "epoch": 74.72972972972973,
+      "grad_norm": 1.2078764484613203e-05,
+      "learning_rate": 1.2577951463938282e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28084592,
+      "step": 35945
+    },
+    {
+      "epoch": 74.74012474012474,
+      "grad_norm": 6.162823638078407e-07,
+      "learning_rate": 1.2547221732104569e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28088528,
+      "step": 35950
+    },
+    {
+      "epoch": 74.75051975051976,
+      "grad_norm": 5.402917850005906e-06,
+      "learning_rate": 1.25165286184733e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28092368,
+      "step": 35955
+    },
+    {
+      "epoch": 74.76091476091476,
+      "grad_norm": 8.038711530389264e-05,
+      "learning_rate": 1.248587212777777e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28096272,
+      "step": 35960
+    },
+    {
+      "epoch": 74.77130977130977,
+      "grad_norm": 2.8596903121069772e-06,
+      "learning_rate": 1.2455252264745532e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28100176,
+      "step": 35965
+    },
+    {
+      "epoch": 74.78170478170478,
+      "grad_norm": 2.6728880584414583e-06,
+      "learning_rate": 1.2424669034098528e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28104080,
+      "step": 35970
+    },
+    {
+      "epoch": 74.79209979209979,
+      "grad_norm": 6.477081115008332e-06,
+      "learning_rate": 1.2394122440553185e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 28108080,
+      "step": 35975
+    },
+    {
+      "epoch": 74.8024948024948,
+      "grad_norm": 3.93638356399606e-06,
+      "learning_rate": 1.2363612488820037e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 28112016,
+      "step": 35980
+    },
+    {
+      "epoch": 74.81288981288981,
+      "grad_norm": 1.302226974075893e-05,
+      "learning_rate": 1.2333139183604208e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28115824,
+      "step": 35985
+    },
+    {
+      "epoch": 74.82328482328482,
+      "grad_norm": 0.15237000584602356,
+      "learning_rate": 1.2302702529604998e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 28119824,
+      "step": 35990
+    },
+    {
+      "epoch": 74.83367983367984,
+      "grad_norm": 7.9890264714777e-07,
+      "learning_rate": 1.227230253151615e-06,
+      "loss": 0.0022,
+      "num_input_tokens_seen": 28123792,
+      "step": 35995
+    },
+    {
+      "epoch": 74.84407484407484,
+      "grad_norm": 4.4794378482038155e-06,
+      "learning_rate": 1.2241939194025748e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28127664,
+      "step": 36000
+    },
+    {
+      "epoch": 74.84407484407484,
+      "eval_loss": 0.5978870987892151,
+      "eval_runtime": 11.7019,
+      "eval_samples_per_second": 73.151,
+      "eval_steps_per_second": 18.288,
+      "num_input_tokens_seen": 28127664,
+      "step": 36000
+    },
+    {
+      "epoch": 74.85446985446985,
+      "grad_norm": 2.5497686237940798e-06,
+      "learning_rate": 1.2211612521816156e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28131696,
+      "step": 36005
+    },
+    {
+      "epoch": 74.86486486486487,
+      "grad_norm": 1.3841198551745038e-06,
+      "learning_rate": 1.2181322519564137e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28135792,
+      "step": 36010
+    },
+    {
+      "epoch": 74.87525987525987,
+      "grad_norm": 2.2834942683402915e-06,
+      "learning_rate": 1.2151069191940839e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28139728,
+      "step": 36015
+    },
+    {
+      "epoch": 74.88565488565489,
+      "grad_norm": 6.377292447723448e-05,
+      "learning_rate": 1.2120852543611644e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 28143664,
+      "step": 36020
+    },
+    {
+      "epoch": 74.8960498960499,
+      "grad_norm": 5.486719965119846e-05,
+      "learning_rate": 1.2090672579236379e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28147568,
+      "step": 36025
+    },
+    {
+      "epoch": 74.9064449064449,
+      "grad_norm": 1.1075580914621241e-06,
+      "learning_rate": 1.2060529303469126e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28151472,
+      "step": 36030
+    },
+    {
+      "epoch": 74.91683991683992,
+      "grad_norm": 3.2794075650599552e-06,
+      "learning_rate": 1.2030422720958445e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28155504,
+      "step": 36035
+    },
+    {
+      "epoch": 74.92723492723492,
+      "grad_norm": 0.15116457641124725,
+      "learning_rate": 1.200035283634704e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 28159344,
+      "step": 36040
+    },
+    {
+      "epoch": 74.93762993762994,
+      "grad_norm": 9.012697432808636e-07,
+      "learning_rate": 1.1970319654272144e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28163184,
+      "step": 36045
+    },
+    {
+      "epoch": 74.94802494802495,
+      "grad_norm": 7.708753173574223e-07,
+      "learning_rate": 1.1940323179365192e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28167184,
+      "step": 36050
+    },
+    {
+      "epoch": 74.95841995841995,
+      "grad_norm": 1.9519436591508565e-06,
+      "learning_rate": 1.1910363416252095e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28171344,
+      "step": 36055
+    },
+    {
+      "epoch": 74.96881496881497,
+      "grad_norm": 7.506309884774964e-07,
+      "learning_rate": 1.1880440369552964e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28175184,
+      "step": 36060
+    },
+    {
+      "epoch": 74.97920997920998,
+      "grad_norm": 0.1438775211572647,
+      "learning_rate": 1.1850554043882328e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 28179248,
+      "step": 36065
+    },
+    {
+      "epoch": 74.98960498960498,
+      "grad_norm": 9.849265552475117e-06,
+      "learning_rate": 1.1820704443849028e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28183248,
+      "step": 36070
+    },
+    {
+      "epoch": 75.0,
+      "grad_norm": 3.4135696296289098e-06,
+      "learning_rate": 1.1790891574056219e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28187008,
+      "step": 36075
+    },
+    {
+      "epoch": 75.01039501039502,
+      "grad_norm": 7.203788118204102e-05,
+      "learning_rate": 1.1761115439101523e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28190912,
+      "step": 36080
+    },
+    {
+      "epoch": 75.02079002079002,
+      "grad_norm": 3.629155980888754e-05,
+      "learning_rate": 1.1731376043576659e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28194784,
+      "step": 36085
+    },
+    {
+      "epoch": 75.03118503118503,
+      "grad_norm": 4.424269263836322e-06,
+      "learning_rate": 1.1701673392067875e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28198816,
+      "step": 36090
+    },
+    {
+      "epoch": 75.04158004158005,
+      "grad_norm": 3.5687244235305116e-06,
+      "learning_rate": 1.1672007489155757e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28202720,
+      "step": 36095
+    },
+    {
+      "epoch": 75.05197505197505,
+      "grad_norm": 2.4162047793652164e-06,
+      "learning_rate": 1.164237833941506e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28206752,
+      "step": 36100
+    },
+    {
+      "epoch": 75.06237006237006,
+      "grad_norm": 4.617772447090829e-06,
+      "learning_rate": 1.1612785947415022e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 28210592,
+      "step": 36105
+    },
+    {
+      "epoch": 75.07276507276508,
+      "grad_norm": 4.397519205667777e-06,
+      "learning_rate": 1.1583230317719185e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28214304,
+      "step": 36110
+    },
+    {
+      "epoch": 75.08316008316008,
+      "grad_norm": 0.00011060433462262154,
+      "learning_rate": 1.1553711454885318e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28218304,
+      "step": 36115
+    },
+    {
+      "epoch": 75.0935550935551,
+      "grad_norm": 6.30840895610163e-06,
+      "learning_rate": 1.152422936346567e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28222176,
+      "step": 36120
+    },
+    {
+      "epoch": 75.1039501039501,
+      "grad_norm": 1.6046765267674346e-06,
+      "learning_rate": 1.1494784048006718e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28226208,
+      "step": 36125
+    },
+    {
+      "epoch": 75.11434511434511,
+      "grad_norm": 8.474182777717942e-07,
+      "learning_rate": 1.1465375513049326e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28230208,
+      "step": 36130
+    },
+    {
+      "epoch": 75.12474012474013,
+      "grad_norm": 5.3226365707814693e-05,
+      "learning_rate": 1.1436003763128616e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28234048,
+      "step": 36135
+    },
+    {
+      "epoch": 75.13513513513513,
+      "grad_norm": 1.926482218550518e-05,
+      "learning_rate": 1.1406668802774106e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28237920,
+      "step": 36140
+    },
+    {
+      "epoch": 75.14553014553015,
+      "grad_norm": 5.829929432366043e-05,
+      "learning_rate": 1.137737063650965e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28241792,
+      "step": 36145
+    },
+    {
+      "epoch": 75.15592515592516,
+      "grad_norm": 1.4675980537504074e-06,
+      "learning_rate": 1.1348109268853323e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28245952,
+      "step": 36150
+    },
+    {
+      "epoch": 75.16632016632016,
+      "grad_norm": 6.958878202567575e-07,
+      "learning_rate": 1.1318884704317634e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28249792,
+      "step": 36155
+    },
+    {
+      "epoch": 75.17671517671518,
+      "grad_norm": 1.899728158605285e-05,
+      "learning_rate": 1.1289696947409417e-06,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 28253728,
+      "step": 36160
+    },
+    {
+      "epoch": 75.18711018711019,
+      "grad_norm": 1.5727852087366045e-06,
+      "learning_rate": 1.126054600262974e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28257536,
+      "step": 36165
+    },
+    {
+      "epoch": 75.1975051975052,
+      "grad_norm": 5.904015779378824e-06,
+      "learning_rate": 1.1231431874474064e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28261504,
+      "step": 36170
+    },
+    {
+      "epoch": 75.20790020790021,
+      "grad_norm": 0.0007000819896347821,
+      "learning_rate": 1.12023545674321e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 28265472,
+      "step": 36175
+    },
+    {
+      "epoch": 75.21829521829522,
+      "grad_norm": 1.2727248304145178e-06,
+      "learning_rate": 1.117331408598804e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 28269344,
+      "step": 36180
+    },
+    {
+      "epoch": 75.22869022869023,
+      "grad_norm": 2.5350134364998667e-06,
+      "learning_rate": 1.1144310434620191e-06,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 28273248,
+      "step": 36185
+    },
+    {
+      "epoch": 75.23908523908524,
+      "grad_norm": 6.412485618056962e-07,
+      "learning_rate": 1.1115343617801365e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28277024,
+      "step": 36190
+    },
+    {
+      "epoch": 75.24948024948024,
+      "grad_norm": 1.398362883264781e-06,
+      "learning_rate": 1.1086413639998515e-06,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 28280960,
+      "step": 36195
+    },
+    {
+      "epoch": 75.25987525987526,
+      "grad_norm": 2.1723303689213935e-06,
+      "learning_rate": 1.1057520505673103e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28284736,
+      "step": 36200
+    },
+    {
+      "epoch": 75.25987525987526,
+      "eval_loss": 0.5975652933120728,
+      "eval_runtime": 11.705,
+      "eval_samples_per_second": 73.131,
+      "eval_steps_per_second": 18.283,
+      "num_input_tokens_seen": 28284736,
+      "step": 36200
+    },
+    {
+      "epoch": 75.27027027027027,
+      "grad_norm": 8.651272196402715e-07,
+      "learning_rate": 1.1028664219280727e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28288384,
+      "step": 36205
+    },
+    {
+      "epoch": 75.28066528066527,
+      "grad_norm": 4.804206128028454e-06,
+      "learning_rate": 1.0999844785271468e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28292352,
+      "step": 36210
+    },
+    {
+      "epoch": 75.29106029106029,
+      "grad_norm": 0.14623244106769562,
+      "learning_rate": 1.097106220808955e-06,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 28296256,
+      "step": 36215
+    },
+    {
+      "epoch": 75.3014553014553,
+      "grad_norm": 1.0850883427337976e-06,
+      "learning_rate": 1.0942316492173698e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28300224,
+      "step": 36220
+    },
+    {
+      "epoch": 75.3118503118503,
+      "grad_norm": 4.786789304489503e-06,
+      "learning_rate": 1.0913607641956841e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 28304320,
+      "step": 36225
+    },
+    {
+      "epoch": 75.32224532224532,
+      "grad_norm": 0.12751242518424988,
+      "learning_rate": 1.0884935661866213e-06,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 28308288,
+      "step": 36230
+    },
+    {
+      "epoch": 75.33264033264034,
+      "grad_norm": 1.4307767060017795e-06,
+      "learning_rate": 1.0856300556323418e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28312224,
+      "step": 36235
+    },
+    {
+      "epoch": 75.34303534303534,
+      "grad_norm": 7.178810506047739e-07,
+      "learning_rate": 1.0827702329744365e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28316064,
+      "step": 36240
+    },
+    {
+      "epoch": 75.35343035343035,
+      "grad_norm": 6.437833235395374e-06,
+      "learning_rate": 1.0799140986539197e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28320032,
+      "step": 36245
+    },
+    {
+      "epoch": 75.36382536382537,
+      "grad_norm": 1.5081628589541651e-05,
+      "learning_rate": 1.0770616531112526e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28324000,
+      "step": 36250
+    },
+    {
+      "epoch": 75.37422037422037,
+      "grad_norm": 4.257819000486052e-06,
+      "learning_rate": 1.0742128967863085e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28327776,
+      "step": 36255
+    },
+    {
+      "epoch": 75.38461538461539,
+      "grad_norm": 0.00037305522710084915,
+      "learning_rate": 1.071367830118411e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28331616,
+      "step": 36260
+    },
+    {
+      "epoch": 75.39501039501039,
+      "grad_norm": 2.4722536181798205e-05,
+      "learning_rate": 1.068526453546298e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28335456,
+      "step": 36265
+    },
+    {
+      "epoch": 75.4054054054054,
+      "grad_norm": 9.48718807194382e-05,
+      "learning_rate": 1.0656887675081467e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28339360,
+      "step": 36270
+    },
+    {
+      "epoch": 75.41580041580042,
+      "grad_norm": 6.418365501303924e-06,
+      "learning_rate": 1.0628547724415628e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28343232,
+      "step": 36275
+    },
+    {
+      "epoch": 75.42619542619542,
+      "grad_norm": 1.7245614571947954e-06,
+      "learning_rate": 1.0600244687835881e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28347008,
+      "step": 36280
+    },
+    {
+      "epoch": 75.43659043659044,
+      "grad_norm": 1.3354372185858665e-06,
+      "learning_rate": 1.0571978569706876e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28350944,
+      "step": 36285
+    },
+    {
+      "epoch": 75.44698544698545,
+      "grad_norm": 1.4071459190745372e-05,
+      "learning_rate": 1.0543749374387652e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28354848,
+      "step": 36290
+    },
+    {
+      "epoch": 75.45738045738045,
+      "grad_norm": 1.78741156560136e-05,
+      "learning_rate": 1.051555710623142e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28358816,
+      "step": 36295
+    },
+    {
+      "epoch": 75.46777546777547,
+      "grad_norm": 1.8050941434921697e-06,
+      "learning_rate": 1.0487401769585847e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28362816,
+      "step": 36300
+    },
+    {
+      "epoch": 75.47817047817048,
+      "grad_norm": 1.3017060155107174e-05,
+      "learning_rate": 1.0459283368792845e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28366656,
+      "step": 36305
+    },
+    {
+      "epoch": 75.48856548856548,
+      "grad_norm": 1.4720104445586912e-05,
+      "learning_rate": 1.043120190818858e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28370528,
+      "step": 36310
+    },
+    {
+      "epoch": 75.4989604989605,
+      "grad_norm": 1.0116614248545375e-05,
+      "learning_rate": 1.0403157392103596e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28374464,
+      "step": 36315
+    },
+    {
+      "epoch": 75.50935550935552,
+      "grad_norm": 0.1416947841644287,
+      "learning_rate": 1.0375149824862735e-06,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 28378432,
+      "step": 36320
+    },
+    {
+      "epoch": 75.51975051975052,
+      "grad_norm": 2.9701707262574928e-06,
+      "learning_rate": 1.034717921078507e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28382432,
+      "step": 36325
+    },
+    {
+      "epoch": 75.53014553014553,
+      "grad_norm": 1.0297134167558397e-06,
+      "learning_rate": 1.0319245554184009e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28386336,
+      "step": 36330
+    },
+    {
+      "epoch": 75.54054054054055,
+      "grad_norm": 2.185761786677176e-06,
+      "learning_rate": 1.0291348859367361e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28390336,
+      "step": 36335
+    },
+    {
+      "epoch": 75.55093555093555,
+      "grad_norm": 1.2072512163285865e-06,
+      "learning_rate": 1.0263489130637016e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28394304,
+      "step": 36340
+    },
+    {
+      "epoch": 75.56133056133056,
+      "grad_norm": 0.00013139126531314105,
+      "learning_rate": 1.0235666372289427e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28398208,
+      "step": 36345
+    },
+    {
+      "epoch": 75.57172557172557,
+      "grad_norm": 4.594686288328376e-06,
+      "learning_rate": 1.0207880588615076e-06,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 28401984,
+      "step": 36350
+    },
+    {
+      "epoch": 75.58212058212058,
+      "grad_norm": 7.1393537837138865e-06,
+      "learning_rate": 1.0180131783898984e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28405888,
+      "step": 36355
+    },
+    {
+      "epoch": 75.5925155925156,
+      "grad_norm": 1.14345502879587e-05,
+      "learning_rate": 1.0152419962420362e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28409824,
+      "step": 36360
+    },
+    {
+      "epoch": 75.6029106029106,
+      "grad_norm": 1.4757226836081827e-06,
+      "learning_rate": 1.0124745128452685e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28413600,
+      "step": 36365
+    },
+    {
+      "epoch": 75.61330561330561,
+      "grad_norm": 1.1798158084275201e-05,
+      "learning_rate": 1.0097107286263758e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28417664,
+      "step": 36370
+    },
+    {
+      "epoch": 75.62370062370063,
+      "grad_norm": 0.000696991803124547,
+      "learning_rate": 1.00695064401157e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28421600,
+      "step": 36375
+    },
+    {
+      "epoch": 75.63409563409563,
+      "grad_norm": 9.09482537281292e-07,
+      "learning_rate": 1.0041942594264886e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28425472,
+      "step": 36380
+    },
+    {
+      "epoch": 75.64449064449065,
+      "grad_norm": 2.641481614773511e-06,
+      "learning_rate": 1.001441575296208e-06,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28429216,
+      "step": 36385
+    },
+    {
+      "epoch": 75.65488565488566,
+      "grad_norm": 1.3573281876233523e-06,
+      "learning_rate": 9.986925920452139e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28432960,
+      "step": 36390
+    },
+    {
+      "epoch": 75.66528066528066,
+      "grad_norm": 7.079875103954691e-07,
+      "learning_rate": 9.959473100974475e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28436800,
+      "step": 36395
+    },
+    {
+      "epoch": 75.67567567567568,
+      "grad_norm": 8.161730420397362e-07,
+      "learning_rate": 9.932057298762564e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28440672,
+      "step": 36400
+    },
+    {
+      "epoch": 75.67567567567568,
+      "eval_loss": 0.6028590798377991,
+      "eval_runtime": 11.6993,
+      "eval_samples_per_second": 73.167,
+      "eval_steps_per_second": 18.292,
+      "num_input_tokens_seen": 28440672,
+      "step": 36400
+    },
+    {
+      "epoch": 75.68607068607069,
+      "grad_norm": 1.4736461935171974e-06,
+      "learning_rate": 9.90467851804433e-07,
+      "loss": 0.003,
+      "num_input_tokens_seen": 28444544,
+      "step": 36405
+    },
+    {
+      "epoch": 75.6964656964657,
+      "grad_norm": 3.2002687930798857e-06,
+      "learning_rate": 9.877336763041895e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28448448,
+      "step": 36410
+    },
+    {
+      "epoch": 75.70686070686071,
+      "grad_norm": 8.864631695360004e-07,
+      "learning_rate": 9.850032037971662e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28452320,
+      "step": 36415
+    },
+    {
+      "epoch": 75.71725571725571,
+      "grad_norm": 9.244502621186257e-07,
+      "learning_rate": 9.822764347044406e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28456224,
+      "step": 36420
+    },
+    {
+      "epoch": 75.72765072765073,
+      "grad_norm": 1.6309806596837007e-05,
+      "learning_rate": 9.795533694465175e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28460128,
+      "step": 36425
+    },
+    {
+      "epoch": 75.73804573804574,
+      "grad_norm": 2.3577169486088678e-05,
+      "learning_rate": 9.768340084433197e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28464128,
+      "step": 36430
+    },
+    {
+      "epoch": 75.74844074844074,
+      "grad_norm": 6.891179964441108e-07,
+      "learning_rate": 9.741183521142143e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 28468192,
+      "step": 36435
+    },
+    {
+      "epoch": 75.75883575883576,
+      "grad_norm": 0.0003321605618111789,
+      "learning_rate": 9.714064008779889e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28472032,
+      "step": 36440
+    },
+    {
+      "epoch": 75.76923076923077,
+      "grad_norm": 4.6485172788379714e-05,
+      "learning_rate": 9.686981551528584e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28476032,
+      "step": 36445
+    },
+    {
+      "epoch": 75.77962577962577,
+      "grad_norm": 4.3128864035679726e-07,
+      "learning_rate": 9.65993615356467e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28480096,
+      "step": 36450
+    },
+    {
+      "epoch": 75.79002079002079,
+      "grad_norm": 2.127660536643816e-06,
+      "learning_rate": 9.632927819058917e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 28484000,
+      "step": 36455
+    },
+    {
+      "epoch": 75.8004158004158,
+      "grad_norm": 9.454190035285137e-07,
+      "learning_rate": 9.605956552176305e-07,
+      "loss": 0.0079,
+      "num_input_tokens_seen": 28487840,
+      "step": 36460
+    },
+    {
+      "epoch": 75.8108108108108,
+      "grad_norm": 7.78141911723651e-05,
+      "learning_rate": 9.579022357076223e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28491680,
+      "step": 36465
+    },
+    {
+      "epoch": 75.82120582120582,
+      "grad_norm": 8.771448483457789e-05,
+      "learning_rate": 9.552125237912158e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28495648,
+      "step": 36470
+    },
+    {
+      "epoch": 75.83160083160084,
+      "grad_norm": 6.158091991892434e-07,
+      "learning_rate": 9.525265198832096e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28499616,
+      "step": 36475
+    },
+    {
+      "epoch": 75.84199584199584,
+      "grad_norm": 9.42622307320562e-07,
+      "learning_rate": 9.498442243978112e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28503584,
+      "step": 36480
+    },
+    {
+      "epoch": 75.85239085239085,
+      "grad_norm": 9.954957249647123e-07,
+      "learning_rate": 9.471656377486649e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28507392,
+      "step": 36485
+    },
+    {
+      "epoch": 75.86278586278586,
+      "grad_norm": 6.02912177782855e-06,
+      "learning_rate": 9.444907603488456e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28511424,
+      "step": 36490
+    },
+    {
+      "epoch": 75.87318087318087,
+      "grad_norm": 2.4991244572447613e-05,
+      "learning_rate": 9.418195926108514e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28515200,
+      "step": 36495
+    },
+    {
+      "epoch": 75.88357588357589,
+      "grad_norm": 2.1651396764355013e-06,
+      "learning_rate": 9.391521349466053e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28518912,
+      "step": 36500
+    },
+    {
+      "epoch": 75.89397089397089,
+      "grad_norm": 7.637333328602836e-05,
+      "learning_rate": 9.364883877674758e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28522976,
+      "step": 36505
+    },
+    {
+      "epoch": 75.9043659043659,
+      "grad_norm": 9.499666475676349e-07,
+      "learning_rate": 9.33828351484231e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 28526816,
+      "step": 36510
+    },
+    {
+      "epoch": 75.91476091476092,
+      "grad_norm": 1.1386438927729614e-05,
+      "learning_rate": 9.311720265070906e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28530752,
+      "step": 36515
+    },
+    {
+      "epoch": 75.92515592515592,
+      "grad_norm": 1.3736698747379705e-05,
+      "learning_rate": 9.285194132456931e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28534560,
+      "step": 36520
+    },
+    {
+      "epoch": 75.93555093555094,
+      "grad_norm": 0.00014631202793680131,
+      "learning_rate": 9.258705121091032e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28538464,
+      "step": 36525
+    },
+    {
+      "epoch": 75.94594594594595,
+      "grad_norm": 2.0605107238225173e-06,
+      "learning_rate": 9.232253235058136e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28542304,
+      "step": 36530
+    },
+    {
+      "epoch": 75.95634095634095,
+      "grad_norm": 7.296271178347524e-07,
+      "learning_rate": 9.205838478437478e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28546240,
+      "step": 36535
+    },
+    {
+      "epoch": 75.96673596673597,
+      "grad_norm": 0.00030590701499022543,
+      "learning_rate": 9.179460855302524e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28550176,
+      "step": 36540
+    },
+    {
+      "epoch": 75.97713097713098,
+      "grad_norm": 0.1583569347858429,
+      "learning_rate": 9.153120369721046e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 28554080,
+      "step": 36545
+    },
+    {
+      "epoch": 75.98752598752598,
+      "grad_norm": 2.7309556571708526e-06,
+      "learning_rate": 9.126817025755103e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28557920,
+      "step": 36550
+    },
+    {
+      "epoch": 75.997920997921,
+      "grad_norm": 4.304966410018096e-07,
+      "learning_rate": 9.100550827460947e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28561760,
+      "step": 36555
+    },
+    {
+      "epoch": 76.00831600831602,
+      "grad_norm": 0.14939789474010468,
+      "learning_rate": 9.0743217788892e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 28565672,
+      "step": 36560
+    },
+    {
+      "epoch": 76.01871101871102,
+      "grad_norm": 1.498858637205558e-06,
+      "learning_rate": 9.048129884084683e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28569608,
+      "step": 36565
+    },
+    {
+      "epoch": 76.02910602910603,
+      "grad_norm": 6.791624036850408e-05,
+      "learning_rate": 9.021975147086553e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28573704,
+      "step": 36570
+    },
+    {
+      "epoch": 76.03950103950103,
+      "grad_norm": 6.797042146899912e-07,
+      "learning_rate": 8.995857571928141e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28577544,
+      "step": 36575
+    },
+    {
+      "epoch": 76.04989604989605,
+      "grad_norm": 2.6359102776041254e-05,
+      "learning_rate": 8.969777162637139e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28581416,
+      "step": 36580
+    },
+    {
+      "epoch": 76.06029106029106,
+      "grad_norm": 3.5539933378458954e-06,
+      "learning_rate": 8.943733923235525e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 28585224,
+      "step": 36585
+    },
+    {
+      "epoch": 76.07068607068607,
+      "grad_norm": 0.00014261544856708497,
+      "learning_rate": 8.917727857739394e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28589128,
+      "step": 36590
+    },
+    {
+      "epoch": 76.08108108108108,
+      "grad_norm": 9.748619049787521e-07,
+      "learning_rate": 8.891758970159258e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28592968,
+      "step": 36595
+    },
+    {
+      "epoch": 76.0914760914761,
+      "grad_norm": 6.761277290934231e-06,
+      "learning_rate": 8.86582726449986e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28596968,
+      "step": 36600
+    },
+    {
+      "epoch": 76.0914760914761,
+      "eval_loss": 0.601006031036377,
+      "eval_runtime": 11.6987,
+      "eval_samples_per_second": 73.171,
+      "eval_steps_per_second": 18.293,
+      "num_input_tokens_seen": 28596968,
+      "step": 36600
+    },
+    {
+      "epoch": 76.1018711018711,
+      "grad_norm": 6.74138300382765e-06,
+      "learning_rate": 8.839932744760165e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28600840,
+      "step": 36605
+    },
+    {
+      "epoch": 76.11226611226611,
+      "grad_norm": 0.15322712063789368,
+      "learning_rate": 8.814075414933482e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 28604744,
+      "step": 36610
+    },
+    {
+      "epoch": 76.12266112266113,
+      "grad_norm": 3.479719225651934e-06,
+      "learning_rate": 8.788255279007257e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28608712,
+      "step": 36615
+    },
+    {
+      "epoch": 76.13305613305613,
+      "grad_norm": 1.1262760608587996e-06,
+      "learning_rate": 8.762472340963362e-07,
+      "loss": 0.003,
+      "num_input_tokens_seen": 28612744,
+      "step": 36620
+    },
+    {
+      "epoch": 76.14345114345114,
+      "grad_norm": 5.75358342302934e-07,
+      "learning_rate": 8.736726604777811e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28616744,
+      "step": 36625
+    },
+    {
+      "epoch": 76.15384615384616,
+      "grad_norm": 3.4180586681031855e-06,
+      "learning_rate": 8.711018074420901e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28620520,
+      "step": 36630
+    },
+    {
+      "epoch": 76.16424116424116,
+      "grad_norm": 6.726837114001682e-07,
+      "learning_rate": 8.685346753857209e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28624392,
+      "step": 36635
+    },
+    {
+      "epoch": 76.17463617463618,
+      "grad_norm": 0.00011586827167775482,
+      "learning_rate": 8.659712647045654e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28628360,
+      "step": 36640
+    },
+    {
+      "epoch": 76.18503118503118,
+      "grad_norm": 0.16344386339187622,
+      "learning_rate": 8.634115757939209e-07,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 28632328,
+      "step": 36645
+    },
+    {
+      "epoch": 76.1954261954262,
+      "grad_norm": 0.0002148225175915286,
+      "learning_rate": 8.608556090485387e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 28636200,
+      "step": 36650
+    },
+    {
+      "epoch": 76.20582120582121,
+      "grad_norm": 7.678953807044309e-06,
+      "learning_rate": 8.583033648625671e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 28639976,
+      "step": 36655
+    },
+    {
+      "epoch": 76.21621621621621,
+      "grad_norm": 1.2741307727992535e-05,
+      "learning_rate": 8.557548436295998e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28643880,
+      "step": 36660
+    },
+    {
+      "epoch": 76.22661122661123,
+      "grad_norm": 1.9057259805776994e-06,
+      "learning_rate": 8.532100457426556e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28647752,
+      "step": 36665
+    },
+    {
+      "epoch": 76.23700623700624,
+      "grad_norm": 1.010191499517532e-05,
+      "learning_rate": 8.506689715941679e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28651688,
+      "step": 36670
+    },
+    {
+      "epoch": 76.24740124740124,
+      "grad_norm": 8.248346148320707e-07,
+      "learning_rate": 8.481316215760011e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28655656,
+      "step": 36675
+    },
+    {
+      "epoch": 76.25779625779626,
+      "grad_norm": 3.8447033148258924e-06,
+      "learning_rate": 8.455979960794558e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28659528,
+      "step": 36680
+    },
+    {
+      "epoch": 76.26819126819127,
+      "grad_norm": 6.466067929977726e-07,
+      "learning_rate": 8.430680954952364e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28663432,
+      "step": 36685
+    },
+    {
+      "epoch": 76.27858627858627,
+      "grad_norm": 1.850234366429504e-05,
+      "learning_rate": 8.405419202134974e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 28667304,
+      "step": 36690
+    },
+    {
+      "epoch": 76.28898128898129,
+      "grad_norm": 0.0006923950277268887,
+      "learning_rate": 8.380194706237993e-07,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 28671144,
+      "step": 36695
+    },
+    {
+      "epoch": 76.2993762993763,
+      "grad_norm": 2.179185139539186e-05,
+      "learning_rate": 8.355007471151366e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28675048,
+      "step": 36700
+    },
+    {
+      "epoch": 76.3097713097713,
+      "grad_norm": 1.027342750603566e-05,
+      "learning_rate": 8.329857500759292e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28679048,
+      "step": 36705
+    },
+    {
+      "epoch": 76.32016632016632,
+      "grad_norm": 8.354780902664061e-07,
+      "learning_rate": 8.304744798940194e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28682920,
+      "step": 36710
+    },
+    {
+      "epoch": 76.33056133056132,
+      "grad_norm": 8.461142328997084e-07,
+      "learning_rate": 8.279669369566756e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28686696,
+      "step": 36715
+    },
+    {
+      "epoch": 76.34095634095634,
+      "grad_norm": 1.06530605989974e-05,
+      "learning_rate": 8.254631216505993e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28690728,
+      "step": 36720
+    },
+    {
+      "epoch": 76.35135135135135,
+      "grad_norm": 2.9954870115034282e-05,
+      "learning_rate": 8.229630343619038e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28694664,
+      "step": 36725
+    },
+    {
+      "epoch": 76.36174636174636,
+      "grad_norm": 2.7506753212946933e-06,
+      "learning_rate": 8.204666754761392e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28698472,
+      "step": 36730
+    },
+    {
+      "epoch": 76.37214137214137,
+      "grad_norm": 6.684096661047079e-06,
+      "learning_rate": 8.179740453782669e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28702408,
+      "step": 36735
+    },
+    {
+      "epoch": 76.38253638253639,
+      "grad_norm": 1.114421365855378e-06,
+      "learning_rate": 8.154851444526907e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28706248,
+      "step": 36740
+    },
+    {
+      "epoch": 76.39293139293139,
+      "grad_norm": 6.120611146798183e-07,
+      "learning_rate": 8.129999730832283e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28710120,
+      "step": 36745
+    },
+    {
+      "epoch": 76.4033264033264,
+      "grad_norm": 2.9381533295236295e-06,
+      "learning_rate": 8.105185316531178e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28713992,
+      "step": 36750
+    },
+    {
+      "epoch": 76.41372141372142,
+      "grad_norm": 1.3367733799896087e-06,
+      "learning_rate": 8.08040820545039e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28717768,
+      "step": 36755
+    },
+    {
+      "epoch": 76.42411642411642,
+      "grad_norm": 3.653909743661643e-06,
+      "learning_rate": 8.055668401410782e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28721576,
+      "step": 36760
+    },
+    {
+      "epoch": 76.43451143451144,
+      "grad_norm": 3.489655864541419e-05,
+      "learning_rate": 8.030965908227578e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28725736,
+      "step": 36765
+    },
+    {
+      "epoch": 76.44490644490645,
+      "grad_norm": 1.064045477505715e-06,
+      "learning_rate": 8.006300729710203e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28729768,
+      "step": 36770
+    },
+    {
+      "epoch": 76.45530145530145,
+      "grad_norm": 3.7088164390297607e-05,
+      "learning_rate": 7.981672869662337e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28733928,
+      "step": 36775
+    },
+    {
+      "epoch": 76.46569646569647,
+      "grad_norm": 1.6128858987940475e-05,
+      "learning_rate": 7.957082331881888e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28737768,
+      "step": 36780
+    },
+    {
+      "epoch": 76.47609147609148,
+      "grad_norm": 5.952760488980857e-07,
+      "learning_rate": 7.932529120161069e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28741800,
+      "step": 36785
+    },
+    {
+      "epoch": 76.48648648648648,
+      "grad_norm": 0.00015157742018345743,
+      "learning_rate": 7.908013238286243e-07,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 28745672,
+      "step": 36790
+    },
+    {
+      "epoch": 76.4968814968815,
+      "grad_norm": 6.68492987188074e-07,
+      "learning_rate": 7.883534690038136e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28749544,
+      "step": 36795
+    },
+    {
+      "epoch": 76.5072765072765,
+      "grad_norm": 2.535499334044289e-06,
+      "learning_rate": 7.859093479191559e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 28753672,
+      "step": 36800
+    },
+    {
+      "epoch": 76.5072765072765,
+      "eval_loss": 0.603899359703064,
+      "eval_runtime": 11.7126,
+      "eval_samples_per_second": 73.084,
+      "eval_steps_per_second": 18.271,
+      "num_input_tokens_seen": 28753672,
+      "step": 36800
+    },
+    {
+      "epoch": 76.51767151767152,
+      "grad_norm": 9.595217989044613e-07,
+      "learning_rate": 7.834689609515722e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28757480,
+      "step": 36805
+    },
+    {
+      "epoch": 76.52806652806653,
+      "grad_norm": 2.3498787413700484e-06,
+      "learning_rate": 7.810323084774002e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28761384,
+      "step": 36810
+    },
+    {
+      "epoch": 76.53846153846153,
+      "grad_norm": 0.0005013669724576175,
+      "learning_rate": 7.785993908723976e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28765224,
+      "step": 36815
+    },
+    {
+      "epoch": 76.54885654885655,
+      "grad_norm": 2.531738118705107e-06,
+      "learning_rate": 7.761702085117534e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28769256,
+      "step": 36820
+    },
+    {
+      "epoch": 76.55925155925156,
+      "grad_norm": 0.1509251594543457,
+      "learning_rate": 7.737447617700844e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 28773096,
+      "step": 36825
+    },
+    {
+      "epoch": 76.56964656964657,
+      "grad_norm": 2.3556046926387353e-06,
+      "learning_rate": 7.713230510214136e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28776904,
+      "step": 36830
+    },
+    {
+      "epoch": 76.58004158004158,
+      "grad_norm": 1.2016485015919898e-06,
+      "learning_rate": 7.689050766392092e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28780648,
+      "step": 36835
+    },
+    {
+      "epoch": 76.5904365904366,
+      "grad_norm": 2.627657067932887e-06,
+      "learning_rate": 7.664908389963477e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28784648,
+      "step": 36840
+    },
+    {
+      "epoch": 76.6008316008316,
+      "grad_norm": 5.042505108576734e-06,
+      "learning_rate": 7.64080338465134e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28788424,
+      "step": 36845
+    },
+    {
+      "epoch": 76.61122661122661,
+      "grad_norm": 1.672932398832927e-06,
+      "learning_rate": 7.616735754173043e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28792200,
+      "step": 36850
+    },
+    {
+      "epoch": 76.62162162162163,
+      "grad_norm": 8.392332347284537e-06,
+      "learning_rate": 7.592705502240005e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28796040,
+      "step": 36855
+    },
+    {
+      "epoch": 76.63201663201663,
+      "grad_norm": 0.00013852986739948392,
+      "learning_rate": 7.568712632558095e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28799880,
+      "step": 36860
+    },
+    {
+      "epoch": 76.64241164241164,
+      "grad_norm": 9.568223049427615e-07,
+      "learning_rate": 7.544757148827297e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28803560,
+      "step": 36865
+    },
+    {
+      "epoch": 76.65280665280665,
+      "grad_norm": 7.323463250941131e-06,
+      "learning_rate": 7.520839054741797e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28807464,
+      "step": 36870
+    },
+    {
+      "epoch": 76.66320166320166,
+      "grad_norm": 3.377648681635037e-06,
+      "learning_rate": 7.496958353990113e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28811464,
+      "step": 36875
+    },
+    {
+      "epoch": 76.67359667359668,
+      "grad_norm": 1.8408956748316996e-05,
+      "learning_rate": 7.473115050254941e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 28815432,
+      "step": 36880
+    },
+    {
+      "epoch": 76.68399168399168,
+      "grad_norm": 1.0570235644991044e-05,
+      "learning_rate": 7.449309147213173e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28819208,
+      "step": 36885
+    },
+    {
+      "epoch": 76.6943866943867,
+      "grad_norm": 2.1928560727246804e-06,
+      "learning_rate": 7.425540648536067e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28823112,
+      "step": 36890
+    },
+    {
+      "epoch": 76.70478170478171,
+      "grad_norm": 4.675663603848079e-06,
+      "learning_rate": 7.40180955788894e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28826984,
+      "step": 36895
+    },
+    {
+      "epoch": 76.71517671517671,
+      "grad_norm": 7.454789283656282e-07,
+      "learning_rate": 7.378115878931474e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28830856,
+      "step": 36900
+    },
+    {
+      "epoch": 76.72557172557173,
+      "grad_norm": 5.964598585705971e-06,
+      "learning_rate": 7.354459615317527e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28834824,
+      "step": 36905
+    },
+    {
+      "epoch": 76.73596673596674,
+      "grad_norm": 5.6501383369322866e-05,
+      "learning_rate": 7.33084077069518e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28838760,
+      "step": 36910
+    },
+    {
+      "epoch": 76.74636174636174,
+      "grad_norm": 2.2954413907427806e-06,
+      "learning_rate": 7.307259348706768e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28842760,
+      "step": 36915
+    },
+    {
+      "epoch": 76.75675675675676,
+      "grad_norm": 6.604265081477934e-07,
+      "learning_rate": 7.283715352988801e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28846728,
+      "step": 36920
+    },
+    {
+      "epoch": 76.76715176715177,
+      "grad_norm": 7.131583288355614e-07,
+      "learning_rate": 7.260208787172068e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28850536,
+      "step": 36925
+    },
+    {
+      "epoch": 76.77754677754677,
+      "grad_norm": 1.0469305379956495e-05,
+      "learning_rate": 7.23673965488167e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 28854536,
+      "step": 36930
+    },
+    {
+      "epoch": 76.78794178794179,
+      "grad_norm": 4.7177622946037445e-06,
+      "learning_rate": 7.213307959736709e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28858504,
+      "step": 36935
+    },
+    {
+      "epoch": 76.7983367983368,
+      "grad_norm": 8.766329847276211e-05,
+      "learning_rate": 7.189913705350715e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28862568,
+      "step": 36940
+    },
+    {
+      "epoch": 76.8087318087318,
+      "grad_norm": 1.3095658459860715e-06,
+      "learning_rate": 7.166556895331411e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28866440,
+      "step": 36945
+    },
+    {
+      "epoch": 76.81912681912682,
+      "grad_norm": 1.881171738205012e-05,
+      "learning_rate": 7.143237533280639e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28870312,
+      "step": 36950
+    },
+    {
+      "epoch": 76.82952182952182,
+      "grad_norm": 0.00016001438780222088,
+      "learning_rate": 7.119955622794578e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28874440,
+      "step": 36955
+    },
+    {
+      "epoch": 76.83991683991684,
+      "grad_norm": 6.303129680418351e-07,
+      "learning_rate": 7.096711167463577e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28878248,
+      "step": 36960
+    },
+    {
+      "epoch": 76.85031185031185,
+      "grad_norm": 0.0004028723342344165,
+      "learning_rate": 7.073504170872213e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28882184,
+      "step": 36965
+    },
+    {
+      "epoch": 76.86070686070686,
+      "grad_norm": 2.3072212570696138e-05,
+      "learning_rate": 7.05033463659932e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28886184,
+      "step": 36970
+    },
+    {
+      "epoch": 76.87110187110187,
+      "grad_norm": 2.3106114895199426e-05,
+      "learning_rate": 7.027202568217928e-07,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 28890152,
+      "step": 36975
+    },
+    {
+      "epoch": 76.88149688149689,
+      "grad_norm": 4.888278908765642e-06,
+      "learning_rate": 7.004107969295293e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28894152,
+      "step": 36980
+    },
+    {
+      "epoch": 76.89189189189189,
+      "grad_norm": 9.834850516199367e-07,
+      "learning_rate": 6.9810508433929e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28898088,
+      "step": 36985
+    },
+    {
+      "epoch": 76.9022869022869,
+      "grad_norm": 5.029110070609022e-07,
+      "learning_rate": 6.958031194066406e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28902024,
+      "step": 36990
+    },
+    {
+      "epoch": 76.91268191268192,
+      "grad_norm": 2.0733554265461862e-05,
+      "learning_rate": 6.935049024865776e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 28905896,
+      "step": 36995
+    },
+    {
+      "epoch": 76.92307692307692,
+      "grad_norm": 0.00011009421723429114,
+      "learning_rate": 6.912104339335118e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 28909800,
+      "step": 37000
+    },
+    {
+      "epoch": 76.92307692307692,
+      "eval_loss": 0.6039987206459045,
+      "eval_runtime": 11.6974,
+      "eval_samples_per_second": 73.179,
+      "eval_steps_per_second": 18.295,
+      "num_input_tokens_seen": 28909800,
+      "step": 37000
+    },
+    {
+      "epoch": 76.93347193347194,
+      "grad_norm": 1.8220556739834137e-05,
+      "learning_rate": 6.889197141012799e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28913704,
+      "step": 37005
+    },
+    {
+      "epoch": 76.94386694386695,
+      "grad_norm": 2.7793237677542493e-05,
+      "learning_rate": 6.866327433431435e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28917608,
+      "step": 37010
+    },
+    {
+      "epoch": 76.95426195426195,
+      "grad_norm": 0.0005856467178091407,
+      "learning_rate": 6.843495220117735e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28921512,
+      "step": 37015
+    },
+    {
+      "epoch": 76.96465696465697,
+      "grad_norm": 6.45062027615495e-05,
+      "learning_rate": 6.820700504592798e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28925448,
+      "step": 37020
+    },
+    {
+      "epoch": 76.97505197505197,
+      "grad_norm": 2.4967102945083752e-06,
+      "learning_rate": 6.797943290371839e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28929384,
+      "step": 37025
+    },
+    {
+      "epoch": 76.98544698544698,
+      "grad_norm": 8.161880941770505e-06,
+      "learning_rate": 6.775223580964274e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28933256,
+      "step": 37030
+    },
+    {
+      "epoch": 76.995841995842,
+      "grad_norm": 6.473717348853825e-07,
+      "learning_rate": 6.7525413798738e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 28937096,
+      "step": 37035
+    },
+    {
+      "epoch": 77.006237006237,
+      "grad_norm": 3.295225951660541e-06,
+      "learning_rate": 6.729896690598259e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28940856,
+      "step": 37040
+    },
+    {
+      "epoch": 77.01663201663202,
+      "grad_norm": 8.634377195448906e-07,
+      "learning_rate": 6.707289516629772e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28944888,
+      "step": 37045
+    },
+    {
+      "epoch": 77.02702702702703,
+      "grad_norm": 0.0011494504287838936,
+      "learning_rate": 6.684719861454692e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28948792,
+      "step": 37050
+    },
+    {
+      "epoch": 77.03742203742203,
+      "grad_norm": 9.155035513686016e-06,
+      "learning_rate": 6.662187728553481e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 28952696,
+      "step": 37055
+    },
+    {
+      "epoch": 77.04781704781705,
+      "grad_norm": 5.2593854888982605e-06,
+      "learning_rate": 6.639693121400892e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28956728,
+      "step": 37060
+    },
+    {
+      "epoch": 77.05821205821206,
+      "grad_norm": 1.7303951608482748e-05,
+      "learning_rate": 6.617236043465868e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28960696,
+      "step": 37065
+    },
+    {
+      "epoch": 77.06860706860707,
+      "grad_norm": 6.28166162641719e-07,
+      "learning_rate": 6.594816498211587e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28964472,
+      "step": 37070
+    },
+    {
+      "epoch": 77.07900207900208,
+      "grad_norm": 1.0502196346351411e-05,
+      "learning_rate": 6.572434489095447e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28968728,
+      "step": 37075
+    },
+    {
+      "epoch": 77.0893970893971,
+      "grad_norm": 1.8042503597825998e-06,
+      "learning_rate": 6.550090019568994e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28972664,
+      "step": 37080
+    },
+    {
+      "epoch": 77.0997920997921,
+      "grad_norm": 1.6032403209464974e-06,
+      "learning_rate": 6.527783093078027e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28976632,
+      "step": 37085
+    },
+    {
+      "epoch": 77.11018711018711,
+      "grad_norm": 6.993391252763104e-07,
+      "learning_rate": 6.5055137130626e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28980472,
+      "step": 37090
+    },
+    {
+      "epoch": 77.12058212058211,
+      "grad_norm": 3.928766091121361e-06,
+      "learning_rate": 6.483281882956854e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28984376,
+      "step": 37095
+    },
+    {
+      "epoch": 77.13097713097713,
+      "grad_norm": 2.0370189304230735e-06,
+      "learning_rate": 6.461087606189298e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28988216,
+      "step": 37100
+    },
+    {
+      "epoch": 77.14137214137214,
+      "grad_norm": 1.4836909940640908e-05,
+      "learning_rate": 6.438930886182554e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28992216,
+      "step": 37105
+    },
+    {
+      "epoch": 77.15176715176715,
+      "grad_norm": 1.0168348580918973e-06,
+      "learning_rate": 6.416811726353417e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28996152,
+      "step": 37110
+    },
+    {
+      "epoch": 77.16216216216216,
+      "grad_norm": 2.537656473577954e-05,
+      "learning_rate": 6.394730130112991e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 28999992,
+      "step": 37115
+    },
+    {
+      "epoch": 77.17255717255718,
+      "grad_norm": 4.392435585032217e-06,
+      "learning_rate": 6.372686100866471e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29003928,
+      "step": 37120
+    },
+    {
+      "epoch": 77.18295218295218,
+      "grad_norm": 4.621275365934707e-06,
+      "learning_rate": 6.350679642013413e-07,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 29007768,
+      "step": 37125
+    },
+    {
+      "epoch": 77.1933471933472,
+      "grad_norm": 4.5697153836954385e-05,
+      "learning_rate": 6.328710756947437e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29011608,
+      "step": 37130
+    },
+    {
+      "epoch": 77.20374220374221,
+      "grad_norm": 3.7178715501795523e-06,
+      "learning_rate": 6.306779449056416e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29015608,
+      "step": 37135
+    },
+    {
+      "epoch": 77.21413721413721,
+      "grad_norm": 1.7190513972309418e-06,
+      "learning_rate": 6.284885721722422e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29019192,
+      "step": 37140
+    },
+    {
+      "epoch": 77.22453222453223,
+      "grad_norm": 3.8216352550080046e-05,
+      "learning_rate": 6.26302957832181e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29023000,
+      "step": 37145
+    },
+    {
+      "epoch": 77.23492723492724,
+      "grad_norm": 2.9406926842057146e-05,
+      "learning_rate": 6.241211022224997e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29027064,
+      "step": 37150
+    },
+    {
+      "epoch": 77.24532224532224,
+      "grad_norm": 0.00031966797541826963,
+      "learning_rate": 6.219430056796732e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 29031000,
+      "step": 37155
+    },
+    {
+      "epoch": 77.25571725571726,
+      "grad_norm": 0.00030560846789740026,
+      "learning_rate": 6.19768668539586e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29034968,
+      "step": 37160
+    },
+    {
+      "epoch": 77.26611226611226,
+      "grad_norm": 7.359858045674628e-06,
+      "learning_rate": 6.175980911375528e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29038712,
+      "step": 37165
+    },
+    {
+      "epoch": 77.27650727650727,
+      "grad_norm": 2.161664497180027e-06,
+      "learning_rate": 6.154312738083034e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29042584,
+      "step": 37170
+    },
+    {
+      "epoch": 77.28690228690229,
+      "grad_norm": 1.2947664345119847e-06,
+      "learning_rate": 6.132682168859843e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29046360,
+      "step": 37175
+    },
+    {
+      "epoch": 77.29729729729729,
+      "grad_norm": 4.3743821152020246e-05,
+      "learning_rate": 6.111089207041704e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29050200,
+      "step": 37180
+    },
+    {
+      "epoch": 77.3076923076923,
+      "grad_norm": 1.1196718332939781e-05,
+      "learning_rate": 6.089533855958507e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29054328,
+      "step": 37185
+    },
+    {
+      "epoch": 77.31808731808732,
+      "grad_norm": 6.967961212467344e-07,
+      "learning_rate": 6.068016118934372e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29058328,
+      "step": 37190
+    },
+    {
+      "epoch": 77.32848232848232,
+      "grad_norm": 1.1337428986735176e-05,
+      "learning_rate": 6.04653599928759e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29062328,
+      "step": 37195
+    },
+    {
+      "epoch": 77.33887733887734,
+      "grad_norm": 1.1668519164231839e-06,
+      "learning_rate": 6.025093500330675e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29066104,
+      "step": 37200
+    },
+    {
+      "epoch": 77.33887733887734,
+      "eval_loss": 0.6020748615264893,
+      "eval_runtime": 11.723,
+      "eval_samples_per_second": 73.019,
+      "eval_steps_per_second": 18.255,
+      "num_input_tokens_seen": 29066104,
+      "step": 37200
+    },
+    {
+      "epoch": 77.34927234927235,
+      "grad_norm": 0.0003951512626372278,
+      "learning_rate": 6.003688625370291e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29070008,
+      "step": 37205
+    },
+    {
+      "epoch": 77.35966735966736,
+      "grad_norm": 8.212603279389441e-05,
+      "learning_rate": 5.982321377707406e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 29073944,
+      "step": 37210
+    },
+    {
+      "epoch": 77.37006237006237,
+      "grad_norm": 4.985207851859741e-06,
+      "learning_rate": 5.96099176063708e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29077752,
+      "step": 37215
+    },
+    {
+      "epoch": 77.38045738045739,
+      "grad_norm": 2.9383802029769868e-05,
+      "learning_rate": 5.93969977744857e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29081528,
+      "step": 37220
+    },
+    {
+      "epoch": 77.39085239085239,
+      "grad_norm": 4.1747247450985014e-05,
+      "learning_rate": 5.918445431425445e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29085464,
+      "step": 37225
+    },
+    {
+      "epoch": 77.4012474012474,
+      "grad_norm": 5.051401603850536e-05,
+      "learning_rate": 5.897228725845333e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29089496,
+      "step": 37230
+    },
+    {
+      "epoch": 77.41164241164242,
+      "grad_norm": 1.310831635237264e-06,
+      "learning_rate": 5.876049663980171e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29093336,
+      "step": 37235
+    },
+    {
+      "epoch": 77.42203742203742,
+      "grad_norm": 7.591681787744164e-05,
+      "learning_rate": 5.854908249095959e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29097304,
+      "step": 37240
+    },
+    {
+      "epoch": 77.43243243243244,
+      "grad_norm": 2.467608283041045e-06,
+      "learning_rate": 5.833804484453031e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29101112,
+      "step": 37245
+    },
+    {
+      "epoch": 77.44282744282744,
+      "grad_norm": 3.2005220873543294e-06,
+      "learning_rate": 5.81273837330587e-07,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 29104952,
+      "step": 37250
+    },
+    {
+      "epoch": 77.45322245322245,
+      "grad_norm": 1.5775492556713289e-06,
+      "learning_rate": 5.791709918903071e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29108856,
+      "step": 37255
+    },
+    {
+      "epoch": 77.46361746361747,
+      "grad_norm": 3.7007794162491336e-07,
+      "learning_rate": 5.770719124487483e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29112696,
+      "step": 37260
+    },
+    {
+      "epoch": 77.47401247401247,
+      "grad_norm": 1.8412400777378934e-06,
+      "learning_rate": 5.749765993296241e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29116600,
+      "step": 37265
+    },
+    {
+      "epoch": 77.48440748440748,
+      "grad_norm": 5.893950856261654e-06,
+      "learning_rate": 5.728850528560509e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29120664,
+      "step": 37270
+    },
+    {
+      "epoch": 77.4948024948025,
+      "grad_norm": 6.99781667208299e-05,
+      "learning_rate": 5.707972733505707e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29124600,
+      "step": 37275
+    },
+    {
+      "epoch": 77.5051975051975,
+      "grad_norm": 9.718397450342309e-06,
+      "learning_rate": 5.687132611351509e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29128536,
+      "step": 37280
+    },
+    {
+      "epoch": 77.51559251559252,
+      "grad_norm": 9.484642941970378e-05,
+      "learning_rate": 5.666330165311651e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29132600,
+      "step": 37285
+    },
+    {
+      "epoch": 77.52598752598753,
+      "grad_norm": 8.687005674801185e-07,
+      "learning_rate": 5.645565398594204e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29136408,
+      "step": 37290
+    },
+    {
+      "epoch": 77.53638253638253,
+      "grad_norm": 9.865281072052312e-07,
+      "learning_rate": 5.624838314401304e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29140280,
+      "step": 37295
+    },
+    {
+      "epoch": 77.54677754677755,
+      "grad_norm": 8.989718480734155e-07,
+      "learning_rate": 5.604148915929336e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29144280,
+      "step": 37300
+    },
+    {
+      "epoch": 77.55717255717256,
+      "grad_norm": 3.558741809683852e-05,
+      "learning_rate": 5.583497206368887e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 29148152,
+      "step": 37305
+    },
+    {
+      "epoch": 77.56756756756756,
+      "grad_norm": 6.167113042465644e-07,
+      "learning_rate": 5.562883188904688e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29151928,
+      "step": 37310
+    },
+    {
+      "epoch": 77.57796257796258,
+      "grad_norm": 1.5060045370773878e-05,
+      "learning_rate": 5.542306866715724e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29155928,
+      "step": 37315
+    },
+    {
+      "epoch": 77.58835758835758,
+      "grad_norm": 1.1077943781856447e-05,
+      "learning_rate": 5.52176824297504e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29159864,
+      "step": 37320
+    },
+    {
+      "epoch": 77.5987525987526,
+      "grad_norm": 1.1003582585544791e-05,
+      "learning_rate": 5.501267320850018e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29163864,
+      "step": 37325
+    },
+    {
+      "epoch": 77.60914760914761,
+      "grad_norm": 1.6498909189976985e-06,
+      "learning_rate": 5.480804103502157e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29167640,
+      "step": 37330
+    },
+    {
+      "epoch": 77.61954261954261,
+      "grad_norm": 2.3082089683157392e-05,
+      "learning_rate": 5.460378594087101e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29171576,
+      "step": 37335
+    },
+    {
+      "epoch": 77.62993762993763,
+      "grad_norm": 7.471901994904329e-07,
+      "learning_rate": 5.439990795754773e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29175448,
+      "step": 37340
+    },
+    {
+      "epoch": 77.64033264033264,
+      "grad_norm": 1.3588858564617112e-05,
+      "learning_rate": 5.419640711649188e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29179224,
+      "step": 37345
+    },
+    {
+      "epoch": 77.65072765072765,
+      "grad_norm": 4.04914408136392e-06,
+      "learning_rate": 5.399328344908583e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29183160,
+      "step": 37350
+    },
+    {
+      "epoch": 77.66112266112266,
+      "grad_norm": 3.779462076636264e-06,
+      "learning_rate": 5.379053698665399e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29187160,
+      "step": 37355
+    },
+    {
+      "epoch": 77.67151767151768,
+      "grad_norm": 2.4950909391918685e-06,
+      "learning_rate": 5.358816776046216e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29191064,
+      "step": 37360
+    },
+    {
+      "epoch": 77.68191268191268,
+      "grad_norm": 1.7270831449422985e-05,
+      "learning_rate": 5.338617580171817e-07,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 29194904,
+      "step": 37365
+    },
+    {
+      "epoch": 77.6923076923077,
+      "grad_norm": 3.1155545912042726e-06,
+      "learning_rate": 5.318456114157239e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29198776,
+      "step": 37370
+    },
+    {
+      "epoch": 77.70270270270271,
+      "grad_norm": 2.013255880228826e-06,
+      "learning_rate": 5.298332381111576e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29202616,
+      "step": 37375
+    },
+    {
+      "epoch": 77.71309771309771,
+      "grad_norm": 8.137975783029106e-06,
+      "learning_rate": 5.27824638413818e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29206648,
+      "step": 37380
+    },
+    {
+      "epoch": 77.72349272349273,
+      "grad_norm": 2.8632830435526557e-06,
+      "learning_rate": 5.258198126334546e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29210616,
+      "step": 37385
+    },
+    {
+      "epoch": 77.73388773388774,
+      "grad_norm": 7.909533223937615e-07,
+      "learning_rate": 5.238187610792367e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29214488,
+      "step": 37390
+    },
+    {
+      "epoch": 77.74428274428274,
+      "grad_norm": 9.516550562693737e-06,
+      "learning_rate": 5.218214840597563e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 29218392,
+      "step": 37395
+    },
+    {
+      "epoch": 77.75467775467776,
+      "grad_norm": 9.329673957836349e-06,
+      "learning_rate": 5.198279818830115e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29222328,
+      "step": 37400
+    },
+    {
+      "epoch": 77.75467775467776,
+      "eval_loss": 0.6024804711341858,
+      "eval_runtime": 11.7404,
+      "eval_samples_per_second": 72.911,
+      "eval_steps_per_second": 18.228,
+      "num_input_tokens_seen": 29222328,
+      "step": 37400
+    },
+    {
+      "epoch": 77.76507276507276,
+      "grad_norm": 1.1673283779600752e-06,
+      "learning_rate": 5.178382548564287e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29226200,
+      "step": 37405
+    },
+    {
+      "epoch": 77.77546777546777,
+      "grad_norm": 7.86171767686028e-07,
+      "learning_rate": 5.15852303286854e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29230104,
+      "step": 37410
+    },
+    {
+      "epoch": 77.78586278586279,
+      "grad_norm": 8.136104952427559e-06,
+      "learning_rate": 5.138701274805396e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29234008,
+      "step": 37415
+    },
+    {
+      "epoch": 77.79625779625779,
+      "grad_norm": 1.5307871308323229e-06,
+      "learning_rate": 5.118917277431606e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29237944,
+      "step": 37420
+    },
+    {
+      "epoch": 77.8066528066528,
+      "grad_norm": 6.718014901707647e-06,
+      "learning_rate": 5.099171043798145e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29241816,
+      "step": 37425
+    },
+    {
+      "epoch": 77.81704781704782,
+      "grad_norm": 0.0001818558812374249,
+      "learning_rate": 5.079462576950133e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29245688,
+      "step": 37430
+    },
+    {
+      "epoch": 77.82744282744282,
+      "grad_norm": 1.2868264320786693e-06,
+      "learning_rate": 5.059791879926862e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29249592,
+      "step": 37435
+    },
+    {
+      "epoch": 77.83783783783784,
+      "grad_norm": 7.59906652092468e-06,
+      "learning_rate": 5.040158955761793e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29253528,
+      "step": 37440
+    },
+    {
+      "epoch": 77.84823284823285,
+      "grad_norm": 2.7111229883303167e-06,
+      "learning_rate": 5.020563807482559e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29257464,
+      "step": 37445
+    },
+    {
+      "epoch": 77.85862785862786,
+      "grad_norm": 5.889025942451553e-06,
+      "learning_rate": 5.001006438110995e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29261304,
+      "step": 37450
+    },
+    {
+      "epoch": 77.86902286902287,
+      "grad_norm": 1.9330932445882354e-06,
+      "learning_rate": 4.981486850663075e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29265176,
+      "step": 37455
+    },
+    {
+      "epoch": 77.87941787941789,
+      "grad_norm": 1.0173959026360535e-06,
+      "learning_rate": 4.962005048149005e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29269016,
+      "step": 37460
+    },
+    {
+      "epoch": 77.88981288981289,
+      "grad_norm": 1.1105261364718899e-05,
+      "learning_rate": 4.942561033573073e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29272856,
+      "step": 37465
+    },
+    {
+      "epoch": 77.9002079002079,
+      "grad_norm": 2.482339732523542e-05,
+      "learning_rate": 4.923154809933827e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29276600,
+      "step": 37470
+    },
+    {
+      "epoch": 77.9106029106029,
+      "grad_norm": 2.016219013967202e-06,
+      "learning_rate": 4.903786380223957e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29280472,
+      "step": 37475
+    },
+    {
+      "epoch": 77.92099792099792,
+      "grad_norm": 7.284896582859801e-06,
+      "learning_rate": 4.884455747430266e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29284408,
+      "step": 37480
+    },
+    {
+      "epoch": 77.93139293139293,
+      "grad_norm": 1.3989062608743552e-05,
+      "learning_rate": 4.865162914533816e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29288344,
+      "step": 37485
+    },
+    {
+      "epoch": 77.94178794178794,
+      "grad_norm": 7.17000693839509e-06,
+      "learning_rate": 4.845907884509809e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 29292248,
+      "step": 37490
+    },
+    {
+      "epoch": 77.95218295218295,
+      "grad_norm": 1.2023232329738676e-06,
+      "learning_rate": 4.82669066032762e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29296184,
+      "step": 37495
+    },
+    {
+      "epoch": 77.96257796257797,
+      "grad_norm": 3.547880851328955e-06,
+      "learning_rate": 4.807511244950768e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29300088,
+      "step": 37500
+    },
+    {
+      "epoch": 77.97297297297297,
+      "grad_norm": 2.863322151824832e-05,
+      "learning_rate": 4.788369641336943e-07,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 29303992,
+      "step": 37505
+    },
+    {
+      "epoch": 77.98336798336798,
+      "grad_norm": 3.004987547683413e-06,
+      "learning_rate": 4.769265852438032e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29307928,
+      "step": 37510
+    },
+    {
+      "epoch": 77.993762993763,
+      "grad_norm": 2.6200675620202674e-06,
+      "learning_rate": 4.750199881200124e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29311864,
+      "step": 37515
+    },
+    {
+      "epoch": 78.004158004158,
+      "grad_norm": 1.8522705431678332e-05,
+      "learning_rate": 4.7311717305633664e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29315624,
+      "step": 37520
+    },
+    {
+      "epoch": 78.01455301455302,
+      "grad_norm": 0.00042007039883174,
+      "learning_rate": 4.7121814034621623e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29319496,
+      "step": 37525
+    },
+    {
+      "epoch": 78.02494802494803,
+      "grad_norm": 0.00020708984811790287,
+      "learning_rate": 4.693228902825114e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29323496,
+      "step": 37530
+    },
+    {
+      "epoch": 78.03534303534303,
+      "grad_norm": 1.7061596736311913e-05,
+      "learning_rate": 4.6743142315748277e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29327464,
+      "step": 37535
+    },
+    {
+      "epoch": 78.04573804573805,
+      "grad_norm": 0.0009342526900582016,
+      "learning_rate": 4.655437392628276e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29331464,
+      "step": 37540
+    },
+    {
+      "epoch": 78.05613305613305,
+      "grad_norm": 1.3612098882731516e-05,
+      "learning_rate": 4.636598388896463e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29335464,
+      "step": 37545
+    },
+    {
+      "epoch": 78.06652806652806,
+      "grad_norm": 5.027347356190148e-07,
+      "learning_rate": 4.6177972232845925e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29339240,
+      "step": 37550
+    },
+    {
+      "epoch": 78.07692307692308,
+      "grad_norm": 3.053649152207072e-06,
+      "learning_rate": 4.5990338986920953e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29343240,
+      "step": 37555
+    },
+    {
+      "epoch": 78.08731808731808,
+      "grad_norm": 3.6086518662159506e-07,
+      "learning_rate": 4.5803084180124633e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29347016,
+      "step": 37560
+    },
+    {
+      "epoch": 78.0977130977131,
+      "grad_norm": 4.6464505430776626e-05,
+      "learning_rate": 4.561620784133386e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29350888,
+      "step": 37565
+    },
+    {
+      "epoch": 78.10810810810811,
+      "grad_norm": 5.8735890888783615e-06,
+      "learning_rate": 4.5429709999367796e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29354792,
+      "step": 37570
+    },
+    {
+      "epoch": 78.11850311850311,
+      "grad_norm": 1.9058235238844645e-06,
+      "learning_rate": 4.5243590682986223e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29358696,
+      "step": 37575
+    },
+    {
+      "epoch": 78.12889812889813,
+      "grad_norm": 0.001614103908650577,
+      "learning_rate": 4.5057849920891735e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29362696,
+      "step": 37580
+    },
+    {
+      "epoch": 78.13929313929314,
+      "grad_norm": 0.0002229598758276552,
+      "learning_rate": 4.487248774172698e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29366600,
+      "step": 37585
+    },
+    {
+      "epoch": 78.14968814968815,
+      "grad_norm": 9.605419472791255e-05,
+      "learning_rate": 4.4687504174077965e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29370408,
+      "step": 37590
+    },
+    {
+      "epoch": 78.16008316008316,
+      "grad_norm": 5.878422939531447e-07,
+      "learning_rate": 4.450289924647133e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29374312,
+      "step": 37595
+    },
+    {
+      "epoch": 78.17047817047818,
+      "grad_norm": 9.09955724637257e-06,
+      "learning_rate": 4.431867298737513e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29378344,
+      "step": 37600
+    },
+    {
+      "epoch": 78.17047817047818,
+      "eval_loss": 0.6012476086616516,
+      "eval_runtime": 11.6995,
+      "eval_samples_per_second": 73.166,
+      "eval_steps_per_second": 18.291,
+      "num_input_tokens_seen": 29378344,
+      "step": 37600
+    },
+    {
+      "epoch": 78.18087318087318,
+      "grad_norm": 9.110849532589782e-07,
+      "learning_rate": 4.41348254251997e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29382280,
+      "step": 37605
+    },
+    {
+      "epoch": 78.1912681912682,
+      "grad_norm": 1.8118513480658294e-06,
+      "learning_rate": 4.395135658829652e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29386248,
+      "step": 37610
+    },
+    {
+      "epoch": 78.20166320166321,
+      "grad_norm": 5.386425254982896e-05,
+      "learning_rate": 4.376826650495852e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29390280,
+      "step": 37615
+    },
+    {
+      "epoch": 78.21205821205821,
+      "grad_norm": 4.964853701494576e-07,
+      "learning_rate": 4.358555520342117e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29394152,
+      "step": 37620
+    },
+    {
+      "epoch": 78.22245322245323,
+      "grad_norm": 1.0560769396761316e-06,
+      "learning_rate": 4.3403222711860257e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29397992,
+      "step": 37625
+    },
+    {
+      "epoch": 78.23284823284823,
+      "grad_norm": 1.3310276472111582e-06,
+      "learning_rate": 4.3221269058394133e-07,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 29401832,
+      "step": 37630
+    },
+    {
+      "epoch": 78.24324324324324,
+      "grad_norm": 7.636040390934795e-05,
+      "learning_rate": 4.303969427108173e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29405672,
+      "step": 37635
+    },
+    {
+      "epoch": 78.25363825363826,
+      "grad_norm": 3.401887443033047e-05,
+      "learning_rate": 4.2858498377924825e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29409448,
+      "step": 37640
+    },
+    {
+      "epoch": 78.26403326403326,
+      "grad_norm": 6.771103358005348e-07,
+      "learning_rate": 4.267768140686579e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29413320,
+      "step": 37645
+    },
+    {
+      "epoch": 78.27442827442827,
+      "grad_norm": 3.3237356547033414e-06,
+      "learning_rate": 4.2497243385788975e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29417352,
+      "step": 37650
+    },
+    {
+      "epoch": 78.28482328482329,
+      "grad_norm": 0.0005004482809454203,
+      "learning_rate": 4.231718434251991e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29421320,
+      "step": 37655
+    },
+    {
+      "epoch": 78.29521829521829,
+      "grad_norm": 1.3426484883893863e-06,
+      "learning_rate": 4.213750430482666e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29425224,
+      "step": 37660
+    },
+    {
+      "epoch": 78.3056133056133,
+      "grad_norm": 6.726186711603077e-06,
+      "learning_rate": 4.1958203300417054e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29429160,
+      "step": 37665
+    },
+    {
+      "epoch": 78.31600831600832,
+      "grad_norm": 1.0221453521808144e-05,
+      "learning_rate": 4.177928135694259e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29433096,
+      "step": 37670
+    },
+    {
+      "epoch": 78.32640332640332,
+      "grad_norm": 1.3336643860384356e-06,
+      "learning_rate": 4.1600738501994807e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29437064,
+      "step": 37675
+    },
+    {
+      "epoch": 78.33679833679834,
+      "grad_norm": 7.370453658950282e-06,
+      "learning_rate": 4.1422574763107237e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29441256,
+      "step": 37680
+    },
+    {
+      "epoch": 78.34719334719335,
+      "grad_norm": 1.9247247564635472e-06,
+      "learning_rate": 4.124479016775512e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29445192,
+      "step": 37685
+    },
+    {
+      "epoch": 78.35758835758836,
+      "grad_norm": 1.6416543076047674e-05,
+      "learning_rate": 4.106738474335514e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29449160,
+      "step": 37690
+    },
+    {
+      "epoch": 78.36798336798337,
+      "grad_norm": 9.802696467886562e-07,
+      "learning_rate": 4.089035851726486e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29453032,
+      "step": 37695
+    },
+    {
+      "epoch": 78.37837837837837,
+      "grad_norm": 5.729139957111329e-06,
+      "learning_rate": 4.0713711516784937e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29456776,
+      "step": 37700
+    },
+    {
+      "epoch": 78.38877338877339,
+      "grad_norm": 4.0760673414297344e-07,
+      "learning_rate": 4.05374437691558e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29460616,
+      "step": 37705
+    },
+    {
+      "epoch": 78.3991683991684,
+      "grad_norm": 4.372502644400811e-07,
+      "learning_rate": 4.036155530156044e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29464520,
+      "step": 37710
+    },
+    {
+      "epoch": 78.4095634095634,
+      "grad_norm": 8.304728908115067e-06,
+      "learning_rate": 4.018604614112298e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29468424,
+      "step": 37715
+    },
+    {
+      "epoch": 78.41995841995842,
+      "grad_norm": 6.975408177822828e-05,
+      "learning_rate": 4.0010916314908996e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29472360,
+      "step": 37720
+    },
+    {
+      "epoch": 78.43035343035343,
+      "grad_norm": 1.1335883755236864e-05,
+      "learning_rate": 3.983616584992578e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29476200,
+      "step": 37725
+    },
+    {
+      "epoch": 78.44074844074844,
+      "grad_norm": 0.1504909247159958,
+      "learning_rate": 3.9661794773122595e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29480168,
+      "step": 37730
+    },
+    {
+      "epoch": 78.45114345114345,
+      "grad_norm": 3.219677455490455e-05,
+      "learning_rate": 3.9487803111388777e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29483944,
+      "step": 37735
+    },
+    {
+      "epoch": 78.46153846153847,
+      "grad_norm": 1.2690702533291187e-05,
+      "learning_rate": 3.9314190891556747e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29487880,
+      "step": 37740
+    },
+    {
+      "epoch": 78.47193347193347,
+      "grad_norm": 0.14241454005241394,
+      "learning_rate": 3.914095814039925e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 29491880,
+      "step": 37745
+    },
+    {
+      "epoch": 78.48232848232848,
+      "grad_norm": 4.1151042751153e-06,
+      "learning_rate": 3.896810488463104e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29495752,
+      "step": 37750
+    },
+    {
+      "epoch": 78.4927234927235,
+      "grad_norm": 1.5382220226456411e-06,
+      "learning_rate": 3.8795631150908565e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29499592,
+      "step": 37755
+    },
+    {
+      "epoch": 78.5031185031185,
+      "grad_norm": 8.253038686234504e-05,
+      "learning_rate": 3.862353696582888e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29503496,
+      "step": 37760
+    },
+    {
+      "epoch": 78.51351351351352,
+      "grad_norm": 4.121566234971397e-06,
+      "learning_rate": 3.8451822355931313e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29507304,
+      "step": 37765
+    },
+    {
+      "epoch": 78.52390852390852,
+      "grad_norm": 3.5391278174756735e-07,
+      "learning_rate": 3.82804873476969e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29511208,
+      "step": 37770
+    },
+    {
+      "epoch": 78.53430353430353,
+      "grad_norm": 9.069503335012996e-07,
+      "learning_rate": 3.810953196754702e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29515336,
+      "step": 37775
+    },
+    {
+      "epoch": 78.54469854469855,
+      "grad_norm": 7.57122179493308e-05,
+      "learning_rate": 3.793895624184529e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29519176,
+      "step": 37780
+    },
+    {
+      "epoch": 78.55509355509355,
+      "grad_norm": 0.00010899108019657433,
+      "learning_rate": 3.776876019689679e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29523272,
+      "step": 37785
+    },
+    {
+      "epoch": 78.56548856548856,
+      "grad_norm": 2.853111709555378e-06,
+      "learning_rate": 3.7598943858947743e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29527112,
+      "step": 37790
+    },
+    {
+      "epoch": 78.57588357588358,
+      "grad_norm": 5.96598283664207e-06,
+      "learning_rate": 3.742950725418637e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29530888,
+      "step": 37795
+    },
+    {
+      "epoch": 78.58627858627858,
+      "grad_norm": 0.0004339931474532932,
+      "learning_rate": 3.726045040874093e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29534888,
+      "step": 37800
+    },
+    {
+      "epoch": 78.58627858627858,
+      "eval_loss": 0.6024923324584961,
+      "eval_runtime": 11.6963,
+      "eval_samples_per_second": 73.185,
+      "eval_steps_per_second": 18.296,
+      "num_input_tokens_seen": 29534888,
+      "step": 37800
+    },
+    {
+      "epoch": 78.5966735966736,
+      "grad_norm": 1.912964307848597e-06,
+      "learning_rate": 3.709177334868308e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29538696,
+      "step": 37805
+    },
+    {
+      "epoch": 78.60706860706861,
+      "grad_norm": 1.452194010198582e-05,
+      "learning_rate": 3.692347610002478e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 29542600,
+      "step": 37810
+    },
+    {
+      "epoch": 78.61746361746361,
+      "grad_norm": 4.3121599446749315e-06,
+      "learning_rate": 3.675555868871916e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29546632,
+      "step": 37815
+    },
+    {
+      "epoch": 78.62785862785863,
+      "grad_norm": 1.0162854096051888e-06,
+      "learning_rate": 3.658802114066162e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29550376,
+      "step": 37820
+    },
+    {
+      "epoch": 78.63825363825364,
+      "grad_norm": 7.65515687817242e-06,
+      "learning_rate": 3.6420863481688437e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29554344,
+      "step": 37825
+    },
+    {
+      "epoch": 78.64864864864865,
+      "grad_norm": 1.4852534150122665e-05,
+      "learning_rate": 3.625408573757705e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29558216,
+      "step": 37830
+    },
+    {
+      "epoch": 78.65904365904366,
+      "grad_norm": 2.083835624944186e-06,
+      "learning_rate": 3.608768793404743e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29562152,
+      "step": 37835
+    },
+    {
+      "epoch": 78.66943866943868,
+      "grad_norm": 2.527964625187451e-06,
+      "learning_rate": 3.592167009675934e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29565960,
+      "step": 37840
+    },
+    {
+      "epoch": 78.67983367983368,
+      "grad_norm": 9.475403203396127e-06,
+      "learning_rate": 3.575603225131563e-07,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 29569736,
+      "step": 37845
+    },
+    {
+      "epoch": 78.6902286902287,
+      "grad_norm": 2.4795683657430345e-06,
+      "learning_rate": 3.55907744232592e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29573544,
+      "step": 37850
+    },
+    {
+      "epoch": 78.7006237006237,
+      "grad_norm": 1.530647409708763e-06,
+      "learning_rate": 3.5425896638075217e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29577352,
+      "step": 37855
+    },
+    {
+      "epoch": 78.71101871101871,
+      "grad_norm": 5.047548938819091e-07,
+      "learning_rate": 3.5261398921189736e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29581160,
+      "step": 37860
+    },
+    {
+      "epoch": 78.72141372141373,
+      "grad_norm": 5.690270086233795e-07,
+      "learning_rate": 3.509728129797024e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29584840,
+      "step": 37865
+    },
+    {
+      "epoch": 78.73180873180873,
+      "grad_norm": 1.6169851733138785e-05,
+      "learning_rate": 3.4933543793725656e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29588776,
+      "step": 37870
+    },
+    {
+      "epoch": 78.74220374220374,
+      "grad_norm": 9.926974598784e-06,
+      "learning_rate": 3.4770186433707163e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29592648,
+      "step": 37875
+    },
+    {
+      "epoch": 78.75259875259876,
+      "grad_norm": 1.0912062862189487e-05,
+      "learning_rate": 3.4607209243105453e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29596648,
+      "step": 37880
+    },
+    {
+      "epoch": 78.76299376299376,
+      "grad_norm": 2.6713410079537425e-06,
+      "learning_rate": 3.444461224705431e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29600424,
+      "step": 37885
+    },
+    {
+      "epoch": 78.77338877338877,
+      "grad_norm": 3.982356702181278e-06,
+      "learning_rate": 3.4282395470628116e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29604104,
+      "step": 37890
+    },
+    {
+      "epoch": 78.78378378378379,
+      "grad_norm": 1.8408256892143982e-06,
+      "learning_rate": 3.4120558938842417e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29607944,
+      "step": 37895
+    },
+    {
+      "epoch": 78.79417879417879,
+      "grad_norm": 1.386531766911503e-05,
+      "learning_rate": 3.395910267665503e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29611816,
+      "step": 37900
+    },
+    {
+      "epoch": 78.8045738045738,
+      "grad_norm": 2.4037806269916473e-06,
+      "learning_rate": 3.3798026708964094e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29615688,
+      "step": 37905
+    },
+    {
+      "epoch": 78.81496881496882,
+      "grad_norm": 6.129638222773792e-06,
+      "learning_rate": 3.3637331060609456e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29619656,
+      "step": 37910
+    },
+    {
+      "epoch": 78.82536382536382,
+      "grad_norm": 2.756995172603638e-06,
+      "learning_rate": 3.3477015756372966e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29623592,
+      "step": 37915
+    },
+    {
+      "epoch": 78.83575883575884,
+      "grad_norm": 5.244598355602648e-07,
+      "learning_rate": 3.3317080820976785e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29627464,
+      "step": 37920
+    },
+    {
+      "epoch": 78.84615384615384,
+      "grad_norm": 2.182886419177521e-06,
+      "learning_rate": 3.315752627908508e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29631464,
+      "step": 37925
+    },
+    {
+      "epoch": 78.85654885654886,
+      "grad_norm": 0.00014059190289117396,
+      "learning_rate": 3.299835215530317e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 29635400,
+      "step": 37930
+    },
+    {
+      "epoch": 78.86694386694387,
+      "grad_norm": 1.5809770047781058e-05,
+      "learning_rate": 3.2839558474177245e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29639336,
+      "step": 37935
+    },
+    {
+      "epoch": 78.87733887733887,
+      "grad_norm": 3.196827674400993e-05,
+      "learning_rate": 3.2681145260196056e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29643240,
+      "step": 37940
+    },
+    {
+      "epoch": 78.88773388773389,
+      "grad_norm": 0.0005057404050603509,
+      "learning_rate": 3.252311253778839e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29647112,
+      "step": 37945
+    },
+    {
+      "epoch": 78.8981288981289,
+      "grad_norm": 2.6952611733577214e-05,
+      "learning_rate": 3.2365460331325034e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 29651048,
+      "step": 37950
+    },
+    {
+      "epoch": 78.9085239085239,
+      "grad_norm": 0.00016448421229142696,
+      "learning_rate": 3.2208188665117934e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29655240,
+      "step": 37955
+    },
+    {
+      "epoch": 78.91891891891892,
+      "grad_norm": 1.214806343341479e-05,
+      "learning_rate": 3.205129756342018e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29659272,
+      "step": 37960
+    },
+    {
+      "epoch": 78.92931392931393,
+      "grad_norm": 2.8098775146645494e-06,
+      "learning_rate": 3.189478705042659e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29663304,
+      "step": 37965
+    },
+    {
+      "epoch": 78.93970893970894,
+      "grad_norm": 0.1371154636144638,
+      "learning_rate": 3.173865715027341e-07,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 29667112,
+      "step": 37970
+    },
+    {
+      "epoch": 78.95010395010395,
+      "grad_norm": 9.83527297648834e-06,
+      "learning_rate": 3.158290788703694e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29671048,
+      "step": 37975
+    },
+    {
+      "epoch": 78.96049896049897,
+      "grad_norm": 7.086055120453238e-07,
+      "learning_rate": 3.1427539284736297e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29675080,
+      "step": 37980
+    },
+    {
+      "epoch": 78.97089397089397,
+      "grad_norm": 1.259014993593155e-06,
+      "learning_rate": 3.127255136733093e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29678952,
+      "step": 37985
+    },
+    {
+      "epoch": 78.98128898128898,
+      "grad_norm": 1.1308104603813263e-06,
+      "learning_rate": 3.1117944158722544e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 29682792,
+      "step": 37990
+    },
+    {
+      "epoch": 78.99168399168398,
+      "grad_norm": 6.121677870396525e-06,
+      "learning_rate": 3.0963717682752635e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29686600,
+      "step": 37995
+    },
+    {
+      "epoch": 79.002079002079,
+      "grad_norm": 4.979322511644568e-06,
+      "learning_rate": 3.080987196320578e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29690392,
+      "step": 38000
+    },
+    {
+      "epoch": 79.002079002079,
+      "eval_loss": 0.5996507406234741,
+      "eval_runtime": 11.7106,
+      "eval_samples_per_second": 73.096,
+      "eval_steps_per_second": 18.274,
+      "num_input_tokens_seen": 29690392,
+      "step": 38000
+    },
+    {
+      "epoch": 79.01247401247402,
+      "grad_norm": 7.035234830254922e-07,
+      "learning_rate": 3.065640702380607e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29694168,
+      "step": 38005
+    },
+    {
+      "epoch": 79.02286902286902,
+      "grad_norm": 0.00019932024588342756,
+      "learning_rate": 3.050332288822011e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29698072,
+      "step": 38010
+    },
+    {
+      "epoch": 79.03326403326403,
+      "grad_norm": 7.282409205799922e-05,
+      "learning_rate": 3.035061958005542e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29702072,
+      "step": 38015
+    },
+    {
+      "epoch": 79.04365904365905,
+      "grad_norm": 1.94699259736808e-05,
+      "learning_rate": 3.019829712286093e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29705880,
+      "step": 38020
+    },
+    {
+      "epoch": 79.05405405405405,
+      "grad_norm": 1.938579043780919e-05,
+      "learning_rate": 3.004635554012647e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29709848,
+      "step": 38025
+    },
+    {
+      "epoch": 79.06444906444906,
+      "grad_norm": 1.7227068838110426e-06,
+      "learning_rate": 2.9894794855283017e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29713784,
+      "step": 38030
+    },
+    {
+      "epoch": 79.07484407484408,
+      "grad_norm": 2.08104461307812e-06,
+      "learning_rate": 2.9743615091703816e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29717752,
+      "step": 38035
+    },
+    {
+      "epoch": 79.08523908523908,
+      "grad_norm": 2.0149884676357033e-06,
+      "learning_rate": 2.959281627270216e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29721560,
+      "step": 38040
+    },
+    {
+      "epoch": 79.0956340956341,
+      "grad_norm": 9.175577133646584e-07,
+      "learning_rate": 2.944239842153362e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29725336,
+      "step": 38045
+    },
+    {
+      "epoch": 79.10602910602911,
+      "grad_norm": 1.2409592500262079e-06,
+      "learning_rate": 2.929236156139381e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29729368,
+      "step": 38050
+    },
+    {
+      "epoch": 79.11642411642411,
+      "grad_norm": 1.200592851091642e-05,
+      "learning_rate": 2.9142705715420883e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29733240,
+      "step": 38055
+    },
+    {
+      "epoch": 79.12681912681913,
+      "grad_norm": 1.8069722500513308e-06,
+      "learning_rate": 2.8993430906693595e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29737240,
+      "step": 38060
+    },
+    {
+      "epoch": 79.13721413721414,
+      "grad_norm": 6.353552635118831e-07,
+      "learning_rate": 2.88445371582316e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29741304,
+      "step": 38065
+    },
+    {
+      "epoch": 79.14760914760915,
+      "grad_norm": 0.14918871223926544,
+      "learning_rate": 2.8696024492996796e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29745112,
+      "step": 38070
+    },
+    {
+      "epoch": 79.15800415800416,
+      "grad_norm": 8.06368698249571e-05,
+      "learning_rate": 2.854789293389115e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29749112,
+      "step": 38075
+    },
+    {
+      "epoch": 79.16839916839916,
+      "grad_norm": 1.756070923875086e-05,
+      "learning_rate": 2.8400142503758606e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 29753048,
+      "step": 38080
+    },
+    {
+      "epoch": 79.17879417879418,
+      "grad_norm": 6.438253421947593e-06,
+      "learning_rate": 2.8252773225384276e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29757016,
+      "step": 38085
+    },
+    {
+      "epoch": 79.1891891891892,
+      "grad_norm": 3.1942677196639124e-06,
+      "learning_rate": 2.8105785121494143e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29760920,
+      "step": 38090
+    },
+    {
+      "epoch": 79.1995841995842,
+      "grad_norm": 1.0770771723400685e-06,
+      "learning_rate": 2.795917821475563e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29764888,
+      "step": 38095
+    },
+    {
+      "epoch": 79.20997920997921,
+      "grad_norm": 3.840274075628258e-05,
+      "learning_rate": 2.78129525277776e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29768760,
+      "step": 38100
+    },
+    {
+      "epoch": 79.22037422037423,
+      "grad_norm": 6.342056622088421e-06,
+      "learning_rate": 2.766710808310952e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29772600,
+      "step": 38105
+    },
+    {
+      "epoch": 79.23076923076923,
+      "grad_norm": 1.784286951078684e-06,
+      "learning_rate": 2.7521644903242827e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29776568,
+      "step": 38110
+    },
+    {
+      "epoch": 79.24116424116424,
+      "grad_norm": 2.067717559839366e-06,
+      "learning_rate": 2.7376563010609593e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29780440,
+      "step": 38115
+    },
+    {
+      "epoch": 79.25155925155926,
+      "grad_norm": 7.326642048610665e-07,
+      "learning_rate": 2.72318624275833e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29784216,
+      "step": 38120
+    },
+    {
+      "epoch": 79.26195426195426,
+      "grad_norm": 2.939106707344763e-06,
+      "learning_rate": 2.7087543176478324e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29788088,
+      "step": 38125
+    },
+    {
+      "epoch": 79.27234927234927,
+      "grad_norm": 1.7973012290894985e-05,
+      "learning_rate": 2.694360527955103e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29792248,
+      "step": 38130
+    },
+    {
+      "epoch": 79.28274428274429,
+      "grad_norm": 0.1437721699476242,
+      "learning_rate": 2.680004875899811e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 29796120,
+      "step": 38135
+    },
+    {
+      "epoch": 79.29313929313929,
+      "grad_norm": 1.4645545888924971e-05,
+      "learning_rate": 2.665687363695768e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29800088,
+      "step": 38140
+    },
+    {
+      "epoch": 79.3035343035343,
+      "grad_norm": 4.342287411418511e-06,
+      "learning_rate": 2.6514079935509584e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29804024,
+      "step": 38145
+    },
+    {
+      "epoch": 79.31392931392931,
+      "grad_norm": 3.6458961858443217e-06,
+      "learning_rate": 2.6371667676673983e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29808088,
+      "step": 38150
+    },
+    {
+      "epoch": 79.32432432432432,
+      "grad_norm": 0.0004344577027950436,
+      "learning_rate": 2.6229636882412755e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29812056,
+      "step": 38155
+    },
+    {
+      "epoch": 79.33471933471934,
+      "grad_norm": 3.9736301005177666e-06,
+      "learning_rate": 2.6087987574628935e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 29815864,
+      "step": 38160
+    },
+    {
+      "epoch": 79.34511434511434,
+      "grad_norm": 6.536622549901949e-06,
+      "learning_rate": 2.5946719775166437e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 29819736,
+      "step": 38165
+    },
+    {
+      "epoch": 79.35550935550935,
+      "grad_norm": 1.2018939514746307e-06,
+      "learning_rate": 2.5805833505810616e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29823672,
+      "step": 38170
+    },
+    {
+      "epoch": 79.36590436590437,
+      "grad_norm": 1.955573225131957e-06,
+      "learning_rate": 2.566532878828798e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 29827416,
+      "step": 38175
+    },
+    {
+      "epoch": 79.37629937629937,
+      "grad_norm": 7.520452527387533e-07,
+      "learning_rate": 2.552520564426619e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29831352,
+      "step": 38180
+    },
+    {
+      "epoch": 79.38669438669439,
+      "grad_norm": 1.3541226735469536e-06,
+      "learning_rate": 2.5385464095353803e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29835256,
+      "step": 38185
+    },
+    {
+      "epoch": 79.3970893970894,
+      "grad_norm": 1.163415981864091e-05,
+      "learning_rate": 2.5246104163100804e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29839096,
+      "step": 38190
+    },
+    {
+      "epoch": 79.4074844074844,
+      "grad_norm": 4.942543796460086e-07,
+      "learning_rate": 2.510712586899833e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29843064,
+      "step": 38195
+    },
+    {
+      "epoch": 79.41787941787942,
+      "grad_norm": 8.521003110217862e-06,
+      "learning_rate": 2.4968529234478124e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29846936,
+      "step": 38200
+    },
+    {
+      "epoch": 79.41787941787942,
+      "eval_loss": 0.6018643975257874,
+      "eval_runtime": 11.7172,
+      "eval_samples_per_second": 73.055,
+      "eval_steps_per_second": 18.264,
+      "num_input_tokens_seen": 29846936,
+      "step": 38200
+    },
+    {
+      "epoch": 79.42827442827443,
+      "grad_norm": 8.731250272830948e-05,
+      "learning_rate": 2.483031428091448e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29850968,
+      "step": 38205
+    },
+    {
+      "epoch": 79.43866943866944,
+      "grad_norm": 7.677322173549328e-06,
+      "learning_rate": 2.469248102962091e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29854936,
+      "step": 38210
+    },
+    {
+      "epoch": 79.44906444906445,
+      "grad_norm": 3.5391360597714083e-07,
+      "learning_rate": 2.4555029501853455e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29858808,
+      "step": 38215
+    },
+    {
+      "epoch": 79.45945945945945,
+      "grad_norm": 2.596345211713924e-06,
+      "learning_rate": 2.441795971880906e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29862456,
+      "step": 38220
+    },
+    {
+      "epoch": 79.46985446985447,
+      "grad_norm": 0.0003836444520857185,
+      "learning_rate": 2.4281271701625255e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29866264,
+      "step": 38225
+    },
+    {
+      "epoch": 79.48024948024948,
+      "grad_norm": 3.87784120903234e-06,
+      "learning_rate": 2.4144965471381007e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29870232,
+      "step": 38230
+    },
+    {
+      "epoch": 79.49064449064448,
+      "grad_norm": 7.338613158935914e-06,
+      "learning_rate": 2.400904104909674e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29874136,
+      "step": 38235
+    },
+    {
+      "epoch": 79.5010395010395,
+      "grad_norm": 5.2902480092598125e-05,
+      "learning_rate": 2.3873498455733725e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29878040,
+      "step": 38240
+    },
+    {
+      "epoch": 79.51143451143452,
+      "grad_norm": 3.351777559146285e-05,
+      "learning_rate": 2.3738337712194137e-07,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 29881944,
+      "step": 38245
+    },
+    {
+      "epoch": 79.52182952182952,
+      "grad_norm": 1.3769724318990484e-05,
+      "learning_rate": 2.3603558839321305e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29885784,
+      "step": 38250
+    },
+    {
+      "epoch": 79.53222453222453,
+      "grad_norm": 2.0543827758956468e-06,
+      "learning_rate": 2.3469161857900267e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29889720,
+      "step": 38255
+    },
+    {
+      "epoch": 79.54261954261955,
+      "grad_norm": 2.5736553652677685e-06,
+      "learning_rate": 2.3335146788656393e-07,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 29893528,
+      "step": 38260
+    },
+    {
+      "epoch": 79.55301455301455,
+      "grad_norm": 9.691514605947305e-06,
+      "learning_rate": 2.3201513652256757e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29897464,
+      "step": 38265
+    },
+    {
+      "epoch": 79.56340956340956,
+      "grad_norm": 0.14656265079975128,
+      "learning_rate": 2.3068262469308766e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 29901240,
+      "step": 38270
+    },
+    {
+      "epoch": 79.57380457380458,
+      "grad_norm": 1.0852199920918792e-06,
+      "learning_rate": 2.2935393260362093e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29905144,
+      "step": 38275
+    },
+    {
+      "epoch": 79.58419958419958,
+      "grad_norm": 6.56845145385887e-07,
+      "learning_rate": 2.2802906045906458e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29908888,
+      "step": 38280
+    },
+    {
+      "epoch": 79.5945945945946,
+      "grad_norm": 3.7732294003944844e-05,
+      "learning_rate": 2.2670800846373018e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29912664,
+      "step": 38285
+    },
+    {
+      "epoch": 79.60498960498961,
+      "grad_norm": 1.7740687781042652e-06,
+      "learning_rate": 2.2539077682134367e-07,
+      "loss": 0.003,
+      "num_input_tokens_seen": 29916696,
+      "step": 38290
+    },
+    {
+      "epoch": 79.61538461538461,
+      "grad_norm": 2.911367346314364e-06,
+      "learning_rate": 2.2407736573503423e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29920664,
+      "step": 38295
+    },
+    {
+      "epoch": 79.62577962577963,
+      "grad_norm": 5.904499857933843e-07,
+      "learning_rate": 2.2276777540735093e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29924568,
+      "step": 38300
+    },
+    {
+      "epoch": 79.63617463617463,
+      "grad_norm": 0.00013188844604883343,
+      "learning_rate": 2.2146200604024613e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29928440,
+      "step": 38305
+    },
+    {
+      "epoch": 79.64656964656965,
+      "grad_norm": 1.0362868124502711e-05,
+      "learning_rate": 2.2016005783508375e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29932504,
+      "step": 38310
+    },
+    {
+      "epoch": 79.65696465696466,
+      "grad_norm": 2.1883786757825874e-05,
+      "learning_rate": 2.1886193099264763e-07,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 29936472,
+      "step": 38315
+    },
+    {
+      "epoch": 79.66735966735966,
+      "grad_norm": 9.919017429638188e-06,
+      "learning_rate": 2.175676257131165e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29940504,
+      "step": 38320
+    },
+    {
+      "epoch": 79.67775467775468,
+      "grad_norm": 1.9135339243803173e-05,
+      "learning_rate": 2.162771421960974e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29944344,
+      "step": 38325
+    },
+    {
+      "epoch": 79.6881496881497,
+      "grad_norm": 2.9785369406454265e-05,
+      "learning_rate": 2.1499048064059224e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29948056,
+      "step": 38330
+    },
+    {
+      "epoch": 79.6985446985447,
+      "grad_norm": 5.849030912941089e-07,
+      "learning_rate": 2.1370764124502285e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29951896,
+      "step": 38335
+    },
+    {
+      "epoch": 79.70893970893971,
+      "grad_norm": 1.212611778100836e-06,
+      "learning_rate": 2.1242862420721988e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29955832,
+      "step": 38340
+    },
+    {
+      "epoch": 79.71933471933473,
+      "grad_norm": 3.0973449156590505e-06,
+      "learning_rate": 2.1115342972442276e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29959672,
+      "step": 38345
+    },
+    {
+      "epoch": 79.72972972972973,
+      "grad_norm": 4.3601512516033836e-06,
+      "learning_rate": 2.0988205799328252e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29963512,
+      "step": 38350
+    },
+    {
+      "epoch": 79.74012474012474,
+      "grad_norm": 4.7578914745827205e-06,
+      "learning_rate": 2.0861450920986182e-07,
+      "loss": 0.0029,
+      "num_input_tokens_seen": 29967384,
+      "step": 38355
+    },
+    {
+      "epoch": 79.75051975051976,
+      "grad_norm": 0.0002273958525620401,
+      "learning_rate": 2.07350783569632e-07,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 29971320,
+      "step": 38360
+    },
+    {
+      "epoch": 79.76091476091476,
+      "grad_norm": 1.3559686067310395e-06,
+      "learning_rate": 2.060908812674761e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29975224,
+      "step": 38365
+    },
+    {
+      "epoch": 79.77130977130977,
+      "grad_norm": 1.6181922546820715e-05,
+      "learning_rate": 2.0483480249768317e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29979000,
+      "step": 38370
+    },
+    {
+      "epoch": 79.78170478170478,
+      "grad_norm": 0.00016278584371320903,
+      "learning_rate": 2.035825474539621e-07,
+      "loss": 0.0055,
+      "num_input_tokens_seen": 29982968,
+      "step": 38375
+    },
+    {
+      "epoch": 79.79209979209979,
+      "grad_norm": 2.4973587642307393e-06,
+      "learning_rate": 2.0233411632942235e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29986904,
+      "step": 38380
+    },
+    {
+      "epoch": 79.8024948024948,
+      "grad_norm": 5.8929683291353285e-05,
+      "learning_rate": 2.0108950931658764e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29990712,
+      "step": 38385
+    },
+    {
+      "epoch": 79.81288981288981,
+      "grad_norm": 3.394402938283747e-06,
+      "learning_rate": 1.998487266073934e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29994616,
+      "step": 38390
+    },
+    {
+      "epoch": 79.82328482328482,
+      "grad_norm": 4.616777459887089e-06,
+      "learning_rate": 1.986117683931865e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 29998488,
+      "step": 38395
+    },
+    {
+      "epoch": 79.83367983367984,
+      "grad_norm": 8.881207236299815e-07,
+      "learning_rate": 1.9737863486471442e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30002424,
+      "step": 38400
+    },
+    {
+      "epoch": 79.83367983367984,
+      "eval_loss": 0.5981683731079102,
+      "eval_runtime": 11.7055,
+      "eval_samples_per_second": 73.128,
+      "eval_steps_per_second": 18.282,
+      "num_input_tokens_seen": 30002424,
+      "step": 38400
+    },
+    {
+      "epoch": 79.84407484407484,
+      "grad_norm": 1.4531028682540637e-05,
+      "learning_rate": 1.9614932621215e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30006328,
+      "step": 38405
+    },
+    {
+      "epoch": 79.85446985446985,
+      "grad_norm": 1.054799326993816e-06,
+      "learning_rate": 1.9492384262506102e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30010328,
+      "step": 38410
+    },
+    {
+      "epoch": 79.86486486486487,
+      "grad_norm": 3.205526581950835e-06,
+      "learning_rate": 1.9370218429243524e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30014200,
+      "step": 38415
+    },
+    {
+      "epoch": 79.87525987525987,
+      "grad_norm": 8.330214768648148e-06,
+      "learning_rate": 1.9248435140267197e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30018072,
+      "step": 38420
+    },
+    {
+      "epoch": 79.88565488565489,
+      "grad_norm": 2.3561060515930876e-05,
+      "learning_rate": 1.9127034414356814e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30021944,
+      "step": 38425
+    },
+    {
+      "epoch": 79.8960498960499,
+      "grad_norm": 1.22569883842516e-06,
+      "learning_rate": 1.9006016270234627e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30025912,
+      "step": 38430
+    },
+    {
+      "epoch": 79.9064449064449,
+      "grad_norm": 3.395252861082554e-05,
+      "learning_rate": 1.888538072656293e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30029752,
+      "step": 38435
+    },
+    {
+      "epoch": 79.91683991683992,
+      "grad_norm": 8.233892572206969e-07,
+      "learning_rate": 1.8765127801944893e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30033624,
+      "step": 38440
+    },
+    {
+      "epoch": 79.92723492723492,
+      "grad_norm": 2.128549385815859e-05,
+      "learning_rate": 1.8645257514925406e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30037592,
+      "step": 38445
+    },
+    {
+      "epoch": 79.93762993762994,
+      "grad_norm": 4.812306087842444e-06,
+      "learning_rate": 1.8525769883989685e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30041432,
+      "step": 38450
+    },
+    {
+      "epoch": 79.94802494802495,
+      "grad_norm": 3.0202893412933918e-06,
+      "learning_rate": 1.8406664927564654e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30045240,
+      "step": 38455
+    },
+    {
+      "epoch": 79.95841995841995,
+      "grad_norm": 0.0001072806553565897,
+      "learning_rate": 1.8287942664017566e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30049144,
+      "step": 38460
+    },
+    {
+      "epoch": 79.96881496881497,
+      "grad_norm": 1.0007880518969614e-05,
+      "learning_rate": 1.8169603111656552e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30053048,
+      "step": 38465
+    },
+    {
+      "epoch": 79.97920997920998,
+      "grad_norm": 5.846392468811246e-06,
+      "learning_rate": 1.805164628873146e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30057016,
+      "step": 38470
+    },
+    {
+      "epoch": 79.98960498960498,
+      "grad_norm": 1.1965101293753833e-05,
+      "learning_rate": 1.793407221343274e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30060952,
+      "step": 38475
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 1.2337935004325118e-05,
+      "learning_rate": 1.781688090389172e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30064936,
+      "step": 38480
+    },
+    {
+      "epoch": 80.01039501039502,
+      "grad_norm": 8.06583500434499e-07,
+      "learning_rate": 1.770007237818061e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30068712,
+      "step": 38485
+    },
+    {
+      "epoch": 80.02079002079002,
+      "grad_norm": 2.989679842357873e-06,
+      "learning_rate": 1.7583646654313059e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30072680,
+      "step": 38490
+    },
+    {
+      "epoch": 80.03118503118503,
+      "grad_norm": 1.2753164810419548e-05,
+      "learning_rate": 1.7467603750242757e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30076616,
+      "step": 38495
+    },
+    {
+      "epoch": 80.04158004158005,
+      "grad_norm": 6.263318937271833e-06,
+      "learning_rate": 1.7351943683865944e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30080392,
+      "step": 38500
+    },
+    {
+      "epoch": 80.05197505197505,
+      "grad_norm": 1.0280641618010122e-06,
+      "learning_rate": 1.723666647301808e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30084424,
+      "step": 38505
+    },
+    {
+      "epoch": 80.06237006237006,
+      "grad_norm": 2.978985503432341e-05,
+      "learning_rate": 1.712177213547661e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 30088264,
+      "step": 38510
+    },
+    {
+      "epoch": 80.07276507276508,
+      "grad_norm": 1.8940743302664487e-06,
+      "learning_rate": 1.7007260688959581e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30092136,
+      "step": 38515
+    },
+    {
+      "epoch": 80.08316008316008,
+      "grad_norm": 1.280782998946961e-05,
+      "learning_rate": 1.68931321511262e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30096072,
+      "step": 38520
+    },
+    {
+      "epoch": 80.0935550935551,
+      "grad_norm": 1.955457446456421e-05,
+      "learning_rate": 1.6779386539576835e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30100008,
+      "step": 38525
+    },
+    {
+      "epoch": 80.1039501039501,
+      "grad_norm": 6.917859423083428e-07,
+      "learning_rate": 1.666602387185162e-07,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30103912,
+      "step": 38530
+    },
+    {
+      "epoch": 80.11434511434511,
+      "grad_norm": 8.134697964123916e-07,
+      "learning_rate": 1.655304416543352e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30107752,
+      "step": 38535
+    },
+    {
+      "epoch": 80.12474012474013,
+      "grad_norm": 7.932948733468947e-07,
+      "learning_rate": 1.6440447437744698e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30111688,
+      "step": 38540
+    },
+    {
+      "epoch": 80.13513513513513,
+      "grad_norm": 2.6080435873154784e-06,
+      "learning_rate": 1.6328233706149332e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30115592,
+      "step": 38545
+    },
+    {
+      "epoch": 80.14553014553015,
+      "grad_norm": 1.9319688817631686e-06,
+      "learning_rate": 1.6216402987951906e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30119432,
+      "step": 38550
+    },
+    {
+      "epoch": 80.15592515592516,
+      "grad_norm": 0.0001105776900658384,
+      "learning_rate": 1.6104955300398627e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30123400,
+      "step": 38555
+    },
+    {
+      "epoch": 80.16632016632016,
+      "grad_norm": 0.00012984433851670474,
+      "learning_rate": 1.5993890660675748e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30127240,
+      "step": 38560
+    },
+    {
+      "epoch": 80.17671517671518,
+      "grad_norm": 3.7002823773946147e-06,
+      "learning_rate": 1.5883209085910678e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30131080,
+      "step": 38565
+    },
+    {
+      "epoch": 80.18711018711019,
+      "grad_norm": 8.57883151184069e-07,
+      "learning_rate": 1.5772910593172264e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30134888,
+      "step": 38570
+    },
+    {
+      "epoch": 80.1975051975052,
+      "grad_norm": 6.055178687347507e-07,
+      "learning_rate": 1.5662995199469954e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30138856,
+      "step": 38575
+    },
+    {
+      "epoch": 80.20790020790021,
+      "grad_norm": 1.4506445040751714e-06,
+      "learning_rate": 1.5553462921753802e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30142728,
+      "step": 38580
+    },
+    {
+      "epoch": 80.21829521829522,
+      "grad_norm": 4.647055538953282e-05,
+      "learning_rate": 1.544431377691502e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30146664,
+      "step": 38585
+    },
+    {
+      "epoch": 80.22869022869023,
+      "grad_norm": 8.478428981106845e-07,
+      "learning_rate": 1.5335547781785975e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30150600,
+      "step": 38590
+    },
+    {
+      "epoch": 80.23908523908524,
+      "grad_norm": 1.413708218933607e-06,
+      "learning_rate": 1.5227164953139917e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30154632,
+      "step": 38595
+    },
+    {
+      "epoch": 80.24948024948024,
+      "grad_norm": 0.0004046724643558264,
+      "learning_rate": 1.511916530769042e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30158536,
+      "step": 38600
+    },
+    {
+      "epoch": 80.24948024948024,
+      "eval_loss": 0.6050472259521484,
+      "eval_runtime": 11.7112,
+      "eval_samples_per_second": 73.093,
+      "eval_steps_per_second": 18.273,
+      "num_input_tokens_seen": 30158536,
+      "step": 38600
+    },
+    {
+      "epoch": 80.25987525987526,
+      "grad_norm": 1.2897722626803443e-05,
+      "learning_rate": 1.5011548862092773e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30162568,
+      "step": 38605
+    },
+    {
+      "epoch": 80.27027027027027,
+      "grad_norm": 6.719499651808292e-06,
+      "learning_rate": 1.490431563294231e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30166440,
+      "step": 38610
+    },
+    {
+      "epoch": 80.28066528066527,
+      "grad_norm": 7.070154879329493e-07,
+      "learning_rate": 1.4797465636776365e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30170344,
+      "step": 38615
+    },
+    {
+      "epoch": 80.29106029106029,
+      "grad_norm": 7.827886292943731e-06,
+      "learning_rate": 1.4690998890072027e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30174184,
+      "step": 38620
+    },
+    {
+      "epoch": 80.3014553014553,
+      "grad_norm": 8.204183927773556e-07,
+      "learning_rate": 1.4584915409248112e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30178120,
+      "step": 38625
+    },
+    {
+      "epoch": 80.3118503118503,
+      "grad_norm": 7.366134013864212e-07,
+      "learning_rate": 1.4479215210663754e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30181928,
+      "step": 38630
+    },
+    {
+      "epoch": 80.32224532224532,
+      "grad_norm": 0.1469605267047882,
+      "learning_rate": 1.4373898310619528e-07,
+      "loss": 0.0054,
+      "num_input_tokens_seen": 30185800,
+      "step": 38635
+    },
+    {
+      "epoch": 80.33264033264034,
+      "grad_norm": 2.9233956411189865e-06,
+      "learning_rate": 1.4268964725356604e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30189736,
+      "step": 38640
+    },
+    {
+      "epoch": 80.34303534303534,
+      "grad_norm": 2.7831006264023017e-06,
+      "learning_rate": 1.4164414471056764e-07,
+      "loss": 0.003,
+      "num_input_tokens_seen": 30193736,
+      "step": 38645
+    },
+    {
+      "epoch": 80.35343035343035,
+      "grad_norm": 2.944150992334471e-06,
+      "learning_rate": 1.4060247563843497e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30197672,
+      "step": 38650
+    },
+    {
+      "epoch": 80.36382536382537,
+      "grad_norm": 1.612689629837405e-06,
+      "learning_rate": 1.3956464019780068e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30201512,
+      "step": 38655
+    },
+    {
+      "epoch": 80.37422037422037,
+      "grad_norm": 2.3033315301290713e-05,
+      "learning_rate": 1.385306385487145e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30205384,
+      "step": 38660
+    },
+    {
+      "epoch": 80.38461538461539,
+      "grad_norm": 7.24604205970536e-06,
+      "learning_rate": 1.3750047085063222e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30209352,
+      "step": 38665
+    },
+    {
+      "epoch": 80.39501039501039,
+      "grad_norm": 1.703915768302977e-05,
+      "learning_rate": 1.3647413726242119e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30213192,
+      "step": 38670
+    },
+    {
+      "epoch": 80.4054054054054,
+      "grad_norm": 6.639341449954372e-07,
+      "learning_rate": 1.3545163794235205e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30217032,
+      "step": 38675
+    },
+    {
+      "epoch": 80.41580041580042,
+      "grad_norm": 5.157331202099158e-07,
+      "learning_rate": 1.3443297304810698e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30221000,
+      "step": 38680
+    },
+    {
+      "epoch": 80.42619542619542,
+      "grad_norm": 4.246921889716759e-06,
+      "learning_rate": 1.3341814273677977e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30224904,
+      "step": 38685
+    },
+    {
+      "epoch": 80.43659043659044,
+      "grad_norm": 0.00012435451208148152,
+      "learning_rate": 1.324071471648647e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30228744,
+      "step": 38690
+    },
+    {
+      "epoch": 80.44698544698545,
+      "grad_norm": 1.1926781553484034e-06,
+      "learning_rate": 1.3139998648827312e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30232680,
+      "step": 38695
+    },
+    {
+      "epoch": 80.45738045738045,
+      "grad_norm": 3.4516340292611858e-06,
+      "learning_rate": 1.3039666086232526e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30236552,
+      "step": 38700
+    },
+    {
+      "epoch": 80.46777546777547,
+      "grad_norm": 7.202547567430884e-05,
+      "learning_rate": 1.2939717044174183e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30240392,
+      "step": 38705
+    },
+    {
+      "epoch": 80.47817047817048,
+      "grad_norm": 1.8491014088795055e-06,
+      "learning_rate": 1.284015153806578e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30244264,
+      "step": 38710
+    },
+    {
+      "epoch": 80.48856548856548,
+      "grad_norm": 9.79519754764624e-05,
+      "learning_rate": 1.274096958326171e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30248232,
+      "step": 38715
+    },
+    {
+      "epoch": 80.4989604989605,
+      "grad_norm": 9.133303137787152e-06,
+      "learning_rate": 1.2642171195056952e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30252072,
+      "step": 38720
+    },
+    {
+      "epoch": 80.50935550935552,
+      "grad_norm": 6.585028131667059e-06,
+      "learning_rate": 1.2543756388687377e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30256104,
+      "step": 38725
+    },
+    {
+      "epoch": 80.51975051975052,
+      "grad_norm": 0.1446106880903244,
+      "learning_rate": 1.2445725179330014e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30260104,
+      "step": 38730
+    },
+    {
+      "epoch": 80.53014553014553,
+      "grad_norm": 5.130360659677535e-05,
+      "learning_rate": 1.2348077582102212e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30263848,
+      "step": 38735
+    },
+    {
+      "epoch": 80.54054054054055,
+      "grad_norm": 1.304430270465673e-06,
+      "learning_rate": 1.2250813612062762e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30267784,
+      "step": 38740
+    },
+    {
+      "epoch": 80.55093555093555,
+      "grad_norm": 2.2393221570382593e-06,
+      "learning_rate": 1.215393328421105e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30271688,
+      "step": 38745
+    },
+    {
+      "epoch": 80.56133056133056,
+      "grad_norm": 5.3405010476126336e-06,
+      "learning_rate": 1.2057436613486796e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30275752,
+      "step": 38750
+    },
+    {
+      "epoch": 80.57172557172557,
+      "grad_norm": 7.283856575668324e-06,
+      "learning_rate": 1.1961323614771424e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30279720,
+      "step": 38755
+    },
+    {
+      "epoch": 80.58212058212058,
+      "grad_norm": 6.468887363553222e-07,
+      "learning_rate": 1.1865594302886418e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30283624,
+      "step": 38760
+    },
+    {
+      "epoch": 80.5925155925156,
+      "grad_norm": 1.2409466307872208e-06,
+      "learning_rate": 1.1770248692594687e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30287592,
+      "step": 38765
+    },
+    {
+      "epoch": 80.6029106029106,
+      "grad_norm": 1.950809746631421e-05,
+      "learning_rate": 1.167528679859975e-07,
+      "loss": 0.003,
+      "num_input_tokens_seen": 30291592,
+      "step": 38770
+    },
+    {
+      "epoch": 80.61330561330561,
+      "grad_norm": 0.0008309173863381147,
+      "learning_rate": 1.1580708635545446e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30295464,
+      "step": 38775
+    },
+    {
+      "epoch": 80.62370062370063,
+      "grad_norm": 0.00030750787118449807,
+      "learning_rate": 1.1486514218017885e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30299432,
+      "step": 38780
+    },
+    {
+      "epoch": 80.63409563409563,
+      "grad_norm": 2.607926944619976e-06,
+      "learning_rate": 1.1392703560542117e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30303272,
+      "step": 38785
+    },
+    {
+      "epoch": 80.64449064449065,
+      "grad_norm": 0.00010818737064255401,
+      "learning_rate": 1.129927667758518e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30307144,
+      "step": 38790
+    },
+    {
+      "epoch": 80.65488565488566,
+      "grad_norm": 2.159204768759082e-06,
+      "learning_rate": 1.1206233583554992e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30311080,
+      "step": 38795
+    },
+    {
+      "epoch": 80.66528066528066,
+      "grad_norm": 3.339767999932519e-06,
+      "learning_rate": 1.1113574292799523e-07,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30314984,
+      "step": 38800
+    },
+    {
+      "epoch": 80.66528066528066,
+      "eval_loss": 0.6021023392677307,
+      "eval_runtime": 11.7075,
+      "eval_samples_per_second": 73.116,
+      "eval_steps_per_second": 18.279,
+      "num_input_tokens_seen": 30314984,
+      "step": 38800
+    },
+    {
+      "epoch": 80.67567567567568,
+      "grad_norm": 0.14977623522281647,
+      "learning_rate": 1.1021298819608449e-07,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30318920,
+      "step": 38805
+    },
+    {
+      "epoch": 80.68607068607069,
+      "grad_norm": 4.89414651383413e-06,
+      "learning_rate": 1.0929407178211226e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30322728,
+      "step": 38810
+    },
+    {
+      "epoch": 80.6964656964657,
+      "grad_norm": 1.2771447472914588e-06,
+      "learning_rate": 1.0837899382779293e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30326696,
+      "step": 38815
+    },
+    {
+      "epoch": 80.70686070686071,
+      "grad_norm": 0.0007792308460921049,
+      "learning_rate": 1.0746775447423862e-07,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 30330696,
+      "step": 38820
+    },
+    {
+      "epoch": 80.71725571725571,
+      "grad_norm": 5.458301757244044e-07,
+      "learning_rate": 1.0656035386197583e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30334632,
+      "step": 38825
+    },
+    {
+      "epoch": 80.72765072765073,
+      "grad_norm": 1.1616673873504624e-05,
+      "learning_rate": 1.0565679213093982e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30338600,
+      "step": 38830
+    },
+    {
+      "epoch": 80.73804573804574,
+      "grad_norm": 8.75268369782134e-07,
+      "learning_rate": 1.0475706942046638e-07,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 30342632,
+      "step": 38835
+    },
+    {
+      "epoch": 80.74844074844074,
+      "grad_norm": 8.83374298155104e-07,
+      "learning_rate": 1.0386118586930282e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30346600,
+      "step": 38840
+    },
+    {
+      "epoch": 80.75883575883576,
+      "grad_norm": 1.2134428288845811e-05,
+      "learning_rate": 1.0296914161561367e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30350568,
+      "step": 38845
+    },
+    {
+      "epoch": 80.76923076923077,
+      "grad_norm": 1.1503139830892906e-05,
+      "learning_rate": 1.0208093679695552e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30354536,
+      "step": 38850
+    },
+    {
+      "epoch": 80.77962577962577,
+      "grad_norm": 7.298595301108435e-05,
+      "learning_rate": 1.0119657155030493e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30358376,
+      "step": 38855
+    },
+    {
+      "epoch": 80.79002079002079,
+      "grad_norm": 2.853929800039623e-05,
+      "learning_rate": 1.003160460120417e-07,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30362248,
+      "step": 38860
+    },
+    {
+      "epoch": 80.8004158004158,
+      "grad_norm": 1.0064555681310594e-05,
+      "learning_rate": 9.943936031795165e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30366184,
+      "step": 38865
+    },
+    {
+      "epoch": 80.8108108108108,
+      "grad_norm": 1.9435910871834494e-05,
+      "learning_rate": 9.856651460323219e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30370088,
+      "step": 38870
+    },
+    {
+      "epoch": 80.82120582120582,
+      "grad_norm": 5.9492413129191846e-06,
+      "learning_rate": 9.769750900248953e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30374024,
+      "step": 38875
+    },
+    {
+      "epoch": 80.83160083160084,
+      "grad_norm": 6.621104148507584e-06,
+      "learning_rate": 9.683234364973038e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30377992,
+      "step": 38880
+    },
+    {
+      "epoch": 80.84199584199584,
+      "grad_norm": 1.2271055993551272e-06,
+      "learning_rate": 9.597101867837854e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30381928,
+      "step": 38885
+    },
+    {
+      "epoch": 80.85239085239085,
+      "grad_norm": 2.9927168725407682e-06,
+      "learning_rate": 9.511353422125835e-08,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30385960,
+      "step": 38890
+    },
+    {
+      "epoch": 80.86278586278586,
+      "grad_norm": 2.3559455257782247e-06,
+      "learning_rate": 9.42598904106029e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30389768,
+      "step": 38895
+    },
+    {
+      "epoch": 80.87318087318087,
+      "grad_norm": 9.857658005785197e-05,
+      "learning_rate": 9.341008737806245e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30393672,
+      "step": 38900
+    },
+    {
+      "epoch": 80.88357588357589,
+      "grad_norm": 4.3590205223154044e-07,
+      "learning_rate": 9.256412525467661e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30397640,
+      "step": 38905
+    },
+    {
+      "epoch": 80.89397089397089,
+      "grad_norm": 1.4516093642669148e-06,
+      "learning_rate": 9.172200417091326e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30401704,
+      "step": 38910
+    },
+    {
+      "epoch": 80.9043659043659,
+      "grad_norm": 4.876939328823937e-06,
+      "learning_rate": 9.088372425663239e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30405672,
+      "step": 38915
+    },
+    {
+      "epoch": 80.91476091476092,
+      "grad_norm": 1.2870023056166247e-06,
+      "learning_rate": 9.004928564110837e-08,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30409544,
+      "step": 38920
+    },
+    {
+      "epoch": 80.92515592515592,
+      "grad_norm": 5.169488872525108e-07,
+      "learning_rate": 8.92186884530244e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30413416,
+      "step": 38925
+    },
+    {
+      "epoch": 80.93555093555094,
+      "grad_norm": 1.897879087664478e-06,
+      "learning_rate": 8.83919328204641e-08,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30417224,
+      "step": 38930
+    },
+    {
+      "epoch": 80.94594594594595,
+      "grad_norm": 5.659511771227699e-06,
+      "learning_rate": 8.756901887093105e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30421032,
+      "step": 38935
+    },
+    {
+      "epoch": 80.95634095634095,
+      "grad_norm": 0.00047625447041355073,
+      "learning_rate": 8.674994673132098e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30424936,
+      "step": 38940
+    },
+    {
+      "epoch": 80.96673596673597,
+      "grad_norm": 1.4017257399245864e-06,
+      "learning_rate": 8.593471652794949e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30428744,
+      "step": 38945
+    },
+    {
+      "epoch": 80.97713097713098,
+      "grad_norm": 1.2952809811395127e-05,
+      "learning_rate": 8.512332838653548e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30432616,
+      "step": 38950
+    },
+    {
+      "epoch": 80.98752598752598,
+      "grad_norm": 2.990364737343043e-05,
+      "learning_rate": 8.431578243220106e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30436520,
+      "step": 38955
+    },
+    {
+      "epoch": 80.997920997921,
+      "grad_norm": 5.571748715738067e-07,
+      "learning_rate": 8.351207878948552e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30440392,
+      "step": 38960
+    },
+    {
+      "epoch": 81.00831600831602,
+      "grad_norm": 3.4935765143018216e-06,
+      "learning_rate": 8.271221758232583e-08,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 30444184,
+      "step": 38965
+    },
+    {
+      "epoch": 81.01871101871102,
+      "grad_norm": 2.9801758500980213e-05,
+      "learning_rate": 8.191619893407332e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30448120,
+      "step": 38970
+    },
+    {
+      "epoch": 81.02910602910603,
+      "grad_norm": 1.0605432180454955e-05,
+      "learning_rate": 8.112402296748534e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30451960,
+      "step": 38975
+    },
+    {
+      "epoch": 81.03950103950103,
+      "grad_norm": 2.633328767842613e-05,
+      "learning_rate": 8.033568980471973e-08,
+      "loss": 0.0051,
+      "num_input_tokens_seen": 30455704,
+      "step": 38980
+    },
+    {
+      "epoch": 81.04989604989605,
+      "grad_norm": 1.8801977148541482e-06,
+      "learning_rate": 7.955119956735146e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30459672,
+      "step": 38985
+    },
+    {
+      "epoch": 81.06029106029106,
+      "grad_norm": 1.4381655091710854e-05,
+      "learning_rate": 7.877055237636155e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30463544,
+      "step": 38990
+    },
+    {
+      "epoch": 81.07068607068607,
+      "grad_norm": 3.453871784131479e-07,
+      "learning_rate": 7.79937483521287e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30467384,
+      "step": 38995
+    },
+    {
+      "epoch": 81.08108108108108,
+      "grad_norm": 7.450391422025859e-06,
+      "learning_rate": 7.722078761444873e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30471288,
+      "step": 39000
+    },
+    {
+      "epoch": 81.08108108108108,
+      "eval_loss": 0.6039025187492371,
+      "eval_runtime": 11.7131,
+      "eval_samples_per_second": 73.081,
+      "eval_steps_per_second": 18.27,
+      "num_input_tokens_seen": 30471288,
+      "step": 39000
+    },
+    {
+      "epoch": 81.0914760914761,
+      "grad_norm": 0.15222442150115967,
+      "learning_rate": 7.645167028252631e-08,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30475160,
+      "step": 39005
+    },
+    {
+      "epoch": 81.1018711018711,
+      "grad_norm": 1.2670636806433322e-06,
+      "learning_rate": 7.568639647496379e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30479000,
+      "step": 39010
+    },
+    {
+      "epoch": 81.11226611226611,
+      "grad_norm": 1.5398888990603155e-06,
+      "learning_rate": 7.492496630977508e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30482968,
+      "step": 39015
+    },
+    {
+      "epoch": 81.12266112266113,
+      "grad_norm": 0.00011588069901335984,
+      "learning_rate": 7.416737990438571e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30486904,
+      "step": 39020
+    },
+    {
+      "epoch": 81.13305613305613,
+      "grad_norm": 1.2429518392309546e-05,
+      "learning_rate": 7.341363737562445e-08,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30490712,
+      "step": 39025
+    },
+    {
+      "epoch": 81.14345114345114,
+      "grad_norm": 1.0047781870525796e-05,
+      "learning_rate": 7.266373883972887e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30494552,
+      "step": 39030
+    },
+    {
+      "epoch": 81.15384615384616,
+      "grad_norm": 7.592749966534029e-07,
+      "learning_rate": 7.191768441233981e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30498488,
+      "step": 39035
+    },
+    {
+      "epoch": 81.16424116424116,
+      "grad_norm": 1.358119880023878e-05,
+      "learning_rate": 7.11754742085069e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30502424,
+      "step": 39040
+    },
+    {
+      "epoch": 81.17463617463618,
+      "grad_norm": 2.3448587853636127e-06,
+      "learning_rate": 7.043710834269413e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30506232,
+      "step": 39045
+    },
+    {
+      "epoch": 81.18503118503118,
+      "grad_norm": 5.469105417432729e-06,
+      "learning_rate": 6.970258692876319e-08,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30510328,
+      "step": 39050
+    },
+    {
+      "epoch": 81.1954261954262,
+      "grad_norm": 5.377985416998854e-06,
+      "learning_rate": 6.897191007998738e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30514232,
+      "step": 39055
+    },
+    {
+      "epoch": 81.20582120582121,
+      "grad_norm": 2.825868250511121e-06,
+      "learning_rate": 6.824507790904599e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30518168,
+      "step": 39060
+    },
+    {
+      "epoch": 81.21621621621621,
+      "grad_norm": 1.7695537053441512e-06,
+      "learning_rate": 6.752209052802439e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30522104,
+      "step": 39065
+    },
+    {
+      "epoch": 81.22661122661123,
+      "grad_norm": 1.0574785846984014e-05,
+      "learning_rate": 6.680294804841946e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30525976,
+      "step": 39070
+    },
+    {
+      "epoch": 81.23700623700624,
+      "grad_norm": 4.364229425846133e-06,
+      "learning_rate": 6.608765058112865e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30529784,
+      "step": 39075
+    },
+    {
+      "epoch": 81.24740124740124,
+      "grad_norm": 6.007100523675035e-07,
+      "learning_rate": 6.537619823646368e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30533656,
+      "step": 39080
+    },
+    {
+      "epoch": 81.25779625779626,
+      "grad_norm": 4.2229145037708804e-05,
+      "learning_rate": 6.466859112413404e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30537560,
+      "step": 39085
+    },
+    {
+      "epoch": 81.26819126819127,
+      "grad_norm": 1.3674539331987035e-06,
+      "learning_rate": 6.39648293532663e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30541368,
+      "step": 39090
+    },
+    {
+      "epoch": 81.27858627858627,
+      "grad_norm": 0.00032720097806304693,
+      "learning_rate": 6.32649130323848e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30545400,
+      "step": 39095
+    },
+    {
+      "epoch": 81.28898128898129,
+      "grad_norm": 9.892441994452383e-07,
+      "learning_rate": 6.256884226943094e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30549336,
+      "step": 39100
+    },
+    {
+      "epoch": 81.2993762993763,
+      "grad_norm": 1.1215237236683606e-06,
+      "learning_rate": 6.187661717174386e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30553400,
+      "step": 39105
+    },
+    {
+      "epoch": 81.3097713097713,
+      "grad_norm": 1.9252461243013386e-06,
+      "learning_rate": 6.118823784607708e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30557304,
+      "step": 39110
+    },
+    {
+      "epoch": 81.32016632016632,
+      "grad_norm": 1.1135786053273478e-06,
+      "learning_rate": 6.050370439858178e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30561304,
+      "step": 39115
+    },
+    {
+      "epoch": 81.33056133056132,
+      "grad_norm": 0.149749293923378,
+      "learning_rate": 5.98230169348235e-08,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30565112,
+      "step": 39120
+    },
+    {
+      "epoch": 81.34095634095634,
+      "grad_norm": 1.1730862752301618e-06,
+      "learning_rate": 5.914617555977664e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30569048,
+      "step": 39125
+    },
+    {
+      "epoch": 81.35135135135135,
+      "grad_norm": 4.351987172412919e-06,
+      "learning_rate": 5.8473180377816017e-08,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30572984,
+      "step": 39130
+    },
+    {
+      "epoch": 81.36174636174636,
+      "grad_norm": 2.02769606403308e-05,
+      "learning_rate": 5.780403149272251e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30576952,
+      "step": 39135
+    },
+    {
+      "epoch": 81.37214137214137,
+      "grad_norm": 4.3530704374461493e-07,
+      "learning_rate": 5.7138729007694126e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30581080,
+      "step": 39140
+    },
+    {
+      "epoch": 81.38253638253639,
+      "grad_norm": 7.762875611661002e-06,
+      "learning_rate": 5.64772730253238e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30584984,
+      "step": 39145
+    },
+    {
+      "epoch": 81.39293139293139,
+      "grad_norm": 1.8065800304611912e-06,
+      "learning_rate": 5.5819663647618814e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30588952,
+      "step": 39150
+    },
+    {
+      "epoch": 81.4033264033264,
+      "grad_norm": 1.596814172444283e-06,
+      "learning_rate": 5.5165900975989723e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30592952,
+      "step": 39155
+    },
+    {
+      "epoch": 81.41372141372142,
+      "grad_norm": 4.7719036956550553e-05,
+      "learning_rate": 5.451598511125311e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30596856,
+      "step": 39160
+    },
+    {
+      "epoch": 81.42411642411642,
+      "grad_norm": 4.517587058217032e-06,
+      "learning_rate": 5.3869916153637124e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30600792,
+      "step": 39165
+    },
+    {
+      "epoch": 81.43451143451144,
+      "grad_norm": 1.6577805581619032e-06,
+      "learning_rate": 5.322769420277318e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30604696,
+      "step": 39170
+    },
+    {
+      "epoch": 81.44490644490645,
+      "grad_norm": 5.375314685807098e-06,
+      "learning_rate": 5.258931935769873e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30608568,
+      "step": 39175
+    },
+    {
+      "epoch": 81.45530145530145,
+      "grad_norm": 2.862288965843618e-05,
+      "learning_rate": 5.19547917168628e-08,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30612408,
+      "step": 39180
+    },
+    {
+      "epoch": 81.46569646569647,
+      "grad_norm": 2.760838469839655e-06,
+      "learning_rate": 5.13241113781121e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30616344,
+      "step": 39185
+    },
+    {
+      "epoch": 81.47609147609148,
+      "grad_norm": 1.0015955012931954e-05,
+      "learning_rate": 5.0697278438707755e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30620248,
+      "step": 39190
+    },
+    {
+      "epoch": 81.48648648648648,
+      "grad_norm": 2.151579565179418e-06,
+      "learning_rate": 5.0074292995316854e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30624120,
+      "step": 39195
+    },
+    {
+      "epoch": 81.4968814968815,
+      "grad_norm": 5.3327476052800193e-05,
+      "learning_rate": 4.945515514400978e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30628024,
+      "step": 39200
+    },
+    {
+      "epoch": 81.4968814968815,
+      "eval_loss": 0.6009311676025391,
+      "eval_runtime": 11.7012,
+      "eval_samples_per_second": 73.155,
+      "eval_steps_per_second": 18.289,
+      "num_input_tokens_seen": 30628024,
+      "step": 39200
+    },
+    {
+      "epoch": 81.5072765072765,
+      "grad_norm": 6.0090584156569093e-05,
+      "learning_rate": 4.883986498026571e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30631960,
+      "step": 39205
+    },
+    {
+      "epoch": 81.51767151767152,
+      "grad_norm": 1.0087595001095906e-05,
+      "learning_rate": 4.822842259896987e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30635896,
+      "step": 39210
+    },
+    {
+      "epoch": 81.52806652806653,
+      "grad_norm": 5.033367415308021e-05,
+      "learning_rate": 4.762082809441626e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30639736,
+      "step": 39215
+    },
+    {
+      "epoch": 81.53846153846153,
+      "grad_norm": 3.968385863117874e-05,
+      "learning_rate": 4.7017081560302156e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30643704,
+      "step": 39220
+    },
+    {
+      "epoch": 81.54885654885655,
+      "grad_norm": 2.0034315184602747e-06,
+      "learning_rate": 4.6417183089730866e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30647608,
+      "step": 39225
+    },
+    {
+      "epoch": 81.55925155925156,
+      "grad_norm": 8.480174074065872e-06,
+      "learning_rate": 4.5821132775217265e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30651576,
+      "step": 39230
+    },
+    {
+      "epoch": 81.56964656964657,
+      "grad_norm": 1.091369085770566e-05,
+      "learning_rate": 4.5228930708679504e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30655544,
+      "step": 39235
+    },
+    {
+      "epoch": 81.58004158004158,
+      "grad_norm": 1.6399342257500393e-06,
+      "learning_rate": 4.464057698144175e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30659576,
+      "step": 39240
+    },
+    {
+      "epoch": 81.5904365904366,
+      "grad_norm": 5.657013844029279e-07,
+      "learning_rate": 4.4056071684236974e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30663320,
+      "step": 39245
+    },
+    {
+      "epoch": 81.6008316008316,
+      "grad_norm": 1.7077874872484244e-05,
+      "learning_rate": 4.347541490719864e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30667384,
+      "step": 39250
+    },
+    {
+      "epoch": 81.61122661122661,
+      "grad_norm": 6.342419283100753e-07,
+      "learning_rate": 4.2898606739877336e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30671192,
+      "step": 39255
+    },
+    {
+      "epoch": 81.62162162162163,
+      "grad_norm": 1.376058321511664e-06,
+      "learning_rate": 4.232564727122135e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30675288,
+      "step": 39260
+    },
+    {
+      "epoch": 81.63201663201663,
+      "grad_norm": 0.15935389697551727,
+      "learning_rate": 4.1756536589585004e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30679160,
+      "step": 39265
+    },
+    {
+      "epoch": 81.64241164241164,
+      "grad_norm": 2.550134013290517e-06,
+      "learning_rate": 4.119127478273976e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30682968,
+      "step": 39270
+    },
+    {
+      "epoch": 81.65280665280665,
+      "grad_norm": 0.0004787758516613394,
+      "learning_rate": 4.062986193784923e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30686744,
+      "step": 39275
+    },
+    {
+      "epoch": 81.66320166320166,
+      "grad_norm": 0.00010772552923299372,
+      "learning_rate": 4.007229814149416e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30690744,
+      "step": 39280
+    },
+    {
+      "epoch": 81.67359667359668,
+      "grad_norm": 1.0775796909001656e-05,
+      "learning_rate": 3.951858347965576e-08,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 30694616,
+      "step": 39285
+    },
+    {
+      "epoch": 81.68399168399168,
+      "grad_norm": 1.996414994209772e-06,
+      "learning_rate": 3.896871803772684e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30698552,
+      "step": 39290
+    },
+    {
+      "epoch": 81.6943866943867,
+      "grad_norm": 5.230502893027733e-07,
+      "learning_rate": 3.842270190050068e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30702360,
+      "step": 39295
+    },
+    {
+      "epoch": 81.70478170478171,
+      "grad_norm": 2.356878758291714e-05,
+      "learning_rate": 3.7880535152179376e-08,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 30706168,
+      "step": 39300
+    },
+    {
+      "epoch": 81.71517671517671,
+      "grad_norm": 1.610215235814394e-06,
+      "learning_rate": 3.734221787637382e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30710136,
+      "step": 39305
+    },
+    {
+      "epoch": 81.72557172557173,
+      "grad_norm": 0.15074963867664337,
+      "learning_rate": 3.680775015609817e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30714072,
+      "step": 39310
+    },
+    {
+      "epoch": 81.73596673596674,
+      "grad_norm": 3.1662152650824282e-06,
+      "learning_rate": 3.627713207377537e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30718040,
+      "step": 39315
+    },
+    {
+      "epoch": 81.74636174636174,
+      "grad_norm": 1.099033170248731e-06,
+      "learning_rate": 3.575036371123164e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30722008,
+      "step": 39320
+    },
+    {
+      "epoch": 81.75675675675676,
+      "grad_norm": 4.395206633489579e-05,
+      "learning_rate": 3.5227445149704776e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30725880,
+      "step": 39325
+    },
+    {
+      "epoch": 81.76715176715177,
+      "grad_norm": 1.319241891906131e-05,
+      "learning_rate": 3.470837646983027e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30729976,
+      "step": 39330
+    },
+    {
+      "epoch": 81.77754677754677,
+      "grad_norm": 1.2907058817290817e-06,
+      "learning_rate": 3.419315775165799e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30733784,
+      "step": 39335
+    },
+    {
+      "epoch": 81.78794178794179,
+      "grad_norm": 1.0288921430401388e-06,
+      "learning_rate": 3.368178907464103e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30737880,
+      "step": 39340
+    },
+    {
+      "epoch": 81.7983367983368,
+      "grad_norm": 2.357782705075806e-06,
+      "learning_rate": 3.317427051763855e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30741816,
+      "step": 39345
+    },
+    {
+      "epoch": 81.8087318087318,
+      "grad_norm": 7.264028226927621e-06,
+      "learning_rate": 3.267060215891571e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30745688,
+      "step": 39350
+    },
+    {
+      "epoch": 81.81912681912682,
+      "grad_norm": 1.4151353298075264e-06,
+      "learning_rate": 3.217078407614649e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30749560,
+      "step": 39355
+    },
+    {
+      "epoch": 81.82952182952182,
+      "grad_norm": 2.9373917641351e-05,
+      "learning_rate": 3.1674816346405345e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30753432,
+      "step": 39360
+    },
+    {
+      "epoch": 81.83991683991684,
+      "grad_norm": 2.8725435186061077e-05,
+      "learning_rate": 3.11826990461811e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30757432,
+      "step": 39365
+    },
+    {
+      "epoch": 81.85031185031185,
+      "grad_norm": 1.5109865216800245e-06,
+      "learning_rate": 3.069443225136304e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30761368,
+      "step": 39370
+    },
+    {
+      "epoch": 81.86070686070686,
+      "grad_norm": 0.15520057082176208,
+      "learning_rate": 3.021001603724372e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30765240,
+      "step": 39375
+    },
+    {
+      "epoch": 81.87110187110187,
+      "grad_norm": 1.5215891835396178e-05,
+      "learning_rate": 2.9729450478532818e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30769080,
+      "step": 39380
+    },
+    {
+      "epoch": 81.88149688149689,
+      "grad_norm": 1.689698183326982e-05,
+      "learning_rate": 2.9252735649337726e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30772888,
+      "step": 39385
+    },
+    {
+      "epoch": 81.89189189189189,
+      "grad_norm": 1.1153420018672477e-06,
+      "learning_rate": 2.8779871623171863e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30776824,
+      "step": 39390
+    },
+    {
+      "epoch": 81.9022869022869,
+      "grad_norm": 3.3140822779387236e-05,
+      "learning_rate": 2.8310858472957448e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30780568,
+      "step": 39395
+    },
+    {
+      "epoch": 81.91268191268192,
+      "grad_norm": 8.918235039345745e-07,
+      "learning_rate": 2.784569627101996e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30784376,
+      "step": 39400
+    },
+    {
+      "epoch": 81.91268191268192,
+      "eval_loss": 0.6022453904151917,
+      "eval_runtime": 11.6978,
+      "eval_samples_per_second": 73.176,
+      "eval_steps_per_second": 18.294,
+      "num_input_tokens_seen": 30784376,
+      "step": 39400
+    },
+    {
+      "epoch": 81.92307692307692,
+      "grad_norm": 0.0002228775410912931,
+      "learning_rate": 2.738438508909924e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30788344,
+      "step": 39405
+    },
+    {
+      "epoch": 81.93347193347194,
+      "grad_norm": 0.1560581922531128,
+      "learning_rate": 2.692692499833005e-08,
+      "loss": 0.0053,
+      "num_input_tokens_seen": 30792280,
+      "step": 39410
+    },
+    {
+      "epoch": 81.94386694386695,
+      "grad_norm": 5.7959972764365375e-05,
+      "learning_rate": 2.647331606926151e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30796216,
+      "step": 39415
+    },
+    {
+      "epoch": 81.95426195426195,
+      "grad_norm": 7.907787562544399e-07,
+      "learning_rate": 2.6023558371843225e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30800120,
+      "step": 39420
+    },
+    {
+      "epoch": 81.96465696465697,
+      "grad_norm": 1.6932307289607706e-06,
+      "learning_rate": 2.557765197543638e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30804120,
+      "step": 39425
+    },
+    {
+      "epoch": 81.97505197505197,
+      "grad_norm": 9.531860314382357e-07,
+      "learning_rate": 2.513559694880263e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30807960,
+      "step": 39430
+    },
+    {
+      "epoch": 81.98544698544698,
+      "grad_norm": 2.7602725367614767e-06,
+      "learning_rate": 2.469739336011523e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30811928,
+      "step": 39435
+    },
+    {
+      "epoch": 81.995841995842,
+      "grad_norm": 7.622684847774508e-07,
+      "learning_rate": 2.4263041276947894e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30815896,
+      "step": 39440
+    },
+    {
+      "epoch": 82.006237006237,
+      "grad_norm": 5.97737152929767e-06,
+      "learning_rate": 2.3832540766283164e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30819880,
+      "step": 39445
+    },
+    {
+      "epoch": 82.01663201663202,
+      "grad_norm": 1.1472284313640557e-06,
+      "learning_rate": 2.3405891894512366e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30823656,
+      "step": 39450
+    },
+    {
+      "epoch": 82.02702702702703,
+      "grad_norm": 4.788317255588481e-06,
+      "learning_rate": 2.29830947274301e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30827528,
+      "step": 39455
+    },
+    {
+      "epoch": 82.03742203742203,
+      "grad_norm": 9.414958412889973e-07,
+      "learning_rate": 2.2564149330231432e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30831368,
+      "step": 39460
+    },
+    {
+      "epoch": 82.04781704781705,
+      "grad_norm": 7.081904982442211e-07,
+      "learning_rate": 2.2149055767528572e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30835304,
+      "step": 39465
+    },
+    {
+      "epoch": 82.05821205821206,
+      "grad_norm": 1.9513901861500926e-05,
+      "learning_rate": 2.1737814103334197e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30839208,
+      "step": 39470
+    },
+    {
+      "epoch": 82.06860706860707,
+      "grad_norm": 5.920231342315674e-05,
+      "learning_rate": 2.1330424401064253e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30843176,
+      "step": 39475
+    },
+    {
+      "epoch": 82.07900207900208,
+      "grad_norm": 3.490590461296961e-05,
+      "learning_rate": 2.092688672354348e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30847080,
+      "step": 39480
+    },
+    {
+      "epoch": 82.0893970893971,
+      "grad_norm": 2.6119716494577006e-05,
+      "learning_rate": 2.0527201133005435e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30850984,
+      "step": 39485
+    },
+    {
+      "epoch": 82.0997920997921,
+      "grad_norm": 1.0538921742409002e-05,
+      "learning_rate": 2.0131367691084148e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30854856,
+      "step": 39490
+    },
+    {
+      "epoch": 82.11018711018711,
+      "grad_norm": 1.4291264960775152e-05,
+      "learning_rate": 1.9739386458819675e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30858664,
+      "step": 39495
+    },
+    {
+      "epoch": 82.12058212058211,
+      "grad_norm": 4.79128675578977e-06,
+      "learning_rate": 1.9351257496666442e-08,
+      "loss": 0.0028,
+      "num_input_tokens_seen": 30862504,
+      "step": 39500
+    },
+    {
+      "epoch": 82.13097713097713,
+      "grad_norm": 5.68214545637602e-07,
+      "learning_rate": 1.896698086447657e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30866440,
+      "step": 39505
+    },
+    {
+      "epoch": 82.14137214137214,
+      "grad_norm": 0.15265494585037231,
+      "learning_rate": 1.8586556621505436e-08,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30870440,
+      "step": 39510
+    },
+    {
+      "epoch": 82.15176715176715,
+      "grad_norm": 1.7770853446563706e-06,
+      "learning_rate": 1.820998482642833e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30874440,
+      "step": 39515
+    },
+    {
+      "epoch": 82.16216216216216,
+      "grad_norm": 6.443277129619673e-07,
+      "learning_rate": 1.7837265537309912e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30878312,
+      "step": 39520
+    },
+    {
+      "epoch": 82.17255717255718,
+      "grad_norm": 1.6472853531013243e-05,
+      "learning_rate": 1.7468398811629206e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30882280,
+      "step": 39525
+    },
+    {
+      "epoch": 82.18295218295218,
+      "grad_norm": 7.77343814206688e-07,
+      "learning_rate": 1.710338470627404e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30886056,
+      "step": 39530
+    },
+    {
+      "epoch": 82.1933471933472,
+      "grad_norm": 1.674332088441588e-05,
+      "learning_rate": 1.6742223277529945e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30889832,
+      "step": 39535
+    },
+    {
+      "epoch": 82.20374220374221,
+      "grad_norm": 0.00018147875380236655,
+      "learning_rate": 1.6384914581094036e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30893832,
+      "step": 39540
+    },
+    {
+      "epoch": 82.21413721413721,
+      "grad_norm": 6.7654532358574215e-06,
+      "learning_rate": 1.6031458672069455e-08,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30897800,
+      "step": 39545
+    },
+    {
+      "epoch": 82.22453222453223,
+      "grad_norm": 4.793621428689221e-07,
+      "learning_rate": 1.5681855604962602e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30901768,
+      "step": 39550
+    },
+    {
+      "epoch": 82.23492723492724,
+      "grad_norm": 1.1416691449994687e-05,
+      "learning_rate": 1.5336105433683135e-08,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 30905640,
+      "step": 39555
+    },
+    {
+      "epoch": 82.24532224532224,
+      "grad_norm": 1.2308872101129964e-05,
+      "learning_rate": 1.499420821155506e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30909480,
+      "step": 39560
+    },
+    {
+      "epoch": 82.25571725571726,
+      "grad_norm": 0.00013180787209421396,
+      "learning_rate": 1.4656163991302874e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30913416,
+      "step": 39565
+    },
+    {
+      "epoch": 82.26611226611226,
+      "grad_norm": 1.0547980764386011e-06,
+      "learning_rate": 1.4321972825051544e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30917352,
+      "step": 39570
+    },
+    {
+      "epoch": 82.27650727650727,
+      "grad_norm": 4.478581467992626e-05,
+      "learning_rate": 1.3991634764345951e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30921288,
+      "step": 39575
+    },
+    {
+      "epoch": 82.28690228690229,
+      "grad_norm": 7.0507521741092205e-06,
+      "learning_rate": 1.3665149860120352e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30925448,
+      "step": 39580
+    },
+    {
+      "epoch": 82.29729729729729,
+      "grad_norm": 0.14551712572574615,
+      "learning_rate": 1.3342518162728912e-08,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30929224,
+      "step": 39585
+    },
+    {
+      "epoch": 82.3076923076923,
+      "grad_norm": 3.429363232498872e-06,
+      "learning_rate": 1.30237397219235e-08,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 30933128,
+      "step": 39590
+    },
+    {
+      "epoch": 82.31808731808732,
+      "grad_norm": 3.6791341244679643e-06,
+      "learning_rate": 1.2708814586862016e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30937000,
+      "step": 39595
+    },
+    {
+      "epoch": 82.32848232848232,
+      "grad_norm": 4.969208748661913e-05,
+      "learning_rate": 1.2397742806111168e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30940904,
+      "step": 39600
+    },
+    {
+      "epoch": 82.32848232848232,
+      "eval_loss": 0.6016472578048706,
+      "eval_runtime": 11.6994,
+      "eval_samples_per_second": 73.166,
+      "eval_steps_per_second": 18.292,
+      "num_input_tokens_seen": 30940904,
+      "step": 39600
+    },
+    {
+      "epoch": 82.33887733887734,
+      "grad_norm": 3.502655090414919e-05,
+      "learning_rate": 1.209052442764369e-08,
+      "loss": 0.0025,
+      "num_input_tokens_seen": 30945000,
+      "step": 39605
+    },
+    {
+      "epoch": 82.34927234927235,
+      "grad_norm": 3.345807999721728e-05,
+      "learning_rate": 1.17871594988328e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30948968,
+      "step": 39610
+    },
+    {
+      "epoch": 82.35966735966736,
+      "grad_norm": 1.4693690673084348e-06,
+      "learning_rate": 1.1487648066466072e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30953032,
+      "step": 39615
+    },
+    {
+      "epoch": 82.37006237006237,
+      "grad_norm": 5.3717256378149614e-05,
+      "learning_rate": 1.1191990176728784e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30956872,
+      "step": 39620
+    },
+    {
+      "epoch": 82.38045738045739,
+      "grad_norm": 1.963275281013921e-05,
+      "learning_rate": 1.0900185875215018e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30960712,
+      "step": 39625
+    },
+    {
+      "epoch": 82.39085239085239,
+      "grad_norm": 8.54919653647812e-06,
+      "learning_rate": 1.0612235206924891e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30964584,
+      "step": 39630
+    },
+    {
+      "epoch": 82.4012474012474,
+      "grad_norm": 1.7558743365952978e-06,
+      "learning_rate": 1.0328138216264549e-08,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30968584,
+      "step": 39635
+    },
+    {
+      "epoch": 82.41164241164242,
+      "grad_norm": 3.857411866192706e-05,
+      "learning_rate": 1.004789494704339e-08,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30972392,
+      "step": 39640
+    },
+    {
+      "epoch": 82.42203742203742,
+      "grad_norm": 1.1313686627545394e-05,
+      "learning_rate": 9.771505442482397e-09,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30976168,
+      "step": 39645
+    },
+    {
+      "epoch": 82.43243243243244,
+      "grad_norm": 1.9126291590509936e-06,
+      "learning_rate": 9.498969745200259e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30980136,
+      "step": 39650
+    },
+    {
+      "epoch": 82.44282744282744,
+      "grad_norm": 3.9689680306764785e-06,
+      "learning_rate": 9.230287897230017e-09,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 30984104,
+      "step": 39655
+    },
+    {
+      "epoch": 82.45322245322245,
+      "grad_norm": 7.913761237432482e-07,
+      "learning_rate": 8.965459940002419e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30988072,
+      "step": 39660
+    },
+    {
+      "epoch": 82.46361746361747,
+      "grad_norm": 5.054369012214011e-07,
+      "learning_rate": 8.704485914357019e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30991880,
+      "step": 39665
+    },
+    {
+      "epoch": 82.47401247401247,
+      "grad_norm": 0.1580316573381424,
+      "learning_rate": 8.447365860539402e-09,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 30996008,
+      "step": 39670
+    },
+    {
+      "epoch": 82.48440748440748,
+      "grad_norm": 8.098963917291258e-06,
+      "learning_rate": 8.194099818201184e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 30999944,
+      "step": 39675
+    },
+    {
+      "epoch": 82.4948024948025,
+      "grad_norm": 1.3562731510319281e-05,
+      "learning_rate": 7.944687826400011e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31003784,
+      "step": 39680
+    },
+    {
+      "epoch": 82.5051975051975,
+      "grad_norm": 1.1686231573548866e-06,
+      "learning_rate": 7.699129923599557e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31007720,
+      "step": 39685
+    },
+    {
+      "epoch": 82.51559251559252,
+      "grad_norm": 7.85016879945033e-07,
+      "learning_rate": 7.457426147663982e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31011592,
+      "step": 39690
+    },
+    {
+      "epoch": 82.52598752598753,
+      "grad_norm": 1.3822638720739633e-05,
+      "learning_rate": 7.219576535871797e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31015432,
+      "step": 39695
+    },
+    {
+      "epoch": 82.53638253638253,
+      "grad_norm": 1.5927885215205606e-06,
+      "learning_rate": 6.985581124896445e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31019176,
+      "step": 39700
+    },
+    {
+      "epoch": 82.54677754677755,
+      "grad_norm": 6.926368314452702e-07,
+      "learning_rate": 6.755439950828501e-09,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 31023144,
+      "step": 39705
+    },
+    {
+      "epoch": 82.55717255717256,
+      "grad_norm": 1.0217391718470026e-05,
+      "learning_rate": 6.5291530491562444e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31027112,
+      "step": 39710
+    },
+    {
+      "epoch": 82.56756756756756,
+      "grad_norm": 1.799012807168765e-06,
+      "learning_rate": 6.3067204547739845e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31031080,
+      "step": 39715
+    },
+    {
+      "epoch": 82.57796257796258,
+      "grad_norm": 2.2455858925241046e-05,
+      "learning_rate": 6.088142201987612e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31035016,
+      "step": 39720
+    },
+    {
+      "epoch": 82.58835758835758,
+      "grad_norm": 5.836528544023167e-06,
+      "learning_rate": 5.873418324503499e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31038984,
+      "step": 39725
+    },
+    {
+      "epoch": 82.5987525987526,
+      "grad_norm": 9.658360795583576e-06,
+      "learning_rate": 5.6625488554340465e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31042952,
+      "step": 39730
+    },
+    {
+      "epoch": 82.60914760914761,
+      "grad_norm": 5.513653604793944e-07,
+      "learning_rate": 5.455533827297688e-09,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 31046696,
+      "step": 39735
+    },
+    {
+      "epoch": 82.61954261954261,
+      "grad_norm": 1.6254894944722764e-05,
+      "learning_rate": 5.252373272018885e-09,
+      "loss": 0.0023,
+      "num_input_tokens_seen": 31050632,
+      "step": 39740
+    },
+    {
+      "epoch": 82.62993762993763,
+      "grad_norm": 1.3412011867330875e-05,
+      "learning_rate": 5.053067220925356e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31054568,
+      "step": 39745
+    },
+    {
+      "epoch": 82.64033264033264,
+      "grad_norm": 1.7408436860932852e-06,
+      "learning_rate": 4.857615704759177e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31058408,
+      "step": 39750
+    },
+    {
+      "epoch": 82.65072765072765,
+      "grad_norm": 2.5952995201805606e-06,
+      "learning_rate": 4.666018753654577e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31062376,
+      "step": 39755
+    },
+    {
+      "epoch": 82.66112266112266,
+      "grad_norm": 7.290113899216522e-07,
+      "learning_rate": 4.478276397162917e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31066248,
+      "step": 39760
+    },
+    {
+      "epoch": 82.67151767151768,
+      "grad_norm": 7.213802177830075e-07,
+      "learning_rate": 4.294388664233262e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31070056,
+      "step": 39765
+    },
+    {
+      "epoch": 82.68191268191268,
+      "grad_norm": 1.276097373192897e-05,
+      "learning_rate": 4.114355583223484e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31073928,
+      "step": 39770
+    },
+    {
+      "epoch": 82.6923076923077,
+      "grad_norm": 1.5402654298668494e-06,
+      "learning_rate": 3.9381771818974845e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31077896,
+      "step": 39775
+    },
+    {
+      "epoch": 82.70270270270271,
+      "grad_norm": 2.2048867322155274e-05,
+      "learning_rate": 3.765853487427973e-09,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 31081928,
+      "step": 39780
+    },
+    {
+      "epoch": 82.71309771309771,
+      "grad_norm": 6.352942364173941e-06,
+      "learning_rate": 3.5973845263825857e-09,
+      "loss": 0.0027,
+      "num_input_tokens_seen": 31085832,
+      "step": 39785
+    },
+    {
+      "epoch": 82.72349272349273,
+      "grad_norm": 2.2280046323430724e-05,
+      "learning_rate": 3.4327703247488684e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31089608,
+      "step": 39790
+    },
+    {
+      "epoch": 82.73388773388774,
+      "grad_norm": 1.992006900763954e-06,
+      "learning_rate": 3.2720109079037443e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31093384,
+      "step": 39795
+    },
+    {
+      "epoch": 82.74428274428274,
+      "grad_norm": 1.1714492757164408e-05,
+      "learning_rate": 3.1151063006468193e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31097352,
+      "step": 39800
+    },
+    {
+      "epoch": 82.74428274428274,
+      "eval_loss": 0.6060176491737366,
+      "eval_runtime": 11.7058,
+      "eval_samples_per_second": 73.126,
+      "eval_steps_per_second": 18.282,
+      "num_input_tokens_seen": 31097352,
+      "step": 39800
+    },
+    {
+      "epoch": 82.75467775467776,
+      "grad_norm": 8.198737191378314e-07,
+      "learning_rate": 2.962056527169854e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31101288,
+      "step": 39805
+    },
+    {
+      "epoch": 82.76507276507276,
+      "grad_norm": 1.3304334970598575e-05,
+      "learning_rate": 2.8128616110761898e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31105096,
+      "step": 39810
+    },
+    {
+      "epoch": 82.77546777546777,
+      "grad_norm": 1.0097631957250996e-06,
+      "learning_rate": 2.6675215753724223e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31108872,
+      "step": 39815
+    },
+    {
+      "epoch": 82.78586278586279,
+      "grad_norm": 6.134525278866931e-07,
+      "learning_rate": 2.5260364424739557e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31112744,
+      "step": 39820
+    },
+    {
+      "epoch": 82.79625779625779,
+      "grad_norm": 2.469371975166723e-05,
+      "learning_rate": 2.3884062341994475e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31116712,
+      "step": 39825
+    },
+    {
+      "epoch": 82.8066528066528,
+      "grad_norm": 1.8095100813297904e-06,
+      "learning_rate": 2.25463097177081e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31120648,
+      "step": 39830
+    },
+    {
+      "epoch": 82.81704781704782,
+      "grad_norm": 0.15067800879478455,
+      "learning_rate": 2.1247106758215397e-09,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 31124648,
+      "step": 39835
+    },
+    {
+      "epoch": 82.82744282744282,
+      "grad_norm": 8.454032922600163e-07,
+      "learning_rate": 1.998645366382834e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31128584,
+      "step": 39840
+    },
+    {
+      "epoch": 82.83783783783784,
+      "grad_norm": 2.3524620701209642e-05,
+      "learning_rate": 1.876435062897475e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31132680,
+      "step": 39845
+    },
+    {
+      "epoch": 82.84823284823285,
+      "grad_norm": 5.918165015827981e-07,
+      "learning_rate": 1.758079784211497e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31136584,
+      "step": 39850
+    },
+    {
+      "epoch": 82.85862785862786,
+      "grad_norm": 1.4155189091979992e-06,
+      "learning_rate": 1.6435795485797434e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31140392,
+      "step": 39855
+    },
+    {
+      "epoch": 82.86902286902287,
+      "grad_norm": 2.6398849968245486e-06,
+      "learning_rate": 1.5329343736547596e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31144360,
+      "step": 39860
+    },
+    {
+      "epoch": 82.87941787941789,
+      "grad_norm": 1.3240689440863207e-05,
+      "learning_rate": 1.4261442765006739e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31148264,
+      "step": 39865
+    },
+    {
+      "epoch": 82.88981288981289,
+      "grad_norm": 3.700486786328838e-06,
+      "learning_rate": 1.3232092735876445e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31152008,
+      "step": 39870
+    },
+    {
+      "epoch": 82.9002079002079,
+      "grad_norm": 3.2825235393829644e-05,
+      "learning_rate": 1.2241293807918607e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31155816,
+      "step": 39875
+    },
+    {
+      "epoch": 82.9106029106029,
+      "grad_norm": 4.002375590062002e-06,
+      "learning_rate": 1.128904613387216e-09,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31159592,
+      "step": 39880
+    },
+    {
+      "epoch": 82.92099792099792,
+      "grad_norm": 1.9379006062081316e-06,
+      "learning_rate": 1.0375349860591853e-09,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 31163528,
+      "step": 39885
+    },
+    {
+      "epoch": 82.93139293139293,
+      "grad_norm": 3.446963319220231e-06,
+      "learning_rate": 9.5002051290205e-10,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 31167464,
+      "step": 39890
+    },
+    {
+      "epoch": 82.94178794178794,
+      "grad_norm": 1.1582612387428526e-05,
+      "learning_rate": 8.663612074077954e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31171496,
+      "step": 39895
+    },
+    {
+      "epoch": 82.95218295218295,
+      "grad_norm": 6.761306394764688e-06,
+      "learning_rate": 7.865570824799884e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31175272,
+      "step": 39900
+    },
+    {
+      "epoch": 82.96257796257797,
+      "grad_norm": 1.5687528502894565e-06,
+      "learning_rate": 7.106081504254514e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31178952,
+      "step": 39905
+    },
+    {
+      "epoch": 82.97297297297297,
+      "grad_norm": 6.119397539805504e-07,
+      "learning_rate": 6.385144229570372e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31182952,
+      "step": 39910
+    },
+    {
+      "epoch": 82.98336798336798,
+      "grad_norm": 2.229292476840783e-06,
+      "learning_rate": 5.70275911190854e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31186888,
+      "step": 39915
+    },
+    {
+      "epoch": 82.993762993763,
+      "grad_norm": 1.4044094314158428e-05,
+      "learning_rate": 5.058926256490403e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31190792,
+      "step": 39920
+    },
+    {
+      "epoch": 83.004158004158,
+      "grad_norm": 4.643121428671293e-06,
+      "learning_rate": 4.4536457626254134e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31194744,
+      "step": 39925
+    },
+    {
+      "epoch": 83.01455301455302,
+      "grad_norm": 3.0122751923045143e-05,
+      "learning_rate": 3.88691772365557e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31198648,
+      "step": 39930
+    },
+    {
+      "epoch": 83.02494802494803,
+      "grad_norm": 7.729611752438359e-06,
+      "learning_rate": 3.358742226955425e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31202584,
+      "step": 39935
+    },
+    {
+      "epoch": 83.03534303534303,
+      "grad_norm": 2.2235115011426387e-06,
+      "learning_rate": 2.8691193539875925e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31206360,
+      "step": 39940
+    },
+    {
+      "epoch": 83.04573804573805,
+      "grad_norm": 5.4961888963589445e-05,
+      "learning_rate": 2.418049180274995e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31210296,
+      "step": 39945
+    },
+    {
+      "epoch": 83.05613305613305,
+      "grad_norm": 1.1087545317423064e-06,
+      "learning_rate": 2.005531775373104e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31214136,
+      "step": 39950
+    },
+    {
+      "epoch": 83.06652806652806,
+      "grad_norm": 2.7739042707253247e-06,
+      "learning_rate": 1.6315672028699435e-10,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31217912,
+      "step": 39955
+    },
+    {
+      "epoch": 83.07692307692308,
+      "grad_norm": 6.471647679973103e-07,
+      "learning_rate": 1.2961555204693555e-10,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 31221848,
+      "step": 39960
+    },
+    {
+      "epoch": 83.08731808731808,
+      "grad_norm": 2.02356659428915e-06,
+      "learning_rate": 9.992967798799768e-11,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31225688,
+      "step": 39965
+    },
+    {
+      "epoch": 83.0977130977131,
+      "grad_norm": 2.3073159809428034e-06,
+      "learning_rate": 7.409910268707521e-11,
+      "loss": 0.0024,
+      "num_input_tokens_seen": 31229688,
+      "step": 39970
+    },
+    {
+      "epoch": 83.10810810810811,
+      "grad_norm": 9.209286872646771e-06,
+      "learning_rate": 5.212383012986877e-11,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31233624,
+      "step": 39975
+    },
+    {
+      "epoch": 83.11850311850311,
+      "grad_norm": 6.353739991027396e-06,
+      "learning_rate": 3.400386370533415e-11,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31237624,
+      "step": 39980
+    },
+    {
+      "epoch": 83.12889812889813,
+      "grad_norm": 0.00015363757847808301,
+      "learning_rate": 1.9739206205682258e-11,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31241560,
+      "step": 39985
+    },
+    {
+      "epoch": 83.13929313929314,
+      "grad_norm": 2.1993571408529533e-06,
+      "learning_rate": 9.329859829154685e-12,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31245464,
+      "step": 39990
+    },
+    {
+      "epoch": 83.14968814968815,
+      "grad_norm": 0.146912083029747,
+      "learning_rate": 2.7758261855748148e-12,
+      "loss": 0.0026,
+      "num_input_tokens_seen": 31249336,
+      "step": 39995
+    },
+    {
+      "epoch": 83.16008316008316,
+      "grad_norm": 1.299550945077499e-06,
+      "learning_rate": 7.710628524559838e-14,
+      "loss": 0.0,
+      "num_input_tokens_seen": 31253176,
+      "step": 40000
+    },
+    {
+      "epoch": 83.16008316008316,
+      "eval_loss": 0.6030600666999817,
+      "eval_runtime": 11.7384,
+      "eval_samples_per_second": 72.923,
+      "eval_steps_per_second": 18.231,
+      "num_input_tokens_seen": 31253176,
+      "step": 40000
+    },
+    {
+      "epoch": 83.16008316008316,
+      "num_input_tokens_seen": 31253176,
+      "step": 40000,
+      "total_flos": 1.3171753735637606e+17,
+      "train_loss": 0.011159961750394992,
+      "train_runtime": 36067.8227,
+      "train_samples_per_second": 17.744,
+      "train_steps_per_second": 1.109
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 40000,
+  "num_input_tokens_seen": 31253176,
+  "num_train_epochs": 84,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3171753735637606e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}