diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,16637 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999939429547479,
+  "eval_steps": 2000,
+  "global_step": 20637,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00048456362017480634,
+      "grad_norm": 1.9635685682296753,
+      "learning_rate": 8.064516129032258e-07,
+      "loss": 3.9647,
+      "num_input_tokens_seen": 396856,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009691272403496127,
+      "grad_norm": 1.8676133155822754,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 3.8773,
+      "num_input_tokens_seen": 797088,
+      "step": 20
+    },
+    {
+      "epoch": 0.001453690860524419,
+      "grad_norm": 1.3513129949569702,
+      "learning_rate": 2.4193548387096776e-06,
+      "loss": 3.7214,
+      "num_input_tokens_seen": 1195348,
+      "step": 30
+    },
+    {
+      "epoch": 0.0019382544806992254,
+      "grad_norm": 1.085821270942688,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 3.608,
+      "num_input_tokens_seen": 1563616,
+      "step": 40
+    },
+    {
+      "epoch": 0.0024228181008740315,
+      "grad_norm": 0.7733827233314514,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 3.4798,
+      "num_input_tokens_seen": 1953880,
+      "step": 50
+    },
+    {
+      "epoch": 0.002907381721048838,
+      "grad_norm": 0.7254723310470581,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 3.3992,
+      "num_input_tokens_seen": 2358888,
+      "step": 60
+    },
+    {
+      "epoch": 0.0033919453412236443,
+      "grad_norm": 0.692788302898407,
+      "learning_rate": 5.64516129032258e-06,
+      "loss": 3.287,
+      "num_input_tokens_seen": 2784136,
+      "step": 70
+    },
+    {
+      "epoch": 0.0038765089613984507,
+      "grad_norm": 0.7256464958190918,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 3.1694,
+      "num_input_tokens_seen": 3174820,
+      "step": 80
+    },
+    {
+      "epoch": 0.004361072581573257,
+      "grad_norm": 0.6375479102134705,
+      "learning_rate": 7.258064516129033e-06,
+      "loss": 3.093,
+      "num_input_tokens_seen": 3579040,
+      "step": 90
+    },
+    {
+      "epoch": 0.004845636201748063,
+      "grad_norm": 0.5901396870613098,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 2.9749,
+      "num_input_tokens_seen": 3959804,
+      "step": 100
+    },
+    {
+      "epoch": 0.00533019982192287,
+      "grad_norm": 0.5945073366165161,
+      "learning_rate": 8.870967741935484e-06,
+      "loss": 2.9903,
+      "num_input_tokens_seen": 4327636,
+      "step": 110
+    },
+    {
+      "epoch": 0.005814763442097676,
+      "grad_norm": 0.5525237917900085,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 2.9102,
+      "num_input_tokens_seen": 4722664,
+      "step": 120
+    },
+    {
+      "epoch": 0.006299327062272482,
+      "grad_norm": 0.5421098470687866,
+      "learning_rate": 1.0483870967741936e-05,
+      "loss": 2.8672,
+      "num_input_tokens_seen": 5108488,
+      "step": 130
+    },
+    {
+      "epoch": 0.006783890682447289,
+      "grad_norm": 0.6057586669921875,
+      "learning_rate": 1.129032258064516e-05,
+      "loss": 2.7504,
+      "num_input_tokens_seen": 5499032,
+      "step": 140
+    },
+    {
+      "epoch": 0.007268454302622095,
+      "grad_norm": 0.5147073864936829,
+      "learning_rate": 1.2096774193548388e-05,
+      "loss": 2.728,
+      "num_input_tokens_seen": 5899468,
+      "step": 150
+    },
+    {
+      "epoch": 0.0077530179227969015,
+      "grad_norm": 0.4975648820400238,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 2.635,
+      "num_input_tokens_seen": 6294784,
+      "step": 160
+    },
+    {
+      "epoch": 0.008237581542971708,
+      "grad_norm": 0.6665644645690918,
+      "learning_rate": 1.3709677419354839e-05,
+      "loss": 2.6568,
+      "num_input_tokens_seen": 6724124,
+      "step": 170
+    },
+    {
+      "epoch": 0.008722145163146514,
+      "grad_norm": 0.5339853763580322,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 2.6569,
+      "num_input_tokens_seen": 7126528,
+      "step": 180
+    },
+    {
+      "epoch": 0.00920670878332132,
+      "grad_norm": 0.6043190956115723,
+      "learning_rate": 1.5322580645161292e-05,
+      "loss": 2.5659,
+      "num_input_tokens_seen": 7539124,
+      "step": 190
+    },
+    {
+      "epoch": 0.009691272403496126,
+      "grad_norm": 0.576831579208374,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 2.6224,
+      "num_input_tokens_seen": 7943312,
+      "step": 200
+    },
+    {
+      "epoch": 0.010175836023670932,
+      "grad_norm": 0.5602587461471558,
+      "learning_rate": 1.693548387096774e-05,
+      "loss": 2.5489,
+      "num_input_tokens_seen": 8345264,
+      "step": 210
+    },
+    {
+      "epoch": 0.01066039964384574,
+      "grad_norm": 0.5199883580207825,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 2.4895,
+      "num_input_tokens_seen": 8724596,
+      "step": 220
+    },
+    {
+      "epoch": 0.011144963264020546,
+      "grad_norm": 0.5386412143707275,
+      "learning_rate": 1.8548387096774193e-05,
+      "loss": 2.4376,
+      "num_input_tokens_seen": 9128556,
+      "step": 230
+    },
+    {
+      "epoch": 0.011629526884195352,
+      "grad_norm": 0.4958534836769104,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 2.4393,
+      "num_input_tokens_seen": 9531512,
+      "step": 240
+    },
+    {
+      "epoch": 0.012114090504370158,
+      "grad_norm": 0.5383809208869934,
+      "learning_rate": 2.0161290322580645e-05,
+      "loss": 2.408,
+      "num_input_tokens_seen": 9926976,
+      "step": 250
+    },
+    {
+      "epoch": 0.012598654124544964,
+      "grad_norm": 0.5953584313392639,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 2.4448,
+      "num_input_tokens_seen": 10307532,
+      "step": 260
+    },
+    {
+      "epoch": 0.013083217744719771,
+      "grad_norm": 0.6360442638397217,
+      "learning_rate": 2.1774193548387097e-05,
+      "loss": 2.4015,
+      "num_input_tokens_seen": 10680264,
+      "step": 270
+    },
+    {
+      "epoch": 0.013567781364894577,
+      "grad_norm": 0.48746633529663086,
+      "learning_rate": 2.258064516129032e-05,
+      "loss": 2.3641,
+      "num_input_tokens_seen": 11070464,
+      "step": 280
+    },
+    {
+      "epoch": 0.014052344985069383,
+      "grad_norm": 0.6602309346199036,
+      "learning_rate": 2.338709677419355e-05,
+      "loss": 2.3722,
+      "num_input_tokens_seen": 11470192,
+      "step": 290
+    },
+    {
+      "epoch": 0.01453690860524419,
+      "grad_norm": 0.5370129346847534,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 2.3605,
+      "num_input_tokens_seen": 11872288,
+      "step": 300
+    },
+    {
+      "epoch": 0.015021472225418997,
+      "grad_norm": 0.575202226638794,
+      "learning_rate": 2.5e-05,
+      "loss": 2.3164,
+      "num_input_tokens_seen": 12282104,
+      "step": 310
+    },
+    {
+      "epoch": 0.015506035845593803,
+      "grad_norm": 0.5220383405685425,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 2.3395,
+      "num_input_tokens_seen": 12670432,
+      "step": 320
+    },
+    {
+      "epoch": 0.01599059946576861,
+      "grad_norm": 0.6971769332885742,
+      "learning_rate": 2.661290322580645e-05,
+      "loss": 2.3035,
+      "num_input_tokens_seen": 13048952,
+      "step": 330
+    },
+    {
+      "epoch": 0.016475163085943417,
+      "grad_norm": 0.5094983577728271,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 2.2668,
+      "num_input_tokens_seen": 13454104,
+      "step": 340
+    },
+    {
+      "epoch": 0.01695972670611822,
+      "grad_norm": 0.5272497534751892,
+      "learning_rate": 2.822580645161291e-05,
+      "loss": 2.2398,
+      "num_input_tokens_seen": 13846768,
+      "step": 350
+    },
+    {
+      "epoch": 0.01744429032629303,
+      "grad_norm": 0.5047647356987,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 2.2096,
+      "num_input_tokens_seen": 14247376,
+      "step": 360
+    },
+    {
+      "epoch": 0.017928853946467833,
+      "grad_norm": 0.5520374178886414,
+      "learning_rate": 2.9838709677419357e-05,
+      "loss": 2.2371,
+      "num_input_tokens_seen": 14644760,
+      "step": 370
+    },
+    {
+      "epoch": 0.01841341756664264,
+      "grad_norm": 0.46873587369918823,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 2.2636,
+      "num_input_tokens_seen": 15030968,
+      "step": 380
+    },
+    {
+      "epoch": 0.018897981186817448,
+      "grad_norm": 0.5534300208091736,
+      "learning_rate": 3.1451612903225806e-05,
+      "loss": 2.1924,
+      "num_input_tokens_seen": 15442684,
+      "step": 390
+    },
+    {
+      "epoch": 0.019382544806992252,
+      "grad_norm": 0.5355961918830872,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 2.1482,
+      "num_input_tokens_seen": 15823228,
+      "step": 400
+    },
+    {
+      "epoch": 0.01986710842716706,
+      "grad_norm": 0.5530371069908142,
+      "learning_rate": 3.306451612903226e-05,
+      "loss": 2.1865,
+      "num_input_tokens_seen": 16206660,
+      "step": 410
+    },
+    {
+      "epoch": 0.020351672047341864,
+      "grad_norm": 0.5505437254905701,
+      "learning_rate": 3.387096774193548e-05,
+      "loss": 2.1665,
+      "num_input_tokens_seen": 16610308,
+      "step": 420
+    },
+    {
+      "epoch": 0.020836235667516672,
+      "grad_norm": 0.5833243727684021,
+      "learning_rate": 3.467741935483872e-05,
+      "loss": 2.1413,
+      "num_input_tokens_seen": 17013200,
+      "step": 430
+    },
+    {
+      "epoch": 0.02132079928769148,
+      "grad_norm": 0.5723456144332886,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 2.149,
+      "num_input_tokens_seen": 17412528,
+      "step": 440
+    },
+    {
+      "epoch": 0.021805362907866284,
+      "grad_norm": 0.49083006381988525,
+      "learning_rate": 3.6290322580645165e-05,
+      "loss": 2.208,
+      "num_input_tokens_seen": 17818648,
+      "step": 450
+    },
+    {
+      "epoch": 0.02228992652804109,
+      "grad_norm": 0.5125194787979126,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 2.1628,
+      "num_input_tokens_seen": 18215112,
+      "step": 460
+    },
+    {
+      "epoch": 0.022774490148215896,
+      "grad_norm": 0.5249300003051758,
+      "learning_rate": 3.7903225806451614e-05,
+      "loss": 2.1179,
+      "num_input_tokens_seen": 18613100,
+      "step": 470
+    },
+    {
+      "epoch": 0.023259053768390704,
+      "grad_norm": 0.5913705825805664,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 2.0918,
+      "num_input_tokens_seen": 19006580,
+      "step": 480
+    },
+    {
+      "epoch": 0.02374361738856551,
+      "grad_norm": 0.5256056785583496,
+      "learning_rate": 3.951612903225806e-05,
+      "loss": 2.1146,
+      "num_input_tokens_seen": 19410508,
+      "step": 490
+    },
+    {
+      "epoch": 0.024228181008740315,
+      "grad_norm": 0.5684739947319031,
+      "learning_rate": 4.032258064516129e-05,
+      "loss": 2.0621,
+      "num_input_tokens_seen": 19801804,
+      "step": 500
+    },
+    {
+      "epoch": 0.024712744628915123,
+      "grad_norm": 0.5903922319412231,
+      "learning_rate": 4.112903225806452e-05,
+      "loss": 2.1035,
+      "num_input_tokens_seen": 20183328,
+      "step": 510
+    },
+    {
+      "epoch": 0.025197308249089927,
+      "grad_norm": 0.5174550414085388,
+      "learning_rate": 4.1935483870967746e-05,
+      "loss": 2.088,
+      "num_input_tokens_seen": 20585364,
+      "step": 520
+    },
+    {
+      "epoch": 0.025681871869264735,
+      "grad_norm": 0.5224788188934326,
+      "learning_rate": 4.2741935483870973e-05,
+      "loss": 2.0708,
+      "num_input_tokens_seen": 20971292,
+      "step": 530
+    },
+    {
+      "epoch": 0.026166435489439543,
+      "grad_norm": 0.5536626577377319,
+      "learning_rate": 4.3548387096774194e-05,
+      "loss": 2.0135,
+      "num_input_tokens_seen": 21390084,
+      "step": 540
+    },
+    {
+      "epoch": 0.026650999109614347,
+      "grad_norm": 0.49406698346138,
+      "learning_rate": 4.435483870967742e-05,
+      "loss": 2.0626,
+      "num_input_tokens_seen": 21790964,
+      "step": 550
+    },
+    {
+      "epoch": 0.027135562729789155,
+      "grad_norm": 0.5543055534362793,
+      "learning_rate": 4.516129032258064e-05,
+      "loss": 2.0395,
+      "num_input_tokens_seen": 22160120,
+      "step": 560
+    },
+    {
+      "epoch": 0.02762012634996396,
+      "grad_norm": 0.5112178325653076,
+      "learning_rate": 4.596774193548387e-05,
+      "loss": 2.0956,
+      "num_input_tokens_seen": 22555328,
+      "step": 570
+    },
+    {
+      "epoch": 0.028104689970138767,
+      "grad_norm": 0.5143571496009827,
+      "learning_rate": 4.67741935483871e-05,
+      "loss": 2.0209,
+      "num_input_tokens_seen": 22941364,
+      "step": 580
+    },
+    {
+      "epoch": 0.028589253590313574,
+      "grad_norm": 0.501890242099762,
+      "learning_rate": 4.7580645161290326e-05,
+      "loss": 2.0326,
+      "num_input_tokens_seen": 23298096,
+      "step": 590
+    },
+    {
+      "epoch": 0.02907381721048838,
+      "grad_norm": 0.47222423553466797,
+      "learning_rate": 4.8387096774193554e-05,
+      "loss": 2.0099,
+      "num_input_tokens_seen": 23698488,
+      "step": 600
+    },
+    {
+      "epoch": 0.029558380830663186,
+      "grad_norm": 0.5555810928344727,
+      "learning_rate": 4.9193548387096775e-05,
+      "loss": 1.9978,
+      "num_input_tokens_seen": 24088604,
+      "step": 610
+    },
+    {
+      "epoch": 0.030042944450837994,
+      "grad_norm": 0.6044710278511047,
+      "learning_rate": 5e-05,
+      "loss": 2.0005,
+      "num_input_tokens_seen": 24477980,
+      "step": 620
+    },
+    {
+      "epoch": 0.030527508071012798,
+      "grad_norm": 0.5720553398132324,
+      "learning_rate": 4.999996920985807e-05,
+      "loss": 1.9575,
+      "num_input_tokens_seen": 24859124,
+      "step": 630
+    },
+    {
+      "epoch": 0.031012071691187606,
+      "grad_norm": 0.5833857655525208,
+      "learning_rate": 4.9999876839508106e-05,
+      "loss": 1.992,
+      "num_input_tokens_seen": 25283928,
+      "step": 640
+    },
+    {
+      "epoch": 0.03149663531136241,
+      "grad_norm": 0.49288150668144226,
+      "learning_rate": 4.999972288917764e-05,
+      "loss": 1.952,
+      "num_input_tokens_seen": 25662068,
+      "step": 650
+    },
+    {
+      "epoch": 0.03198119893153722,
+      "grad_norm": 0.5453637838363647,
+      "learning_rate": 4.999950735924589e-05,
+      "loss": 2.0213,
+      "num_input_tokens_seen": 26066280,
+      "step": 660
+    },
+    {
+      "epoch": 0.032465762551712025,
+      "grad_norm": 0.4919157922267914,
+      "learning_rate": 4.9999230250243744e-05,
+      "loss": 1.9531,
+      "num_input_tokens_seen": 26446308,
+      "step": 670
+    },
+    {
+      "epoch": 0.03295032617188683,
+      "grad_norm": 0.5228652954101562,
+      "learning_rate": 4.999889156285379e-05,
+      "loss": 1.9746,
+      "num_input_tokens_seen": 26862780,
+      "step": 680
+    },
+    {
+      "epoch": 0.033434889792061634,
+      "grad_norm": 0.5261794328689575,
+      "learning_rate": 4.999849129791028e-05,
+      "loss": 1.9717,
+      "num_input_tokens_seen": 27239428,
+      "step": 690
+    },
+    {
+      "epoch": 0.03391945341223644,
+      "grad_norm": 0.5125803351402283,
+      "learning_rate": 4.9998029456399144e-05,
+      "loss": 1.975,
+      "num_input_tokens_seen": 27633548,
+      "step": 700
+    },
+    {
+      "epoch": 0.03440401703241125,
+      "grad_norm": 0.48046156764030457,
+      "learning_rate": 4.999750603945801e-05,
+      "loss": 1.925,
+      "num_input_tokens_seen": 28046196,
+      "step": 710
+    },
+    {
+      "epoch": 0.03488858065258606,
+      "grad_norm": 0.5283321738243103,
+      "learning_rate": 4.999692104837615e-05,
+      "loss": 1.9442,
+      "num_input_tokens_seen": 28423160,
+      "step": 720
+    },
+    {
+      "epoch": 0.035373144272760865,
+      "grad_norm": 0.4812624156475067,
+      "learning_rate": 4.999627448459453e-05,
+      "loss": 1.8945,
+      "num_input_tokens_seen": 28842052,
+      "step": 730
+    },
+    {
+      "epoch": 0.035857707892935665,
+      "grad_norm": 0.5417799353599548,
+      "learning_rate": 4.999556634970578e-05,
+      "loss": 1.9628,
+      "num_input_tokens_seen": 29251476,
+      "step": 740
+    },
+    {
+      "epoch": 0.03634227151311047,
+      "grad_norm": 0.48694953322410583,
+      "learning_rate": 4.999479664545417e-05,
+      "loss": 1.9061,
+      "num_input_tokens_seen": 29660076,
+      "step": 750
+    },
+    {
+      "epoch": 0.03682683513328528,
+      "grad_norm": 0.5140628814697266,
+      "learning_rate": 4.999396537373565e-05,
+      "loss": 1.9304,
+      "num_input_tokens_seen": 30049624,
+      "step": 760
+    },
+    {
+      "epoch": 0.03731139875346009,
+      "grad_norm": 0.48269200325012207,
+      "learning_rate": 4.9993072536597816e-05,
+      "loss": 1.9188,
+      "num_input_tokens_seen": 30466320,
+      "step": 770
+    },
+    {
+      "epoch": 0.037795962373634896,
+      "grad_norm": 0.5299521684646606,
+      "learning_rate": 4.999211813623993e-05,
+      "loss": 1.8832,
+      "num_input_tokens_seen": 30853272,
+      "step": 780
+    },
+    {
+      "epoch": 0.0382805259938097,
+      "grad_norm": 0.5212485790252686,
+      "learning_rate": 4.999110217501286e-05,
+      "loss": 1.8968,
+      "num_input_tokens_seen": 31245900,
+      "step": 790
+    },
+    {
+      "epoch": 0.038765089613984505,
+      "grad_norm": 0.5231190323829651,
+      "learning_rate": 4.9990024655419146e-05,
+      "loss": 1.8859,
+      "num_input_tokens_seen": 31662092,
+      "step": 800
+    },
+    {
+      "epoch": 0.03924965323415931,
+      "grad_norm": 0.49723413586616516,
+      "learning_rate": 4.998888558011295e-05,
+      "loss": 1.872,
+      "num_input_tokens_seen": 32059116,
+      "step": 810
+    },
+    {
+      "epoch": 0.03973421685433412,
+      "grad_norm": 0.6407498121261597,
+      "learning_rate": 4.9987684951900036e-05,
+      "loss": 1.8606,
+      "num_input_tokens_seen": 32484640,
+      "step": 820
+    },
+    {
+      "epoch": 0.04021878047450893,
+      "grad_norm": 0.5182517766952515,
+      "learning_rate": 4.998642277373783e-05,
+      "loss": 1.8906,
+      "num_input_tokens_seen": 32857036,
+      "step": 830
+    },
+    {
+      "epoch": 0.04070334409468373,
+      "grad_norm": 0.46163928508758545,
+      "learning_rate": 4.998509904873533e-05,
+      "loss": 1.8566,
+      "num_input_tokens_seen": 33227492,
+      "step": 840
+    },
+    {
+      "epoch": 0.041187907714858536,
+      "grad_norm": 0.47485044598579407,
+      "learning_rate": 4.998371378015314e-05,
+      "loss": 1.8759,
+      "num_input_tokens_seen": 33614644,
+      "step": 850
+    },
+    {
+      "epoch": 0.041672471335033344,
+      "grad_norm": 0.4748258888721466,
+      "learning_rate": 4.998226697140349e-05,
+      "loss": 1.877,
+      "num_input_tokens_seen": 33984924,
+      "step": 860
+    },
+    {
+      "epoch": 0.04215703495520815,
+      "grad_norm": 0.5473403930664062,
+      "learning_rate": 4.998075862605017e-05,
+      "loss": 1.8403,
+      "num_input_tokens_seen": 34379932,
+      "step": 870
+    },
+    {
+      "epoch": 0.04264159857538296,
+      "grad_norm": 0.4143201410770416,
+      "learning_rate": 4.9979188747808545e-05,
+      "loss": 1.87,
+      "num_input_tokens_seen": 34769220,
+      "step": 880
+    },
+    {
+      "epoch": 0.04312616219555776,
+      "grad_norm": 0.43778666853904724,
+      "learning_rate": 4.997755734054557e-05,
+      "loss": 1.9247,
+      "num_input_tokens_seen": 35124728,
+      "step": 890
+    },
+    {
+      "epoch": 0.04361072581573257,
+      "grad_norm": 0.532357394695282,
+      "learning_rate": 4.9975864408279725e-05,
+      "loss": 1.8527,
+      "num_input_tokens_seen": 35526992,
+      "step": 900
+    },
+    {
+      "epoch": 0.044095289435907375,
+      "grad_norm": 0.4756541848182678,
+      "learning_rate": 4.997410995518108e-05,
+      "loss": 1.8328,
+      "num_input_tokens_seen": 35932440,
+      "step": 910
+    },
+    {
+      "epoch": 0.04457985305608218,
+      "grad_norm": 0.4680721163749695,
+      "learning_rate": 4.997229398557122e-05,
+      "loss": 1.804,
+      "num_input_tokens_seen": 36352128,
+      "step": 920
+    },
+    {
+      "epoch": 0.04506441667625699,
+      "grad_norm": 0.48043662309646606,
+      "learning_rate": 4.9970416503923254e-05,
+      "loss": 1.8609,
+      "num_input_tokens_seen": 36746228,
+      "step": 930
+    },
+    {
+      "epoch": 0.04554898029643179,
+      "grad_norm": 0.4995608627796173,
+      "learning_rate": 4.996847751486182e-05,
+      "loss": 1.8493,
+      "num_input_tokens_seen": 37138932,
+      "step": 940
+    },
+    {
+      "epoch": 0.0460335439166066,
+      "grad_norm": 0.5137046575546265,
+      "learning_rate": 4.996647702316306e-05,
+      "loss": 1.8672,
+      "num_input_tokens_seen": 37519424,
+      "step": 950
+    },
+    {
+      "epoch": 0.04651810753678141,
+      "grad_norm": 0.48803040385246277,
+      "learning_rate": 4.996441503375461e-05,
+      "loss": 1.8371,
+      "num_input_tokens_seen": 37895416,
+      "step": 960
+    },
+    {
+      "epoch": 0.047002671156956215,
+      "grad_norm": 0.4492471516132355,
+      "learning_rate": 4.996229155171558e-05,
+      "loss": 1.8697,
+      "num_input_tokens_seen": 38307728,
+      "step": 970
+    },
+    {
+      "epoch": 0.04748723477713102,
+      "grad_norm": 0.4708845913410187,
+      "learning_rate": 4.9960106582276556e-05,
+      "loss": 1.8481,
+      "num_input_tokens_seen": 38697636,
+      "step": 980
+    },
+    {
+      "epoch": 0.04797179839730582,
+      "grad_norm": 0.49647071957588196,
+      "learning_rate": 4.995786013081958e-05,
+      "loss": 1.8812,
+      "num_input_tokens_seen": 39088844,
+      "step": 990
+    },
+    {
+      "epoch": 0.04845636201748063,
+      "grad_norm": 0.5095282793045044,
+      "learning_rate": 4.995555220287814e-05,
+      "loss": 1.8549,
+      "num_input_tokens_seen": 39495924,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04894092563765544,
+      "grad_norm": 0.47867298126220703,
+      "learning_rate": 4.9953182804137144e-05,
+      "loss": 1.783,
+      "num_input_tokens_seen": 39909712,
+      "step": 1010
+    },
+    {
+      "epoch": 0.049425489257830246,
+      "grad_norm": 0.4481337070465088,
+      "learning_rate": 4.9950751940432935e-05,
+      "loss": 1.7918,
+      "num_input_tokens_seen": 40301008,
+      "step": 1020
+    },
+    {
+      "epoch": 0.049910052878005054,
+      "grad_norm": 0.4755343496799469,
+      "learning_rate": 4.994825961775323e-05,
+      "loss": 1.7714,
+      "num_input_tokens_seen": 40676476,
+      "step": 1030
+    },
+    {
+      "epoch": 0.050394616498179855,
+      "grad_norm": 0.5189250707626343,
+      "learning_rate": 4.994570584223715e-05,
+      "loss": 1.8269,
+      "num_input_tokens_seen": 41064096,
+      "step": 1040
+    },
+    {
+      "epoch": 0.05087918011835466,
+      "grad_norm": 0.47439032793045044,
+      "learning_rate": 4.994309062017519e-05,
+      "loss": 1.7933,
+      "num_input_tokens_seen": 41445416,
+      "step": 1050
+    },
+    {
+      "epoch": 0.05136374373852947,
+      "grad_norm": 0.4340980052947998,
+      "learning_rate": 4.994041395800918e-05,
+      "loss": 1.7959,
+      "num_input_tokens_seen": 41850336,
+      "step": 1060
+    },
+    {
+      "epoch": 0.05184830735870428,
+      "grad_norm": 0.4619131088256836,
+      "learning_rate": 4.993767586233232e-05,
+      "loss": 1.836,
+      "num_input_tokens_seen": 42265248,
+      "step": 1070
+    },
+    {
+      "epoch": 0.052332870978879086,
+      "grad_norm": 0.5889285206794739,
+      "learning_rate": 4.993487633988912e-05,
+      "loss": 1.8257,
+      "num_input_tokens_seen": 42676764,
+      "step": 1080
+    },
+    {
+      "epoch": 0.052817434599053886,
+      "grad_norm": 0.4897553026676178,
+      "learning_rate": 4.993201539757538e-05,
+      "loss": 1.8588,
+      "num_input_tokens_seen": 43053880,
+      "step": 1090
+    },
+    {
+      "epoch": 0.053301998219228694,
+      "grad_norm": 0.47212594747543335,
+      "learning_rate": 4.992909304243822e-05,
+      "loss": 1.8094,
+      "num_input_tokens_seen": 43430872,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0537865618394035,
+      "grad_norm": 0.44554567337036133,
+      "learning_rate": 4.992610928167601e-05,
+      "loss": 1.7834,
+      "num_input_tokens_seen": 43851752,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05427112545957831,
+      "grad_norm": 0.45216086506843567,
+      "learning_rate": 4.992306412263839e-05,
+      "loss": 1.8522,
+      "num_input_tokens_seen": 44254964,
+      "step": 1120
+    },
+    {
+      "epoch": 0.05475568907975312,
+      "grad_norm": 0.5445901155471802,
+      "learning_rate": 4.9919957572826216e-05,
+      "loss": 1.839,
+      "num_input_tokens_seen": 44656772,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05524025269992792,
+      "grad_norm": 0.5541059970855713,
+      "learning_rate": 4.9916789639891595e-05,
+      "loss": 1.7898,
+      "num_input_tokens_seen": 45058836,
+      "step": 1140
+    },
+    {
+      "epoch": 0.055724816320102726,
+      "grad_norm": 0.5577613711357117,
+      "learning_rate": 4.99135603316378e-05,
+      "loss": 1.8336,
+      "num_input_tokens_seen": 45471592,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05620937994027753,
+      "grad_norm": 0.47430625557899475,
+      "learning_rate": 4.991026965601932e-05,
+      "loss": 1.7995,
+      "num_input_tokens_seen": 45846124,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05669394356045234,
+      "grad_norm": 0.5153654217720032,
+      "learning_rate": 4.990691762114176e-05,
+      "loss": 1.8068,
+      "num_input_tokens_seen": 46213352,
+      "step": 1170
+    },
+    {
+      "epoch": 0.05717850718062715,
+      "grad_norm": 0.5673589110374451,
+      "learning_rate": 4.99035042352619e-05,
+      "loss": 1.773,
+      "num_input_tokens_seen": 46626636,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05766307080080195,
+      "grad_norm": 0.505126416683197,
+      "learning_rate": 4.9900029506787645e-05,
+      "loss": 1.7516,
+      "num_input_tokens_seen": 47031172,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05814763442097676,
+      "grad_norm": 0.4553806483745575,
+      "learning_rate": 4.989649344427796e-05,
+      "loss": 1.7862,
+      "num_input_tokens_seen": 47425060,
+      "step": 1200
+    },
+    {
+      "epoch": 0.058632198041151565,
+      "grad_norm": 0.4654844105243683,
+      "learning_rate": 4.989289605644294e-05,
+      "loss": 1.7957,
+      "num_input_tokens_seen": 47824468,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05911676166132637,
+      "grad_norm": 0.4679258167743683,
+      "learning_rate": 4.988923735214369e-05,
+      "loss": 1.7706,
+      "num_input_tokens_seen": 48220624,
+      "step": 1220
+    },
+    {
+      "epoch": 0.05960132528150118,
+      "grad_norm": 0.4447259306907654,
+      "learning_rate": 4.988551734039239e-05,
+      "loss": 1.7788,
+      "num_input_tokens_seen": 48597080,
+      "step": 1230
+    },
+    {
+      "epoch": 0.06008588890167599,
+      "grad_norm": 0.5004059672355652,
+      "learning_rate": 4.98817360303522e-05,
+      "loss": 1.7994,
+      "num_input_tokens_seen": 49002120,
+      "step": 1240
+    },
+    {
+      "epoch": 0.06057045252185079,
+      "grad_norm": 0.4713003635406494,
+      "learning_rate": 4.98778934313373e-05,
+      "loss": 1.729,
+      "num_input_tokens_seen": 49399124,
+      "step": 1250
+    },
+    {
+      "epoch": 0.061055016142025596,
+      "grad_norm": 0.4393056035041809,
+      "learning_rate": 4.987398955281281e-05,
+      "loss": 1.7764,
+      "num_input_tokens_seen": 49820492,
+      "step": 1260
+    },
+    {
+      "epoch": 0.061539579762200404,
+      "grad_norm": 0.49267756938934326,
+      "learning_rate": 4.987002440439481e-05,
+      "loss": 1.7541,
+      "num_input_tokens_seen": 50240940,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06202414338237521,
+      "grad_norm": 0.4592071771621704,
+      "learning_rate": 4.986599799585031e-05,
+      "loss": 1.8162,
+      "num_input_tokens_seen": 50644136,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06250870700255001,
+      "grad_norm": 0.4428696930408478,
+      "learning_rate": 4.98619103370972e-05,
+      "loss": 1.7786,
+      "num_input_tokens_seen": 51030808,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06299327062272482,
+      "grad_norm": 0.4989221394062042,
+      "learning_rate": 4.985776143820423e-05,
+      "loss": 1.7498,
+      "num_input_tokens_seen": 51408072,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06347783424289963,
+      "grad_norm": 0.46731674671173096,
+      "learning_rate": 4.985355130939104e-05,
+      "loss": 1.7556,
+      "num_input_tokens_seen": 51788056,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06396239786307444,
+      "grad_norm": 0.4602036476135254,
+      "learning_rate": 4.984927996102806e-05,
+      "loss": 1.7395,
+      "num_input_tokens_seen": 52195616,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06444696148324924,
+      "grad_norm": 0.49800294637680054,
+      "learning_rate": 4.984494740363651e-05,
+      "loss": 1.7324,
+      "num_input_tokens_seen": 52605440,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06493152510342405,
+      "grad_norm": 0.4798908233642578,
+      "learning_rate": 4.984055364788842e-05,
+      "loss": 1.7495,
+      "num_input_tokens_seen": 53005000,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06541608872359886,
+      "grad_norm": 0.510181188583374,
+      "learning_rate": 4.9836098704606515e-05,
+      "loss": 1.7238,
+      "num_input_tokens_seen": 53428392,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06590065234377367,
+      "grad_norm": 0.4547858238220215,
+      "learning_rate": 4.983158258476427e-05,
+      "loss": 1.7711,
+      "num_input_tokens_seen": 53787676,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06638521596394846,
+      "grad_norm": 0.486717164516449,
+      "learning_rate": 4.982700529948585e-05,
+      "loss": 1.7099,
+      "num_input_tokens_seen": 54228120,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06686977958412327,
+      "grad_norm": 0.4875461459159851,
+      "learning_rate": 4.982236686004606e-05,
+      "loss": 1.7542,
+      "num_input_tokens_seen": 54637396,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06735434320429808,
+      "grad_norm": 0.4732173681259155,
+      "learning_rate": 4.9817667277870384e-05,
+      "loss": 1.725,
+      "num_input_tokens_seen": 55042360,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06783890682447288,
+      "grad_norm": 0.4369388818740845,
+      "learning_rate": 4.981290656453486e-05,
+      "loss": 1.7288,
+      "num_input_tokens_seen": 55430008,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06832347044464769,
+      "grad_norm": 0.42465418577194214,
+      "learning_rate": 4.9808084731766134e-05,
+      "loss": 1.7026,
+      "num_input_tokens_seen": 55815764,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0688080340648225,
+      "grad_norm": 0.4628385007381439,
+      "learning_rate": 4.980320179144141e-05,
+      "loss": 1.7732,
+      "num_input_tokens_seen": 56164316,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0692925976849973,
+      "grad_norm": 0.5061572194099426,
+      "learning_rate": 4.97982577555884e-05,
+      "loss": 1.7107,
+      "num_input_tokens_seen": 56523812,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06977716130517211,
+      "grad_norm": 0.49572449922561646,
+      "learning_rate": 4.9793252636385305e-05,
+      "loss": 1.6988,
+      "num_input_tokens_seen": 56892024,
+      "step": 1440
+    },
+    {
+      "epoch": 0.07026172492534692,
+      "grad_norm": 0.5096593499183655,
+      "learning_rate": 4.9788186446160795e-05,
+      "loss": 1.7888,
+      "num_input_tokens_seen": 57270864,
+      "step": 1450
+    },
+    {
+      "epoch": 0.07074628854552173,
+      "grad_norm": 0.4694225490093231,
+      "learning_rate": 4.978305919739396e-05,
+      "loss": 1.7126,
+      "num_input_tokens_seen": 57643264,
+      "step": 1460
+    },
+    {
+      "epoch": 0.07123085216569652,
+      "grad_norm": 0.43490421772003174,
+      "learning_rate": 4.9777870902714306e-05,
+      "loss": 1.6902,
+      "num_input_tokens_seen": 58035896,
+      "step": 1470
+    },
+    {
+      "epoch": 0.07171541578587133,
+      "grad_norm": 0.42695242166519165,
+      "learning_rate": 4.97726215749017e-05,
+      "loss": 1.7368,
+      "num_input_tokens_seen": 58412320,
+      "step": 1480
+    },
+    {
+      "epoch": 0.07219997940604614,
+      "grad_norm": 0.48507386445999146,
+      "learning_rate": 4.976731122688634e-05,
+      "loss": 1.6676,
+      "num_input_tokens_seen": 58792164,
+      "step": 1490
+    },
+    {
+      "epoch": 0.07268454302622095,
+      "grad_norm": 0.5080244541168213,
+      "learning_rate": 4.9761939871748734e-05,
+      "loss": 1.6863,
+      "num_input_tokens_seen": 59187108,
+      "step": 1500
+    },
+    {
+      "epoch": 0.07316910664639575,
+      "grad_norm": 0.4605392813682556,
+      "learning_rate": 4.975650752271967e-05,
+      "loss": 1.7126,
+      "num_input_tokens_seen": 59576276,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07365367026657056,
+      "grad_norm": 0.5134192705154419,
+      "learning_rate": 4.9751014193180165e-05,
+      "loss": 1.7108,
+      "num_input_tokens_seen": 59993792,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07413823388674537,
+      "grad_norm": 0.4663729667663574,
+      "learning_rate": 4.974545989666147e-05,
+      "loss": 1.7352,
+      "num_input_tokens_seen": 60408024,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07462279750692018,
+      "grad_norm": 0.4669811427593231,
+      "learning_rate": 4.973984464684497e-05,
+      "loss": 1.7421,
+      "num_input_tokens_seen": 60767824,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07510736112709498,
+      "grad_norm": 0.46583324670791626,
+      "learning_rate": 4.973416845756221e-05,
+      "loss": 1.7187,
+      "num_input_tokens_seen": 61143088,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07559192474726979,
+      "grad_norm": 0.4906235635280609,
+      "learning_rate": 4.9728431342794865e-05,
+      "loss": 1.7195,
+      "num_input_tokens_seen": 61531644,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07607648836744459,
+      "grad_norm": 0.4702226221561432,
+      "learning_rate": 4.972263331667465e-05,
+      "loss": 1.6975,
+      "num_input_tokens_seen": 61937832,
+      "step": 1570
+    },
+    {
+      "epoch": 0.0765610519876194,
+      "grad_norm": 0.4586687386035919,
+      "learning_rate": 4.971677439348332e-05,
+      "loss": 1.7196,
+      "num_input_tokens_seen": 62332144,
+      "step": 1580
+    },
+    {
+      "epoch": 0.0770456156077942,
+      "grad_norm": 0.4383968710899353,
+      "learning_rate": 4.9710854587652654e-05,
+      "loss": 1.6704,
+      "num_input_tokens_seen": 62703152,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07753017922796901,
+      "grad_norm": 0.4787232577800751,
+      "learning_rate": 4.970487391376438e-05,
+      "loss": 1.7729,
+      "num_input_tokens_seen": 63098040,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07801474284814382,
+      "grad_norm": 0.47267040610313416,
+      "learning_rate": 4.969883238655017e-05,
+      "loss": 1.6948,
+      "num_input_tokens_seen": 63500456,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07849930646831862,
+      "grad_norm": 0.4243592917919159,
+      "learning_rate": 4.969273002089157e-05,
+      "loss": 1.6758,
+      "num_input_tokens_seen": 63899700,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07898387008849343,
+      "grad_norm": 0.5044109225273132,
+      "learning_rate": 4.968656683181999e-05,
+      "loss": 1.7079,
+      "num_input_tokens_seen": 64275644,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07946843370866824,
+      "grad_norm": 0.4761982858181,
+      "learning_rate": 4.968034283451669e-05,
+      "loss": 1.6974,
+      "num_input_tokens_seen": 64658944,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07995299732884305,
+      "grad_norm": 0.46890366077423096,
+      "learning_rate": 4.967405804431267e-05,
+      "loss": 1.6835,
+      "num_input_tokens_seen": 65053324,
+      "step": 1650
+    },
+    {
+      "epoch": 0.08043756094901786,
+      "grad_norm": 0.4691985547542572,
+      "learning_rate": 4.966771247668871e-05,
+      "loss": 1.6842,
+      "num_input_tokens_seen": 65471828,
+      "step": 1660
+    },
+    {
+      "epoch": 0.08092212456919266,
+      "grad_norm": 0.5264722108840942,
+      "learning_rate": 4.966130614727529e-05,
+      "loss": 1.6913,
+      "num_input_tokens_seen": 65888112,
+      "step": 1670
+    },
+    {
+      "epoch": 0.08140668818936746,
+      "grad_norm": 0.46564558148384094,
+      "learning_rate": 4.9654839071852535e-05,
+      "loss": 1.6894,
+      "num_input_tokens_seen": 66301484,
+      "step": 1680
+    },
+    {
+      "epoch": 0.08189125180954226,
+      "grad_norm": 0.4879053831100464,
+      "learning_rate": 4.964831126635022e-05,
+      "loss": 1.7088,
+      "num_input_tokens_seen": 66683244,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08237581542971707,
+      "grad_norm": 0.39720138907432556,
+      "learning_rate": 4.964172274684772e-05,
+      "loss": 1.7227,
+      "num_input_tokens_seen": 67102092,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08286037904989188,
+      "grad_norm": 0.4953685998916626,
+      "learning_rate": 4.963507352957395e-05,
+      "loss": 1.7142,
+      "num_input_tokens_seen": 67495072,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08334494267006669,
+      "grad_norm": 0.4297001361846924,
+      "learning_rate": 4.962836363090734e-05,
+      "loss": 1.6934,
+      "num_input_tokens_seen": 67872712,
+      "step": 1720
+    },
+    {
+      "epoch": 0.0838295062902415,
+      "grad_norm": 0.4302520155906677,
+      "learning_rate": 4.962159306737578e-05,
+      "loss": 1.6898,
+      "num_input_tokens_seen": 68264844,
+      "step": 1730
+    },
+    {
+      "epoch": 0.0843140699104163,
+      "grad_norm": 0.47304287552833557,
+      "learning_rate": 4.96147618556566e-05,
+      "loss": 1.6631,
+      "num_input_tokens_seen": 68650516,
+      "step": 1740
+    },
+    {
+      "epoch": 0.08479863353059111,
+      "grad_norm": 0.47196635603904724,
+      "learning_rate": 4.960787001257652e-05,
+      "loss": 1.6963,
+      "num_input_tokens_seen": 69003456,
+      "step": 1750
+    },
+    {
+      "epoch": 0.08528319715076592,
+      "grad_norm": 0.46627262234687805,
+      "learning_rate": 4.9600917555111615e-05,
+      "loss": 1.6733,
+      "num_input_tokens_seen": 69386088,
+      "step": 1760
+    },
+    {
+      "epoch": 0.08576776077094073,
+      "grad_norm": 0.4163726568222046,
+      "learning_rate": 4.9593904500387245e-05,
+      "loss": 1.6907,
+      "num_input_tokens_seen": 69762188,
+      "step": 1770
+    },
+    {
+      "epoch": 0.08625232439111552,
+      "grad_norm": 0.46610623598098755,
+      "learning_rate": 4.9586830865678046e-05,
+      "loss": 1.6722,
+      "num_input_tokens_seen": 70170884,
+      "step": 1780
+    },
+    {
+      "epoch": 0.08673688801129033,
+      "grad_norm": 0.5110335946083069,
+      "learning_rate": 4.957969666840788e-05,
+      "loss": 1.6884,
+      "num_input_tokens_seen": 70544880,
+      "step": 1790
+    },
+    {
+      "epoch": 0.08722145163146514,
+      "grad_norm": 0.4962434768676758,
+      "learning_rate": 4.957250192614979e-05,
+      "loss": 1.659,
+      "num_input_tokens_seen": 70937756,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08770601525163994,
+      "grad_norm": 0.4480324387550354,
+      "learning_rate": 4.956524665662593e-05,
+      "loss": 1.6428,
+      "num_input_tokens_seen": 71295700,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08819057887181475,
+      "grad_norm": 0.46165645122528076,
+      "learning_rate": 4.955793087770758e-05,
+      "loss": 1.7055,
+      "num_input_tokens_seen": 71677740,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08867514249198956,
+      "grad_norm": 0.4461020827293396,
+      "learning_rate": 4.955055460741503e-05,
+      "loss": 1.7052,
+      "num_input_tokens_seen": 72088472,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08915970611216437,
+      "grad_norm": 0.46780285239219666,
+      "learning_rate": 4.9543117863917624e-05,
+      "loss": 1.6684,
+      "num_input_tokens_seen": 72483916,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08964426973233917,
+      "grad_norm": 0.4738537073135376,
+      "learning_rate": 4.953562066553359e-05,
+      "loss": 1.6663,
+      "num_input_tokens_seen": 72895204,
+      "step": 1850
+    },
+    {
+      "epoch": 0.09012883335251398,
+      "grad_norm": 0.41888901591300964,
+      "learning_rate": 4.952806303073015e-05,
+      "loss": 1.7064,
+      "num_input_tokens_seen": 73289656,
+      "step": 1860
+    },
+    {
+      "epoch": 0.09061339697268879,
+      "grad_norm": 0.4899497330188751,
+      "learning_rate": 4.952044497812334e-05,
+      "loss": 1.7179,
+      "num_input_tokens_seen": 73724044,
+      "step": 1870
+    },
+    {
+      "epoch": 0.09109796059286358,
+      "grad_norm": 0.4161849319934845,
+      "learning_rate": 4.951276652647803e-05,
+      "loss": 1.6841,
+      "num_input_tokens_seen": 74111024,
+      "step": 1880
+    },
+    {
+      "epoch": 0.09158252421303839,
+      "grad_norm": 0.44479620456695557,
+      "learning_rate": 4.950502769470787e-05,
+      "loss": 1.6294,
+      "num_input_tokens_seen": 74502300,
+      "step": 1890
+    },
+    {
+      "epoch": 0.0920670878332132,
+      "grad_norm": 0.48210737109184265,
+      "learning_rate": 4.949722850187525e-05,
+      "loss": 1.6615,
+      "num_input_tokens_seen": 74920152,
+      "step": 1900
+    },
+    {
+      "epoch": 0.092551651453388,
+      "grad_norm": 0.5125983357429504,
+      "learning_rate": 4.948936896719121e-05,
+      "loss": 1.6056,
+      "num_input_tokens_seen": 75318632,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09303621507356281,
+      "grad_norm": 0.45514819025993347,
+      "learning_rate": 4.948144911001546e-05,
+      "loss": 1.6808,
+      "num_input_tokens_seen": 75682924,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09352077869373762,
+      "grad_norm": 0.45765024423599243,
+      "learning_rate": 4.9473468949856295e-05,
+      "loss": 1.6256,
+      "num_input_tokens_seen": 76084200,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09400534231391243,
+      "grad_norm": 0.46034979820251465,
+      "learning_rate": 4.946542850637051e-05,
+      "loss": 1.6342,
+      "num_input_tokens_seen": 76509976,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09448990593408724,
+      "grad_norm": 0.47426116466522217,
+      "learning_rate": 4.945732779936343e-05,
+      "loss": 1.6783,
+      "num_input_tokens_seen": 76925836,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09497446955426204,
+      "grad_norm": 0.44276344776153564,
+      "learning_rate": 4.944916684878881e-05,
+      "loss": 1.6505,
+      "num_input_tokens_seen": 77309476,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09545903317443685,
+      "grad_norm": 0.4252452552318573,
+      "learning_rate": 4.944094567474878e-05,
+      "loss": 1.6174,
+      "num_input_tokens_seen": 77719060,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09594359679461165,
+      "grad_norm": 0.4776414930820465,
+      "learning_rate": 4.9432664297493855e-05,
+      "loss": 1.6953,
+      "num_input_tokens_seen": 78109656,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09642816041478645,
+      "grad_norm": 0.5169130563735962,
+      "learning_rate": 4.94243227374228e-05,
+      "loss": 1.6782,
+      "num_input_tokens_seen": 78511580,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09691272403496126,
+      "grad_norm": 0.41066092252731323,
+      "learning_rate": 4.941592101508264e-05,
+      "loss": 1.6487,
+      "num_input_tokens_seen": 78885660,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09691272403496126,
+      "eval_loss": 1.7664850950241089,
+      "eval_runtime": 3.9496,
+      "eval_samples_per_second": 37.979,
+      "eval_steps_per_second": 4.811,
+      "num_input_tokens_seen": 78885660,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09739728765513607,
+      "grad_norm": 0.4655153751373291,
+      "learning_rate": 4.94074591511686e-05,
+      "loss": 1.6776,
+      "num_input_tokens_seen": 79304256,
+      "step": 2010
+    },
+    {
+      "epoch": 0.09788185127531088,
+      "grad_norm": 0.47283700108528137,
+      "learning_rate": 4.939893716652404e-05,
+      "loss": 1.6605,
+      "num_input_tokens_seen": 79720364,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09836641489548568,
+      "grad_norm": 0.433597207069397,
+      "learning_rate": 4.93903550821404e-05,
+      "loss": 1.6256,
+      "num_input_tokens_seen": 80133808,
+      "step": 2030
+    },
+    {
+      "epoch": 0.09885097851566049,
+      "grad_norm": 0.5006552934646606,
+      "learning_rate": 4.9381712919157174e-05,
+      "loss": 1.672,
+      "num_input_tokens_seen": 80519820,
+      "step": 2040
+    },
+    {
+      "epoch": 0.0993355421358353,
+      "grad_norm": 0.4227645993232727,
+      "learning_rate": 4.937301069886184e-05,
+      "loss": 1.5913,
+      "num_input_tokens_seen": 80921068,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09982010575601011,
+      "grad_norm": 0.4382697641849518,
+      "learning_rate": 4.93642484426898e-05,
+      "loss": 1.6587,
+      "num_input_tokens_seen": 81330628,
+      "step": 2060
+    },
+    {
+      "epoch": 0.10030466937618492,
+      "grad_norm": 0.45559099316596985,
+      "learning_rate": 4.935542617222434e-05,
+      "loss": 1.6698,
+      "num_input_tokens_seen": 81744300,
+      "step": 2070
+    },
+    {
+      "epoch": 0.10078923299635971,
+      "grad_norm": 0.44068899750709534,
+      "learning_rate": 4.9346543909196584e-05,
+      "loss": 1.6527,
+      "num_input_tokens_seen": 82151876,
+      "step": 2080
+    },
+    {
+      "epoch": 0.10127379661653452,
+      "grad_norm": 0.4670219123363495,
+      "learning_rate": 4.933760167548542e-05,
+      "loss": 1.66,
+      "num_input_tokens_seen": 82539776,
+      "step": 2090
+    },
+    {
+      "epoch": 0.10175836023670932,
+      "grad_norm": 0.4150504171848297,
+      "learning_rate": 4.932859949311745e-05,
+      "loss": 1.6959,
+      "num_input_tokens_seen": 82918640,
+      "step": 2100
+    },
+    {
+      "epoch": 0.10224292385688413,
+      "grad_norm": 0.46419015526771545,
+      "learning_rate": 4.931953738426698e-05,
+      "loss": 1.6897,
+      "num_input_tokens_seen": 83299820,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10272748747705894,
+      "grad_norm": 0.4582190215587616,
+      "learning_rate": 4.931041537125587e-05,
+      "loss": 1.6196,
+      "num_input_tokens_seen": 83659764,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10321205109723375,
+      "grad_norm": 0.4939762055873871,
+      "learning_rate": 4.930123347655358e-05,
+      "loss": 1.6231,
+      "num_input_tokens_seen": 84052080,
+      "step": 2130
+    },
+    {
+      "epoch": 0.10369661471740856,
+      "grad_norm": 0.43949180841445923,
+      "learning_rate": 4.929199172277705e-05,
+      "loss": 1.5858,
+      "num_input_tokens_seen": 84431660,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10418117833758336,
+      "grad_norm": 0.44852331280708313,
+      "learning_rate": 4.928269013269069e-05,
+      "loss": 1.678,
+      "num_input_tokens_seen": 84828416,
+      "step": 2150
+    },
+    {
+      "epoch": 0.10466574195775817,
+      "grad_norm": 0.4769739508628845,
+      "learning_rate": 4.927332872920626e-05,
+      "loss": 1.6463,
+      "num_input_tokens_seen": 85231908,
+      "step": 2160
+    },
+    {
+      "epoch": 0.10515030557793298,
+      "grad_norm": 0.41337233781814575,
+      "learning_rate": 4.926390753538288e-05,
+      "loss": 1.6137,
+      "num_input_tokens_seen": 85644796,
+      "step": 2170
+    },
+    {
+      "epoch": 0.10563486919810777,
+      "grad_norm": 0.4415737986564636,
+      "learning_rate": 4.925442657442696e-05,
+      "loss": 1.638,
+      "num_input_tokens_seen": 86024508,
+      "step": 2180
+    },
+    {
+      "epoch": 0.10611943281828258,
+      "grad_norm": 0.4723414480686188,
+      "learning_rate": 4.924488586969208e-05,
+      "loss": 1.6074,
+      "num_input_tokens_seen": 86425788,
+      "step": 2190
+    },
+    {
+      "epoch": 0.10660399643845739,
+      "grad_norm": 0.4304730296134949,
+      "learning_rate": 4.923528544467905e-05,
+      "loss": 1.6436,
+      "num_input_tokens_seen": 86797336,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1070885600586322,
+      "grad_norm": 0.4499008059501648,
+      "learning_rate": 4.9225625323035706e-05,
+      "loss": 1.6196,
+      "num_input_tokens_seen": 87182044,
+      "step": 2210
+    },
+    {
+      "epoch": 0.107573123678807,
+      "grad_norm": 0.415367066860199,
+      "learning_rate": 4.9215905528557e-05,
+      "loss": 1.6363,
+      "num_input_tokens_seen": 87604236,
+      "step": 2220
+    },
+    {
+      "epoch": 0.10805768729898181,
+      "grad_norm": 0.4271062910556793,
+      "learning_rate": 4.9206126085184824e-05,
+      "loss": 1.6101,
+      "num_input_tokens_seen": 87995600,
+      "step": 2230
+    },
+    {
+      "epoch": 0.10854225091915662,
+      "grad_norm": 0.4146898686885834,
+      "learning_rate": 4.919628701700802e-05,
+      "loss": 1.5793,
+      "num_input_tokens_seen": 88393892,
+      "step": 2240
+    },
+    {
+      "epoch": 0.10902681453933143,
+      "grad_norm": 0.43309634923934937,
+      "learning_rate": 4.918638834826229e-05,
+      "loss": 1.5593,
+      "num_input_tokens_seen": 88770548,
+      "step": 2250
+    },
+    {
+      "epoch": 0.10951137815950623,
+      "grad_norm": 0.42651528120040894,
+      "learning_rate": 4.917643010333015e-05,
+      "loss": 1.6263,
+      "num_input_tokens_seen": 89160792,
+      "step": 2260
+    },
+    {
+      "epoch": 0.10999594177968104,
+      "grad_norm": 0.4150625467300415,
+      "learning_rate": 4.916641230674086e-05,
+      "loss": 1.6196,
+      "num_input_tokens_seen": 89548084,
+      "step": 2270
+    },
+    {
+      "epoch": 0.11048050539985584,
+      "grad_norm": 0.4922536313533783,
+      "learning_rate": 4.915633498317037e-05,
+      "loss": 1.6234,
+      "num_input_tokens_seen": 89957316,
+      "step": 2280
+    },
+    {
+      "epoch": 0.11096506902003064,
+      "grad_norm": 0.436235636472702,
+      "learning_rate": 4.914619815744126e-05,
+      "loss": 1.602,
+      "num_input_tokens_seen": 90352888,
+      "step": 2290
+    },
+    {
+      "epoch": 0.11144963264020545,
+      "grad_norm": 0.44822990894317627,
+      "learning_rate": 4.913600185452267e-05,
+      "loss": 1.6393,
+      "num_input_tokens_seen": 90729132,
+      "step": 2300
+    },
+    {
+      "epoch": 0.11193419626038026,
+      "grad_norm": 0.4682275652885437,
+      "learning_rate": 4.912574609953026e-05,
+      "loss": 1.5937,
+      "num_input_tokens_seen": 91133560,
+      "step": 2310
+    },
+    {
+      "epoch": 0.11241875988055507,
+      "grad_norm": 0.44868946075439453,
+      "learning_rate": 4.911543091772611e-05,
+      "loss": 1.6376,
+      "num_input_tokens_seen": 91514000,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11290332350072987,
+      "grad_norm": 0.4826831817626953,
+      "learning_rate": 4.910505633451869e-05,
+      "loss": 1.6667,
+      "num_input_tokens_seen": 91919504,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11338788712090468,
+      "grad_norm": 0.5310072898864746,
+      "learning_rate": 4.9094622375462806e-05,
+      "loss": 1.6148,
+      "num_input_tokens_seen": 92332568,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11387245074107949,
+      "grad_norm": 0.43625608086586,
+      "learning_rate": 4.90841290662595e-05,
+      "loss": 1.6217,
+      "num_input_tokens_seen": 92716832,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1143570143612543,
+      "grad_norm": 0.46015968918800354,
+      "learning_rate": 4.9073576432755995e-05,
+      "loss": 1.6222,
+      "num_input_tokens_seen": 93138708,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1148415779814291,
+      "grad_norm": 0.4389876425266266,
+      "learning_rate": 4.906296450094568e-05,
+      "loss": 1.617,
+      "num_input_tokens_seen": 93552848,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1153261416016039,
+      "grad_norm": 0.46346965432167053,
+      "learning_rate": 4.9052293296967975e-05,
+      "loss": 1.5736,
+      "num_input_tokens_seen": 93947928,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1158107052217787,
+      "grad_norm": 0.4337058961391449,
+      "learning_rate": 4.9041562847108304e-05,
+      "loss": 1.5778,
+      "num_input_tokens_seen": 94323152,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11629526884195351,
+      "grad_norm": 0.4524190425872803,
+      "learning_rate": 4.903077317779805e-05,
+      "loss": 1.5926,
+      "num_input_tokens_seen": 94707744,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11677983246212832,
+      "grad_norm": 0.4808904826641083,
+      "learning_rate": 4.901992431561443e-05,
+      "loss": 1.6427,
+      "num_input_tokens_seen": 95084712,
+      "step": 2410
+    },
+    {
+      "epoch": 0.11726439608230313,
+      "grad_norm": 0.4621996581554413,
+      "learning_rate": 4.9009016287280496e-05,
+      "loss": 1.5838,
+      "num_input_tokens_seen": 95498392,
+      "step": 2420
+    },
+    {
+      "epoch": 0.11774895970247794,
+      "grad_norm": 0.41997554898262024,
+      "learning_rate": 4.899804911966502e-05,
+      "loss": 1.6226,
+      "num_input_tokens_seen": 95880184,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11823352332265274,
+      "grad_norm": 0.475421279668808,
+      "learning_rate": 4.898702283978247e-05,
+      "loss": 1.6259,
+      "num_input_tokens_seen": 96265332,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11871808694282755,
+      "grad_norm": 0.45337820053100586,
+      "learning_rate": 4.8975937474792895e-05,
+      "loss": 1.6884,
+      "num_input_tokens_seen": 96681320,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11920265056300236,
+      "grad_norm": 0.42129579186439514,
+      "learning_rate": 4.896479305200188e-05,
+      "loss": 1.5833,
+      "num_input_tokens_seen": 97088276,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11968721418317717,
+      "grad_norm": 0.42145466804504395,
+      "learning_rate": 4.895358959886051e-05,
+      "loss": 1.6114,
+      "num_input_tokens_seen": 97501600,
+      "step": 2470
+    },
+    {
+      "epoch": 0.12017177780335198,
+      "grad_norm": 0.4644761085510254,
+      "learning_rate": 4.8942327142965244e-05,
+      "loss": 1.6331,
+      "num_input_tokens_seen": 97893056,
+      "step": 2480
+    },
+    {
+      "epoch": 0.12065634142352677,
+      "grad_norm": 0.4692575931549072,
+      "learning_rate": 4.8931005712057905e-05,
+      "loss": 1.591,
+      "num_input_tokens_seen": 98303764,
+      "step": 2490
+    },
+    {
+      "epoch": 0.12114090504370158,
+      "grad_norm": 0.4318110942840576,
+      "learning_rate": 4.891962533402556e-05,
+      "loss": 1.5659,
+      "num_input_tokens_seen": 98672376,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12162546866387638,
+      "grad_norm": 0.44102174043655396,
+      "learning_rate": 4.890818603690049e-05,
+      "loss": 1.5804,
+      "num_input_tokens_seen": 99057612,
+      "step": 2510
+    },
+    {
+      "epoch": 0.12211003228405119,
+      "grad_norm": 0.45612481236457825,
+      "learning_rate": 4.88966878488601e-05,
+      "loss": 1.6189,
+      "num_input_tokens_seen": 99443028,
+      "step": 2520
+    },
+    {
+      "epoch": 0.122594595904226,
+      "grad_norm": 0.502396821975708,
+      "learning_rate": 4.888513079822686e-05,
+      "loss": 1.542,
+      "num_input_tokens_seen": 99861476,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12307915952440081,
+      "grad_norm": 0.438915491104126,
+      "learning_rate": 4.887351491346822e-05,
+      "loss": 1.554,
+      "num_input_tokens_seen": 100294464,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12356372314457562,
+      "grad_norm": 0.474985808134079,
+      "learning_rate": 4.886184022319657e-05,
+      "loss": 1.6263,
+      "num_input_tokens_seen": 100653652,
+      "step": 2550
+    },
+    {
+      "epoch": 0.12404828676475042,
+      "grad_norm": 0.43759259581565857,
+      "learning_rate": 4.8850106756169146e-05,
+      "loss": 1.5717,
+      "num_input_tokens_seen": 101039388,
+      "step": 2560
+    },
+    {
+      "epoch": 0.12453285038492523,
+      "grad_norm": 0.4843815267086029,
+      "learning_rate": 4.8838314541287936e-05,
+      "loss": 1.6216,
+      "num_input_tokens_seen": 101443224,
+      "step": 2570
+    },
+    {
+      "epoch": 0.12501741400510002,
+      "grad_norm": 0.5523076057434082,
+      "learning_rate": 4.882646360759967e-05,
+      "loss": 1.5746,
+      "num_input_tokens_seen": 101857288,
+      "step": 2580
+    },
+    {
+      "epoch": 0.12550197762527485,
+      "grad_norm": 0.43904462456703186,
+      "learning_rate": 4.88145539842957e-05,
+      "loss": 1.5948,
+      "num_input_tokens_seen": 102260496,
+      "step": 2590
+    },
+    {
+      "epoch": 0.12598654124544964,
+      "grad_norm": 0.4333835244178772,
+      "learning_rate": 4.880258570071194e-05,
+      "loss": 1.6429,
+      "num_input_tokens_seen": 102644332,
+      "step": 2600
+    },
+    {
+      "epoch": 0.12647110486562446,
+      "grad_norm": 0.42234155535697937,
+      "learning_rate": 4.879055878632881e-05,
+      "loss": 1.5685,
+      "num_input_tokens_seen": 103032436,
+      "step": 2610
+    },
+    {
+      "epoch": 0.12695566848579926,
+      "grad_norm": 0.40778931975364685,
+      "learning_rate": 4.8778473270771144e-05,
+      "loss": 1.5875,
+      "num_input_tokens_seen": 103422688,
+      "step": 2620
+    },
+    {
+      "epoch": 0.12744023210597408,
+      "grad_norm": 0.45513054728507996,
+      "learning_rate": 4.8766329183808115e-05,
+      "loss": 1.5896,
+      "num_input_tokens_seen": 103798060,
+      "step": 2630
+    },
+    {
+      "epoch": 0.12792479572614887,
+      "grad_norm": 0.4449268579483032,
+      "learning_rate": 4.875412655535319e-05,
+      "loss": 1.6182,
+      "num_input_tokens_seen": 104187120,
+      "step": 2640
+    },
+    {
+      "epoch": 0.12840935934632366,
+      "grad_norm": 0.4176337718963623,
+      "learning_rate": 4.8741865415463995e-05,
+      "loss": 1.5704,
+      "num_input_tokens_seen": 104566676,
+      "step": 2650
+    },
+    {
+      "epoch": 0.1288939229664985,
+      "grad_norm": 0.4052715003490448,
+      "learning_rate": 4.8729545794342326e-05,
+      "loss": 1.5684,
+      "num_input_tokens_seen": 104970872,
+      "step": 2660
+    },
+    {
+      "epoch": 0.12937848658667328,
+      "grad_norm": 0.43383073806762695,
+      "learning_rate": 4.871716772233401e-05,
+      "loss": 1.5695,
+      "num_input_tokens_seen": 105341948,
+      "step": 2670
+    },
+    {
+      "epoch": 0.1298630502068481,
+      "grad_norm": 0.4440796673297882,
+      "learning_rate": 4.870473122992886e-05,
+      "loss": 1.5893,
+      "num_input_tokens_seen": 105747904,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1303476138270229,
+      "grad_norm": 0.47211089730262756,
+      "learning_rate": 4.869223634776059e-05,
+      "loss": 1.5961,
+      "num_input_tokens_seen": 106128328,
+      "step": 2690
+    },
+    {
+      "epoch": 0.13083217744719772,
+      "grad_norm": 0.4548238515853882,
+      "learning_rate": 4.867968310660671e-05,
+      "loss": 1.6209,
+      "num_input_tokens_seen": 106507300,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1313167410673725,
+      "grad_norm": 0.4506012797355652,
+      "learning_rate": 4.8667071537388535e-05,
+      "loss": 1.6299,
+      "num_input_tokens_seen": 106889452,
+      "step": 2710
+    },
+    {
+      "epoch": 0.13180130468754733,
+      "grad_norm": 0.42806220054626465,
+      "learning_rate": 4.8654401671171014e-05,
+      "loss": 1.5991,
+      "num_input_tokens_seen": 107252004,
+      "step": 2720
+    },
+    {
+      "epoch": 0.13228586830772213,
+      "grad_norm": 0.4180905222892761,
+      "learning_rate": 4.86416735391627e-05,
+      "loss": 1.5844,
+      "num_input_tokens_seen": 107629528,
+      "step": 2730
+    },
+    {
+      "epoch": 0.13277043192789692,
+      "grad_norm": 0.4253162145614624,
+      "learning_rate": 4.862888717271568e-05,
+      "loss": 1.5904,
+      "num_input_tokens_seen": 107996228,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13325499554807174,
+      "grad_norm": 0.44208529591560364,
+      "learning_rate": 4.861604260332547e-05,
+      "loss": 1.5569,
+      "num_input_tokens_seen": 108380192,
+      "step": 2750
+    },
+    {
+      "epoch": 0.13373955916824654,
+      "grad_norm": 0.40524107217788696,
+      "learning_rate": 4.8603139862630966e-05,
+      "loss": 1.6175,
+      "num_input_tokens_seen": 108779284,
+      "step": 2760
+    },
+    {
+      "epoch": 0.13422412278842136,
+      "grad_norm": 0.5582908391952515,
+      "learning_rate": 4.8590178982414346e-05,
+      "loss": 1.5905,
+      "num_input_tokens_seen": 109193100,
+      "step": 2770
+    },
+    {
+      "epoch": 0.13470868640859615,
+      "grad_norm": 0.4604218900203705,
+      "learning_rate": 4.8577159994600995e-05,
+      "loss": 1.5554,
+      "num_input_tokens_seen": 109569728,
+      "step": 2780
+    },
+    {
+      "epoch": 0.13519325002877097,
+      "grad_norm": 0.41278553009033203,
+      "learning_rate": 4.8564082931259426e-05,
+      "loss": 1.5784,
+      "num_input_tokens_seen": 109984452,
+      "step": 2790
+    },
+    {
+      "epoch": 0.13567781364894577,
+      "grad_norm": 0.4047918915748596,
+      "learning_rate": 4.8550947824601216e-05,
+      "loss": 1.5317,
+      "num_input_tokens_seen": 110389128,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1361623772691206,
+      "grad_norm": 0.4299299716949463,
+      "learning_rate": 4.853775470698091e-05,
+      "loss": 1.5954,
+      "num_input_tokens_seen": 110788072,
+      "step": 2810
+    },
+    {
+      "epoch": 0.13664694088929538,
+      "grad_norm": 0.4523586332798004,
+      "learning_rate": 4.8524503610895944e-05,
+      "loss": 1.5679,
+      "num_input_tokens_seen": 111191500,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1371315045094702,
+      "grad_norm": 0.4595218002796173,
+      "learning_rate": 4.8511194568986563e-05,
+      "loss": 1.5974,
+      "num_input_tokens_seen": 111580564,
+      "step": 2830
+    },
+    {
+      "epoch": 0.137616068129645,
+      "grad_norm": 0.4451237916946411,
+      "learning_rate": 4.8497827614035755e-05,
+      "loss": 1.5807,
+      "num_input_tokens_seen": 111974908,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1381006317498198,
+      "grad_norm": 0.4401971995830536,
+      "learning_rate": 4.848440277896915e-05,
+      "loss": 1.5301,
+      "num_input_tokens_seen": 112379284,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1385851953699946,
+      "grad_norm": 0.4478752017021179,
+      "learning_rate": 4.847092009685496e-05,
+      "loss": 1.5541,
+      "num_input_tokens_seen": 112797012,
+      "step": 2860
+    },
+    {
+      "epoch": 0.1390697589901694,
+      "grad_norm": 0.47638776898384094,
+      "learning_rate": 4.8457379600903886e-05,
+      "loss": 1.5861,
+      "num_input_tokens_seen": 113177816,
+      "step": 2870
+    },
+    {
+      "epoch": 0.13955432261034423,
+      "grad_norm": 0.4315278232097626,
+      "learning_rate": 4.844378132446903e-05,
+      "loss": 1.527,
+      "num_input_tokens_seen": 113585668,
+      "step": 2880
+    },
+    {
+      "epoch": 0.14003888623051902,
+      "grad_norm": 0.45071589946746826,
+      "learning_rate": 4.843012530104581e-05,
+      "loss": 1.5542,
+      "num_input_tokens_seen": 114012516,
+      "step": 2890
+    },
+    {
+      "epoch": 0.14052344985069384,
+      "grad_norm": 0.4602771997451782,
+      "learning_rate": 4.841641156427189e-05,
+      "loss": 1.543,
+      "num_input_tokens_seen": 114413536,
+      "step": 2900
+    },
+    {
+      "epoch": 0.14100801347086864,
+      "grad_norm": 0.4230845272541046,
+      "learning_rate": 4.8402640147927134e-05,
+      "loss": 1.5379,
+      "num_input_tokens_seen": 114815632,
+      "step": 2910
+    },
+    {
+      "epoch": 0.14149257709104346,
+      "grad_norm": 0.4051437973976135,
+      "learning_rate": 4.838881108593342e-05,
+      "loss": 1.5231,
+      "num_input_tokens_seen": 115202344,
+      "step": 2920
+    },
+    {
+      "epoch": 0.14197714071121825,
+      "grad_norm": 0.5333889126777649,
+      "learning_rate": 4.837492441235467e-05,
+      "loss": 1.5378,
+      "num_input_tokens_seen": 115622044,
+      "step": 2930
+    },
+    {
+      "epoch": 0.14246170433139305,
+      "grad_norm": 0.4066469967365265,
+      "learning_rate": 4.8360980161396685e-05,
+      "loss": 1.584,
+      "num_input_tokens_seen": 115983972,
+      "step": 2940
+    },
+    {
+      "epoch": 0.14294626795156787,
+      "grad_norm": 0.4050688147544861,
+      "learning_rate": 4.834697836740712e-05,
+      "loss": 1.5642,
+      "num_input_tokens_seen": 116370664,
+      "step": 2950
+    },
+    {
+      "epoch": 0.14343083157174266,
+      "grad_norm": 0.4575314521789551,
+      "learning_rate": 4.833291906487533e-05,
+      "loss": 1.512,
+      "num_input_tokens_seen": 116802032,
+      "step": 2960
+    },
+    {
+      "epoch": 0.14391539519191748,
+      "grad_norm": 0.43021175265312195,
+      "learning_rate": 4.831880228843235e-05,
+      "loss": 1.5795,
+      "num_input_tokens_seen": 117193496,
+      "step": 2970
+    },
+    {
+      "epoch": 0.14439995881209228,
+      "grad_norm": 0.4586018919944763,
+      "learning_rate": 4.83046280728508e-05,
+      "loss": 1.556,
+      "num_input_tokens_seen": 117616412,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1448845224322671,
+      "grad_norm": 0.47244206070899963,
+      "learning_rate": 4.8290396453044764e-05,
+      "loss": 1.5146,
+      "num_input_tokens_seen": 117972728,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1453690860524419,
+      "grad_norm": 0.3958747088909149,
+      "learning_rate": 4.827610746406972e-05,
+      "loss": 1.5349,
+      "num_input_tokens_seen": 118356720,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14585364967261671,
+      "grad_norm": 0.45024189352989197,
+      "learning_rate": 4.826176114112247e-05,
+      "loss": 1.6121,
+      "num_input_tokens_seen": 118769664,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1463382132927915,
+      "grad_norm": 0.44841715693473816,
+      "learning_rate": 4.824735751954106e-05,
+      "loss": 1.5356,
+      "num_input_tokens_seen": 119180436,
+      "step": 3020
+    },
+    {
+      "epoch": 0.14682277691296633,
+      "grad_norm": 0.4641439616680145,
+      "learning_rate": 4.8232896634804634e-05,
+      "loss": 1.6029,
+      "num_input_tokens_seen": 119553244,
+      "step": 3030
+    },
+    {
+      "epoch": 0.14730734053314112,
+      "grad_norm": 0.4200107455253601,
+      "learning_rate": 4.8218378522533404e-05,
+      "loss": 1.5575,
+      "num_input_tokens_seen": 119933780,
+      "step": 3040
+    },
+    {
+      "epoch": 0.14779190415331592,
+      "grad_norm": 0.47531604766845703,
+      "learning_rate": 4.8203803218488567e-05,
+      "loss": 1.5486,
+      "num_input_tokens_seen": 120345892,
+      "step": 3050
+    },
+    {
+      "epoch": 0.14827646777349074,
+      "grad_norm": 0.42880284786224365,
+      "learning_rate": 4.8189170758572154e-05,
+      "loss": 1.5223,
+      "num_input_tokens_seen": 120711848,
+      "step": 3060
+    },
+    {
+      "epoch": 0.14876103139366553,
+      "grad_norm": 0.4171142280101776,
+      "learning_rate": 4.817448117882703e-05,
+      "loss": 1.5771,
+      "num_input_tokens_seen": 121113568,
+      "step": 3070
+    },
+    {
+      "epoch": 0.14924559501384035,
+      "grad_norm": 0.43676143884658813,
+      "learning_rate": 4.815973451543672e-05,
+      "loss": 1.5337,
+      "num_input_tokens_seen": 121540304,
+      "step": 3080
+    },
+    {
+      "epoch": 0.14973015863401515,
+      "grad_norm": 0.4129098057746887,
+      "learning_rate": 4.814493080472538e-05,
+      "loss": 1.5162,
+      "num_input_tokens_seen": 121912956,
+      "step": 3090
+    },
+    {
+      "epoch": 0.15021472225418997,
+      "grad_norm": 0.4339870810508728,
+      "learning_rate": 4.8130070083157676e-05,
+      "loss": 1.5347,
+      "num_input_tokens_seen": 122302884,
+      "step": 3100
+    },
+    {
+      "epoch": 0.15069928587436476,
+      "grad_norm": 0.4579829275608063,
+      "learning_rate": 4.8115152387338705e-05,
+      "loss": 1.5682,
+      "num_input_tokens_seen": 122729188,
+      "step": 3110
+    },
+    {
+      "epoch": 0.15118384949453958,
+      "grad_norm": 0.44452232122421265,
+      "learning_rate": 4.81001777540139e-05,
+      "loss": 1.5084,
+      "num_input_tokens_seen": 123099424,
+      "step": 3120
+    },
+    {
+      "epoch": 0.15166841311471438,
+      "grad_norm": 0.4334086775779724,
+      "learning_rate": 4.8085146220068955e-05,
+      "loss": 1.5089,
+      "num_input_tokens_seen": 123501724,
+      "step": 3130
+    },
+    {
+      "epoch": 0.15215297673488917,
+      "grad_norm": 0.4431312680244446,
+      "learning_rate": 4.8070057822529715e-05,
+      "loss": 1.4841,
+      "num_input_tokens_seen": 123897884,
+      "step": 3140
+    },
+    {
+      "epoch": 0.152637540355064,
+      "grad_norm": 0.42815452814102173,
+      "learning_rate": 4.8054912598562086e-05,
+      "loss": 1.5234,
+      "num_input_tokens_seen": 124295448,
+      "step": 3150
+    },
+    {
+      "epoch": 0.1531221039752388,
+      "grad_norm": 0.44743165373802185,
+      "learning_rate": 4.8039710585471966e-05,
+      "loss": 1.5925,
+      "num_input_tokens_seen": 124680420,
+      "step": 3160
+    },
+    {
+      "epoch": 0.1536066675954136,
+      "grad_norm": 0.41514381766319275,
+      "learning_rate": 4.802445182070511e-05,
+      "loss": 1.5505,
+      "num_input_tokens_seen": 125076112,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1540912312155884,
+      "grad_norm": 0.4845329225063324,
+      "learning_rate": 4.8009136341847094e-05,
+      "loss": 1.5366,
+      "num_input_tokens_seen": 125473256,
+      "step": 3180
+    },
+    {
+      "epoch": 0.15457579483576322,
+      "grad_norm": 0.4526010751724243,
+      "learning_rate": 4.799376418662318e-05,
+      "loss": 1.5357,
+      "num_input_tokens_seen": 125882064,
+      "step": 3190
+    },
+    {
+      "epoch": 0.15506035845593802,
+      "grad_norm": 0.42789971828460693,
+      "learning_rate": 4.7978335392898235e-05,
+      "loss": 1.4749,
+      "num_input_tokens_seen": 126280000,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15554492207611284,
+      "grad_norm": 0.4213913083076477,
+      "learning_rate": 4.796284999867663e-05,
+      "loss": 1.5433,
+      "num_input_tokens_seen": 126699892,
+      "step": 3210
+    },
+    {
+      "epoch": 0.15602948569628763,
+      "grad_norm": 0.39674410223960876,
+      "learning_rate": 4.794730804210217e-05,
+      "loss": 1.5379,
+      "num_input_tokens_seen": 127115872,
+      "step": 3220
+    },
+    {
+      "epoch": 0.15651404931646246,
+      "grad_norm": 0.45900386571884155,
+      "learning_rate": 4.793170956145798e-05,
+      "loss": 1.5696,
+      "num_input_tokens_seen": 127499596,
+      "step": 3230
+    },
+    {
+      "epoch": 0.15699861293663725,
+      "grad_norm": 0.43314629793167114,
+      "learning_rate": 4.791605459516641e-05,
+      "loss": 1.5543,
+      "num_input_tokens_seen": 127906492,
+      "step": 3240
+    },
+    {
+      "epoch": 0.15748317655681204,
+      "grad_norm": 0.408204585313797,
+      "learning_rate": 4.7900343181788955e-05,
+      "loss": 1.5537,
+      "num_input_tokens_seen": 128323396,
+      "step": 3250
+    },
+    {
+      "epoch": 0.15796774017698686,
+      "grad_norm": 0.45687878131866455,
+      "learning_rate": 4.788457536002614e-05,
+      "loss": 1.5473,
+      "num_input_tokens_seen": 128719820,
+      "step": 3260
+    },
+    {
+      "epoch": 0.15845230379716166,
+      "grad_norm": 0.423566997051239,
+      "learning_rate": 4.7868751168717454e-05,
+      "loss": 1.6038,
+      "num_input_tokens_seen": 129126672,
+      "step": 3270
+    },
+    {
+      "epoch": 0.15893686741733648,
+      "grad_norm": 0.4187241196632385,
+      "learning_rate": 4.785287064684122e-05,
+      "loss": 1.5793,
+      "num_input_tokens_seen": 129537864,
+      "step": 3280
+    },
+    {
+      "epoch": 0.15942143103751127,
+      "grad_norm": 0.45594966411590576,
+      "learning_rate": 4.783693383351452e-05,
+      "loss": 1.5753,
+      "num_input_tokens_seen": 129948744,
+      "step": 3290
+    },
+    {
+      "epoch": 0.1599059946576861,
+      "grad_norm": 0.4248170554637909,
+      "learning_rate": 4.782094076799308e-05,
+      "loss": 1.5443,
+      "num_input_tokens_seen": 130350392,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1603905582778609,
+      "grad_norm": 0.491043359041214,
+      "learning_rate": 4.780489148967122e-05,
+      "loss": 1.5265,
+      "num_input_tokens_seen": 130744872,
+      "step": 3310
+    },
+    {
+      "epoch": 0.1608751218980357,
+      "grad_norm": 0.44143855571746826,
+      "learning_rate": 4.77887860380817e-05,
+      "loss": 1.4956,
+      "num_input_tokens_seen": 131164480,
+      "step": 3320
+    },
+    {
+      "epoch": 0.1613596855182105,
+      "grad_norm": 0.4861333668231964,
+      "learning_rate": 4.777262445289565e-05,
+      "loss": 1.5403,
+      "num_input_tokens_seen": 131523208,
+      "step": 3330
+    },
+    {
+      "epoch": 0.16184424913838533,
+      "grad_norm": 0.44125744700431824,
+      "learning_rate": 4.775640677392246e-05,
+      "loss": 1.5067,
+      "num_input_tokens_seen": 131941532,
+      "step": 3340
+    },
+    {
+      "epoch": 0.16232881275856012,
+      "grad_norm": 0.43593043088912964,
+      "learning_rate": 4.774013304110972e-05,
+      "loss": 1.5443,
+      "num_input_tokens_seen": 132309260,
+      "step": 3350
+    },
+    {
+      "epoch": 0.16281337637873491,
+      "grad_norm": 0.37986883521080017,
+      "learning_rate": 4.7723803294543056e-05,
+      "loss": 1.5141,
+      "num_input_tokens_seen": 132674688,
+      "step": 3360
+    },
+    {
+      "epoch": 0.16329793999890974,
+      "grad_norm": 0.4436347484588623,
+      "learning_rate": 4.7707417574446086e-05,
+      "loss": 1.5817,
+      "num_input_tokens_seen": 133074724,
+      "step": 3370
+    },
+    {
+      "epoch": 0.16378250361908453,
+      "grad_norm": 0.48060745000839233,
+      "learning_rate": 4.769097592118033e-05,
+      "loss": 1.5664,
+      "num_input_tokens_seen": 133457192,
+      "step": 3380
+    },
+    {
+      "epoch": 0.16426706723925935,
+      "grad_norm": 0.4321390688419342,
+      "learning_rate": 4.7674478375245013e-05,
+      "loss": 1.5077,
+      "num_input_tokens_seen": 133844504,
+      "step": 3390
+    },
+    {
+      "epoch": 0.16475163085943414,
+      "grad_norm": 0.4549488425254822,
+      "learning_rate": 4.76579249772771e-05,
+      "loss": 1.5348,
+      "num_input_tokens_seen": 134258676,
+      "step": 3400
+    },
+    {
+      "epoch": 0.16523619447960897,
+      "grad_norm": 0.47234755754470825,
+      "learning_rate": 4.764131576805111e-05,
+      "loss": 1.55,
+      "num_input_tokens_seen": 134615472,
+      "step": 3410
+    },
+    {
+      "epoch": 0.16572075809978376,
+      "grad_norm": 0.4680884778499603,
+      "learning_rate": 4.762465078847903e-05,
+      "loss": 1.4957,
+      "num_input_tokens_seen": 135018208,
+      "step": 3420
+    },
+    {
+      "epoch": 0.16620532171995858,
+      "grad_norm": 0.4551864564418793,
+      "learning_rate": 4.760793007961023e-05,
+      "loss": 1.4933,
+      "num_input_tokens_seen": 135419696,
+      "step": 3430
+    },
+    {
+      "epoch": 0.16668988534013338,
+      "grad_norm": 0.4141683578491211,
+      "learning_rate": 4.759115368263135e-05,
+      "loss": 1.5556,
+      "num_input_tokens_seen": 135800204,
+      "step": 3440
+    },
+    {
+      "epoch": 0.16717444896030817,
+      "grad_norm": 0.3936518430709839,
+      "learning_rate": 4.75743216388662e-05,
+      "loss": 1.5483,
+      "num_input_tokens_seen": 136190572,
+      "step": 3450
+    },
+    {
+      "epoch": 0.167659012580483,
+      "grad_norm": 0.41701188683509827,
+      "learning_rate": 4.7557433989775654e-05,
+      "loss": 1.5062,
+      "num_input_tokens_seen": 136592572,
+      "step": 3460
+    },
+    {
+      "epoch": 0.16814357620065778,
+      "grad_norm": 0.41708120703697205,
+      "learning_rate": 4.754049077695758e-05,
+      "loss": 1.557,
+      "num_input_tokens_seen": 136951096,
+      "step": 3470
+    },
+    {
+      "epoch": 0.1686281398208326,
+      "grad_norm": 0.5057613253593445,
+      "learning_rate": 4.752349204214668e-05,
+      "loss": 1.5247,
+      "num_input_tokens_seen": 137320496,
+      "step": 3480
+    },
+    {
+      "epoch": 0.1691127034410074,
+      "grad_norm": 0.4021996855735779,
+      "learning_rate": 4.750643782721442e-05,
+      "loss": 1.548,
+      "num_input_tokens_seen": 137671948,
+      "step": 3490
+    },
+    {
+      "epoch": 0.16959726706118222,
+      "grad_norm": 0.4202756881713867,
+      "learning_rate": 4.7489328174168964e-05,
+      "loss": 1.529,
+      "num_input_tokens_seen": 138105600,
+      "step": 3500
+    },
+    {
+      "epoch": 0.17008183068135702,
+      "grad_norm": 0.4259394407272339,
+      "learning_rate": 4.747216312515498e-05,
+      "loss": 1.5241,
+      "num_input_tokens_seen": 138487120,
+      "step": 3510
+    },
+    {
+      "epoch": 0.17056639430153184,
+      "grad_norm": 0.3875693678855896,
+      "learning_rate": 4.745494272245361e-05,
+      "loss": 1.524,
+      "num_input_tokens_seen": 138885020,
+      "step": 3520
+    },
+    {
+      "epoch": 0.17105095792170663,
+      "grad_norm": 0.3988937735557556,
+      "learning_rate": 4.743766700848237e-05,
+      "loss": 1.5126,
+      "num_input_tokens_seen": 139288928,
+      "step": 3530
+    },
+    {
+      "epoch": 0.17153552154188145,
+      "grad_norm": 0.41776183247566223,
+      "learning_rate": 4.742033602579497e-05,
+      "loss": 1.5061,
+      "num_input_tokens_seen": 139667980,
+      "step": 3540
+    },
+    {
+      "epoch": 0.17202008516205625,
+      "grad_norm": 0.4490506649017334,
+      "learning_rate": 4.740294981708129e-05,
+      "loss": 1.5089,
+      "num_input_tokens_seen": 140092556,
+      "step": 3550
+    },
+    {
+      "epoch": 0.17250464878223104,
+      "grad_norm": 0.44959497451782227,
+      "learning_rate": 4.738550842516724e-05,
+      "loss": 1.5266,
+      "num_input_tokens_seen": 140485860,
+      "step": 3560
+    },
+    {
+      "epoch": 0.17298921240240586,
+      "grad_norm": 0.42410293221473694,
+      "learning_rate": 4.736801189301466e-05,
+      "loss": 1.5103,
+      "num_input_tokens_seen": 140883128,
+      "step": 3570
+    },
+    {
+      "epoch": 0.17347377602258066,
+      "grad_norm": 0.43540823459625244,
+      "learning_rate": 4.73504602637212e-05,
+      "loss": 1.5119,
+      "num_input_tokens_seen": 141262836,
+      "step": 3580
+    },
+    {
+      "epoch": 0.17395833964275548,
+      "grad_norm": 0.3838653862476349,
+      "learning_rate": 4.733285358052022e-05,
+      "loss": 1.4833,
+      "num_input_tokens_seen": 141659460,
+      "step": 3590
+    },
+    {
+      "epoch": 0.17444290326293027,
+      "grad_norm": 0.40834611654281616,
+      "learning_rate": 4.7315191886780727e-05,
+      "loss": 1.5645,
+      "num_input_tokens_seen": 142062380,
+      "step": 3600
+    },
+    {
+      "epoch": 0.1749274668831051,
+      "grad_norm": 0.39526131749153137,
+      "learning_rate": 4.729747522600719e-05,
+      "loss": 1.5083,
+      "num_input_tokens_seen": 142433008,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1754120305032799,
+      "grad_norm": 0.48167720437049866,
+      "learning_rate": 4.727970364183949e-05,
+      "loss": 1.5694,
+      "num_input_tokens_seen": 142814724,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1758965941234547,
+      "grad_norm": 0.40276476740837097,
+      "learning_rate": 4.72618771780528e-05,
+      "loss": 1.4578,
+      "num_input_tokens_seen": 143208888,
+      "step": 3630
+    },
+    {
+      "epoch": 0.1763811577436295,
+      "grad_norm": 0.46488335728645325,
+      "learning_rate": 4.724399587855747e-05,
+      "loss": 1.5344,
+      "num_input_tokens_seen": 143612984,
+      "step": 3640
+    },
+    {
+      "epoch": 0.1768657213638043,
+      "grad_norm": 0.4321140944957733,
+      "learning_rate": 4.7226059787398914e-05,
+      "loss": 1.5565,
+      "num_input_tokens_seen": 144004420,
+      "step": 3650
+    },
+    {
+      "epoch": 0.17735028498397912,
+      "grad_norm": 0.42687898874282837,
+      "learning_rate": 4.720806894875751e-05,
+      "loss": 1.4911,
+      "num_input_tokens_seen": 144385500,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1778348486041539,
+      "grad_norm": 0.39949852228164673,
+      "learning_rate": 4.7190023406948506e-05,
+      "loss": 1.4925,
+      "num_input_tokens_seen": 144786456,
+      "step": 3670
+    },
+    {
+      "epoch": 0.17831941222432873,
+      "grad_norm": 0.45870667695999146,
+      "learning_rate": 4.7171923206421886e-05,
+      "loss": 1.5279,
+      "num_input_tokens_seen": 145150804,
+      "step": 3680
+    },
+    {
+      "epoch": 0.17880397584450353,
+      "grad_norm": 0.3944126069545746,
+      "learning_rate": 4.715376839176226e-05,
+      "loss": 1.5185,
+      "num_input_tokens_seen": 145542392,
+      "step": 3690
+    },
+    {
+      "epoch": 0.17928853946467835,
+      "grad_norm": 0.41185203194618225,
+      "learning_rate": 4.713555900768879e-05,
+      "loss": 1.4337,
+      "num_input_tokens_seen": 145911624,
+      "step": 3700
+    },
+    {
+      "epoch": 0.17977310308485314,
+      "grad_norm": 0.39053046703338623,
+      "learning_rate": 4.711729509905501e-05,
+      "loss": 1.5279,
+      "num_input_tokens_seen": 146287476,
+      "step": 3710
+    },
+    {
+      "epoch": 0.18025766670502796,
+      "grad_norm": 0.41288021206855774,
+      "learning_rate": 4.709897671084881e-05,
+      "loss": 1.5093,
+      "num_input_tokens_seen": 146688592,
+      "step": 3720
+    },
+    {
+      "epoch": 0.18074223032520276,
+      "grad_norm": 0.400301992893219,
+      "learning_rate": 4.7080603888192256e-05,
+      "loss": 1.4321,
+      "num_input_tokens_seen": 147100860,
+      "step": 3730
+    },
+    {
+      "epoch": 0.18122679394537758,
+      "grad_norm": 0.39671310782432556,
+      "learning_rate": 4.7062176676341476e-05,
+      "loss": 1.4853,
+      "num_input_tokens_seen": 147506564,
+      "step": 3740
+    },
+    {
+      "epoch": 0.18171135756555237,
+      "grad_norm": 0.4283420741558075,
+      "learning_rate": 4.7043695120686594e-05,
+      "loss": 1.4895,
+      "num_input_tokens_seen": 147906624,
+      "step": 3750
+    },
+    {
+      "epoch": 0.18219592118572717,
+      "grad_norm": 0.44357410073280334,
+      "learning_rate": 4.7025159266751586e-05,
+      "loss": 1.5736,
+      "num_input_tokens_seen": 148321696,
+      "step": 3760
+    },
+    {
+      "epoch": 0.182680484805902,
+      "grad_norm": 0.4820682406425476,
+      "learning_rate": 4.7006569160194185e-05,
+      "loss": 1.4884,
+      "num_input_tokens_seen": 148710300,
+      "step": 3770
+    },
+    {
+      "epoch": 0.18316504842607678,
+      "grad_norm": 0.46169939637184143,
+      "learning_rate": 4.698792484680574e-05,
+      "loss": 1.4714,
+      "num_input_tokens_seen": 149111440,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1836496120462516,
+      "grad_norm": 0.42826372385025024,
+      "learning_rate": 4.6969226372511153e-05,
+      "loss": 1.5092,
+      "num_input_tokens_seen": 149498596,
+      "step": 3790
+    },
+    {
+      "epoch": 0.1841341756664264,
+      "grad_norm": 0.4284881055355072,
+      "learning_rate": 4.695047378336871e-05,
+      "loss": 1.5269,
+      "num_input_tokens_seen": 149928100,
+      "step": 3800
+    },
+    {
+      "epoch": 0.18461873928660122,
+      "grad_norm": 0.41511744260787964,
+      "learning_rate": 4.693166712556999e-05,
+      "loss": 1.5135,
+      "num_input_tokens_seen": 150349972,
+      "step": 3810
+    },
+    {
+      "epoch": 0.185103302906776,
+      "grad_norm": 0.43388912081718445,
+      "learning_rate": 4.6912806445439786e-05,
+      "loss": 1.5093,
+      "num_input_tokens_seen": 150745692,
+      "step": 3820
+    },
+    {
+      "epoch": 0.18558786652695083,
+      "grad_norm": 0.4395964741706848,
+      "learning_rate": 4.689389178943593e-05,
+      "loss": 1.5737,
+      "num_input_tokens_seen": 151156868,
+      "step": 3830
+    },
+    {
+      "epoch": 0.18607243014712563,
+      "grad_norm": 0.4370187819004059,
+      "learning_rate": 4.6874923204149215e-05,
+      "loss": 1.5126,
+      "num_input_tokens_seen": 151565152,
+      "step": 3840
+    },
+    {
+      "epoch": 0.18655699376730042,
+      "grad_norm": 0.4095444083213806,
+      "learning_rate": 4.685590073630327e-05,
+      "loss": 1.4713,
+      "num_input_tokens_seen": 151965476,
+      "step": 3850
+    },
+    {
+      "epoch": 0.18704155738747524,
+      "grad_norm": 0.3944699764251709,
+      "learning_rate": 4.683682443275447e-05,
+      "loss": 1.5337,
+      "num_input_tokens_seen": 152338220,
+      "step": 3860
+    },
+    {
+      "epoch": 0.18752612100765004,
+      "grad_norm": 0.4351947009563446,
+      "learning_rate": 4.681769434049177e-05,
+      "loss": 1.491,
+      "num_input_tokens_seen": 152724792,
+      "step": 3870
+    },
+    {
+      "epoch": 0.18801068462782486,
+      "grad_norm": 0.5131449699401855,
+      "learning_rate": 4.6798510506636626e-05,
+      "loss": 1.5185,
+      "num_input_tokens_seen": 153099052,
+      "step": 3880
+    },
+    {
+      "epoch": 0.18849524824799965,
+      "grad_norm": 0.43832552433013916,
+      "learning_rate": 4.6779272978442904e-05,
+      "loss": 1.5165,
+      "num_input_tokens_seen": 153488160,
+      "step": 3890
+    },
+    {
+      "epoch": 0.18897981186817447,
+      "grad_norm": 0.4259849488735199,
+      "learning_rate": 4.6759981803296666e-05,
+      "loss": 1.4512,
+      "num_input_tokens_seen": 153864880,
+      "step": 3900
+    },
+    {
+      "epoch": 0.18946437548834927,
+      "grad_norm": 0.46592122316360474,
+      "learning_rate": 4.674063702871617e-05,
+      "loss": 1.4473,
+      "num_input_tokens_seen": 154258796,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1899489391085241,
+      "grad_norm": 0.4338749349117279,
+      "learning_rate": 4.672123870235169e-05,
+      "loss": 1.5149,
+      "num_input_tokens_seen": 154659628,
+      "step": 3920
+    },
+    {
+      "epoch": 0.19043350272869888,
+      "grad_norm": 0.4241959750652313,
+      "learning_rate": 4.6701786871985395e-05,
+      "loss": 1.546,
+      "num_input_tokens_seen": 155065544,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1909180663488737,
+      "grad_norm": 0.39792391657829285,
+      "learning_rate": 4.6682281585531264e-05,
+      "loss": 1.5081,
+      "num_input_tokens_seen": 155439252,
+      "step": 3940
+    },
+    {
+      "epoch": 0.1914026299690485,
+      "grad_norm": 0.4416780173778534,
+      "learning_rate": 4.666272289103492e-05,
+      "loss": 1.5018,
+      "num_input_tokens_seen": 155848612,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1918871935892233,
+      "grad_norm": 0.42491331696510315,
+      "learning_rate": 4.664311083667359e-05,
+      "loss": 1.5042,
+      "num_input_tokens_seen": 156231544,
+      "step": 3960
+    },
+    {
+      "epoch": 0.19237175720939811,
+      "grad_norm": 0.4376972019672394,
+      "learning_rate": 4.6623445470755875e-05,
+      "loss": 1.5095,
+      "num_input_tokens_seen": 156622212,
+      "step": 3970
+    },
+    {
+      "epoch": 0.1928563208295729,
+      "grad_norm": 0.4348180890083313,
+      "learning_rate": 4.660372684172176e-05,
+      "loss": 1.4884,
+      "num_input_tokens_seen": 157011628,
+      "step": 3980
+    },
+    {
+      "epoch": 0.19334088444974773,
+      "grad_norm": 0.4411153793334961,
+      "learning_rate": 4.658395499814238e-05,
+      "loss": 1.5275,
+      "num_input_tokens_seen": 157391712,
+      "step": 3990
+    },
+    {
+      "epoch": 0.19382544806992252,
+      "grad_norm": 0.4086403548717499,
+      "learning_rate": 4.656412998871996e-05,
+      "loss": 1.4957,
+      "num_input_tokens_seen": 157778628,
+      "step": 4000
+    },
+    {
+      "epoch": 0.19382544806992252,
+      "eval_loss": 1.6085283756256104,
+      "eval_runtime": 3.8392,
+      "eval_samples_per_second": 39.07,
+      "eval_steps_per_second": 4.949,
+      "num_input_tokens_seen": 157778628,
+      "step": 4000
+    },
+    {
+      "epoch": 0.19431001169009735,
+      "grad_norm": 0.4232029914855957,
+      "learning_rate": 4.65442518622877e-05,
+      "loss": 1.4919,
+      "num_input_tokens_seen": 158177392,
+      "step": 4010
+    },
+    {
+      "epoch": 0.19479457531027214,
+      "grad_norm": 0.4746312201023102,
+      "learning_rate": 4.652432066780962e-05,
+      "loss": 1.4801,
+      "num_input_tokens_seen": 158587736,
+      "step": 4020
+    },
+    {
+      "epoch": 0.19527913893044696,
+      "grad_norm": 0.4055451452732086,
+      "learning_rate": 4.650433645438047e-05,
+      "loss": 1.5079,
+      "num_input_tokens_seen": 158997624,
+      "step": 4030
+    },
+    {
+      "epoch": 0.19576370255062175,
+      "grad_norm": 0.47638824582099915,
+      "learning_rate": 4.648429927122558e-05,
+      "loss": 1.5269,
+      "num_input_tokens_seen": 159406580,
+      "step": 4040
+    },
+    {
+      "epoch": 0.19624826617079658,
+      "grad_norm": 0.426217257976532,
+      "learning_rate": 4.646420916770078e-05,
+      "loss": 1.4881,
+      "num_input_tokens_seen": 159776464,
+      "step": 4050
+    },
+    {
+      "epoch": 0.19673282979097137,
+      "grad_norm": 0.3979707360267639,
+      "learning_rate": 4.644406619329223e-05,
+      "loss": 1.5584,
+      "num_input_tokens_seen": 160136328,
+      "step": 4060
+    },
+    {
+      "epoch": 0.19721739341114616,
+      "grad_norm": 0.37950029969215393,
+      "learning_rate": 4.642387039761635e-05,
+      "loss": 1.5176,
+      "num_input_tokens_seen": 160532608,
+      "step": 4070
+    },
+    {
+      "epoch": 0.19770195703132099,
+      "grad_norm": 0.4001169502735138,
+      "learning_rate": 4.6403621830419644e-05,
+      "loss": 1.4584,
+      "num_input_tokens_seen": 160930792,
+      "step": 4080
+    },
+    {
+      "epoch": 0.19818652065149578,
+      "grad_norm": 0.4448148012161255,
+      "learning_rate": 4.6383320541578604e-05,
+      "loss": 1.5309,
+      "num_input_tokens_seen": 161311248,
+      "step": 4090
+    },
+    {
+      "epoch": 0.1986710842716706,
+      "grad_norm": 0.4266856014728546,
+      "learning_rate": 4.63629665810996e-05,
+      "loss": 1.4883,
+      "num_input_tokens_seen": 161718084,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1991556478918454,
+      "grad_norm": 0.4387410581111908,
+      "learning_rate": 4.6342559999118736e-05,
+      "loss": 1.4552,
+      "num_input_tokens_seen": 162101052,
+      "step": 4110
+    },
+    {
+      "epoch": 0.19964021151202022,
+      "grad_norm": 0.3797076344490051,
+      "learning_rate": 4.632210084590175e-05,
+      "loss": 1.5668,
+      "num_input_tokens_seen": 162513632,
+      "step": 4120
+    },
+    {
+      "epoch": 0.200124775132195,
+      "grad_norm": 0.4161923825740814,
+      "learning_rate": 4.630158917184385e-05,
+      "loss": 1.5182,
+      "num_input_tokens_seen": 162905012,
+      "step": 4130
+    },
+    {
+      "epoch": 0.20060933875236983,
+      "grad_norm": 0.41969460248947144,
+      "learning_rate": 4.6281025027469625e-05,
+      "loss": 1.4945,
+      "num_input_tokens_seen": 163281652,
+      "step": 4140
+    },
+    {
+      "epoch": 0.20109390237254463,
+      "grad_norm": 0.4282332956790924,
+      "learning_rate": 4.626040846343291e-05,
+      "loss": 1.5109,
+      "num_input_tokens_seen": 163652028,
+      "step": 4150
+    },
+    {
+      "epoch": 0.20157846599271942,
+      "grad_norm": 0.42534980177879333,
+      "learning_rate": 4.623973953051667e-05,
+      "loss": 1.4954,
+      "num_input_tokens_seen": 164028336,
+      "step": 4160
+    },
+    {
+      "epoch": 0.20206302961289424,
+      "grad_norm": 0.4472944736480713,
+      "learning_rate": 4.621901827963283e-05,
+      "loss": 1.4857,
+      "num_input_tokens_seen": 164392624,
+      "step": 4170
+    },
+    {
+      "epoch": 0.20254759323306903,
+      "grad_norm": 0.42009228467941284,
+      "learning_rate": 4.619824476182223e-05,
+      "loss": 1.5229,
+      "num_input_tokens_seen": 164753004,
+      "step": 4180
+    },
+    {
+      "epoch": 0.20303215685324386,
+      "grad_norm": 0.5307909250259399,
+      "learning_rate": 4.617741902825443e-05,
+      "loss": 1.4997,
+      "num_input_tokens_seen": 165122832,
+      "step": 4190
+    },
+    {
+      "epoch": 0.20351672047341865,
+      "grad_norm": 0.42677533626556396,
+      "learning_rate": 4.615654113022761e-05,
+      "loss": 1.4825,
+      "num_input_tokens_seen": 165469868,
+      "step": 4200
+    },
+    {
+      "epoch": 0.20400128409359347,
+      "grad_norm": 0.38291576504707336,
+      "learning_rate": 4.6135611119168465e-05,
+      "loss": 1.5681,
+      "num_input_tokens_seen": 165869628,
+      "step": 4210
+    },
+    {
+      "epoch": 0.20448584771376827,
+      "grad_norm": 0.40809643268585205,
+      "learning_rate": 4.611462904663201e-05,
+      "loss": 1.4854,
+      "num_input_tokens_seen": 166262208,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2049704113339431,
+      "grad_norm": 0.43991562724113464,
+      "learning_rate": 4.6093594964301534e-05,
+      "loss": 1.4577,
+      "num_input_tokens_seen": 166640892,
+      "step": 4230
+    },
+    {
+      "epoch": 0.20545497495411788,
+      "grad_norm": 0.432851642370224,
+      "learning_rate": 4.607250892398843e-05,
+      "loss": 1.4723,
+      "num_input_tokens_seen": 167080484,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2059395385742927,
+      "grad_norm": 0.41228345036506653,
+      "learning_rate": 4.605137097763207e-05,
+      "loss": 1.4305,
+      "num_input_tokens_seen": 167486976,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2064241021944675,
+      "grad_norm": 0.4661470949649811,
+      "learning_rate": 4.603018117729968e-05,
+      "loss": 1.4348,
+      "num_input_tokens_seen": 167871328,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2069086658146423,
+      "grad_norm": 0.39085134863853455,
+      "learning_rate": 4.600893957518622e-05,
+      "loss": 1.4937,
+      "num_input_tokens_seen": 168255252,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2073932294348171,
+      "grad_norm": 0.3891197443008423,
+      "learning_rate": 4.5987646223614244e-05,
+      "loss": 1.4864,
+      "num_input_tokens_seen": 168634480,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2078777930549919,
+      "grad_norm": 0.40408509969711304,
+      "learning_rate": 4.5966301175033785e-05,
+      "loss": 1.4056,
+      "num_input_tokens_seen": 169049616,
+      "step": 4290
+    },
+    {
+      "epoch": 0.20836235667516673,
+      "grad_norm": 0.4421607553958893,
+      "learning_rate": 4.59449044820222e-05,
+      "loss": 1.491,
+      "num_input_tokens_seen": 169439824,
+      "step": 4300
+    },
+    {
+      "epoch": 0.20884692029534152,
+      "grad_norm": 0.3921188712120056,
+      "learning_rate": 4.5923456197284065e-05,
+      "loss": 1.4783,
+      "num_input_tokens_seen": 169796656,
+      "step": 4310
+    },
+    {
+      "epoch": 0.20933148391551634,
+      "grad_norm": 0.4339124262332916,
+      "learning_rate": 4.590195637365105e-05,
+      "loss": 1.4668,
+      "num_input_tokens_seen": 170178796,
+      "step": 4320
+    },
+    {
+      "epoch": 0.20981604753569114,
+      "grad_norm": 0.41062822937965393,
+      "learning_rate": 4.588040506408176e-05,
+      "loss": 1.4628,
+      "num_input_tokens_seen": 170578188,
+      "step": 4330
+    },
+    {
+      "epoch": 0.21030061115586596,
+      "grad_norm": 0.40424343943595886,
+      "learning_rate": 4.5858802321661616e-05,
+      "loss": 1.4596,
+      "num_input_tokens_seen": 170964776,
+      "step": 4340
+    },
+    {
+      "epoch": 0.21078517477604075,
+      "grad_norm": 0.4466400742530823,
+      "learning_rate": 4.5837148199602745e-05,
+      "loss": 1.5198,
+      "num_input_tokens_seen": 171343616,
+      "step": 4350
+    },
+    {
+      "epoch": 0.21126973839621555,
+      "grad_norm": 0.45615527033805847,
+      "learning_rate": 4.581544275124383e-05,
+      "loss": 1.4517,
+      "num_input_tokens_seen": 171718620,
+      "step": 4360
+    },
+    {
+      "epoch": 0.21175430201639037,
+      "grad_norm": 0.45989635586738586,
+      "learning_rate": 4.5793686030049974e-05,
+      "loss": 1.4469,
+      "num_input_tokens_seen": 172101080,
+      "step": 4370
+    },
+    {
+      "epoch": 0.21223886563656516,
+      "grad_norm": 0.3968257009983063,
+      "learning_rate": 4.577187808961258e-05,
+      "loss": 1.4241,
+      "num_input_tokens_seen": 172502216,
+      "step": 4380
+    },
+    {
+      "epoch": 0.21272342925673998,
+      "grad_norm": 0.4084954559803009,
+      "learning_rate": 4.5750018983649214e-05,
+      "loss": 1.4621,
+      "num_input_tokens_seen": 172905268,
+      "step": 4390
+    },
+    {
+      "epoch": 0.21320799287691478,
+      "grad_norm": 0.42288362979888916,
+      "learning_rate": 4.5728108766003474e-05,
+      "loss": 1.4642,
+      "num_input_tokens_seen": 173303252,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2136925564970896,
+      "grad_norm": 0.40454548597335815,
+      "learning_rate": 4.570614749064486e-05,
+      "loss": 1.4608,
+      "num_input_tokens_seen": 173696264,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2141771201172644,
+      "grad_norm": 0.416332870721817,
+      "learning_rate": 4.568413521166863e-05,
+      "loss": 1.4757,
+      "num_input_tokens_seen": 174067360,
+      "step": 4420
+    },
+    {
+      "epoch": 0.2146616837374392,
+      "grad_norm": 0.4173552095890045,
+      "learning_rate": 4.566207198329568e-05,
+      "loss": 1.4377,
+      "num_input_tokens_seen": 174474464,
+      "step": 4430
+    },
+    {
+      "epoch": 0.215146247357614,
+      "grad_norm": 0.4277855455875397,
+      "learning_rate": 4.563995785987241e-05,
+      "loss": 1.4673,
+      "num_input_tokens_seen": 174876064,
+      "step": 4440
+    },
+    {
+      "epoch": 0.21563081097778883,
+      "grad_norm": 0.45221036672592163,
+      "learning_rate": 4.561779289587058e-05,
+      "loss": 1.4808,
+      "num_input_tokens_seen": 175281952,
+      "step": 4450
+    },
+    {
+      "epoch": 0.21611537459796362,
+      "grad_norm": 0.3993636667728424,
+      "learning_rate": 4.559557714588717e-05,
+      "loss": 1.5006,
+      "num_input_tokens_seen": 175657752,
+      "step": 4460
+    },
+    {
+      "epoch": 0.21659993821813842,
+      "grad_norm": 0.42281702160835266,
+      "learning_rate": 4.557331066464428e-05,
+      "loss": 1.4489,
+      "num_input_tokens_seen": 176038544,
+      "step": 4470
+    },
+    {
+      "epoch": 0.21708450183831324,
+      "grad_norm": 0.4503469467163086,
+      "learning_rate": 4.555099350698895e-05,
+      "loss": 1.5062,
+      "num_input_tokens_seen": 176415768,
+      "step": 4480
+    },
+    {
+      "epoch": 0.21756906545848803,
+      "grad_norm": 0.4171935021877289,
+      "learning_rate": 4.5528625727893065e-05,
+      "loss": 1.4701,
+      "num_input_tokens_seen": 176823204,
+      "step": 4490
+    },
+    {
+      "epoch": 0.21805362907866285,
+      "grad_norm": 0.41661641001701355,
+      "learning_rate": 4.5506207382453184e-05,
+      "loss": 1.485,
+      "num_input_tokens_seen": 177218572,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21853819269883765,
+      "grad_norm": 0.4115769863128662,
+      "learning_rate": 4.548373852589044e-05,
+      "loss": 1.4757,
+      "num_input_tokens_seen": 177591148,
+      "step": 4510
+    },
+    {
+      "epoch": 0.21902275631901247,
+      "grad_norm": 0.42897987365722656,
+      "learning_rate": 4.5461219213550365e-05,
+      "loss": 1.4452,
+      "num_input_tokens_seen": 177978272,
+      "step": 4520
+    },
+    {
+      "epoch": 0.21950731993918726,
+      "grad_norm": 0.4455084204673767,
+      "learning_rate": 4.5438649500902796e-05,
+      "loss": 1.5184,
+      "num_input_tokens_seen": 178398380,
+      "step": 4530
+    },
+    {
+      "epoch": 0.21999188355936208,
+      "grad_norm": 0.3757345974445343,
+      "learning_rate": 4.5416029443541694e-05,
+      "loss": 1.4825,
+      "num_input_tokens_seen": 178807360,
+      "step": 4540
+    },
+    {
+      "epoch": 0.22047644717953688,
+      "grad_norm": 0.4476206302642822,
+      "learning_rate": 4.539335909718504e-05,
+      "loss": 1.5054,
+      "num_input_tokens_seen": 179185000,
+      "step": 4550
+    },
+    {
+      "epoch": 0.22096101079971167,
+      "grad_norm": 0.3936362862586975,
+      "learning_rate": 4.5370638517674716e-05,
+      "loss": 1.477,
+      "num_input_tokens_seen": 179593392,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2214455744198865,
+      "grad_norm": 0.41432103514671326,
+      "learning_rate": 4.534786776097627e-05,
+      "loss": 1.4547,
+      "num_input_tokens_seen": 179985908,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2219301380400613,
+      "grad_norm": 0.3996790051460266,
+      "learning_rate": 4.532504688317891e-05,
+      "loss": 1.4691,
+      "num_input_tokens_seen": 180414548,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2224147016602361,
+      "grad_norm": 0.4232152998447418,
+      "learning_rate": 4.530217594049529e-05,
+      "loss": 1.4798,
+      "num_input_tokens_seen": 180809424,
+      "step": 4590
+    },
+    {
+      "epoch": 0.2228992652804109,
+      "grad_norm": 0.4331175684928894,
+      "learning_rate": 4.527925498926136e-05,
+      "loss": 1.4674,
+      "num_input_tokens_seen": 181195556,
+      "step": 4600
+    },
+    {
+      "epoch": 0.22338382890058572,
+      "grad_norm": 0.4475395083427429,
+      "learning_rate": 4.5256284085936275e-05,
+      "loss": 1.4507,
+      "num_input_tokens_seen": 181583096,
+      "step": 4610
+    },
+    {
+      "epoch": 0.22386839252076052,
+      "grad_norm": 0.42866751551628113,
+      "learning_rate": 4.523326328710222e-05,
+      "loss": 1.4855,
+      "num_input_tokens_seen": 181999512,
+      "step": 4620
+    },
+    {
+      "epoch": 0.22435295614093534,
+      "grad_norm": 0.41574034094810486,
+      "learning_rate": 4.5210192649464296e-05,
+      "loss": 1.464,
+      "num_input_tokens_seen": 182381672,
+      "step": 4630
+    },
+    {
+      "epoch": 0.22483751976111013,
+      "grad_norm": 0.39740654826164246,
+      "learning_rate": 4.518707222985035e-05,
+      "loss": 1.4681,
+      "num_input_tokens_seen": 182796928,
+      "step": 4640
+    },
+    {
+      "epoch": 0.22532208338128495,
+      "grad_norm": 0.41937920451164246,
+      "learning_rate": 4.5163902085210866e-05,
+      "loss": 1.4463,
+      "num_input_tokens_seen": 183201244,
+      "step": 4650
+    },
+    {
+      "epoch": 0.22580664700145975,
+      "grad_norm": 0.39213669300079346,
+      "learning_rate": 4.514068227261882e-05,
+      "loss": 1.423,
+      "num_input_tokens_seen": 183580392,
+      "step": 4660
+    },
+    {
+      "epoch": 0.22629121062163454,
+      "grad_norm": 0.3789583742618561,
+      "learning_rate": 4.511741284926949e-05,
+      "loss": 1.4835,
+      "num_input_tokens_seen": 183979192,
+      "step": 4670
+    },
+    {
+      "epoch": 0.22677577424180936,
+      "grad_norm": 0.4139784872531891,
+      "learning_rate": 4.5094093872480405e-05,
+      "loss": 1.51,
+      "num_input_tokens_seen": 184365440,
+      "step": 4680
+    },
+    {
+      "epoch": 0.22726033786198416,
+      "grad_norm": 0.414145290851593,
+      "learning_rate": 4.507072539969114e-05,
+      "loss": 1.4815,
+      "num_input_tokens_seen": 184735664,
+      "step": 4690
+    },
+    {
+      "epoch": 0.22774490148215898,
+      "grad_norm": 0.4188004434108734,
+      "learning_rate": 4.5047307488463156e-05,
+      "loss": 1.4549,
+      "num_input_tokens_seen": 185132600,
+      "step": 4700
+    },
+    {
+      "epoch": 0.22822946510233377,
+      "grad_norm": 0.4184124767780304,
+      "learning_rate": 4.502384019647974e-05,
+      "loss": 1.5019,
+      "num_input_tokens_seen": 185517396,
+      "step": 4710
+    },
+    {
+      "epoch": 0.2287140287225086,
+      "grad_norm": 0.3586943745613098,
+      "learning_rate": 4.5000323581545784e-05,
+      "loss": 1.4141,
+      "num_input_tokens_seen": 185908324,
+      "step": 4720
+    },
+    {
+      "epoch": 0.2291985923426834,
+      "grad_norm": 0.40754491090774536,
+      "learning_rate": 4.497675770158768e-05,
+      "loss": 1.4299,
+      "num_input_tokens_seen": 186287004,
+      "step": 4730
+    },
+    {
+      "epoch": 0.2296831559628582,
+      "grad_norm": 0.38670873641967773,
+      "learning_rate": 4.4953142614653175e-05,
+      "loss": 1.4442,
+      "num_input_tokens_seen": 186713356,
+      "step": 4740
+    },
+    {
+      "epoch": 0.230167719583033,
+      "grad_norm": 0.40814658999443054,
+      "learning_rate": 4.4929478378911214e-05,
+      "loss": 1.5235,
+      "num_input_tokens_seen": 187111252,
+      "step": 4750
+    },
+    {
+      "epoch": 0.2306522832032078,
+      "grad_norm": 0.4206823408603668,
+      "learning_rate": 4.490576505265182e-05,
+      "loss": 1.4814,
+      "num_input_tokens_seen": 187499448,
+      "step": 4760
+    },
+    {
+      "epoch": 0.23113684682338262,
+      "grad_norm": 0.5019489526748657,
+      "learning_rate": 4.488200269428592e-05,
+      "loss": 1.424,
+      "num_input_tokens_seen": 187884064,
+      "step": 4770
+    },
+    {
+      "epoch": 0.2316214104435574,
+      "grad_norm": 0.39812424778938293,
+      "learning_rate": 4.4858191362345224e-05,
+      "loss": 1.4475,
+      "num_input_tokens_seen": 188262996,
+      "step": 4780
+    },
+    {
+      "epoch": 0.23210597406373223,
+      "grad_norm": 0.43520602583885193,
+      "learning_rate": 4.483433111548208e-05,
+      "loss": 1.4525,
+      "num_input_tokens_seen": 188660764,
+      "step": 4790
+    },
+    {
+      "epoch": 0.23259053768390703,
+      "grad_norm": 0.40959790349006653,
+      "learning_rate": 4.4810422012469315e-05,
+      "loss": 1.4299,
+      "num_input_tokens_seen": 189046960,
+      "step": 4800
+    },
+    {
+      "epoch": 0.23307510130408185,
+      "grad_norm": 0.42006418108940125,
+      "learning_rate": 4.478646411220011e-05,
+      "loss": 1.5022,
+      "num_input_tokens_seen": 189438452,
+      "step": 4810
+    },
+    {
+      "epoch": 0.23355966492425664,
+      "grad_norm": 0.410160094499588,
+      "learning_rate": 4.476245747368783e-05,
+      "loss": 1.4756,
+      "num_input_tokens_seen": 189823928,
+      "step": 4820
+    },
+    {
+      "epoch": 0.23404422854443147,
+      "grad_norm": 0.4331664741039276,
+      "learning_rate": 4.473840215606589e-05,
+      "loss": 1.4821,
+      "num_input_tokens_seen": 190244028,
+      "step": 4830
+    },
+    {
+      "epoch": 0.23452879216460626,
+      "grad_norm": 0.4130837619304657,
+      "learning_rate": 4.4714298218587654e-05,
+      "loss": 1.4746,
+      "num_input_tokens_seen": 190647640,
+      "step": 4840
+    },
+    {
+      "epoch": 0.23501335578478108,
+      "grad_norm": 0.392727792263031,
+      "learning_rate": 4.469014572062618e-05,
+      "loss": 1.4636,
+      "num_input_tokens_seen": 191073988,
+      "step": 4850
+    },
+    {
+      "epoch": 0.23549791940495587,
+      "grad_norm": 0.4510425627231598,
+      "learning_rate": 4.466594472167419e-05,
+      "loss": 1.4326,
+      "num_input_tokens_seen": 191476524,
+      "step": 4860
+    },
+    {
+      "epoch": 0.23598248302513067,
+      "grad_norm": 0.40726661682128906,
+      "learning_rate": 4.4641695281343867e-05,
+      "loss": 1.4472,
+      "num_input_tokens_seen": 191852232,
+      "step": 4870
+    },
+    {
+      "epoch": 0.2364670466453055,
+      "grad_norm": 0.46587809920310974,
+      "learning_rate": 4.461739745936669e-05,
+      "loss": 1.4811,
+      "num_input_tokens_seen": 192231868,
+      "step": 4880
+    },
+    {
+      "epoch": 0.23695161026548028,
+      "grad_norm": 0.37719038128852844,
+      "learning_rate": 4.459305131559334e-05,
+      "loss": 1.4508,
+      "num_input_tokens_seen": 192629400,
+      "step": 4890
+    },
+    {
+      "epoch": 0.2374361738856551,
+      "grad_norm": 0.4082610309123993,
+      "learning_rate": 4.4568656909993515e-05,
+      "loss": 1.4626,
+      "num_input_tokens_seen": 193036008,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2379207375058299,
+      "grad_norm": 0.4302626848220825,
+      "learning_rate": 4.454421430265579e-05,
+      "loss": 1.4371,
+      "num_input_tokens_seen": 193442156,
+      "step": 4910
+    },
+    {
+      "epoch": 0.23840530112600472,
+      "grad_norm": 0.42628800868988037,
+      "learning_rate": 4.451972355378748e-05,
+      "loss": 1.4682,
+      "num_input_tokens_seen": 193831248,
+      "step": 4920
+    },
+    {
+      "epoch": 0.23888986474617951,
+      "grad_norm": 0.3756248950958252,
+      "learning_rate": 4.449518472371447e-05,
+      "loss": 1.4266,
+      "num_input_tokens_seen": 194205320,
+      "step": 4930
+    },
+    {
+      "epoch": 0.23937442836635434,
+      "grad_norm": 0.41749587655067444,
+      "learning_rate": 4.447059787288107e-05,
+      "loss": 1.4397,
+      "num_input_tokens_seen": 194598988,
+      "step": 4940
+    },
+    {
+      "epoch": 0.23985899198652913,
+      "grad_norm": 0.39159849286079407,
+      "learning_rate": 4.444596306184992e-05,
+      "loss": 1.3895,
+      "num_input_tokens_seen": 195006708,
+      "step": 4950
+    },
+    {
+      "epoch": 0.24034355560670395,
+      "grad_norm": 0.44320693612098694,
+      "learning_rate": 4.4421280351301744e-05,
+      "loss": 1.4094,
+      "num_input_tokens_seen": 195416052,
+      "step": 4960
+    },
+    {
+      "epoch": 0.24082811922687875,
+      "grad_norm": 0.4559500813484192,
+      "learning_rate": 4.439654980203527e-05,
+      "loss": 1.4498,
+      "num_input_tokens_seen": 195806840,
+      "step": 4970
+    },
+    {
+      "epoch": 0.24131268284705354,
+      "grad_norm": 0.4020100235939026,
+      "learning_rate": 4.437177147496709e-05,
+      "loss": 1.4628,
+      "num_input_tokens_seen": 196212692,
+      "step": 4980
+    },
+    {
+      "epoch": 0.24179724646722836,
+      "grad_norm": 0.39954742789268494,
+      "learning_rate": 4.434694543113145e-05,
+      "loss": 1.4211,
+      "num_input_tokens_seen": 196623736,
+      "step": 4990
+    },
+    {
+      "epoch": 0.24228181008740315,
+      "grad_norm": 0.4263516366481781,
+      "learning_rate": 4.4322071731680146e-05,
+      "loss": 1.453,
+      "num_input_tokens_seen": 197006512,
+      "step": 5000
+    },
+    {
+      "epoch": 0.24276637370757798,
+      "grad_norm": 0.4143199026584625,
+      "learning_rate": 4.429715043788235e-05,
+      "loss": 1.4952,
+      "num_input_tokens_seen": 197391012,
+      "step": 5010
+    },
+    {
+      "epoch": 0.24325093732775277,
+      "grad_norm": 0.4607667028903961,
+      "learning_rate": 4.427218161112449e-05,
+      "loss": 1.4704,
+      "num_input_tokens_seen": 197751048,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2437355009479276,
+      "grad_norm": 0.4273362159729004,
+      "learning_rate": 4.4247165312910034e-05,
+      "loss": 1.4698,
+      "num_input_tokens_seen": 198159508,
+      "step": 5030
+    },
+    {
+      "epoch": 0.24422006456810239,
+      "grad_norm": 0.39402100443840027,
+      "learning_rate": 4.4222101604859445e-05,
+      "loss": 1.4527,
+      "num_input_tokens_seen": 198550724,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2447046281882772,
+      "grad_norm": 0.42816653847694397,
+      "learning_rate": 4.419699054870992e-05,
+      "loss": 1.4948,
+      "num_input_tokens_seen": 198905612,
+      "step": 5050
+    },
+    {
+      "epoch": 0.245189191808452,
+      "grad_norm": 0.4140118360519409,
+      "learning_rate": 4.417183220631528e-05,
+      "loss": 1.4628,
+      "num_input_tokens_seen": 199308252,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2456737554286268,
+      "grad_norm": 0.3967956304550171,
+      "learning_rate": 4.4146626639645874e-05,
+      "loss": 1.4542,
+      "num_input_tokens_seen": 199703748,
+      "step": 5070
+    },
+    {
+      "epoch": 0.24615831904880162,
+      "grad_norm": 0.4145415425300598,
+      "learning_rate": 4.412137391078832e-05,
+      "loss": 1.4535,
+      "num_input_tokens_seen": 200118988,
+      "step": 5080
+    },
+    {
+      "epoch": 0.2466428826689764,
+      "grad_norm": 0.39303335547447205,
+      "learning_rate": 4.4096074081945425e-05,
+      "loss": 1.4249,
+      "num_input_tokens_seen": 200527544,
+      "step": 5090
+    },
+    {
+      "epoch": 0.24712744628915123,
+      "grad_norm": 0.4353494942188263,
+      "learning_rate": 4.4070727215436025e-05,
+      "loss": 1.4521,
+      "num_input_tokens_seen": 200930660,
+      "step": 5100
+    },
+    {
+      "epoch": 0.24761200990932603,
+      "grad_norm": 0.4420648515224457,
+      "learning_rate": 4.4045333373694795e-05,
+      "loss": 1.4591,
+      "num_input_tokens_seen": 201323040,
+      "step": 5110
+    },
+    {
+      "epoch": 0.24809657352950085,
+      "grad_norm": 0.40382125973701477,
+      "learning_rate": 4.4019892619272144e-05,
+      "loss": 1.4383,
+      "num_input_tokens_seen": 201734800,
+      "step": 5120
+    },
+    {
+      "epoch": 0.24858113714967564,
+      "grad_norm": 0.4114893078804016,
+      "learning_rate": 4.399440501483403e-05,
+      "loss": 1.415,
+      "num_input_tokens_seen": 202108596,
+      "step": 5130
+    },
+    {
+      "epoch": 0.24906570076985046,
+      "grad_norm": 0.4096544086933136,
+      "learning_rate": 4.3968870623161804e-05,
+      "loss": 1.4244,
+      "num_input_tokens_seen": 202487992,
+      "step": 5140
+    },
+    {
+      "epoch": 0.24955026439002526,
+      "grad_norm": 0.40928128361701965,
+      "learning_rate": 4.3943289507152066e-05,
+      "loss": 1.3944,
+      "num_input_tokens_seen": 202886772,
+      "step": 5150
+    },
+    {
+      "epoch": 0.25003482801020005,
+      "grad_norm": 0.4127601087093353,
+      "learning_rate": 4.391766172981653e-05,
+      "loss": 1.4302,
+      "num_input_tokens_seen": 203310128,
+      "step": 5160
+    },
+    {
+      "epoch": 0.25051939163037484,
+      "grad_norm": 0.4010636508464813,
+      "learning_rate": 4.38919873542818e-05,
+      "loss": 1.4494,
+      "num_input_tokens_seen": 203716500,
+      "step": 5170
+    },
+    {
+      "epoch": 0.2510039552505497,
+      "grad_norm": 0.479033499956131,
+      "learning_rate": 4.38662664437893e-05,
+      "loss": 1.466,
+      "num_input_tokens_seen": 204104804,
+      "step": 5180
+    },
+    {
+      "epoch": 0.2514885188707245,
+      "grad_norm": 0.4351840317249298,
+      "learning_rate": 4.384049906169509e-05,
+      "loss": 1.4538,
+      "num_input_tokens_seen": 204478748,
+      "step": 5190
+    },
+    {
+      "epoch": 0.2519730824908993,
+      "grad_norm": 0.45871713757514954,
+      "learning_rate": 4.381468527146965e-05,
+      "loss": 1.4958,
+      "num_input_tokens_seen": 204854344,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2524576461110741,
+      "grad_norm": 0.42551910877227783,
+      "learning_rate": 4.378882513669782e-05,
+      "loss": 1.4588,
+      "num_input_tokens_seen": 205245060,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2529422097312489,
+      "grad_norm": 0.39652207493782043,
+      "learning_rate": 4.376291872107856e-05,
+      "loss": 1.4118,
+      "num_input_tokens_seen": 205643280,
+      "step": 5220
+    },
+    {
+      "epoch": 0.2534267733514237,
+      "grad_norm": 0.4601559042930603,
+      "learning_rate": 4.373696608842486e-05,
+      "loss": 1.4046,
+      "num_input_tokens_seen": 206032812,
+      "step": 5230
+    },
+    {
+      "epoch": 0.2539113369715985,
+      "grad_norm": 0.4585094153881073,
+      "learning_rate": 4.371096730266354e-05,
+      "loss": 1.4753,
+      "num_input_tokens_seen": 206405924,
+      "step": 5240
+    },
+    {
+      "epoch": 0.2543959005917733,
+      "grad_norm": 0.39976418018341064,
+      "learning_rate": 4.3684922427835094e-05,
+      "loss": 1.5263,
+      "num_input_tokens_seen": 206774840,
+      "step": 5250
+    },
+    {
+      "epoch": 0.25488046421194815,
+      "grad_norm": 0.41982370615005493,
+      "learning_rate": 4.365883152809356e-05,
+      "loss": 1.4208,
+      "num_input_tokens_seen": 207172068,
+      "step": 5260
+    },
+    {
+      "epoch": 0.25536502783212295,
+      "grad_norm": 0.4455938935279846,
+      "learning_rate": 4.3632694667706345e-05,
+      "loss": 1.4489,
+      "num_input_tokens_seen": 207533668,
+      "step": 5270
+    },
+    {
+      "epoch": 0.25584959145229774,
+      "grad_norm": 0.46062058210372925,
+      "learning_rate": 4.360651191105405e-05,
+      "loss": 1.4252,
+      "num_input_tokens_seen": 207962000,
+      "step": 5280
+    },
+    {
+      "epoch": 0.25633415507247254,
+      "grad_norm": 0.46039023995399475,
+      "learning_rate": 4.358028332263034e-05,
+      "loss": 1.4279,
+      "num_input_tokens_seen": 208396128,
+      "step": 5290
+    },
+    {
+      "epoch": 0.25681871869264733,
+      "grad_norm": 0.4030589759349823,
+      "learning_rate": 4.355400896704177e-05,
+      "loss": 1.4416,
+      "num_input_tokens_seen": 208771020,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2573032823128222,
+      "grad_norm": 0.46707648038864136,
+      "learning_rate": 4.3527688909007645e-05,
+      "loss": 1.4595,
+      "num_input_tokens_seen": 209175816,
+      "step": 5310
+    },
+    {
+      "epoch": 0.257787845932997,
+      "grad_norm": 0.39080390334129333,
+      "learning_rate": 4.350132321335982e-05,
+      "loss": 1.4442,
+      "num_input_tokens_seen": 209573164,
+      "step": 5320
+    },
+    {
+      "epoch": 0.25827240955317177,
+      "grad_norm": 0.3796825706958771,
+      "learning_rate": 4.347491194504257e-05,
+      "loss": 1.3894,
+      "num_input_tokens_seen": 209973236,
+      "step": 5330
+    },
+    {
+      "epoch": 0.25875697317334656,
+      "grad_norm": 0.413033127784729,
+      "learning_rate": 4.344845516911244e-05,
+      "loss": 1.4657,
+      "num_input_tokens_seen": 210345012,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2592415367935214,
+      "grad_norm": 0.4229408800601959,
+      "learning_rate": 4.342195295073806e-05,
+      "loss": 1.4725,
+      "num_input_tokens_seen": 210769268,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2597261004136962,
+      "grad_norm": 0.39224693179130554,
+      "learning_rate": 4.339540535519999e-05,
+      "loss": 1.4123,
+      "num_input_tokens_seen": 211153696,
+      "step": 5360
+    },
+    {
+      "epoch": 0.260210664033871,
+      "grad_norm": 0.4225060045719147,
+      "learning_rate": 4.3368812447890575e-05,
+      "loss": 1.4308,
+      "num_input_tokens_seen": 211552748,
+      "step": 5370
+    },
+    {
+      "epoch": 0.2606952276540458,
+      "grad_norm": 0.43410396575927734,
+      "learning_rate": 4.334217429431376e-05,
+      "loss": 1.4755,
+      "num_input_tokens_seen": 211943156,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2611797912742206,
+      "grad_norm": 0.3979720175266266,
+      "learning_rate": 4.331549096008496e-05,
+      "loss": 1.4245,
+      "num_input_tokens_seen": 212315716,
+      "step": 5390
+    },
+    {
+      "epoch": 0.26166435489439543,
+      "grad_norm": 0.3867861032485962,
+      "learning_rate": 4.328876251093086e-05,
+      "loss": 1.4195,
+      "num_input_tokens_seen": 212702260,
+      "step": 5400
+    },
+    {
+      "epoch": 0.26214891851457023,
+      "grad_norm": 0.4254813492298126,
+      "learning_rate": 4.326198901268926e-05,
+      "loss": 1.4395,
+      "num_input_tokens_seen": 213112036,
+      "step": 5410
+    },
+    {
+      "epoch": 0.262633482134745,
+      "grad_norm": 0.37997907400131226,
+      "learning_rate": 4.323517053130898e-05,
+      "loss": 1.4331,
+      "num_input_tokens_seen": 213509980,
+      "step": 5420
+    },
+    {
+      "epoch": 0.2631180457549198,
+      "grad_norm": 0.4054989218711853,
+      "learning_rate": 4.320830713284958e-05,
+      "loss": 1.4607,
+      "num_input_tokens_seen": 213898376,
+      "step": 5430
+    },
+    {
+      "epoch": 0.26360260937509467,
+      "grad_norm": 0.5132128596305847,
+      "learning_rate": 4.3181398883481304e-05,
+      "loss": 1.4274,
+      "num_input_tokens_seen": 214290036,
+      "step": 5440
+    },
+    {
+      "epoch": 0.26408717299526946,
+      "grad_norm": 0.40065979957580566,
+      "learning_rate": 4.315444584948485e-05,
+      "loss": 1.4295,
+      "num_input_tokens_seen": 214675368,
+      "step": 5450
+    },
+    {
+      "epoch": 0.26457173661544425,
+      "grad_norm": 0.39053869247436523,
+      "learning_rate": 4.3127448097251235e-05,
+      "loss": 1.4356,
+      "num_input_tokens_seen": 215066532,
+      "step": 5460
+    },
+    {
+      "epoch": 0.26505630023561905,
+      "grad_norm": 0.413308709859848,
+      "learning_rate": 4.310040569328164e-05,
+      "loss": 1.4318,
+      "num_input_tokens_seen": 215432972,
+      "step": 5470
+    },
+    {
+      "epoch": 0.26554086385579384,
+      "grad_norm": 0.38464999198913574,
+      "learning_rate": 4.3073318704187206e-05,
+      "loss": 1.4088,
+      "num_input_tokens_seen": 215839824,
+      "step": 5480
+    },
+    {
+      "epoch": 0.2660254274759687,
+      "grad_norm": 0.39387422800064087,
+      "learning_rate": 4.3046187196688923e-05,
+      "loss": 1.446,
+      "num_input_tokens_seen": 216225156,
+      "step": 5490
+    },
+    {
+      "epoch": 0.2665099910961435,
+      "grad_norm": 0.3859797716140747,
+      "learning_rate": 4.3019011237617434e-05,
+      "loss": 1.4121,
+      "num_input_tokens_seen": 216625012,
+      "step": 5500
+    },
+    {
+      "epoch": 0.2669945547163183,
+      "grad_norm": 0.41637757420539856,
+      "learning_rate": 4.2991790893912856e-05,
+      "loss": 1.4053,
+      "num_input_tokens_seen": 217002304,
+      "step": 5510
+    },
+    {
+      "epoch": 0.26747911833649307,
+      "grad_norm": 0.4308488368988037,
+      "learning_rate": 4.296452623262465e-05,
+      "loss": 1.4176,
+      "num_input_tokens_seen": 217381264,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2679636819566679,
+      "grad_norm": 0.42404836416244507,
+      "learning_rate": 4.293721732091145e-05,
+      "loss": 1.4504,
+      "num_input_tokens_seen": 217798076,
+      "step": 5530
+    },
+    {
+      "epoch": 0.2684482455768427,
+      "grad_norm": 0.42752760648727417,
+      "learning_rate": 4.290986422604087e-05,
+      "loss": 1.4442,
+      "num_input_tokens_seen": 218198560,
+      "step": 5540
+    },
+    {
+      "epoch": 0.2689328091970175,
+      "grad_norm": 0.4021061956882477,
+      "learning_rate": 4.288246701538936e-05,
+      "loss": 1.4238,
+      "num_input_tokens_seen": 218581244,
+      "step": 5550
+    },
+    {
+      "epoch": 0.2694173728171923,
+      "grad_norm": 0.3673078119754791,
+      "learning_rate": 4.285502575644206e-05,
+      "loss": 1.4188,
+      "num_input_tokens_seen": 218957676,
+      "step": 5560
+    },
+    {
+      "epoch": 0.2699019364373671,
+      "grad_norm": 0.40478530526161194,
+      "learning_rate": 4.282754051679256e-05,
+      "loss": 1.4707,
+      "num_input_tokens_seen": 219324792,
+      "step": 5570
+    },
+    {
+      "epoch": 0.27038650005754195,
+      "grad_norm": 0.41956064105033875,
+      "learning_rate": 4.280001136414283e-05,
+      "loss": 1.4583,
+      "num_input_tokens_seen": 219696608,
+      "step": 5580
+    },
+    {
+      "epoch": 0.27087106367771674,
+      "grad_norm": 0.4060947597026825,
+      "learning_rate": 4.2772438366303004e-05,
+      "loss": 1.4013,
+      "num_input_tokens_seen": 220080788,
+      "step": 5590
+    },
+    {
+      "epoch": 0.27135562729789153,
+      "grad_norm": 0.3806227445602417,
+      "learning_rate": 4.274482159119119e-05,
+      "loss": 1.4192,
+      "num_input_tokens_seen": 220493132,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2718401909180663,
+      "grad_norm": 0.3854431211948395,
+      "learning_rate": 4.2717161106833336e-05,
+      "loss": 1.4181,
+      "num_input_tokens_seen": 220880964,
+      "step": 5610
+    },
+    {
+      "epoch": 0.2723247545382412,
+      "grad_norm": 0.40384823083877563,
+      "learning_rate": 4.2689456981363074e-05,
+      "loss": 1.4363,
+      "num_input_tokens_seen": 221269648,
+      "step": 5620
+    },
+    {
+      "epoch": 0.27280931815841597,
+      "grad_norm": 0.3799000680446625,
+      "learning_rate": 4.2661709283021514e-05,
+      "loss": 1.4415,
+      "num_input_tokens_seen": 221654464,
+      "step": 5630
+    },
+    {
+      "epoch": 0.27329388177859076,
+      "grad_norm": 0.4527069628238678,
+      "learning_rate": 4.26339180801571e-05,
+      "loss": 1.4193,
+      "num_input_tokens_seen": 222030672,
+      "step": 5640
+    },
+    {
+      "epoch": 0.27377844539876556,
+      "grad_norm": 0.41296154260635376,
+      "learning_rate": 4.260608344122544e-05,
+      "loss": 1.3981,
+      "num_input_tokens_seen": 222409816,
+      "step": 5650
+    },
+    {
+      "epoch": 0.2742630090189404,
+      "grad_norm": 0.37670767307281494,
+      "learning_rate": 4.257820543478913e-05,
+      "loss": 1.4171,
+      "num_input_tokens_seen": 222803080,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2747475726391152,
+      "grad_norm": 0.3829585611820221,
+      "learning_rate": 4.255028412951761e-05,
+      "loss": 1.4174,
+      "num_input_tokens_seen": 223208420,
+      "step": 5670
+    },
+    {
+      "epoch": 0.27523213625929,
+      "grad_norm": 0.41849249601364136,
+      "learning_rate": 4.2522319594186934e-05,
+      "loss": 1.4399,
+      "num_input_tokens_seen": 223609392,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2757166998794648,
+      "grad_norm": 0.39865440130233765,
+      "learning_rate": 4.2494311897679664e-05,
+      "loss": 1.4429,
+      "num_input_tokens_seen": 223997220,
+      "step": 5690
+    },
+    {
+      "epoch": 0.2762012634996396,
+      "grad_norm": 0.43166518211364746,
+      "learning_rate": 4.246626110898469e-05,
+      "loss": 1.4143,
+      "num_input_tokens_seen": 224395400,
+      "step": 5700
+    },
+    {
+      "epoch": 0.27668582711981443,
+      "grad_norm": 0.41193270683288574,
+      "learning_rate": 4.2438167297197027e-05,
+      "loss": 1.4019,
+      "num_input_tokens_seen": 224779756,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2771703907399892,
+      "grad_norm": 0.38215872645378113,
+      "learning_rate": 4.2410030531517665e-05,
+      "loss": 1.4117,
+      "num_input_tokens_seen": 225155936,
+      "step": 5720
+    },
+    {
+      "epoch": 0.277654954360164,
+      "grad_norm": 0.3913954794406891,
+      "learning_rate": 4.2381850881253415e-05,
+      "loss": 1.4038,
+      "num_input_tokens_seen": 225541516,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2781395179803388,
+      "grad_norm": 0.4006546139717102,
+      "learning_rate": 4.23536284158167e-05,
+      "loss": 1.4241,
+      "num_input_tokens_seen": 225902060,
+      "step": 5740
+    },
+    {
+      "epoch": 0.27862408160051366,
+      "grad_norm": 0.39373400807380676,
+      "learning_rate": 4.232536320472543e-05,
+      "loss": 1.422,
+      "num_input_tokens_seen": 226304908,
+      "step": 5750
+    },
+    {
+      "epoch": 0.27910864522068846,
+      "grad_norm": 0.40954354405403137,
+      "learning_rate": 4.2297055317602785e-05,
+      "loss": 1.4074,
+      "num_input_tokens_seen": 226677752,
+      "step": 5760
+    },
+    {
+      "epoch": 0.27959320884086325,
+      "grad_norm": 0.41480016708374023,
+      "learning_rate": 4.226870482417707e-05,
+      "loss": 1.4142,
+      "num_input_tokens_seen": 227050196,
+      "step": 5770
+    },
+    {
+      "epoch": 0.28007777246103804,
+      "grad_norm": 0.3755822479724884,
+      "learning_rate": 4.2240311794281564e-05,
+      "loss": 1.4397,
+      "num_input_tokens_seen": 227446460,
+      "step": 5780
+    },
+    {
+      "epoch": 0.28056233608121284,
+      "grad_norm": 0.44134095311164856,
+      "learning_rate": 4.221187629785428e-05,
+      "loss": 1.4487,
+      "num_input_tokens_seen": 227831552,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2810468997013877,
+      "grad_norm": 0.4083091616630554,
+      "learning_rate": 4.218339840493786e-05,
+      "loss": 1.414,
+      "num_input_tokens_seen": 228221528,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2815314633215625,
+      "grad_norm": 0.4252268671989441,
+      "learning_rate": 4.215487818567937e-05,
+      "loss": 1.3964,
+      "num_input_tokens_seen": 228590960,
+      "step": 5810
+    },
+    {
+      "epoch": 0.2820160269417373,
+      "grad_norm": 0.3926514983177185,
+      "learning_rate": 4.212631571033015e-05,
+      "loss": 1.3921,
+      "num_input_tokens_seen": 228975244,
+      "step": 5820
+    },
+    {
+      "epoch": 0.28250059056191207,
+      "grad_norm": 0.41333383321762085,
+      "learning_rate": 4.20977110492456e-05,
+      "loss": 1.4569,
+      "num_input_tokens_seen": 229370872,
+      "step": 5830
+    },
+    {
+      "epoch": 0.2829851541820869,
+      "grad_norm": 0.41047653555870056,
+      "learning_rate": 4.206906427288506e-05,
+      "loss": 1.465,
+      "num_input_tokens_seen": 229788924,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2834697178022617,
+      "grad_norm": 0.4507644474506378,
+      "learning_rate": 4.204037545181158e-05,
+      "loss": 1.4222,
+      "num_input_tokens_seen": 230192956,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2839542814224365,
+      "grad_norm": 0.4235022962093353,
+      "learning_rate": 4.201164465669179e-05,
+      "loss": 1.4538,
+      "num_input_tokens_seen": 230540380,
+      "step": 5860
+    },
+    {
+      "epoch": 0.2844388450426113,
+      "grad_norm": 0.3815031051635742,
+      "learning_rate": 4.1982871958295734e-05,
+      "loss": 1.44,
+      "num_input_tokens_seen": 230931500,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2849234086627861,
+      "grad_norm": 0.4607156813144684,
+      "learning_rate": 4.1954057427496615e-05,
+      "loss": 1.3994,
+      "num_input_tokens_seen": 231318500,
+      "step": 5880
+    },
+    {
+      "epoch": 0.28540797228296094,
+      "grad_norm": 0.4481658637523651,
+      "learning_rate": 4.192520113527075e-05,
+      "loss": 1.4693,
+      "num_input_tokens_seen": 231716316,
+      "step": 5890
+    },
+    {
+      "epoch": 0.28589253590313574,
+      "grad_norm": 0.41151443123817444,
+      "learning_rate": 4.1896303152697254e-05,
+      "loss": 1.4007,
+      "num_input_tokens_seen": 232105992,
+      "step": 5900
+    },
+    {
+      "epoch": 0.28637709952331053,
+      "grad_norm": 0.3963596522808075,
+      "learning_rate": 4.186736355095798e-05,
+      "loss": 1.3917,
+      "num_input_tokens_seen": 232505400,
+      "step": 5910
+    },
+    {
+      "epoch": 0.2868616631434853,
+      "grad_norm": 0.42216241359710693,
+      "learning_rate": 4.183838240133728e-05,
+      "loss": 1.43,
+      "num_input_tokens_seen": 232922304,
+      "step": 5920
+    },
+    {
+      "epoch": 0.2873462267636602,
+      "grad_norm": 0.43040603399276733,
+      "learning_rate": 4.1809359775221854e-05,
+      "loss": 1.4388,
+      "num_input_tokens_seen": 233313888,
+      "step": 5930
+    },
+    {
+      "epoch": 0.28783079038383497,
+      "grad_norm": 0.4640505611896515,
+      "learning_rate": 4.178029574410056e-05,
+      "loss": 1.4517,
+      "num_input_tokens_seen": 233693256,
+      "step": 5940
+    },
+    {
+      "epoch": 0.28831535400400976,
+      "grad_norm": 0.40719953179359436,
+      "learning_rate": 4.175119037956425e-05,
+      "loss": 1.4023,
+      "num_input_tokens_seen": 234072408,
+      "step": 5950
+    },
+    {
+      "epoch": 0.28879991762418455,
+      "grad_norm": 0.3908650279045105,
+      "learning_rate": 4.17220437533056e-05,
+      "loss": 1.4172,
+      "num_input_tokens_seen": 234464800,
+      "step": 5960
+    },
+    {
+      "epoch": 0.2892844812443594,
+      "grad_norm": 0.4344137907028198,
+      "learning_rate": 4.169285593711889e-05,
+      "loss": 1.394,
+      "num_input_tokens_seen": 234870688,
+      "step": 5970
+    },
+    {
+      "epoch": 0.2897690448645342,
+      "grad_norm": 0.4607277512550354,
+      "learning_rate": 4.16636270028999e-05,
+      "loss": 1.428,
+      "num_input_tokens_seen": 235291024,
+      "step": 5980
+    },
+    {
+      "epoch": 0.290253608484709,
+      "grad_norm": 0.4087435007095337,
+      "learning_rate": 4.163435702264567e-05,
+      "loss": 1.4582,
+      "num_input_tokens_seen": 235698736,
+      "step": 5990
+    },
+    {
+      "epoch": 0.2907381721048838,
+      "grad_norm": 0.4173283576965332,
+      "learning_rate": 4.160504606845432e-05,
+      "loss": 1.4224,
+      "num_input_tokens_seen": 236103764,
+      "step": 6000
+    },
+    {
+      "epoch": 0.2907381721048838,
+      "eval_loss": 1.5239336490631104,
+      "eval_runtime": 4.4295,
+      "eval_samples_per_second": 33.864,
+      "eval_steps_per_second": 4.289,
+      "num_input_tokens_seen": 236103764,
+      "step": 6000
+    },
+    {
+      "epoch": 0.2912227357250586,
+      "grad_norm": 0.42691048979759216,
+      "learning_rate": 4.157569421252496e-05,
+      "loss": 1.4339,
+      "num_input_tokens_seen": 236508456,
+      "step": 6010
+    },
+    {
+      "epoch": 0.29170729934523343,
+      "grad_norm": 0.4078880548477173,
+      "learning_rate": 4.15463015271574e-05,
+      "loss": 1.4053,
+      "num_input_tokens_seen": 236915292,
+      "step": 6020
+    },
+    {
+      "epoch": 0.2921918629654082,
+      "grad_norm": 0.5139662623405457,
+      "learning_rate": 4.151686808475204e-05,
+      "loss": 1.4309,
+      "num_input_tokens_seen": 237303592,
+      "step": 6030
+    },
+    {
+      "epoch": 0.292676426585583,
+      "grad_norm": 0.40330371260643005,
+      "learning_rate": 4.1487393957809664e-05,
+      "loss": 1.4339,
+      "num_input_tokens_seen": 237688980,
+      "step": 6040
+    },
+    {
+      "epoch": 0.2931609902057578,
+      "grad_norm": 0.45068076252937317,
+      "learning_rate": 4.145787921893128e-05,
+      "loss": 1.4589,
+      "num_input_tokens_seen": 238057980,
+      "step": 6050
+    },
+    {
+      "epoch": 0.29364555382593266,
+      "grad_norm": 0.38412150740623474,
+      "learning_rate": 4.1428323940817933e-05,
+      "loss": 1.4576,
+      "num_input_tokens_seen": 238451220,
+      "step": 6060
+    },
+    {
+      "epoch": 0.29413011744610745,
+      "grad_norm": 0.385979026556015,
+      "learning_rate": 4.139872819627051e-05,
+      "loss": 1.4162,
+      "num_input_tokens_seen": 238829748,
+      "step": 6070
+    },
+    {
+      "epoch": 0.29461468106628225,
+      "grad_norm": 0.4156942069530487,
+      "learning_rate": 4.1369092058189586e-05,
+      "loss": 1.3992,
+      "num_input_tokens_seen": 239203752,
+      "step": 6080
+    },
+    {
+      "epoch": 0.29509924468645704,
+      "grad_norm": 0.4118397533893585,
+      "learning_rate": 4.133941559957524e-05,
+      "loss": 1.4029,
+      "num_input_tokens_seen": 239590056,
+      "step": 6090
+    },
+    {
+      "epoch": 0.29558380830663183,
+      "grad_norm": 0.47945645451545715,
+      "learning_rate": 4.130969889352686e-05,
+      "loss": 1.3959,
+      "num_input_tokens_seen": 239985824,
+      "step": 6100
+    },
+    {
+      "epoch": 0.2960683719268067,
+      "grad_norm": 0.36734700202941895,
+      "learning_rate": 4.1279942013242966e-05,
+      "loss": 1.3805,
+      "num_input_tokens_seen": 240361856,
+      "step": 6110
+    },
+    {
+      "epoch": 0.2965529355469815,
+      "grad_norm": 0.41420719027519226,
+      "learning_rate": 4.125014503202106e-05,
+      "loss": 1.3727,
+      "num_input_tokens_seen": 240780880,
+      "step": 6120
+    },
+    {
+      "epoch": 0.29703749916715627,
+      "grad_norm": 0.43215250968933105,
+      "learning_rate": 4.122030802325738e-05,
+      "loss": 1.4033,
+      "num_input_tokens_seen": 241189188,
+      "step": 6130
+    },
+    {
+      "epoch": 0.29752206278733107,
+      "grad_norm": 0.42517751455307007,
+      "learning_rate": 4.119043106044681e-05,
+      "loss": 1.4514,
+      "num_input_tokens_seen": 241610008,
+      "step": 6140
+    },
+    {
+      "epoch": 0.2980066264075059,
+      "grad_norm": 0.38953685760498047,
+      "learning_rate": 4.116051421718261e-05,
+      "loss": 1.4399,
+      "num_input_tokens_seen": 242006608,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2984911900276807,
+      "grad_norm": 0.4071796238422394,
+      "learning_rate": 4.113055756715628e-05,
+      "loss": 1.3734,
+      "num_input_tokens_seen": 242416040,
+      "step": 6160
+    },
+    {
+      "epoch": 0.2989757536478555,
+      "grad_norm": 0.42148080468177795,
+      "learning_rate": 4.11005611841574e-05,
+      "loss": 1.4508,
+      "num_input_tokens_seen": 242794628,
+      "step": 6170
+    },
+    {
+      "epoch": 0.2994603172680303,
+      "grad_norm": 0.4180959463119507,
+      "learning_rate": 4.107052514207339e-05,
+      "loss": 1.4159,
+      "num_input_tokens_seen": 243164260,
+      "step": 6180
+    },
+    {
+      "epoch": 0.2999448808882051,
+      "grad_norm": 0.36293745040893555,
+      "learning_rate": 4.1040449514889375e-05,
+      "loss": 1.3633,
+      "num_input_tokens_seen": 243550040,
+      "step": 6190
+    },
+    {
+      "epoch": 0.30042944450837994,
+      "grad_norm": 0.4095574617385864,
+      "learning_rate": 4.1010334376687975e-05,
+      "loss": 1.4784,
+      "num_input_tokens_seen": 243908832,
+      "step": 6200
+    },
+    {
+      "epoch": 0.30091400812855473,
+      "grad_norm": 0.4131157100200653,
+      "learning_rate": 4.0980179801649146e-05,
+      "loss": 1.4184,
+      "num_input_tokens_seen": 244298944,
+      "step": 6210
+    },
+    {
+      "epoch": 0.3013985717487295,
+      "grad_norm": 0.39260920882225037,
+      "learning_rate": 4.094998586404998e-05,
+      "loss": 1.4639,
+      "num_input_tokens_seen": 244692220,
+      "step": 6220
+    },
+    {
+      "epoch": 0.3018831353689043,
+      "grad_norm": 0.43091267347335815,
+      "learning_rate": 4.0919752638264516e-05,
+      "loss": 1.4488,
+      "num_input_tokens_seen": 245050952,
+      "step": 6230
+    },
+    {
+      "epoch": 0.30236769898907917,
+      "grad_norm": 0.42741450667381287,
+      "learning_rate": 4.088948019876359e-05,
+      "loss": 1.4434,
+      "num_input_tokens_seen": 245465356,
+      "step": 6240
+    },
+    {
+      "epoch": 0.30285226260925396,
+      "grad_norm": 0.4109672009944916,
+      "learning_rate": 4.085916862011463e-05,
+      "loss": 1.4374,
+      "num_input_tokens_seen": 245880668,
+      "step": 6250
+    },
+    {
+      "epoch": 0.30333682622942876,
+      "grad_norm": 0.4069707691669464,
+      "learning_rate": 4.082881797698143e-05,
+      "loss": 1.3555,
+      "num_input_tokens_seen": 246284348,
+      "step": 6260
+    },
+    {
+      "epoch": 0.30382138984960355,
+      "grad_norm": 0.4159133732318878,
+      "learning_rate": 4.0798428344124064e-05,
+      "loss": 1.3921,
+      "num_input_tokens_seen": 246662396,
+      "step": 6270
+    },
+    {
+      "epoch": 0.30430595346977835,
+      "grad_norm": 0.42207905650138855,
+      "learning_rate": 4.07679997963986e-05,
+      "loss": 1.3996,
+      "num_input_tokens_seen": 247049844,
+      "step": 6280
+    },
+    {
+      "epoch": 0.3047905170899532,
+      "grad_norm": 0.3645390272140503,
+      "learning_rate": 4.0737532408757014e-05,
+      "loss": 1.4054,
+      "num_input_tokens_seen": 247441300,
+      "step": 6290
+    },
+    {
+      "epoch": 0.305275080710128,
+      "grad_norm": 0.41653063893318176,
+      "learning_rate": 4.0707026256246894e-05,
+      "loss": 1.3984,
+      "num_input_tokens_seen": 247851160,
+      "step": 6300
+    },
+    {
+      "epoch": 0.3057596443303028,
+      "grad_norm": 0.41180601716041565,
+      "learning_rate": 4.0676481414011345e-05,
+      "loss": 1.3664,
+      "num_input_tokens_seen": 248245884,
+      "step": 6310
+    },
+    {
+      "epoch": 0.3062442079504776,
+      "grad_norm": 0.4273717403411865,
+      "learning_rate": 4.064589795728878e-05,
+      "loss": 1.4068,
+      "num_input_tokens_seen": 248649064,
+      "step": 6320
+    },
+    {
+      "epoch": 0.3067287715706524,
+      "grad_norm": 0.3938922882080078,
+      "learning_rate": 4.06152759614127e-05,
+      "loss": 1.4184,
+      "num_input_tokens_seen": 249030280,
+      "step": 6330
+    },
+    {
+      "epoch": 0.3072133351908272,
+      "grad_norm": 0.4099946916103363,
+      "learning_rate": 4.0584615501811577e-05,
+      "loss": 1.3773,
+      "num_input_tokens_seen": 249430164,
+      "step": 6340
+    },
+    {
+      "epoch": 0.307697898811002,
+      "grad_norm": 0.4208473265171051,
+      "learning_rate": 4.055391665400858e-05,
+      "loss": 1.3534,
+      "num_input_tokens_seen": 249806996,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3081824624311768,
+      "grad_norm": 0.4177727997303009,
+      "learning_rate": 4.052317949362147e-05,
+      "loss": 1.3888,
+      "num_input_tokens_seen": 250196904,
+      "step": 6360
+    },
+    {
+      "epoch": 0.30866702605135166,
+      "grad_norm": 0.4083334505558014,
+      "learning_rate": 4.049240409636237e-05,
+      "loss": 1.3849,
+      "num_input_tokens_seen": 250581560,
+      "step": 6370
+    },
+    {
+      "epoch": 0.30915158967152645,
+      "grad_norm": 0.4475148022174835,
+      "learning_rate": 4.046159053803758e-05,
+      "loss": 1.3855,
+      "num_input_tokens_seen": 250966948,
+      "step": 6380
+    },
+    {
+      "epoch": 0.30963615329170124,
+      "grad_norm": 0.39256733655929565,
+      "learning_rate": 4.0430738894547426e-05,
+      "loss": 1.4499,
+      "num_input_tokens_seen": 251353488,
+      "step": 6390
+    },
+    {
+      "epoch": 0.31012071691187604,
+      "grad_norm": 0.3913814127445221,
+      "learning_rate": 4.0399849241886e-05,
+      "loss": 1.3994,
+      "num_input_tokens_seen": 251742612,
+      "step": 6400
+    },
+    {
+      "epoch": 0.31060528053205083,
+      "grad_norm": 0.41286328434944153,
+      "learning_rate": 4.0368921656141065e-05,
+      "loss": 1.3918,
+      "num_input_tokens_seen": 252141148,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3110898441522257,
+      "grad_norm": 0.39640408754348755,
+      "learning_rate": 4.03379562134938e-05,
+      "loss": 1.3757,
+      "num_input_tokens_seen": 252532656,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3115744077724005,
+      "grad_norm": 0.40248820185661316,
+      "learning_rate": 4.030695299021863e-05,
+      "loss": 1.3941,
+      "num_input_tokens_seen": 252899616,
+      "step": 6430
+    },
+    {
+      "epoch": 0.31205897139257527,
+      "grad_norm": 0.3976843059062958,
+      "learning_rate": 4.027591206268304e-05,
+      "loss": 1.3953,
+      "num_input_tokens_seen": 253271420,
+      "step": 6440
+    },
+    {
+      "epoch": 0.31254353501275006,
+      "grad_norm": 0.42651301622390747,
+      "learning_rate": 4.02448335073474e-05,
+      "loss": 1.4118,
+      "num_input_tokens_seen": 253688464,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3130280986329249,
+      "grad_norm": 0.4326588213443756,
+      "learning_rate": 4.0213717400764766e-05,
+      "loss": 1.3611,
+      "num_input_tokens_seen": 254091036,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3135126622530997,
+      "grad_norm": 0.43498626351356506,
+      "learning_rate": 4.018256381958068e-05,
+      "loss": 1.3872,
+      "num_input_tokens_seen": 254494504,
+      "step": 6470
+    },
+    {
+      "epoch": 0.3139972258732745,
+      "grad_norm": 0.3556174039840698,
+      "learning_rate": 4.0151372840533e-05,
+      "loss": 1.4149,
+      "num_input_tokens_seen": 254877836,
+      "step": 6480
+    },
+    {
+      "epoch": 0.3144817894934493,
+      "grad_norm": 0.40494486689567566,
+      "learning_rate": 4.0120144540451706e-05,
+      "loss": 1.4471,
+      "num_input_tokens_seen": 255276796,
+      "step": 6490
+    },
+    {
+      "epoch": 0.3149663531136241,
+      "grad_norm": 0.42257654666900635,
+      "learning_rate": 4.008887899625868e-05,
+      "loss": 1.3782,
+      "num_input_tokens_seen": 255671500,
+      "step": 6500
+    },
+    {
+      "epoch": 0.31545091673379894,
+      "grad_norm": 0.4325736165046692,
+      "learning_rate": 4.005757628496759e-05,
+      "loss": 1.4032,
+      "num_input_tokens_seen": 256064776,
+      "step": 6510
+    },
+    {
+      "epoch": 0.31593548035397373,
+      "grad_norm": 0.4220854341983795,
+      "learning_rate": 4.002623648368361e-05,
+      "loss": 1.3634,
+      "num_input_tokens_seen": 256461304,
+      "step": 6520
+    },
+    {
+      "epoch": 0.3164200439741485,
+      "grad_norm": 0.38566797971725464,
+      "learning_rate": 3.9994859669603316e-05,
+      "loss": 1.3689,
+      "num_input_tokens_seen": 256882232,
+      "step": 6530
+    },
+    {
+      "epoch": 0.3169046075943233,
+      "grad_norm": 0.38369229435920715,
+      "learning_rate": 3.99634459200144e-05,
+      "loss": 1.4368,
+      "num_input_tokens_seen": 257264380,
+      "step": 6540
+    },
+    {
+      "epoch": 0.31738917121449817,
+      "grad_norm": 0.432253897190094,
+      "learning_rate": 3.9931995312295596e-05,
+      "loss": 1.4141,
+      "num_input_tokens_seen": 257641472,
+      "step": 6550
+    },
+    {
+      "epoch": 0.31787373483467296,
+      "grad_norm": 0.3968162536621094,
+      "learning_rate": 3.9900507923916394e-05,
+      "loss": 1.428,
+      "num_input_tokens_seen": 258040272,
+      "step": 6560
+    },
+    {
+      "epoch": 0.31835829845484775,
+      "grad_norm": 0.44724634289741516,
+      "learning_rate": 3.9868983832436876e-05,
+      "loss": 1.402,
+      "num_input_tokens_seen": 258430376,
+      "step": 6570
+    },
+    {
+      "epoch": 0.31884286207502255,
+      "grad_norm": 0.39584001898765564,
+      "learning_rate": 3.983742311550755e-05,
+      "loss": 1.3927,
+      "num_input_tokens_seen": 258847780,
+      "step": 6580
+    },
+    {
+      "epoch": 0.31932742569519734,
+      "grad_norm": 0.446869432926178,
+      "learning_rate": 3.9805825850869125e-05,
+      "loss": 1.4192,
+      "num_input_tokens_seen": 259228264,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3198119893153722,
+      "grad_norm": 0.3873324394226074,
+      "learning_rate": 3.977419211635235e-05,
+      "loss": 1.37,
+      "num_input_tokens_seen": 259636616,
+      "step": 6600
+    },
+    {
+      "epoch": 0.320296552935547,
+      "grad_norm": 0.4095400869846344,
+      "learning_rate": 3.9742521989877795e-05,
+      "loss": 1.3664,
+      "num_input_tokens_seen": 260029232,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3207811165557218,
+      "grad_norm": 0.3852662146091461,
+      "learning_rate": 3.971081554945568e-05,
+      "loss": 1.3546,
+      "num_input_tokens_seen": 260456224,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3212656801758966,
+      "grad_norm": 0.3864857256412506,
+      "learning_rate": 3.967907287318566e-05,
+      "loss": 1.3756,
+      "num_input_tokens_seen": 260837824,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3217502437960714,
+      "grad_norm": 0.3785705268383026,
+      "learning_rate": 3.964729403925666e-05,
+      "loss": 1.4206,
+      "num_input_tokens_seen": 261235928,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3222348074162462,
+      "grad_norm": 0.37969115376472473,
+      "learning_rate": 3.961547912594667e-05,
+      "loss": 1.406,
+      "num_input_tokens_seen": 261622640,
+      "step": 6650
+    },
+    {
+      "epoch": 0.322719371036421,
+      "grad_norm": 0.40466952323913574,
+      "learning_rate": 3.958362821162254e-05,
+      "loss": 1.393,
+      "num_input_tokens_seen": 262061520,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3232039346565958,
+      "grad_norm": 0.40950527787208557,
+      "learning_rate": 3.955174137473979e-05,
+      "loss": 1.4314,
+      "num_input_tokens_seen": 262449712,
+      "step": 6670
+    },
+    {
+      "epoch": 0.32368849827677065,
+      "grad_norm": 0.4362260699272156,
+      "learning_rate": 3.951981869384247e-05,
+      "loss": 1.4466,
+      "num_input_tokens_seen": 262842776,
+      "step": 6680
+    },
+    {
+      "epoch": 0.32417306189694545,
+      "grad_norm": 0.4015698730945587,
+      "learning_rate": 3.948786024756287e-05,
+      "loss": 1.4056,
+      "num_input_tokens_seen": 263186520,
+      "step": 6690
+    },
+    {
+      "epoch": 0.32465762551712024,
+      "grad_norm": 0.3874407112598419,
+      "learning_rate": 3.9455866114621396e-05,
+      "loss": 1.379,
+      "num_input_tokens_seen": 263570516,
+      "step": 6700
+    },
+    {
+      "epoch": 0.32514218913729503,
+      "grad_norm": 0.3963698744773865,
+      "learning_rate": 3.9423836373826375e-05,
+      "loss": 1.4146,
+      "num_input_tokens_seen": 263984720,
+      "step": 6710
+    },
+    {
+      "epoch": 0.32562675275746983,
+      "grad_norm": 0.38951539993286133,
+      "learning_rate": 3.9391771104073805e-05,
+      "loss": 1.3588,
+      "num_input_tokens_seen": 264376356,
+      "step": 6720
+    },
+    {
+      "epoch": 0.3261113163776447,
+      "grad_norm": 0.4249687194824219,
+      "learning_rate": 3.9359670384347244e-05,
+      "loss": 1.4055,
+      "num_input_tokens_seen": 264737708,
+      "step": 6730
+    },
+    {
+      "epoch": 0.32659587999781947,
+      "grad_norm": 0.38605302572250366,
+      "learning_rate": 3.9327534293717537e-05,
+      "loss": 1.3578,
+      "num_input_tokens_seen": 265138996,
+      "step": 6740
+    },
+    {
+      "epoch": 0.32708044361799427,
+      "grad_norm": 0.41331490874290466,
+      "learning_rate": 3.929536291134267e-05,
+      "loss": 1.3902,
+      "num_input_tokens_seen": 265509020,
+      "step": 6750
+    },
+    {
+      "epoch": 0.32756500723816906,
+      "grad_norm": 0.372742623090744,
+      "learning_rate": 3.926315631646756e-05,
+      "loss": 1.3639,
+      "num_input_tokens_seen": 265879672,
+      "step": 6760
+    },
+    {
+      "epoch": 0.3280495708583439,
+      "grad_norm": 0.5427610874176025,
+      "learning_rate": 3.9230914588423864e-05,
+      "loss": 1.3371,
+      "num_input_tokens_seen": 266275292,
+      "step": 6770
+    },
+    {
+      "epoch": 0.3285341344785187,
+      "grad_norm": 0.4112084209918976,
+      "learning_rate": 3.9198637806629756e-05,
+      "loss": 1.4216,
+      "num_input_tokens_seen": 266671844,
+      "step": 6780
+    },
+    {
+      "epoch": 0.3290186980986935,
+      "grad_norm": 0.3875449299812317,
+      "learning_rate": 3.916632605058978e-05,
+      "loss": 1.3905,
+      "num_input_tokens_seen": 267057556,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3295032617188683,
+      "grad_norm": 0.3956202566623688,
+      "learning_rate": 3.913397939989461e-05,
+      "loss": 1.3913,
+      "num_input_tokens_seen": 267444300,
+      "step": 6800
+    },
+    {
+      "epoch": 0.3299878253390431,
+      "grad_norm": 0.41076987981796265,
+      "learning_rate": 3.910159793422091e-05,
+      "loss": 1.4032,
+      "num_input_tokens_seen": 267811660,
+      "step": 6810
+    },
+    {
+      "epoch": 0.33047238895921793,
+      "grad_norm": 0.41515880823135376,
+      "learning_rate": 3.9069181733331056e-05,
+      "loss": 1.4098,
+      "num_input_tokens_seen": 268195916,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3309569525793927,
+      "grad_norm": 0.5097044110298157,
+      "learning_rate": 3.9036730877073e-05,
+      "loss": 1.3921,
+      "num_input_tokens_seen": 268578008,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3314415161995675,
+      "grad_norm": 0.43721285462379456,
+      "learning_rate": 3.900424544538006e-05,
+      "loss": 1.3893,
+      "num_input_tokens_seen": 269003192,
+      "step": 6840
+    },
+    {
+      "epoch": 0.3319260798197423,
+      "grad_norm": 0.42796212434768677,
+      "learning_rate": 3.897172551827073e-05,
+      "loss": 1.4325,
+      "num_input_tokens_seen": 269393648,
+      "step": 6850
+    },
+    {
+      "epoch": 0.33241064343991716,
+      "grad_norm": 0.36507412791252136,
+      "learning_rate": 3.8939171175848447e-05,
+      "loss": 1.4346,
+      "num_input_tokens_seen": 269809016,
+      "step": 6860
+    },
+    {
+      "epoch": 0.33289520706009196,
+      "grad_norm": 0.39873602986335754,
+      "learning_rate": 3.8906582498301455e-05,
+      "loss": 1.4412,
+      "num_input_tokens_seen": 270180224,
+      "step": 6870
+    },
+    {
+      "epoch": 0.33337977068026675,
+      "grad_norm": 0.3825407326221466,
+      "learning_rate": 3.887395956590254e-05,
+      "loss": 1.3762,
+      "num_input_tokens_seen": 270581856,
+      "step": 6880
+    },
+    {
+      "epoch": 0.33386433430044155,
+      "grad_norm": 0.40756458044052124,
+      "learning_rate": 3.884130245900889e-05,
+      "loss": 1.4066,
+      "num_input_tokens_seen": 270962900,
+      "step": 6890
+    },
+    {
+      "epoch": 0.33434889792061634,
+      "grad_norm": 0.41066256165504456,
+      "learning_rate": 3.880861125806186e-05,
+      "loss": 1.3903,
+      "num_input_tokens_seen": 271374220,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3348334615407912,
+      "grad_norm": 0.4207150340080261,
+      "learning_rate": 3.877588604358678e-05,
+      "loss": 1.3709,
+      "num_input_tokens_seen": 271765116,
+      "step": 6910
+    },
+    {
+      "epoch": 0.335318025160966,
+      "grad_norm": 0.4337269067764282,
+      "learning_rate": 3.8743126896192784e-05,
+      "loss": 1.3251,
+      "num_input_tokens_seen": 272147924,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3358025887811408,
+      "grad_norm": 0.4367486536502838,
+      "learning_rate": 3.871033389657255e-05,
+      "loss": 1.3525,
+      "num_input_tokens_seen": 272511972,
+      "step": 6930
+    },
+    {
+      "epoch": 0.33628715240131557,
+      "grad_norm": 0.43631139397621155,
+      "learning_rate": 3.867750712550219e-05,
+      "loss": 1.3855,
+      "num_input_tokens_seen": 272876876,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3367717160214904,
+      "grad_norm": 0.40306615829467773,
+      "learning_rate": 3.8644646663840976e-05,
+      "loss": 1.3868,
+      "num_input_tokens_seen": 273287016,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3372562796416652,
+      "grad_norm": 0.3835316598415375,
+      "learning_rate": 3.861175259253117e-05,
+      "loss": 1.3667,
+      "num_input_tokens_seen": 273674052,
+      "step": 6960
+    },
+    {
+      "epoch": 0.33774084326184,
+      "grad_norm": 0.38782796263694763,
+      "learning_rate": 3.857882499259782e-05,
+      "loss": 1.353,
+      "num_input_tokens_seen": 274083796,
+      "step": 6970
+    },
+    {
+      "epoch": 0.3382254068820148,
+      "grad_norm": 0.395175039768219,
+      "learning_rate": 3.854586394514855e-05,
+      "loss": 1.3281,
+      "num_input_tokens_seen": 274493468,
+      "step": 6980
+    },
+    {
+      "epoch": 0.3387099705021896,
+      "grad_norm": 0.45938950777053833,
+      "learning_rate": 3.851286953137341e-05,
+      "loss": 1.4144,
+      "num_input_tokens_seen": 274858868,
+      "step": 6990
+    },
+    {
+      "epoch": 0.33919453412236444,
+      "grad_norm": 0.40472185611724854,
+      "learning_rate": 3.847984183254461e-05,
+      "loss": 1.3877,
+      "num_input_tokens_seen": 275233524,
+      "step": 7000
+    },
+    {
+      "epoch": 0.33967909774253924,
+      "grad_norm": 0.4177361726760864,
+      "learning_rate": 3.8446780930016336e-05,
+      "loss": 1.3834,
+      "num_input_tokens_seen": 275600504,
+      "step": 7010
+    },
+    {
+      "epoch": 0.34016366136271403,
+      "grad_norm": 0.4230976700782776,
+      "learning_rate": 3.8413686905224595e-05,
+      "loss": 1.3601,
+      "num_input_tokens_seen": 275991592,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3406482249828888,
+      "grad_norm": 0.3965121805667877,
+      "learning_rate": 3.838055983968695e-05,
+      "loss": 1.3801,
+      "num_input_tokens_seen": 276355724,
+      "step": 7030
+    },
+    {
+      "epoch": 0.3411327886030637,
+      "grad_norm": 0.39479291439056396,
+      "learning_rate": 3.8347399815002385e-05,
+      "loss": 1.3976,
+      "num_input_tokens_seen": 276749052,
+      "step": 7040
+    },
+    {
+      "epoch": 0.34161735222323847,
+      "grad_norm": 0.37649762630462646,
+      "learning_rate": 3.8314206912851036e-05,
+      "loss": 1.3981,
+      "num_input_tokens_seen": 277163328,
+      "step": 7050
+    },
+    {
+      "epoch": 0.34210191584341326,
+      "grad_norm": 0.389952152967453,
+      "learning_rate": 3.828098121499404e-05,
+      "loss": 1.3929,
+      "num_input_tokens_seen": 277532872,
+      "step": 7060
+    },
+    {
+      "epoch": 0.34258647946358806,
+      "grad_norm": 0.4221923053264618,
+      "learning_rate": 3.82477228032733e-05,
+      "loss": 1.3894,
+      "num_input_tokens_seen": 277930776,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3430710430837629,
+      "grad_norm": 0.4203970730304718,
+      "learning_rate": 3.821443175961134e-05,
+      "loss": 1.3611,
+      "num_input_tokens_seen": 278313072,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3435556067039377,
+      "grad_norm": 0.44928234815597534,
+      "learning_rate": 3.818110816601101e-05,
+      "loss": 1.4055,
+      "num_input_tokens_seen": 278692232,
+      "step": 7090
+    },
+    {
+      "epoch": 0.3440401703241125,
+      "grad_norm": 0.40566375851631165,
+      "learning_rate": 3.814775210455538e-05,
+      "loss": 1.3874,
+      "num_input_tokens_seen": 279070076,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3445247339442873,
+      "grad_norm": 0.4111347794532776,
+      "learning_rate": 3.811436365740748e-05,
+      "loss": 1.4077,
+      "num_input_tokens_seen": 279468284,
+      "step": 7110
+    },
+    {
+      "epoch": 0.3450092975644621,
+      "grad_norm": 0.3757016956806183,
+      "learning_rate": 3.808094290681011e-05,
+      "loss": 1.3623,
+      "num_input_tokens_seen": 279858588,
+      "step": 7120
+    },
+    {
+      "epoch": 0.34549386118463693,
+      "grad_norm": 0.41814279556274414,
+      "learning_rate": 3.8047489935085635e-05,
+      "loss": 1.3726,
+      "num_input_tokens_seen": 280269536,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3459784248048117,
+      "grad_norm": 0.3829779624938965,
+      "learning_rate": 3.801400482463581e-05,
+      "loss": 1.3783,
+      "num_input_tokens_seen": 280677220,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3464629884249865,
+      "grad_norm": 0.40414533019065857,
+      "learning_rate": 3.798048765794151e-05,
+      "loss": 1.3914,
+      "num_input_tokens_seen": 281062416,
+      "step": 7150
+    },
+    {
+      "epoch": 0.3469475520451613,
+      "grad_norm": 0.4087967574596405,
+      "learning_rate": 3.7946938517562635e-05,
+      "loss": 1.4013,
+      "num_input_tokens_seen": 281462400,
+      "step": 7160
+    },
+    {
+      "epoch": 0.34743211566533616,
+      "grad_norm": 0.3836404085159302,
+      "learning_rate": 3.791335748613779e-05,
+      "loss": 1.4126,
+      "num_input_tokens_seen": 281857064,
+      "step": 7170
+    },
+    {
+      "epoch": 0.34791667928551095,
+      "grad_norm": 0.4010114073753357,
+      "learning_rate": 3.7879744646384154e-05,
+      "loss": 1.4138,
+      "num_input_tokens_seen": 282243900,
+      "step": 7180
+    },
+    {
+      "epoch": 0.34840124290568575,
+      "grad_norm": 0.42484843730926514,
+      "learning_rate": 3.7846100081097255e-05,
+      "loss": 1.3773,
+      "num_input_tokens_seen": 282661384,
+      "step": 7190
+    },
+    {
+      "epoch": 0.34888580652586054,
+      "grad_norm": 0.42864614725112915,
+      "learning_rate": 3.7812423873150775e-05,
+      "loss": 1.4116,
+      "num_input_tokens_seen": 283074744,
+      "step": 7200
+    },
+    {
+      "epoch": 0.34937037014603534,
+      "grad_norm": 0.4581463634967804,
+      "learning_rate": 3.777871610549632e-05,
+      "loss": 1.355,
+      "num_input_tokens_seen": 283452024,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3498549337662102,
+      "grad_norm": 0.4017016887664795,
+      "learning_rate": 3.774497686116327e-05,
+      "loss": 1.393,
+      "num_input_tokens_seen": 283843012,
+      "step": 7220
+    },
+    {
+      "epoch": 0.350339497386385,
+      "grad_norm": 0.387157142162323,
+      "learning_rate": 3.7711206223258493e-05,
+      "loss": 1.3497,
+      "num_input_tokens_seen": 284225640,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3508240610065598,
+      "grad_norm": 0.42208153009414673,
+      "learning_rate": 3.767740427496621e-05,
+      "loss": 1.4008,
+      "num_input_tokens_seen": 284608708,
+      "step": 7240
+    },
+    {
+      "epoch": 0.35130862462673457,
+      "grad_norm": 0.4280160367488861,
+      "learning_rate": 3.764357109954777e-05,
+      "loss": 1.326,
+      "num_input_tokens_seen": 285029224,
+      "step": 7250
+    },
+    {
+      "epoch": 0.3517931882469094,
+      "grad_norm": 0.39940527081489563,
+      "learning_rate": 3.7609706780341425e-05,
+      "loss": 1.3688,
+      "num_input_tokens_seen": 285413332,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3522777518670842,
+      "grad_norm": 0.40198463201522827,
+      "learning_rate": 3.757581140076217e-05,
+      "loss": 1.3835,
+      "num_input_tokens_seen": 285805028,
+      "step": 7270
+    },
+    {
+      "epoch": 0.352762315487259,
+      "grad_norm": 0.4302620589733124,
+      "learning_rate": 3.754188504430147e-05,
+      "loss": 1.3317,
+      "num_input_tokens_seen": 286217520,
+      "step": 7280
+    },
+    {
+      "epoch": 0.3532468791074338,
+      "grad_norm": 0.41558027267456055,
+      "learning_rate": 3.750792779452712e-05,
+      "loss": 1.3584,
+      "num_input_tokens_seen": 286618988,
+      "step": 7290
+    },
+    {
+      "epoch": 0.3537314427276086,
+      "grad_norm": 0.4010438621044159,
+      "learning_rate": 3.7473939735082995e-05,
+      "loss": 1.3638,
+      "num_input_tokens_seen": 286993540,
+      "step": 7300
+    },
+    {
+      "epoch": 0.35421600634778344,
+      "grad_norm": 0.4007185995578766,
+      "learning_rate": 3.743992094968888e-05,
+      "loss": 1.3662,
+      "num_input_tokens_seen": 287379072,
+      "step": 7310
+    },
+    {
+      "epoch": 0.35470056996795823,
+      "grad_norm": 0.4279842674732208,
+      "learning_rate": 3.740587152214022e-05,
+      "loss": 1.3124,
+      "num_input_tokens_seen": 287779160,
+      "step": 7320
+    },
+    {
+      "epoch": 0.35518513358813303,
+      "grad_norm": 0.4510299563407898,
+      "learning_rate": 3.737179153630797e-05,
+      "loss": 1.4197,
+      "num_input_tokens_seen": 288150168,
+      "step": 7330
+    },
+    {
+      "epoch": 0.3556696972083078,
+      "grad_norm": 0.4274813234806061,
+      "learning_rate": 3.733768107613832e-05,
+      "loss": 1.3953,
+      "num_input_tokens_seen": 288526312,
+      "step": 7340
+    },
+    {
+      "epoch": 0.35615426082848267,
+      "grad_norm": 0.4202810525894165,
+      "learning_rate": 3.730354022565257e-05,
+      "loss": 1.4117,
+      "num_input_tokens_seen": 288933160,
+      "step": 7350
+    },
+    {
+      "epoch": 0.35663882444865747,
+      "grad_norm": 0.38731688261032104,
+      "learning_rate": 3.7269369068946816e-05,
+      "loss": 1.3829,
+      "num_input_tokens_seen": 289312352,
+      "step": 7360
+    },
+    {
+      "epoch": 0.35712338806883226,
+      "grad_norm": 0.39851441979408264,
+      "learning_rate": 3.7235167690191856e-05,
+      "loss": 1.354,
+      "num_input_tokens_seen": 289728744,
+      "step": 7370
+    },
+    {
+      "epoch": 0.35760795168900705,
+      "grad_norm": 0.35968631505966187,
+      "learning_rate": 3.7200936173632915e-05,
+      "loss": 1.4089,
+      "num_input_tokens_seen": 290118632,
+      "step": 7380
+    },
+    {
+      "epoch": 0.3580925153091819,
+      "grad_norm": 0.41429969668388367,
+      "learning_rate": 3.716667460358945e-05,
+      "loss": 1.3695,
+      "num_input_tokens_seen": 290507048,
+      "step": 7390
+    },
+    {
+      "epoch": 0.3585770789293567,
+      "grad_norm": 0.37059974670410156,
+      "learning_rate": 3.7132383064454956e-05,
+      "loss": 1.3864,
+      "num_input_tokens_seen": 290871436,
+      "step": 7400
+    },
+    {
+      "epoch": 0.3590616425495315,
+      "grad_norm": 0.3918827474117279,
+      "learning_rate": 3.7098061640696734e-05,
+      "loss": 1.3879,
+      "num_input_tokens_seen": 291230652,
+      "step": 7410
+    },
+    {
+      "epoch": 0.3595462061697063,
+      "grad_norm": 0.36059802770614624,
+      "learning_rate": 3.706371041685571e-05,
+      "loss": 1.3579,
+      "num_input_tokens_seen": 291641780,
+      "step": 7420
+    },
+    {
+      "epoch": 0.3600307697898811,
+      "grad_norm": 0.4405132234096527,
+      "learning_rate": 3.70293294775462e-05,
+      "loss": 1.3851,
+      "num_input_tokens_seen": 292025908,
+      "step": 7430
+    },
+    {
+      "epoch": 0.3605153334100559,
+      "grad_norm": 0.4112297594547272,
+      "learning_rate": 3.6994918907455734e-05,
+      "loss": 1.3806,
+      "num_input_tokens_seen": 292393396,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3609998970302307,
+      "grad_norm": 0.39005085825920105,
+      "learning_rate": 3.696047879134481e-05,
+      "loss": 1.3836,
+      "num_input_tokens_seen": 292779588,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3614844606504055,
+      "grad_norm": 0.4453318417072296,
+      "learning_rate": 3.692600921404672e-05,
+      "loss": 1.37,
+      "num_input_tokens_seen": 293187468,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3619690242705803,
+      "grad_norm": 0.4039851725101471,
+      "learning_rate": 3.689151026046732e-05,
+      "loss": 1.3835,
+      "num_input_tokens_seen": 293571348,
+      "step": 7470
+    },
+    {
+      "epoch": 0.36245358789075516,
+      "grad_norm": 0.4166390001773834,
+      "learning_rate": 3.685698201558482e-05,
+      "loss": 1.3745,
+      "num_input_tokens_seen": 293971468,
+      "step": 7480
+    },
+    {
+      "epoch": 0.36293815151092995,
+      "grad_norm": 0.41377830505371094,
+      "learning_rate": 3.6822424564449584e-05,
+      "loss": 1.3585,
+      "num_input_tokens_seen": 294316832,
+      "step": 7490
+    },
+    {
+      "epoch": 0.36342271513110475,
+      "grad_norm": 0.3810221254825592,
+      "learning_rate": 3.6787837992183916e-05,
+      "loss": 1.3045,
+      "num_input_tokens_seen": 294725576,
+      "step": 7500
+    },
+    {
+      "epoch": 0.36390727875127954,
+      "grad_norm": 0.3837166428565979,
+      "learning_rate": 3.675322238398186e-05,
+      "loss": 1.3864,
+      "num_input_tokens_seen": 295122992,
+      "step": 7510
+    },
+    {
+      "epoch": 0.36439184237145433,
+      "grad_norm": 0.4303184151649475,
+      "learning_rate": 3.671857782510897e-05,
+      "loss": 1.4116,
+      "num_input_tokens_seen": 295537492,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3648764059916292,
+      "grad_norm": 0.4150845408439636,
+      "learning_rate": 3.668390440090212e-05,
+      "loss": 1.3549,
+      "num_input_tokens_seen": 295910304,
+      "step": 7530
+    },
+    {
+      "epoch": 0.365360969611804,
+      "grad_norm": 0.39363113045692444,
+      "learning_rate": 3.6649202196769284e-05,
+      "loss": 1.3322,
+      "num_input_tokens_seen": 296280276,
+      "step": 7540
+    },
+    {
+      "epoch": 0.36584553323197877,
+      "grad_norm": 0.395766943693161,
+      "learning_rate": 3.6614471298189323e-05,
+      "loss": 1.3772,
+      "num_input_tokens_seen": 296664476,
+      "step": 7550
+    },
+    {
+      "epoch": 0.36633009685215356,
+      "grad_norm": 0.3787136971950531,
+      "learning_rate": 3.6579711790711777e-05,
+      "loss": 1.3246,
+      "num_input_tokens_seen": 297062148,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3668146604723284,
+      "grad_norm": 0.44149187207221985,
+      "learning_rate": 3.654492375995666e-05,
+      "loss": 1.3614,
+      "num_input_tokens_seen": 297468140,
+      "step": 7570
+    },
+    {
+      "epoch": 0.3672992240925032,
+      "grad_norm": 0.4064597189426422,
+      "learning_rate": 3.6510107291614254e-05,
+      "loss": 1.3648,
+      "num_input_tokens_seen": 297882592,
+      "step": 7580
+    },
+    {
+      "epoch": 0.367783787712678,
+      "grad_norm": 0.425382137298584,
+      "learning_rate": 3.647526247144486e-05,
+      "loss": 1.3475,
+      "num_input_tokens_seen": 298248144,
+      "step": 7590
+    },
+    {
+      "epoch": 0.3682683513328528,
+      "grad_norm": 0.4232633411884308,
+      "learning_rate": 3.644038938527866e-05,
+      "loss": 1.3726,
+      "num_input_tokens_seen": 298645072,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3687529149530276,
+      "grad_norm": 0.4248599708080292,
+      "learning_rate": 3.640548811901541e-05,
+      "loss": 1.3719,
+      "num_input_tokens_seen": 299051544,
+      "step": 7610
+    },
+    {
+      "epoch": 0.36923747857320244,
+      "grad_norm": 0.4088067412376404,
+      "learning_rate": 3.637055875862433e-05,
+      "loss": 1.3693,
+      "num_input_tokens_seen": 299444432,
+      "step": 7620
+    },
+    {
+      "epoch": 0.36972204219337723,
+      "grad_norm": 0.4372110366821289,
+      "learning_rate": 3.6335601390143797e-05,
+      "loss": 1.3766,
+      "num_input_tokens_seen": 299838044,
+      "step": 7630
+    },
+    {
+      "epoch": 0.370206605813552,
+      "grad_norm": 0.391181081533432,
+      "learning_rate": 3.630061609968121e-05,
+      "loss": 1.3456,
+      "num_input_tokens_seen": 300211940,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3706911694337268,
+      "grad_norm": 0.4158838391304016,
+      "learning_rate": 3.6265602973412736e-05,
+      "loss": 1.3756,
+      "num_input_tokens_seen": 300598508,
+      "step": 7650
+    },
+    {
+      "epoch": 0.37117573305390167,
+      "grad_norm": 0.39379456639289856,
+      "learning_rate": 3.623056209758309e-05,
+      "loss": 1.413,
+      "num_input_tokens_seen": 301012928,
+      "step": 7660
+    },
+    {
+      "epoch": 0.37166029667407646,
+      "grad_norm": 0.40179499983787537,
+      "learning_rate": 3.619549355850536e-05,
+      "loss": 1.367,
+      "num_input_tokens_seen": 301381712,
+      "step": 7670
+    },
+    {
+      "epoch": 0.37214486029425126,
+      "grad_norm": 0.3952077627182007,
+      "learning_rate": 3.616039744256078e-05,
+      "loss": 1.3613,
+      "num_input_tokens_seen": 301797060,
+      "step": 7680
+    },
+    {
+      "epoch": 0.37262942391442605,
+      "grad_norm": 0.41166162490844727,
+      "learning_rate": 3.61252738361985e-05,
+      "loss": 1.3913,
+      "num_input_tokens_seen": 302205268,
+      "step": 7690
+    },
+    {
+      "epoch": 0.37311398753460084,
+      "grad_norm": 0.38384559750556946,
+      "learning_rate": 3.609012282593538e-05,
+      "loss": 1.3422,
+      "num_input_tokens_seen": 302592424,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3735985511547757,
+      "grad_norm": 0.4590049982070923,
+      "learning_rate": 3.605494449835578e-05,
+      "loss": 1.3665,
+      "num_input_tokens_seen": 302963964,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3740831147749505,
+      "grad_norm": 0.43561649322509766,
+      "learning_rate": 3.601973894011137e-05,
+      "loss": 1.3385,
+      "num_input_tokens_seen": 303350216,
+      "step": 7720
+    },
+    {
+      "epoch": 0.3745676783951253,
+      "grad_norm": 0.42796000838279724,
+      "learning_rate": 3.598450623792088e-05,
+      "loss": 1.362,
+      "num_input_tokens_seen": 303774160,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3750522420153001,
+      "grad_norm": 0.37895604968070984,
+      "learning_rate": 3.5949246478569885e-05,
+      "loss": 1.3766,
+      "num_input_tokens_seen": 304184292,
+      "step": 7740
+    },
+    {
+      "epoch": 0.3755368056354749,
+      "grad_norm": 0.3759874105453491,
+      "learning_rate": 3.591395974891065e-05,
+      "loss": 1.3544,
+      "num_input_tokens_seen": 304615976,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3760213692556497,
+      "grad_norm": 0.4173840880393982,
+      "learning_rate": 3.5878646135861826e-05,
+      "loss": 1.3919,
+      "num_input_tokens_seen": 305014124,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3765059328758245,
+      "grad_norm": 0.45354533195495605,
+      "learning_rate": 3.5843305726408323e-05,
+      "loss": 1.3588,
+      "num_input_tokens_seen": 305416284,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3769904964959993,
+      "grad_norm": 0.3822251260280609,
+      "learning_rate": 3.580793860760103e-05,
+      "loss": 1.3708,
+      "num_input_tokens_seen": 305811776,
+      "step": 7780
+    },
+    {
+      "epoch": 0.37747506011617415,
+      "grad_norm": 0.3847945034503937,
+      "learning_rate": 3.5772544866556634e-05,
+      "loss": 1.3743,
+      "num_input_tokens_seen": 306210288,
+      "step": 7790
+    },
+    {
+      "epoch": 0.37795962373634895,
+      "grad_norm": 0.3933156132698059,
+      "learning_rate": 3.5737124590457404e-05,
+      "loss": 1.3817,
+      "num_input_tokens_seen": 306612992,
+      "step": 7800
+    },
+    {
+      "epoch": 0.37844418735652374,
+      "grad_norm": 0.42493316531181335,
+      "learning_rate": 3.570167786655096e-05,
+      "loss": 1.3504,
+      "num_input_tokens_seen": 306981668,
+      "step": 7810
+    },
+    {
+      "epoch": 0.37892875097669854,
+      "grad_norm": 0.4432273507118225,
+      "learning_rate": 3.566620478215008e-05,
+      "loss": 1.3901,
+      "num_input_tokens_seen": 307387500,
+      "step": 7820
+    },
+    {
+      "epoch": 0.37941331459687333,
+      "grad_norm": 0.4004375636577606,
+      "learning_rate": 3.5630705424632475e-05,
+      "loss": 1.4102,
+      "num_input_tokens_seen": 307810852,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3798978782170482,
+      "grad_norm": 0.38001713156700134,
+      "learning_rate": 3.5595179881440554e-05,
+      "loss": 1.4202,
+      "num_input_tokens_seen": 308189112,
+      "step": 7840
+    },
+    {
+      "epoch": 0.380382441837223,
+      "grad_norm": 0.40970560908317566,
+      "learning_rate": 3.5559628240081244e-05,
+      "loss": 1.3796,
+      "num_input_tokens_seen": 308593308,
+      "step": 7850
+    },
+    {
+      "epoch": 0.38086700545739777,
+      "grad_norm": 0.4416448473930359,
+      "learning_rate": 3.5524050588125744e-05,
+      "loss": 1.3305,
+      "num_input_tokens_seen": 308977396,
+      "step": 7860
+    },
+    {
+      "epoch": 0.38135156907757256,
+      "grad_norm": 0.3899739682674408,
+      "learning_rate": 3.548844701320934e-05,
+      "loss": 1.3648,
+      "num_input_tokens_seen": 309358540,
+      "step": 7870
+    },
+    {
+      "epoch": 0.3818361326977474,
+      "grad_norm": 0.4026392698287964,
+      "learning_rate": 3.545281760303116e-05,
+      "loss": 1.3735,
+      "num_input_tokens_seen": 309781936,
+      "step": 7880
+    },
+    {
+      "epoch": 0.3823206963179222,
+      "grad_norm": 0.41612690687179565,
+      "learning_rate": 3.5417162445353965e-05,
+      "loss": 1.346,
+      "num_input_tokens_seen": 310174720,
+      "step": 7890
+    },
+    {
+      "epoch": 0.382805259938097,
+      "grad_norm": 0.3857744336128235,
+      "learning_rate": 3.5381481628003964e-05,
+      "loss": 1.386,
+      "num_input_tokens_seen": 310557832,
+      "step": 7900
+    },
+    {
+      "epoch": 0.3832898235582718,
+      "grad_norm": 0.4121117889881134,
+      "learning_rate": 3.534577523887053e-05,
+      "loss": 1.3303,
+      "num_input_tokens_seen": 310940032,
+      "step": 7910
+    },
+    {
+      "epoch": 0.3837743871784466,
+      "grad_norm": 0.45025011897087097,
+      "learning_rate": 3.5310043365906046e-05,
+      "loss": 1.3234,
+      "num_input_tokens_seen": 311335136,
+      "step": 7920
+    },
+    {
+      "epoch": 0.38425895079862143,
+      "grad_norm": 0.3884856700897217,
+      "learning_rate": 3.527428609712569e-05,
+      "loss": 1.3849,
+      "num_input_tokens_seen": 311699972,
+      "step": 7930
+    },
+    {
+      "epoch": 0.38474351441879623,
+      "grad_norm": 0.37907713651657104,
+      "learning_rate": 3.5238503520607144e-05,
+      "loss": 1.3602,
+      "num_input_tokens_seen": 312078664,
+      "step": 7940
+    },
+    {
+      "epoch": 0.385228078038971,
+      "grad_norm": 0.37585073709487915,
+      "learning_rate": 3.520269572449047e-05,
+      "loss": 1.3895,
+      "num_input_tokens_seen": 312453588,
+      "step": 7950
+    },
+    {
+      "epoch": 0.3857126416591458,
+      "grad_norm": 0.395000696182251,
+      "learning_rate": 3.516686279697784e-05,
+      "loss": 1.3542,
+      "num_input_tokens_seen": 312865692,
+      "step": 7960
+    },
+    {
+      "epoch": 0.38619720527932067,
+      "grad_norm": 0.40303224325180054,
+      "learning_rate": 3.513100482633332e-05,
+      "loss": 1.3497,
+      "num_input_tokens_seen": 313258996,
+      "step": 7970
+    },
+    {
+      "epoch": 0.38668176889949546,
+      "grad_norm": 0.4608016312122345,
+      "learning_rate": 3.509512190088269e-05,
+      "loss": 1.3596,
+      "num_input_tokens_seen": 313656952,
+      "step": 7980
+    },
+    {
+      "epoch": 0.38716633251967025,
+      "grad_norm": 0.393044114112854,
+      "learning_rate": 3.505921410901316e-05,
+      "loss": 1.3828,
+      "num_input_tokens_seen": 314048964,
+      "step": 7990
+    },
+    {
+      "epoch": 0.38765089613984505,
+      "grad_norm": 0.41909295320510864,
+      "learning_rate": 3.50232815391732e-05,
+      "loss": 1.3764,
+      "num_input_tokens_seen": 314442716,
+      "step": 8000
+    },
+    {
+      "epoch": 0.38765089613984505,
+      "eval_loss": 1.471493124961853,
+      "eval_runtime": 4.7345,
+      "eval_samples_per_second": 31.683,
+      "eval_steps_per_second": 4.013,
+      "num_input_tokens_seen": 314442716,
+      "step": 8000
+    },
+    {
+      "epoch": 0.38813545976001984,
+      "grad_norm": 0.4128382205963135,
+      "learning_rate": 3.498732427987236e-05,
+      "loss": 1.3531,
+      "num_input_tokens_seen": 314818980,
+      "step": 8010
+    },
+    {
+      "epoch": 0.3886200233801947,
+      "grad_norm": 0.4025222659111023,
+      "learning_rate": 3.4951342419680946e-05,
+      "loss": 1.3588,
+      "num_input_tokens_seen": 315239696,
+      "step": 8020
+    },
+    {
+      "epoch": 0.3891045870003695,
+      "grad_norm": 0.4064643383026123,
+      "learning_rate": 3.491533604722987e-05,
+      "loss": 1.3689,
+      "num_input_tokens_seen": 315645796,
+      "step": 8030
+    },
+    {
+      "epoch": 0.3895891506205443,
+      "grad_norm": 0.36004194617271423,
+      "learning_rate": 3.4879305251210474e-05,
+      "loss": 1.3705,
+      "num_input_tokens_seen": 316049168,
+      "step": 8040
+    },
+    {
+      "epoch": 0.39007371424071907,
+      "grad_norm": 0.36490750312805176,
+      "learning_rate": 3.4843250120374206e-05,
+      "loss": 1.3596,
+      "num_input_tokens_seen": 316455320,
+      "step": 8050
+    },
+    {
+      "epoch": 0.3905582778608939,
+      "grad_norm": 0.3944610357284546,
+      "learning_rate": 3.4807170743532466e-05,
+      "loss": 1.3421,
+      "num_input_tokens_seen": 316841708,
+      "step": 8060
+    },
+    {
+      "epoch": 0.3910428414810687,
+      "grad_norm": 0.3791704475879669,
+      "learning_rate": 3.4771067209556405e-05,
+      "loss": 1.3642,
+      "num_input_tokens_seen": 317231160,
+      "step": 8070
+    },
+    {
+      "epoch": 0.3915274051012435,
+      "grad_norm": 0.39615359902381897,
+      "learning_rate": 3.4734939607376635e-05,
+      "loss": 1.3042,
+      "num_input_tokens_seen": 317616996,
+      "step": 8080
+    },
+    {
+      "epoch": 0.3920119687214183,
+      "grad_norm": 0.3907454013824463,
+      "learning_rate": 3.469878802598308e-05,
+      "loss": 1.3768,
+      "num_input_tokens_seen": 318013328,
+      "step": 8090
+    },
+    {
+      "epoch": 0.39249653234159315,
+      "grad_norm": 0.46760374307632446,
+      "learning_rate": 3.466261255442473e-05,
+      "loss": 1.3638,
+      "num_input_tokens_seen": 318396864,
+      "step": 8100
+    },
+    {
+      "epoch": 0.39298109596176795,
+      "grad_norm": 0.38027724623680115,
+      "learning_rate": 3.4626413281809434e-05,
+      "loss": 1.3292,
+      "num_input_tokens_seen": 318748480,
+      "step": 8110
+    },
+    {
+      "epoch": 0.39346565958194274,
+      "grad_norm": 0.38730815052986145,
+      "learning_rate": 3.4590190297303623e-05,
+      "loss": 1.3718,
+      "num_input_tokens_seen": 319146300,
+      "step": 8120
+    },
+    {
+      "epoch": 0.39395022320211753,
+      "grad_norm": 0.39373862743377686,
+      "learning_rate": 3.455394369013218e-05,
+      "loss": 1.384,
+      "num_input_tokens_seen": 319537280,
+      "step": 8130
+    },
+    {
+      "epoch": 0.3944347868222923,
+      "grad_norm": 0.40906721353530884,
+      "learning_rate": 3.4517673549578154e-05,
+      "loss": 1.3672,
+      "num_input_tokens_seen": 319904616,
+      "step": 8140
+    },
+    {
+      "epoch": 0.3949193504424672,
+      "grad_norm": 0.4988723397254944,
+      "learning_rate": 3.448137996498258e-05,
+      "loss": 1.3505,
+      "num_input_tokens_seen": 320278336,
+      "step": 8150
+    },
+    {
+      "epoch": 0.39540391406264197,
+      "grad_norm": 0.4247932434082031,
+      "learning_rate": 3.44450630257442e-05,
+      "loss": 1.3632,
+      "num_input_tokens_seen": 320673012,
+      "step": 8160
+    },
+    {
+      "epoch": 0.39588847768281676,
+      "grad_norm": 0.37421149015426636,
+      "learning_rate": 3.440872282131934e-05,
+      "loss": 1.3142,
+      "num_input_tokens_seen": 321069804,
+      "step": 8170
+    },
+    {
+      "epoch": 0.39637304130299156,
+      "grad_norm": 0.41044512391090393,
+      "learning_rate": 3.4372359441221594e-05,
+      "loss": 1.3794,
+      "num_input_tokens_seen": 321501056,
+      "step": 8180
+    },
+    {
+      "epoch": 0.3968576049231664,
+      "grad_norm": 0.3953944742679596,
+      "learning_rate": 3.4335972975021646e-05,
+      "loss": 1.3451,
+      "num_input_tokens_seen": 321913760,
+      "step": 8190
+    },
+    {
+      "epoch": 0.3973421685433412,
+      "grad_norm": 0.44824719429016113,
+      "learning_rate": 3.429956351234705e-05,
+      "loss": 1.4203,
+      "num_input_tokens_seen": 322313132,
+      "step": 8200
+    },
+    {
+      "epoch": 0.397826732163516,
+      "grad_norm": 0.4102291166782379,
+      "learning_rate": 3.426313114288203e-05,
+      "loss": 1.3424,
+      "num_input_tokens_seen": 322675076,
+      "step": 8210
+    },
+    {
+      "epoch": 0.3983112957836908,
+      "grad_norm": 0.375482976436615,
+      "learning_rate": 3.4226675956367195e-05,
+      "loss": 1.3142,
+      "num_input_tokens_seen": 323090684,
+      "step": 8220
+    },
+    {
+      "epoch": 0.3987958594038656,
+      "grad_norm": 0.42846444249153137,
+      "learning_rate": 3.419019804259937e-05,
+      "loss": 1.3529,
+      "num_input_tokens_seen": 323463428,
+      "step": 8230
+    },
+    {
+      "epoch": 0.39928042302404043,
+      "grad_norm": 0.42404618859291077,
+      "learning_rate": 3.4153697491431375e-05,
+      "loss": 1.3657,
+      "num_input_tokens_seen": 323846056,
+      "step": 8240
+    },
+    {
+      "epoch": 0.3997649866442152,
+      "grad_norm": 0.40639814734458923,
+      "learning_rate": 3.411717439277178e-05,
+      "loss": 1.3446,
+      "num_input_tokens_seen": 324257868,
+      "step": 8250
+    },
+    {
+      "epoch": 0.40024955026439,
+      "grad_norm": 0.41451844573020935,
+      "learning_rate": 3.40806288365847e-05,
+      "loss": 1.379,
+      "num_input_tokens_seen": 324644548,
+      "step": 8260
+    },
+    {
+      "epoch": 0.4007341138845648,
+      "grad_norm": 0.42741525173187256,
+      "learning_rate": 3.404406091288956e-05,
+      "loss": 1.383,
+      "num_input_tokens_seen": 325051824,
+      "step": 8270
+    },
+    {
+      "epoch": 0.40121867750473966,
+      "grad_norm": 0.35797035694122314,
+      "learning_rate": 3.4007470711760885e-05,
+      "loss": 1.3184,
+      "num_input_tokens_seen": 325435196,
+      "step": 8280
+    },
+    {
+      "epoch": 0.40170324112491446,
+      "grad_norm": 0.44313570857048035,
+      "learning_rate": 3.397085832332808e-05,
+      "loss": 1.3761,
+      "num_input_tokens_seen": 325815308,
+      "step": 8290
+    },
+    {
+      "epoch": 0.40218780474508925,
+      "grad_norm": 0.4352870285511017,
+      "learning_rate": 3.393422383777518e-05,
+      "loss": 1.3331,
+      "num_input_tokens_seen": 326215248,
+      "step": 8300
+    },
+    {
+      "epoch": 0.40267236836526404,
+      "grad_norm": 0.427741140127182,
+      "learning_rate": 3.389756734534069e-05,
+      "loss": 1.3546,
+      "num_input_tokens_seen": 326587588,
+      "step": 8310
+    },
+    {
+      "epoch": 0.40315693198543884,
+      "grad_norm": 0.39280158281326294,
+      "learning_rate": 3.386088893631727e-05,
+      "loss": 1.3251,
+      "num_input_tokens_seen": 326969008,
+      "step": 8320
+    },
+    {
+      "epoch": 0.4036414956056137,
+      "grad_norm": 0.37476789951324463,
+      "learning_rate": 3.382418870105161e-05,
+      "loss": 1.3944,
+      "num_input_tokens_seen": 327337052,
+      "step": 8330
+    },
+    {
+      "epoch": 0.4041260592257885,
+      "grad_norm": 0.3965238332748413,
+      "learning_rate": 3.3787466729944156e-05,
+      "loss": 1.3314,
+      "num_input_tokens_seen": 327726100,
+      "step": 8340
+    },
+    {
+      "epoch": 0.4046106228459633,
+      "grad_norm": 0.3928758203983307,
+      "learning_rate": 3.375072311344887e-05,
+      "loss": 1.3753,
+      "num_input_tokens_seen": 328097960,
+      "step": 8350
+    },
+    {
+      "epoch": 0.40509518646613807,
+      "grad_norm": 0.4109261631965637,
+      "learning_rate": 3.371395794207304e-05,
+      "loss": 1.3061,
+      "num_input_tokens_seen": 328495520,
+      "step": 8360
+    },
+    {
+      "epoch": 0.4055797500863129,
+      "grad_norm": 0.3748038113117218,
+      "learning_rate": 3.3677171306377066e-05,
+      "loss": 1.3259,
+      "num_input_tokens_seen": 328901644,
+      "step": 8370
+    },
+    {
+      "epoch": 0.4060643137064877,
+      "grad_norm": 0.40184059739112854,
+      "learning_rate": 3.36403632969742e-05,
+      "loss": 1.372,
+      "num_input_tokens_seen": 329298316,
+      "step": 8380
+    },
+    {
+      "epoch": 0.4065488773266625,
+      "grad_norm": 0.42977675795555115,
+      "learning_rate": 3.360353400453035e-05,
+      "loss": 1.3228,
+      "num_input_tokens_seen": 329697248,
+      "step": 8390
+    },
+    {
+      "epoch": 0.4070334409468373,
+      "grad_norm": 0.42152881622314453,
+      "learning_rate": 3.356668351976385e-05,
+      "loss": 1.3153,
+      "num_input_tokens_seen": 330092604,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4075180045670121,
+      "grad_norm": 0.42566007375717163,
+      "learning_rate": 3.352981193344523e-05,
+      "loss": 1.3555,
+      "num_input_tokens_seen": 330476288,
+      "step": 8410
+    },
+    {
+      "epoch": 0.40800256818718694,
+      "grad_norm": 0.4720443785190582,
+      "learning_rate": 3.349291933639701e-05,
+      "loss": 1.4095,
+      "num_input_tokens_seen": 330852500,
+      "step": 8420
+    },
+    {
+      "epoch": 0.40848713180736174,
+      "grad_norm": 0.4077743887901306,
+      "learning_rate": 3.345600581949344e-05,
+      "loss": 1.3875,
+      "num_input_tokens_seen": 331243588,
+      "step": 8430
+    },
+    {
+      "epoch": 0.40897169542753653,
+      "grad_norm": 0.37246859073638916,
+      "learning_rate": 3.3419071473660316e-05,
+      "loss": 1.34,
+      "num_input_tokens_seen": 331625036,
+      "step": 8440
+    },
+    {
+      "epoch": 0.4094562590477113,
+      "grad_norm": 0.40044400095939636,
+      "learning_rate": 3.338211638987475e-05,
+      "loss": 1.3883,
+      "num_input_tokens_seen": 332007224,
+      "step": 8450
+    },
+    {
+      "epoch": 0.4099408226678862,
+      "grad_norm": 0.40075942873954773,
+      "learning_rate": 3.33451406591649e-05,
+      "loss": 1.3145,
+      "num_input_tokens_seen": 332377820,
+      "step": 8460
+    },
+    {
+      "epoch": 0.41042538628806097,
+      "grad_norm": 0.41642826795578003,
+      "learning_rate": 3.330814437260983e-05,
+      "loss": 1.3723,
+      "num_input_tokens_seen": 332768892,
+      "step": 8470
+    },
+    {
+      "epoch": 0.41090994990823576,
+      "grad_norm": 0.45073971152305603,
+      "learning_rate": 3.32711276213392e-05,
+      "loss": 1.3439,
+      "num_input_tokens_seen": 333169136,
+      "step": 8480
+    },
+    {
+      "epoch": 0.41139451352841055,
+      "grad_norm": 0.4418103098869324,
+      "learning_rate": 3.32340904965331e-05,
+      "loss": 1.3355,
+      "num_input_tokens_seen": 333560236,
+      "step": 8490
+    },
+    {
+      "epoch": 0.4118790771485854,
+      "grad_norm": 0.38057634234428406,
+      "learning_rate": 3.3197033089421794e-05,
+      "loss": 1.3748,
+      "num_input_tokens_seen": 333908344,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4123636407687602,
+      "grad_norm": 0.38415491580963135,
+      "learning_rate": 3.31599554912855e-05,
+      "loss": 1.3595,
+      "num_input_tokens_seen": 334331680,
+      "step": 8510
+    },
+    {
+      "epoch": 0.412848204388935,
+      "grad_norm": 0.4075462818145752,
+      "learning_rate": 3.3122857793454186e-05,
+      "loss": 1.349,
+      "num_input_tokens_seen": 334721924,
+      "step": 8520
+    },
+    {
+      "epoch": 0.4133327680091098,
+      "grad_norm": 0.4361826479434967,
+      "learning_rate": 3.308574008730732e-05,
+      "loss": 1.3445,
+      "num_input_tokens_seen": 335116152,
+      "step": 8530
+    },
+    {
+      "epoch": 0.4138173316292846,
+      "grad_norm": 0.4111301898956299,
+      "learning_rate": 3.304860246427366e-05,
+      "loss": 1.388,
+      "num_input_tokens_seen": 335498760,
+      "step": 8540
+    },
+    {
+      "epoch": 0.41430189524945943,
+      "grad_norm": 0.4673983156681061,
+      "learning_rate": 3.301144501583102e-05,
+      "loss": 1.3624,
+      "num_input_tokens_seen": 335894792,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4147864588696342,
+      "grad_norm": 0.4051591157913208,
+      "learning_rate": 3.297426783350606e-05,
+      "loss": 1.2684,
+      "num_input_tokens_seen": 336281536,
+      "step": 8560
+    },
+    {
+      "epoch": 0.415271022489809,
+      "grad_norm": 0.4352741241455078,
+      "learning_rate": 3.293707100887401e-05,
+      "loss": 1.336,
+      "num_input_tokens_seen": 336691052,
+      "step": 8570
+    },
+    {
+      "epoch": 0.4157555861099838,
+      "grad_norm": 0.43101462721824646,
+      "learning_rate": 3.2899854633558534e-05,
+      "loss": 1.3345,
+      "num_input_tokens_seen": 337098468,
+      "step": 8580
+    },
+    {
+      "epoch": 0.41624014973015866,
+      "grad_norm": 0.43431204557418823,
+      "learning_rate": 3.2862618799231424e-05,
+      "loss": 1.3372,
+      "num_input_tokens_seen": 337498280,
+      "step": 8590
+    },
+    {
+      "epoch": 0.41672471335033345,
+      "grad_norm": 0.3821309804916382,
+      "learning_rate": 3.2825363597612405e-05,
+      "loss": 1.3787,
+      "num_input_tokens_seen": 337889884,
+      "step": 8600
+    },
+    {
+      "epoch": 0.41720927697050825,
+      "grad_norm": 0.39203059673309326,
+      "learning_rate": 3.2788089120468924e-05,
+      "loss": 1.3589,
+      "num_input_tokens_seen": 338259116,
+      "step": 8610
+    },
+    {
+      "epoch": 0.41769384059068304,
+      "grad_norm": 0.39438045024871826,
+      "learning_rate": 3.275079545961588e-05,
+      "loss": 1.2828,
+      "num_input_tokens_seen": 338648980,
+      "step": 8620
+    },
+    {
+      "epoch": 0.41817840421085783,
+      "grad_norm": 0.4104815423488617,
+      "learning_rate": 3.271348270691546e-05,
+      "loss": 1.3274,
+      "num_input_tokens_seen": 339078392,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4186629678310327,
+      "grad_norm": 0.39212706685066223,
+      "learning_rate": 3.2676150954276846e-05,
+      "loss": 1.3259,
+      "num_input_tokens_seen": 339453980,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4191475314512075,
+      "grad_norm": 0.3868629038333893,
+      "learning_rate": 3.263880029365604e-05,
+      "loss": 1.3381,
+      "num_input_tokens_seen": 339852160,
+      "step": 8650
+    },
+    {
+      "epoch": 0.41963209507138227,
+      "grad_norm": 0.3927762806415558,
+      "learning_rate": 3.260143081705561e-05,
+      "loss": 1.3366,
+      "num_input_tokens_seen": 340245664,
+      "step": 8660
+    },
+    {
+      "epoch": 0.42011665869155707,
+      "grad_norm": 0.4036239683628082,
+      "learning_rate": 3.256404261652449e-05,
+      "loss": 1.3157,
+      "num_input_tokens_seen": 340637260,
+      "step": 8670
+    },
+    {
+      "epoch": 0.4206012223117319,
+      "grad_norm": 0.41092178225517273,
+      "learning_rate": 3.2526635784157695e-05,
+      "loss": 1.3727,
+      "num_input_tokens_seen": 341032376,
+      "step": 8680
+    },
+    {
+      "epoch": 0.4210857859319067,
+      "grad_norm": 0.40871837735176086,
+      "learning_rate": 3.248921041209618e-05,
+      "loss": 1.3247,
+      "num_input_tokens_seen": 341418744,
+      "step": 8690
+    },
+    {
+      "epoch": 0.4215703495520815,
+      "grad_norm": 0.41571468114852905,
+      "learning_rate": 3.245176659252654e-05,
+      "loss": 1.3083,
+      "num_input_tokens_seen": 341804060,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4220549131722563,
+      "grad_norm": 0.38524648547172546,
+      "learning_rate": 3.241430441768081e-05,
+      "loss": 1.2899,
+      "num_input_tokens_seen": 342183932,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4225394767924311,
+      "grad_norm": 0.3792508542537689,
+      "learning_rate": 3.2376823979836256e-05,
+      "loss": 1.3544,
+      "num_input_tokens_seen": 342580412,
+      "step": 8720
+    },
+    {
+      "epoch": 0.42302404041260594,
+      "grad_norm": 0.3552215099334717,
+      "learning_rate": 3.233932537131511e-05,
+      "loss": 1.3571,
+      "num_input_tokens_seen": 342982212,
+      "step": 8730
+    },
+    {
+      "epoch": 0.42350860403278073,
+      "grad_norm": 0.40093865990638733,
+      "learning_rate": 3.230180868448437e-05,
+      "loss": 1.3159,
+      "num_input_tokens_seen": 343375004,
+      "step": 8740
+    },
+    {
+      "epoch": 0.4239931676529555,
+      "grad_norm": 0.42487427592277527,
+      "learning_rate": 3.2264274011755575e-05,
+      "loss": 1.3574,
+      "num_input_tokens_seen": 343760048,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4244777312731303,
+      "grad_norm": 0.370417058467865,
+      "learning_rate": 3.222672144558455e-05,
+      "loss": 1.3853,
+      "num_input_tokens_seen": 344142832,
+      "step": 8760
+    },
+    {
+      "epoch": 0.42496229489330517,
+      "grad_norm": 0.42524397373199463,
+      "learning_rate": 3.21891510784712e-05,
+      "loss": 1.3497,
+      "num_input_tokens_seen": 344534952,
+      "step": 8770
+    },
+    {
+      "epoch": 0.42544685851347996,
+      "grad_norm": 0.36123722791671753,
+      "learning_rate": 3.215156300295928e-05,
+      "loss": 1.3211,
+      "num_input_tokens_seen": 344914096,
+      "step": 8780
+    },
+    {
+      "epoch": 0.42593142213365476,
+      "grad_norm": 0.38714513182640076,
+      "learning_rate": 3.2113957311636154e-05,
+      "loss": 1.3753,
+      "num_input_tokens_seen": 345315972,
+      "step": 8790
+    },
+    {
+      "epoch": 0.42641598575382955,
+      "grad_norm": 0.38927286863327026,
+      "learning_rate": 3.207633409713262e-05,
+      "loss": 1.3256,
+      "num_input_tokens_seen": 345704748,
+      "step": 8800
+    },
+    {
+      "epoch": 0.42690054937400435,
+      "grad_norm": 0.3460540771484375,
+      "learning_rate": 3.203869345212258e-05,
+      "loss": 1.358,
+      "num_input_tokens_seen": 346089592,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4273851129941792,
+      "grad_norm": 0.39570000767707825,
+      "learning_rate": 3.20010354693229e-05,
+      "loss": 1.3397,
+      "num_input_tokens_seen": 346495852,
+      "step": 8820
+    },
+    {
+      "epoch": 0.427869676614354,
+      "grad_norm": 0.5098708271980286,
+      "learning_rate": 3.196336024149316e-05,
+      "loss": 1.3545,
+      "num_input_tokens_seen": 346912816,
+      "step": 8830
+    },
+    {
+      "epoch": 0.4283542402345288,
+      "grad_norm": 0.40923282504081726,
+      "learning_rate": 3.192566786143541e-05,
+      "loss": 1.3697,
+      "num_input_tokens_seen": 347337700,
+      "step": 8840
+    },
+    {
+      "epoch": 0.4288388038547036,
+      "grad_norm": 0.40802377462387085,
+      "learning_rate": 3.1887958421993944e-05,
+      "loss": 1.3471,
+      "num_input_tokens_seen": 347738832,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4293233674748784,
+      "grad_norm": 0.4220910668373108,
+      "learning_rate": 3.185023201605508e-05,
+      "loss": 1.3619,
+      "num_input_tokens_seen": 348131200,
+      "step": 8860
+    },
+    {
+      "epoch": 0.4298079310950532,
+      "grad_norm": 0.4093485474586487,
+      "learning_rate": 3.181248873654693e-05,
+      "loss": 1.3157,
+      "num_input_tokens_seen": 348533868,
+      "step": 8870
+    },
+    {
+      "epoch": 0.430292494715228,
+      "grad_norm": 0.4146769940853119,
+      "learning_rate": 3.177472867643917e-05,
+      "loss": 1.3552,
+      "num_input_tokens_seen": 348915400,
+      "step": 8880
+    },
+    {
+      "epoch": 0.4307770583354028,
+      "grad_norm": 0.4098985493183136,
+      "learning_rate": 3.1736951928742804e-05,
+      "loss": 1.2922,
+      "num_input_tokens_seen": 349298476,
+      "step": 8890
+    },
+    {
+      "epoch": 0.43126162195557766,
+      "grad_norm": 0.3871714174747467,
+      "learning_rate": 3.169915858650996e-05,
+      "loss": 1.3978,
+      "num_input_tokens_seen": 349662284,
+      "step": 8900
+    },
+    {
+      "epoch": 0.43174618557575245,
+      "grad_norm": 0.4378563463687897,
+      "learning_rate": 3.166134874283361e-05,
+      "loss": 1.3481,
+      "num_input_tokens_seen": 350084716,
+      "step": 8910
+    },
+    {
+      "epoch": 0.43223074919592724,
+      "grad_norm": 0.3912814259529114,
+      "learning_rate": 3.16235224908474e-05,
+      "loss": 1.3555,
+      "num_input_tokens_seen": 350479224,
+      "step": 8920
+    },
+    {
+      "epoch": 0.43271531281610204,
+      "grad_norm": 0.41960766911506653,
+      "learning_rate": 3.158567992372538e-05,
+      "loss": 1.3159,
+      "num_input_tokens_seen": 350878796,
+      "step": 8930
+    },
+    {
+      "epoch": 0.43319987643627683,
+      "grad_norm": 0.3939763009548187,
+      "learning_rate": 3.154782113468179e-05,
+      "loss": 1.3948,
+      "num_input_tokens_seen": 351259368,
+      "step": 8940
+    },
+    {
+      "epoch": 0.4336844400564517,
+      "grad_norm": 0.37786418199539185,
+      "learning_rate": 3.1509946216970844e-05,
+      "loss": 1.3577,
+      "num_input_tokens_seen": 351642444,
+      "step": 8950
+    },
+    {
+      "epoch": 0.4341690036766265,
+      "grad_norm": 0.41558992862701416,
+      "learning_rate": 3.1472055263886443e-05,
+      "loss": 1.336,
+      "num_input_tokens_seen": 352081616,
+      "step": 8960
+    },
+    {
+      "epoch": 0.43465356729680127,
+      "grad_norm": 0.407986044883728,
+      "learning_rate": 3.143414836876204e-05,
+      "loss": 1.3638,
+      "num_input_tokens_seen": 352492396,
+      "step": 8970
+    },
+    {
+      "epoch": 0.43513813091697606,
+      "grad_norm": 0.3700858950614929,
+      "learning_rate": 3.13962256249703e-05,
+      "loss": 1.3362,
+      "num_input_tokens_seen": 352908204,
+      "step": 8980
+    },
+    {
+      "epoch": 0.4356226945371509,
+      "grad_norm": 0.43616652488708496,
+      "learning_rate": 3.1358287125922986e-05,
+      "loss": 1.3545,
+      "num_input_tokens_seen": 353322600,
+      "step": 8990
+    },
+    {
+      "epoch": 0.4361072581573257,
+      "grad_norm": 0.40746721625328064,
+      "learning_rate": 3.132033296507063e-05,
+      "loss": 1.3484,
+      "num_input_tokens_seen": 353715168,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4365918217775005,
+      "grad_norm": 0.3873213529586792,
+      "learning_rate": 3.128236323590234e-05,
+      "loss": 1.338,
+      "num_input_tokens_seen": 354097720,
+      "step": 9010
+    },
+    {
+      "epoch": 0.4370763853976753,
+      "grad_norm": 0.37148579955101013,
+      "learning_rate": 3.1244378031945585e-05,
+      "loss": 1.3348,
+      "num_input_tokens_seen": 354523140,
+      "step": 9020
+    },
+    {
+      "epoch": 0.4375609490178501,
+      "grad_norm": 0.4213094711303711,
+      "learning_rate": 3.1206377446765966e-05,
+      "loss": 1.304,
+      "num_input_tokens_seen": 354914168,
+      "step": 9030
+    },
+    {
+      "epoch": 0.43804551263802494,
+      "grad_norm": 0.4143025875091553,
+      "learning_rate": 3.1168361573966945e-05,
+      "loss": 1.3531,
+      "num_input_tokens_seen": 355288384,
+      "step": 9040
+    },
+    {
+      "epoch": 0.43853007625819973,
+      "grad_norm": 0.4011501669883728,
+      "learning_rate": 3.113033050718966e-05,
+      "loss": 1.3663,
+      "num_input_tokens_seen": 355652940,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4390146398783745,
+      "grad_norm": 0.37911659479141235,
+      "learning_rate": 3.109228434011265e-05,
+      "loss": 1.3144,
+      "num_input_tokens_seen": 356034868,
+      "step": 9060
+    },
+    {
+      "epoch": 0.4394992034985493,
+      "grad_norm": 0.4045577049255371,
+      "learning_rate": 3.105422316645169e-05,
+      "loss": 1.3398,
+      "num_input_tokens_seen": 356432868,
+      "step": 9070
+    },
+    {
+      "epoch": 0.43998376711872417,
+      "grad_norm": 0.4345623254776001,
+      "learning_rate": 3.101614707995948e-05,
+      "loss": 1.3056,
+      "num_input_tokens_seen": 356832276,
+      "step": 9080
+    },
+    {
+      "epoch": 0.44046833073889896,
+      "grad_norm": 0.42983555793762207,
+      "learning_rate": 3.097805617442546e-05,
+      "loss": 1.3681,
+      "num_input_tokens_seen": 357220024,
+      "step": 9090
+    },
+    {
+      "epoch": 0.44095289435907375,
+      "grad_norm": 0.3999558389186859,
+      "learning_rate": 3.09399505436756e-05,
+      "loss": 1.3268,
+      "num_input_tokens_seen": 357614528,
+      "step": 9100
+    },
+    {
+      "epoch": 0.44143745797924855,
+      "grad_norm": 0.4042194187641144,
+      "learning_rate": 3.090183028157211e-05,
+      "loss": 1.3262,
+      "num_input_tokens_seen": 357978768,
+      "step": 9110
+    },
+    {
+      "epoch": 0.44192202159942334,
+      "grad_norm": 0.4250580370426178,
+      "learning_rate": 3.086369548201326e-05,
+      "loss": 1.3547,
+      "num_input_tokens_seen": 358366236,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4424065852195982,
+      "grad_norm": 0.394369900226593,
+      "learning_rate": 3.082554623893312e-05,
+      "loss": 1.3799,
+      "num_input_tokens_seen": 358751508,
+      "step": 9130
+    },
+    {
+      "epoch": 0.442891148839773,
+      "grad_norm": 0.38379648327827454,
+      "learning_rate": 3.0787382646301324e-05,
+      "loss": 1.3512,
+      "num_input_tokens_seen": 359143272,
+      "step": 9140
+    },
+    {
+      "epoch": 0.4433757124599478,
+      "grad_norm": 0.3755399286746979,
+      "learning_rate": 3.074920479812289e-05,
+      "loss": 1.3386,
+      "num_input_tokens_seen": 359526224,
+      "step": 9150
+    },
+    {
+      "epoch": 0.4438602760801226,
+      "grad_norm": 0.4327252209186554,
+      "learning_rate": 3.0711012788437916e-05,
+      "loss": 1.3552,
+      "num_input_tokens_seen": 359890840,
+      "step": 9160
+    },
+    {
+      "epoch": 0.4443448397002974,
+      "grad_norm": 0.4325873851776123,
+      "learning_rate": 3.067280671132139e-05,
+      "loss": 1.2917,
+      "num_input_tokens_seen": 360291064,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4448294033204722,
+      "grad_norm": 0.45133915543556213,
+      "learning_rate": 3.063458666088296e-05,
+      "loss": 1.3375,
+      "num_input_tokens_seen": 360685988,
+      "step": 9180
+    },
+    {
+      "epoch": 0.445313966940647,
+      "grad_norm": 0.42375510931015015,
+      "learning_rate": 3.0596352731266684e-05,
+      "loss": 1.3736,
+      "num_input_tokens_seen": 361111956,
+      "step": 9190
+    },
+    {
+      "epoch": 0.4457985305608218,
+      "grad_norm": 0.3929903209209442,
+      "learning_rate": 3.055810501665082e-05,
+      "loss": 1.3429,
+      "num_input_tokens_seen": 361512340,
+      "step": 9200
+    },
+    {
+      "epoch": 0.44628309418099665,
+      "grad_norm": 0.4164799749851227,
+      "learning_rate": 3.051984361124756e-05,
+      "loss": 1.3907,
+      "num_input_tokens_seen": 361904800,
+      "step": 9210
+    },
+    {
+      "epoch": 0.44676765780117145,
+      "grad_norm": 0.39341410994529724,
+      "learning_rate": 3.0481568609302846e-05,
+      "loss": 1.3494,
+      "num_input_tokens_seen": 362296512,
+      "step": 9220
+    },
+    {
+      "epoch": 0.44725222142134624,
+      "grad_norm": 0.37080663442611694,
+      "learning_rate": 3.0443280105096096e-05,
+      "loss": 1.3476,
+      "num_input_tokens_seen": 362693456,
+      "step": 9230
+    },
+    {
+      "epoch": 0.44773678504152103,
+      "grad_norm": 0.436410129070282,
+      "learning_rate": 3.0404978192939974e-05,
+      "loss": 1.3445,
+      "num_input_tokens_seen": 363121352,
+      "step": 9240
+    },
+    {
+      "epoch": 0.44822134866169583,
+      "grad_norm": 0.39854753017425537,
+      "learning_rate": 3.0366662967180198e-05,
+      "loss": 1.3764,
+      "num_input_tokens_seen": 363511128,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4487059122818707,
+      "grad_norm": 0.4296603202819824,
+      "learning_rate": 3.0328334522195262e-05,
+      "loss": 1.3258,
+      "num_input_tokens_seen": 363885672,
+      "step": 9260
+    },
+    {
+      "epoch": 0.44919047590204547,
+      "grad_norm": 0.42518150806427,
+      "learning_rate": 3.0289992952396234e-05,
+      "loss": 1.2693,
+      "num_input_tokens_seen": 364269780,
+      "step": 9270
+    },
+    {
+      "epoch": 0.44967503952222027,
+      "grad_norm": 0.4305862784385681,
+      "learning_rate": 3.0251638352226495e-05,
+      "loss": 1.3471,
+      "num_input_tokens_seen": 364672480,
+      "step": 9280
+    },
+    {
+      "epoch": 0.45015960314239506,
+      "grad_norm": 0.3815818727016449,
+      "learning_rate": 3.0213270816161536e-05,
+      "loss": 1.3163,
+      "num_input_tokens_seen": 365041504,
+      "step": 9290
+    },
+    {
+      "epoch": 0.4506441667625699,
+      "grad_norm": 0.40457382798194885,
+      "learning_rate": 3.0174890438708715e-05,
+      "loss": 1.3232,
+      "num_input_tokens_seen": 365438776,
+      "step": 9300
+    },
+    {
+      "epoch": 0.4511287303827447,
+      "grad_norm": 0.43254563212394714,
+      "learning_rate": 3.0136497314406992e-05,
+      "loss": 1.3224,
+      "num_input_tokens_seen": 365820180,
+      "step": 9310
+    },
+    {
+      "epoch": 0.4516132940029195,
+      "grad_norm": 0.39054685831069946,
+      "learning_rate": 3.0098091537826766e-05,
+      "loss": 1.3043,
+      "num_input_tokens_seen": 366203692,
+      "step": 9320
+    },
+    {
+      "epoch": 0.4520978576230943,
+      "grad_norm": 0.3737923502922058,
+      "learning_rate": 3.0059673203569572e-05,
+      "loss": 1.3191,
+      "num_input_tokens_seen": 366605944,
+      "step": 9330
+    },
+    {
+      "epoch": 0.4525824212432691,
+      "grad_norm": 0.47919556498527527,
+      "learning_rate": 3.0021242406267892e-05,
+      "loss": 1.3303,
+      "num_input_tokens_seen": 366992400,
+      "step": 9340
+    },
+    {
+      "epoch": 0.45306698486344393,
+      "grad_norm": 0.44770705699920654,
+      "learning_rate": 2.9982799240584907e-05,
+      "loss": 1.3101,
+      "num_input_tokens_seen": 367423000,
+      "step": 9350
+    },
+    {
+      "epoch": 0.4535515484836187,
+      "grad_norm": 0.3613060712814331,
+      "learning_rate": 2.9944343801214253e-05,
+      "loss": 1.3585,
+      "num_input_tokens_seen": 367812296,
+      "step": 9360
+    },
+    {
+      "epoch": 0.4540361121037935,
+      "grad_norm": 0.40847277641296387,
+      "learning_rate": 2.9905876182879806e-05,
+      "loss": 1.2867,
+      "num_input_tokens_seen": 368210368,
+      "step": 9370
+    },
+    {
+      "epoch": 0.4545206757239683,
+      "grad_norm": 0.4815332889556885,
+      "learning_rate": 2.986739648033544e-05,
+      "loss": 1.4084,
+      "num_input_tokens_seen": 368581832,
+      "step": 9380
+    },
+    {
+      "epoch": 0.45500523934414316,
+      "grad_norm": 0.4206528663635254,
+      "learning_rate": 2.9828904788364785e-05,
+      "loss": 1.3391,
+      "num_input_tokens_seen": 368993896,
+      "step": 9390
+    },
+    {
+      "epoch": 0.45548980296431796,
+      "grad_norm": 0.43159064650535583,
+      "learning_rate": 2.9790401201781037e-05,
+      "loss": 1.3808,
+      "num_input_tokens_seen": 369375788,
+      "step": 9400
+    },
+    {
+      "epoch": 0.45597436658449275,
+      "grad_norm": 0.43968769907951355,
+      "learning_rate": 2.975188581542665e-05,
+      "loss": 1.3115,
+      "num_input_tokens_seen": 369796992,
+      "step": 9410
+    },
+    {
+      "epoch": 0.45645893020466755,
+      "grad_norm": 0.40895915031433105,
+      "learning_rate": 2.9713358724173167e-05,
+      "loss": 1.3758,
+      "num_input_tokens_seen": 370171640,
+      "step": 9420
+    },
+    {
+      "epoch": 0.45694349382484234,
+      "grad_norm": 0.3732263147830963,
+      "learning_rate": 2.9674820022920953e-05,
+      "loss": 1.3431,
+      "num_input_tokens_seen": 370532156,
+      "step": 9430
+    },
+    {
+      "epoch": 0.4574280574450172,
+      "grad_norm": 0.38699179887771606,
+      "learning_rate": 2.963626980659898e-05,
+      "loss": 1.295,
+      "num_input_tokens_seen": 370950576,
+      "step": 9440
+    },
+    {
+      "epoch": 0.457912621065192,
+      "grad_norm": 0.4024085998535156,
+      "learning_rate": 2.9597708170164567e-05,
+      "loss": 1.3202,
+      "num_input_tokens_seen": 371331076,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4583971846853668,
+      "grad_norm": 0.3956678509712219,
+      "learning_rate": 2.955913520860319e-05,
+      "loss": 1.3289,
+      "num_input_tokens_seen": 371714060,
+      "step": 9460
+    },
+    {
+      "epoch": 0.45888174830554157,
+      "grad_norm": 0.41495439410209656,
+      "learning_rate": 2.9520551016928193e-05,
+      "loss": 1.3285,
+      "num_input_tokens_seen": 372137132,
+      "step": 9470
+    },
+    {
+      "epoch": 0.4593663119257164,
+      "grad_norm": 0.3949216604232788,
+      "learning_rate": 2.9481955690180606e-05,
+      "loss": 1.2749,
+      "num_input_tokens_seen": 372529012,
+      "step": 9480
+    },
+    {
+      "epoch": 0.4598508755458912,
+      "grad_norm": 0.3913583755493164,
+      "learning_rate": 2.9443349323428876e-05,
+      "loss": 1.3173,
+      "num_input_tokens_seen": 372945044,
+      "step": 9490
+    },
+    {
+      "epoch": 0.460335439166066,
+      "grad_norm": 0.4193466901779175,
+      "learning_rate": 2.9404732011768632e-05,
+      "loss": 1.2946,
+      "num_input_tokens_seen": 373331840,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4608200027862408,
+      "grad_norm": 0.4153105914592743,
+      "learning_rate": 2.936610385032249e-05,
+      "loss": 1.3241,
+      "num_input_tokens_seen": 373726396,
+      "step": 9510
+    },
+    {
+      "epoch": 0.4613045664064156,
+      "grad_norm": 0.41479748487472534,
+      "learning_rate": 2.932746493423976e-05,
+      "loss": 1.3622,
+      "num_input_tokens_seen": 374113504,
+      "step": 9520
+    },
+    {
+      "epoch": 0.46178913002659044,
+      "grad_norm": 0.4534710645675659,
+      "learning_rate": 2.9288815358696265e-05,
+      "loss": 1.3537,
+      "num_input_tokens_seen": 374495524,
+      "step": 9530
+    },
+    {
+      "epoch": 0.46227369364676524,
+      "grad_norm": 0.36162427067756653,
+      "learning_rate": 2.9250155218894083e-05,
+      "loss": 1.2787,
+      "num_input_tokens_seen": 374880396,
+      "step": 9540
+    },
+    {
+      "epoch": 0.46275825726694003,
+      "grad_norm": 0.36983734369277954,
+      "learning_rate": 2.9211484610061307e-05,
+      "loss": 1.3314,
+      "num_input_tokens_seen": 375287296,
+      "step": 9550
+    },
+    {
+      "epoch": 0.4632428208871148,
+      "grad_norm": 0.4066306948661804,
+      "learning_rate": 2.9172803627451817e-05,
+      "loss": 1.3169,
+      "num_input_tokens_seen": 375663144,
+      "step": 9560
+    },
+    {
+      "epoch": 0.4637273845072897,
+      "grad_norm": 0.3863866329193115,
+      "learning_rate": 2.9134112366345055e-05,
+      "loss": 1.3676,
+      "num_input_tokens_seen": 376030684,
+      "step": 9570
+    },
+    {
+      "epoch": 0.46421194812746447,
+      "grad_norm": 0.4485596716403961,
+      "learning_rate": 2.909541092204576e-05,
+      "loss": 1.3345,
+      "num_input_tokens_seen": 376376536,
+      "step": 9580
+    },
+    {
+      "epoch": 0.46469651174763926,
+      "grad_norm": 0.4069863259792328,
+      "learning_rate": 2.9056699389883783e-05,
+      "loss": 1.2957,
+      "num_input_tokens_seen": 376751920,
+      "step": 9590
+    },
+    {
+      "epoch": 0.46518107536781406,
+      "grad_norm": 0.41476011276245117,
+      "learning_rate": 2.9017977865213814e-05,
+      "loss": 1.323,
+      "num_input_tokens_seen": 377165456,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4656656389879889,
+      "grad_norm": 0.38815972208976746,
+      "learning_rate": 2.8979246443415132e-05,
+      "loss": 1.2949,
+      "num_input_tokens_seen": 377554888,
+      "step": 9610
+    },
+    {
+      "epoch": 0.4661502026081637,
+      "grad_norm": 0.39285364747047424,
+      "learning_rate": 2.8940505219891432e-05,
+      "loss": 1.2969,
+      "num_input_tokens_seen": 377948944,
+      "step": 9620
+    },
+    {
+      "epoch": 0.4666347662283385,
+      "grad_norm": 0.43776214122772217,
+      "learning_rate": 2.890175429007054e-05,
+      "loss": 1.3136,
+      "num_input_tokens_seen": 378360736,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4671193298485133,
+      "grad_norm": 0.3744518458843231,
+      "learning_rate": 2.8862993749404166e-05,
+      "loss": 1.3181,
+      "num_input_tokens_seen": 378760840,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4676038934686881,
+      "grad_norm": 0.3794598877429962,
+      "learning_rate": 2.8824223693367724e-05,
+      "loss": 1.3395,
+      "num_input_tokens_seen": 379173080,
+      "step": 9650
+    },
+    {
+      "epoch": 0.46808845708886293,
+      "grad_norm": 0.37090277671813965,
+      "learning_rate": 2.8785444217460067e-05,
+      "loss": 1.315,
+      "num_input_tokens_seen": 379569804,
+      "step": 9660
+    },
+    {
+      "epoch": 0.4685730207090377,
+      "grad_norm": 0.4118629992008209,
+      "learning_rate": 2.8746655417203216e-05,
+      "loss": 1.3049,
+      "num_input_tokens_seen": 379963480,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4690575843292125,
+      "grad_norm": 0.4254318177700043,
+      "learning_rate": 2.8707857388142212e-05,
+      "loss": 1.331,
+      "num_input_tokens_seen": 380393192,
+      "step": 9680
+    },
+    {
+      "epoch": 0.4695421479493873,
+      "grad_norm": 0.4114190638065338,
+      "learning_rate": 2.866905022584478e-05,
+      "loss": 1.2986,
+      "num_input_tokens_seen": 380779764,
+      "step": 9690
+    },
+    {
+      "epoch": 0.47002671156956216,
+      "grad_norm": 0.4168798625469208,
+      "learning_rate": 2.8630234025901175e-05,
+      "loss": 1.2768,
+      "num_input_tokens_seen": 381172796,
+      "step": 9700
+    },
+    {
+      "epoch": 0.47051127518973695,
+      "grad_norm": 0.39690959453582764,
+      "learning_rate": 2.8591408883923892e-05,
+      "loss": 1.2874,
+      "num_input_tokens_seen": 381548028,
+      "step": 9710
+    },
+    {
+      "epoch": 0.47099583880991175,
+      "grad_norm": 0.4413483440876007,
+      "learning_rate": 2.8552574895547468e-05,
+      "loss": 1.3259,
+      "num_input_tokens_seen": 381920092,
+      "step": 9720
+    },
+    {
+      "epoch": 0.47148040243008654,
+      "grad_norm": 0.3835400640964508,
+      "learning_rate": 2.8513732156428224e-05,
+      "loss": 1.3392,
+      "num_input_tokens_seen": 382302052,
+      "step": 9730
+    },
+    {
+      "epoch": 0.47196496605026134,
+      "grad_norm": 0.3888629972934723,
+      "learning_rate": 2.8474880762244034e-05,
+      "loss": 1.3213,
+      "num_input_tokens_seen": 382707108,
+      "step": 9740
+    },
+    {
+      "epoch": 0.4724495296704362,
+      "grad_norm": 0.42624571919441223,
+      "learning_rate": 2.8436020808694086e-05,
+      "loss": 1.3378,
+      "num_input_tokens_seen": 383075220,
+      "step": 9750
+    },
+    {
+      "epoch": 0.472934093290611,
+      "grad_norm": 0.40198907256126404,
+      "learning_rate": 2.8397152391498677e-05,
+      "loss": 1.3228,
+      "num_input_tokens_seen": 383491136,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4734186569107858,
+      "grad_norm": 0.4753158390522003,
+      "learning_rate": 2.835827560639892e-05,
+      "loss": 1.3172,
+      "num_input_tokens_seen": 383869604,
+      "step": 9770
+    },
+    {
+      "epoch": 0.47390322053096057,
+      "grad_norm": 0.4025874733924866,
+      "learning_rate": 2.831939054915656e-05,
+      "loss": 1.3099,
+      "num_input_tokens_seen": 384263944,
+      "step": 9780
+    },
+    {
+      "epoch": 0.4743877841511354,
+      "grad_norm": 0.4010298550128937,
+      "learning_rate": 2.8280497315553705e-05,
+      "loss": 1.2988,
+      "num_input_tokens_seen": 384659820,
+      "step": 9790
+    },
+    {
+      "epoch": 0.4748723477713102,
+      "grad_norm": 0.3888191878795624,
+      "learning_rate": 2.8241596001392617e-05,
+      "loss": 1.3535,
+      "num_input_tokens_seen": 385047304,
+      "step": 9800
+    },
+    {
+      "epoch": 0.475356911391485,
+      "grad_norm": 0.4061359763145447,
+      "learning_rate": 2.8202686702495447e-05,
+      "loss": 1.3124,
+      "num_input_tokens_seen": 385440160,
+      "step": 9810
+    },
+    {
+      "epoch": 0.4758414750116598,
+      "grad_norm": 0.3900851607322693,
+      "learning_rate": 2.816376951470402e-05,
+      "loss": 1.3093,
+      "num_input_tokens_seen": 385848164,
+      "step": 9820
+    },
+    {
+      "epoch": 0.4763260386318346,
+      "grad_norm": 0.39700040221214294,
+      "learning_rate": 2.8124844533879607e-05,
+      "loss": 1.31,
+      "num_input_tokens_seen": 386286824,
+      "step": 9830
+    },
+    {
+      "epoch": 0.47681060225200944,
+      "grad_norm": 0.4107182025909424,
+      "learning_rate": 2.808591185590265e-05,
+      "loss": 1.2721,
+      "num_input_tokens_seen": 386666264,
+      "step": 9840
+    },
+    {
+      "epoch": 0.47729516587218423,
+      "grad_norm": 0.392866313457489,
+      "learning_rate": 2.8046971576672582e-05,
+      "loss": 1.3607,
+      "num_input_tokens_seen": 387054376,
+      "step": 9850
+    },
+    {
+      "epoch": 0.47777972949235903,
+      "grad_norm": 0.380993127822876,
+      "learning_rate": 2.8008023792107512e-05,
+      "loss": 1.3442,
+      "num_input_tokens_seen": 387450084,
+      "step": 9860
+    },
+    {
+      "epoch": 0.4782642931125338,
+      "grad_norm": 0.4111970365047455,
+      "learning_rate": 2.7969068598144095e-05,
+      "loss": 1.3341,
+      "num_input_tokens_seen": 387853932,
+      "step": 9870
+    },
+    {
+      "epoch": 0.47874885673270867,
+      "grad_norm": 0.385023832321167,
+      "learning_rate": 2.793010609073719e-05,
+      "loss": 1.3521,
+      "num_input_tokens_seen": 388260636,
+      "step": 9880
+    },
+    {
+      "epoch": 0.47923342035288347,
+      "grad_norm": 0.3947872519493103,
+      "learning_rate": 2.7891136365859683e-05,
+      "loss": 1.3337,
+      "num_input_tokens_seen": 388613256,
+      "step": 9890
+    },
+    {
+      "epoch": 0.47971798397305826,
+      "grad_norm": 0.36505380272865295,
+      "learning_rate": 2.7852159519502263e-05,
+      "loss": 1.3526,
+      "num_input_tokens_seen": 388996424,
+      "step": 9900
+    },
+    {
+      "epoch": 0.48020254759323305,
+      "grad_norm": 0.40470102429389954,
+      "learning_rate": 2.7813175647673123e-05,
+      "loss": 1.2916,
+      "num_input_tokens_seen": 389343832,
+      "step": 9910
+    },
+    {
+      "epoch": 0.4806871112134079,
+      "grad_norm": 0.44250667095184326,
+      "learning_rate": 2.777418484639779e-05,
+      "loss": 1.3489,
+      "num_input_tokens_seen": 389719140,
+      "step": 9920
+    },
+    {
+      "epoch": 0.4811716748335827,
+      "grad_norm": 0.42761561274528503,
+      "learning_rate": 2.773518721171884e-05,
+      "loss": 1.2865,
+      "num_input_tokens_seen": 390108812,
+      "step": 9930
+    },
+    {
+      "epoch": 0.4816562384537575,
+      "grad_norm": 0.3793517053127289,
+      "learning_rate": 2.769618283969569e-05,
+      "loss": 1.3011,
+      "num_input_tokens_seen": 390505056,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4821408020739323,
+      "grad_norm": 0.39107024669647217,
+      "learning_rate": 2.765717182640436e-05,
+      "loss": 1.3088,
+      "num_input_tokens_seen": 390913668,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4826253656941071,
+      "grad_norm": 0.3844979703426361,
+      "learning_rate": 2.7618154267937206e-05,
+      "loss": 1.2989,
+      "num_input_tokens_seen": 391319096,
+      "step": 9960
+    },
+    {
+      "epoch": 0.4831099293142819,
+      "grad_norm": 0.4240836203098297,
+      "learning_rate": 2.7579130260402736e-05,
+      "loss": 1.3159,
+      "num_input_tokens_seen": 391707568,
+      "step": 9970
+    },
+    {
+      "epoch": 0.4835944929344567,
+      "grad_norm": 0.3745054006576538,
+      "learning_rate": 2.7540099899925325e-05,
+      "loss": 1.2823,
+      "num_input_tokens_seen": 392088720,
+      "step": 9980
+    },
+    {
+      "epoch": 0.4840790565546315,
+      "grad_norm": 0.41763851046562195,
+      "learning_rate": 2.750106328264499e-05,
+      "loss": 1.3461,
+      "num_input_tokens_seen": 392488252,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4845636201748063,
+      "grad_norm": 0.371481716632843,
+      "learning_rate": 2.746202050471719e-05,
+      "loss": 1.3553,
+      "num_input_tokens_seen": 392909044,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4845636201748063,
+      "eval_loss": 1.4267897605895996,
+      "eval_runtime": 3.7798,
+      "eval_samples_per_second": 39.684,
+      "eval_steps_per_second": 5.027,
+      "num_input_tokens_seen": 392909044,
+      "step": 10000
+    },
+    {
+      "epoch": 0.48504818379498116,
+      "grad_norm": 0.4022182822227478,
+      "learning_rate": 2.742297166231252e-05,
+      "loss": 1.3079,
+      "num_input_tokens_seen": 393315936,
+      "step": 10010
+    },
+    {
+      "epoch": 0.48553274741515595,
+      "grad_norm": 0.3892512917518616,
+      "learning_rate": 2.738391685161654e-05,
+      "loss": 1.324,
+      "num_input_tokens_seen": 393713452,
+      "step": 10020
+    },
+    {
+      "epoch": 0.48601731103533075,
+      "grad_norm": 0.43548429012298584,
+      "learning_rate": 2.7344856168829502e-05,
+      "loss": 1.2849,
+      "num_input_tokens_seen": 394098204,
+      "step": 10030
+    },
+    {
+      "epoch": 0.48650187465550554,
+      "grad_norm": 0.3741033375263214,
+      "learning_rate": 2.7305789710166123e-05,
+      "loss": 1.256,
+      "num_input_tokens_seen": 394469340,
+      "step": 10040
+    },
+    {
+      "epoch": 0.48698643827568033,
+      "grad_norm": 0.3614325225353241,
+      "learning_rate": 2.726671757185535e-05,
+      "loss": 1.3384,
+      "num_input_tokens_seen": 394882124,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4874710018958552,
+      "grad_norm": 0.4203681945800781,
+      "learning_rate": 2.7227639850140118e-05,
+      "loss": 1.3698,
+      "num_input_tokens_seen": 395262528,
+      "step": 10060
+    },
+    {
+      "epoch": 0.48795556551603,
+      "grad_norm": 0.42467623949050903,
+      "learning_rate": 2.7188556641277107e-05,
+      "loss": 1.3216,
+      "num_input_tokens_seen": 395660708,
+      "step": 10070
+    },
+    {
+      "epoch": 0.48844012913620477,
+      "grad_norm": 0.40550142526626587,
+      "learning_rate": 2.7149468041536535e-05,
+      "loss": 1.3211,
+      "num_input_tokens_seen": 396068784,
+      "step": 10080
+    },
+    {
+      "epoch": 0.48892469275637956,
+      "grad_norm": 0.4358334541320801,
+      "learning_rate": 2.711037414720187e-05,
+      "loss": 1.2643,
+      "num_input_tokens_seen": 396441268,
+      "step": 10090
+    },
+    {
+      "epoch": 0.4894092563765544,
+      "grad_norm": 0.38662582635879517,
+      "learning_rate": 2.7071275054569638e-05,
+      "loss": 1.2977,
+      "num_input_tokens_seen": 396849108,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4898938199967292,
+      "grad_norm": 0.4204176664352417,
+      "learning_rate": 2.703217085994918e-05,
+      "loss": 1.2823,
+      "num_input_tokens_seen": 397241560,
+      "step": 10110
+    },
+    {
+      "epoch": 0.490378383616904,
+      "grad_norm": 0.41810983419418335,
+      "learning_rate": 2.699306165966238e-05,
+      "loss": 1.2884,
+      "num_input_tokens_seen": 397641548,
+      "step": 10120
+    },
+    {
+      "epoch": 0.4908629472370788,
+      "grad_norm": 0.43055158853530884,
+      "learning_rate": 2.695394755004347e-05,
+      "loss": 1.3254,
+      "num_input_tokens_seen": 398025916,
+      "step": 10130
+    },
+    {
+      "epoch": 0.4913475108572536,
+      "grad_norm": 0.43699413537979126,
+      "learning_rate": 2.691482862743877e-05,
+      "loss": 1.2906,
+      "num_input_tokens_seen": 398442156,
+      "step": 10140
+    },
+    {
+      "epoch": 0.49183207447742844,
+      "grad_norm": 0.461458295583725,
+      "learning_rate": 2.6875704988206457e-05,
+      "loss": 1.335,
+      "num_input_tokens_seen": 398846656,
+      "step": 10150
+    },
+    {
+      "epoch": 0.49231663809760323,
+      "grad_norm": 0.3919227123260498,
+      "learning_rate": 2.6836576728716313e-05,
+      "loss": 1.3533,
+      "num_input_tokens_seen": 399230044,
+      "step": 10160
+    },
+    {
+      "epoch": 0.492801201717778,
+      "grad_norm": 0.3971085548400879,
+      "learning_rate": 2.679744394534952e-05,
+      "loss": 1.2967,
+      "num_input_tokens_seen": 399626988,
+      "step": 10170
+    },
+    {
+      "epoch": 0.4932857653379528,
+      "grad_norm": 0.40311896800994873,
+      "learning_rate": 2.6758306734498383e-05,
+      "loss": 1.3335,
+      "num_input_tokens_seen": 399992148,
+      "step": 10180
+    },
+    {
+      "epoch": 0.49377032895812767,
+      "grad_norm": 0.4395916759967804,
+      "learning_rate": 2.6719165192566138e-05,
+      "loss": 1.2957,
+      "num_input_tokens_seen": 400375556,
+      "step": 10190
+    },
+    {
+      "epoch": 0.49425489257830246,
+      "grad_norm": 0.41459786891937256,
+      "learning_rate": 2.6680019415966673e-05,
+      "loss": 1.313,
+      "num_input_tokens_seen": 400752968,
+      "step": 10200
+    },
+    {
+      "epoch": 0.49473945619847726,
+      "grad_norm": 0.43075108528137207,
+      "learning_rate": 2.6640869501124305e-05,
+      "loss": 1.3392,
+      "num_input_tokens_seen": 401136408,
+      "step": 10210
+    },
+    {
+      "epoch": 0.49522401981865205,
+      "grad_norm": 0.43251320719718933,
+      "learning_rate": 2.660171554447355e-05,
+      "loss": 1.267,
+      "num_input_tokens_seen": 401538004,
+      "step": 10220
+    },
+    {
+      "epoch": 0.49570858343882684,
+      "grad_norm": 0.3789636492729187,
+      "learning_rate": 2.6562557642458872e-05,
+      "loss": 1.3561,
+      "num_input_tokens_seen": 401913896,
+      "step": 10230
+    },
+    {
+      "epoch": 0.4961931470590017,
+      "grad_norm": 0.410152405500412,
+      "learning_rate": 2.652339589153447e-05,
+      "loss": 1.3323,
+      "num_input_tokens_seen": 402303296,
+      "step": 10240
+    },
+    {
+      "epoch": 0.4966777106791765,
+      "grad_norm": 0.4302387237548828,
+      "learning_rate": 2.648423038816401e-05,
+      "loss": 1.3612,
+      "num_input_tokens_seen": 402673940,
+      "step": 10250
+    },
+    {
+      "epoch": 0.4971622742993513,
+      "grad_norm": 0.4012112021446228,
+      "learning_rate": 2.6445061228820406e-05,
+      "loss": 1.3009,
+      "num_input_tokens_seen": 403069012,
+      "step": 10260
+    },
+    {
+      "epoch": 0.4976468379195261,
+      "grad_norm": 0.42036718130111694,
+      "learning_rate": 2.6405888509985576e-05,
+      "loss": 1.3463,
+      "num_input_tokens_seen": 403464332,
+      "step": 10270
+    },
+    {
+      "epoch": 0.4981314015397009,
+      "grad_norm": 0.42190247774124146,
+      "learning_rate": 2.6366712328150205e-05,
+      "loss": 1.3038,
+      "num_input_tokens_seen": 403863168,
+      "step": 10280
+    },
+    {
+      "epoch": 0.4986159651598757,
+      "grad_norm": 0.4091307520866394,
+      "learning_rate": 2.6327532779813506e-05,
+      "loss": 1.2853,
+      "num_input_tokens_seen": 404242308,
+      "step": 10290
+    },
+    {
+      "epoch": 0.4991005287800505,
+      "grad_norm": 0.43492767214775085,
+      "learning_rate": 2.6288349961482993e-05,
+      "loss": 1.3446,
+      "num_input_tokens_seen": 404634860,
+      "step": 10300
+    },
+    {
+      "epoch": 0.4995850924002253,
+      "grad_norm": 0.4058671295642853,
+      "learning_rate": 2.624916396967423e-05,
+      "loss": 1.2642,
+      "num_input_tokens_seen": 405016164,
+      "step": 10310
+    },
+    {
+      "epoch": 0.5000696560204001,
+      "grad_norm": 0.3747016191482544,
+      "learning_rate": 2.620997490091058e-05,
+      "loss": 1.3063,
+      "num_input_tokens_seen": 405389624,
+      "step": 10320
+    },
+    {
+      "epoch": 0.5005542196405749,
+      "grad_norm": 0.4174031913280487,
+      "learning_rate": 2.617078285172302e-05,
+      "loss": 1.3075,
+      "num_input_tokens_seen": 405768816,
+      "step": 10330
+    },
+    {
+      "epoch": 0.5010387832607497,
+      "grad_norm": 0.42868369817733765,
+      "learning_rate": 2.6131587918649854e-05,
+      "loss": 1.3153,
+      "num_input_tokens_seen": 406177564,
+      "step": 10340
+    },
+    {
+      "epoch": 0.5015233468809246,
+      "grad_norm": 0.4055063724517822,
+      "learning_rate": 2.6092390198236468e-05,
+      "loss": 1.3084,
+      "num_input_tokens_seen": 406576960,
+      "step": 10350
+    },
+    {
+      "epoch": 0.5020079105010994,
+      "grad_norm": 0.36954593658447266,
+      "learning_rate": 2.6053189787035147e-05,
+      "loss": 1.2724,
+      "num_input_tokens_seen": 407003236,
+      "step": 10360
+    },
+    {
+      "epoch": 0.5024924741212742,
+      "grad_norm": 0.46568143367767334,
+      "learning_rate": 2.6013986781604782e-05,
+      "loss": 1.2893,
+      "num_input_tokens_seen": 407393488,
+      "step": 10370
+    },
+    {
+      "epoch": 0.502977037741449,
+      "grad_norm": 0.40515366196632385,
+      "learning_rate": 2.5974781278510656e-05,
+      "loss": 1.3126,
+      "num_input_tokens_seen": 407794284,
+      "step": 10380
+    },
+    {
+      "epoch": 0.5034616013616238,
+      "grad_norm": 0.4094693064689636,
+      "learning_rate": 2.5935573374324228e-05,
+      "loss": 1.288,
+      "num_input_tokens_seen": 408172776,
+      "step": 10390
+    },
+    {
+      "epoch": 0.5039461649817986,
+      "grad_norm": 0.4908318519592285,
+      "learning_rate": 2.5896363165622833e-05,
+      "loss": 1.3331,
+      "num_input_tokens_seen": 408581868,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5044307286019734,
+      "grad_norm": 0.42689284682273865,
+      "learning_rate": 2.585715074898951e-05,
+      "loss": 1.2885,
+      "num_input_tokens_seen": 408941632,
+      "step": 10410
+    },
+    {
+      "epoch": 0.5049152922221481,
+      "grad_norm": 0.40442904829978943,
+      "learning_rate": 2.5817936221012733e-05,
+      "loss": 1.3529,
+      "num_input_tokens_seen": 409363076,
+      "step": 10420
+    },
+    {
+      "epoch": 0.5053998558423229,
+      "grad_norm": 0.38854020833969116,
+      "learning_rate": 2.5778719678286172e-05,
+      "loss": 1.2953,
+      "num_input_tokens_seen": 409759340,
+      "step": 10430
+    },
+    {
+      "epoch": 0.5058844194624978,
+      "grad_norm": 0.41985204815864563,
+      "learning_rate": 2.5739501217408457e-05,
+      "loss": 1.3149,
+      "num_input_tokens_seen": 410179976,
+      "step": 10440
+    },
+    {
+      "epoch": 0.5063689830826726,
+      "grad_norm": 0.42030394077301025,
+      "learning_rate": 2.5700280934982947e-05,
+      "loss": 1.3282,
+      "num_input_tokens_seen": 410588712,
+      "step": 10450
+    },
+    {
+      "epoch": 0.5068535467028474,
+      "grad_norm": 0.4288428723812103,
+      "learning_rate": 2.5661058927617476e-05,
+      "loss": 1.2864,
+      "num_input_tokens_seen": 410998856,
+      "step": 10460
+    },
+    {
+      "epoch": 0.5073381103230222,
+      "grad_norm": 0.4092228412628174,
+      "learning_rate": 2.5621835291924157e-05,
+      "loss": 1.2977,
+      "num_input_tokens_seen": 411403584,
+      "step": 10470
+    },
+    {
+      "epoch": 0.507822673943197,
+      "grad_norm": 0.40826088190078735,
+      "learning_rate": 2.5582610124519087e-05,
+      "loss": 1.2867,
+      "num_input_tokens_seen": 411806684,
+      "step": 10480
+    },
+    {
+      "epoch": 0.5083072375633718,
+      "grad_norm": 0.44615307450294495,
+      "learning_rate": 2.5543383522022137e-05,
+      "loss": 1.3498,
+      "num_input_tokens_seen": 412178432,
+      "step": 10490
+    },
+    {
+      "epoch": 0.5087918011835466,
+      "grad_norm": 0.382753849029541,
+      "learning_rate": 2.5504155581056734e-05,
+      "loss": 1.2994,
+      "num_input_tokens_seen": 412554676,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5092763648037214,
+      "grad_norm": 0.4127599000930786,
+      "learning_rate": 2.546492639824957e-05,
+      "loss": 1.3228,
+      "num_input_tokens_seen": 412919116,
+      "step": 10510
+    },
+    {
+      "epoch": 0.5097609284238963,
+      "grad_norm": 0.3876650333404541,
+      "learning_rate": 2.542569607023042e-05,
+      "loss": 1.3474,
+      "num_input_tokens_seen": 413304516,
+      "step": 10520
+    },
+    {
+      "epoch": 0.5102454920440711,
+      "grad_norm": 0.3939809799194336,
+      "learning_rate": 2.5386464693631885e-05,
+      "loss": 1.3047,
+      "num_input_tokens_seen": 413691892,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5107300556642459,
+      "grad_norm": 0.4118193984031677,
+      "learning_rate": 2.5347232365089125e-05,
+      "loss": 1.2911,
+      "num_input_tokens_seen": 414112144,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5112146192844207,
+      "grad_norm": 0.40481314063072205,
+      "learning_rate": 2.530799918123966e-05,
+      "loss": 1.341,
+      "num_input_tokens_seen": 414501584,
+      "step": 10550
+    },
+    {
+      "epoch": 0.5116991829045955,
+      "grad_norm": 0.4449983537197113,
+      "learning_rate": 2.526876523872312e-05,
+      "loss": 1.3046,
+      "num_input_tokens_seen": 414863372,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5121837465247703,
+      "grad_norm": 0.37632080912590027,
+      "learning_rate": 2.5229530634180986e-05,
+      "loss": 1.3226,
+      "num_input_tokens_seen": 415256912,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5126683101449451,
+      "grad_norm": 0.4017525613307953,
+      "learning_rate": 2.519029546425639e-05,
+      "loss": 1.2999,
+      "num_input_tokens_seen": 415645068,
+      "step": 10580
+    },
+    {
+      "epoch": 0.5131528737651199,
+      "grad_norm": 0.405073881149292,
+      "learning_rate": 2.5151059825593847e-05,
+      "loss": 1.3299,
+      "num_input_tokens_seen": 416025852,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5136374373852947,
+      "grad_norm": 0.3685546815395355,
+      "learning_rate": 2.511182381483902e-05,
+      "loss": 1.3465,
+      "num_input_tokens_seen": 416410872,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5141220010054696,
+      "grad_norm": 0.40560224652290344,
+      "learning_rate": 2.507258752863851e-05,
+      "loss": 1.3055,
+      "num_input_tokens_seen": 416840448,
+      "step": 10610
+    },
+    {
+      "epoch": 0.5146065646256444,
+      "grad_norm": 0.402498334646225,
+      "learning_rate": 2.503335106363957e-05,
+      "loss": 1.3216,
+      "num_input_tokens_seen": 417239552,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5150911282458192,
+      "grad_norm": 0.40907391905784607,
+      "learning_rate": 2.4994114516489917e-05,
+      "loss": 1.3089,
+      "num_input_tokens_seen": 417648304,
+      "step": 10630
+    },
+    {
+      "epoch": 0.515575691865994,
+      "grad_norm": 0.4574085474014282,
+      "learning_rate": 2.4954877983837446e-05,
+      "loss": 1.2659,
+      "num_input_tokens_seen": 418049068,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5160602554861687,
+      "grad_norm": 0.4318368434906006,
+      "learning_rate": 2.491564156233005e-05,
+      "loss": 1.3243,
+      "num_input_tokens_seen": 418453168,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5165448191063435,
+      "grad_norm": 0.3779464364051819,
+      "learning_rate": 2.4876405348615303e-05,
+      "loss": 1.3325,
+      "num_input_tokens_seen": 418845328,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5170293827265183,
+      "grad_norm": 0.4640369713306427,
+      "learning_rate": 2.483716943934031e-05,
+      "loss": 1.3597,
+      "num_input_tokens_seen": 419240400,
+      "step": 10670
+    },
+    {
+      "epoch": 0.5175139463466931,
+      "grad_norm": 0.39406731724739075,
+      "learning_rate": 2.47979339311514e-05,
+      "loss": 1.348,
+      "num_input_tokens_seen": 419604196,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5179985099668679,
+      "grad_norm": 0.45061835646629333,
+      "learning_rate": 2.4758698920693933e-05,
+      "loss": 1.3107,
+      "num_input_tokens_seen": 420001636,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5184830735870428,
+      "grad_norm": 0.3857516050338745,
+      "learning_rate": 2.4719464504612015e-05,
+      "loss": 1.2952,
+      "num_input_tokens_seen": 420412924,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5189676372072176,
+      "grad_norm": 0.40518441796302795,
+      "learning_rate": 2.4680230779548325e-05,
+      "loss": 1.3131,
+      "num_input_tokens_seen": 420817072,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5194522008273924,
+      "grad_norm": 0.3917674422264099,
+      "learning_rate": 2.4640997842143797e-05,
+      "loss": 1.3628,
+      "num_input_tokens_seen": 421217132,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5199367644475672,
+      "grad_norm": 0.41887742280960083,
+      "learning_rate": 2.4601765789037465e-05,
+      "loss": 1.2873,
+      "num_input_tokens_seen": 421590708,
+      "step": 10730
+    },
+    {
+      "epoch": 0.520421328067742,
+      "grad_norm": 0.4056337773799896,
+      "learning_rate": 2.456253471686617e-05,
+      "loss": 1.3099,
+      "num_input_tokens_seen": 421991544,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5209058916879168,
+      "grad_norm": 0.4133414924144745,
+      "learning_rate": 2.452330472226432e-05,
+      "loss": 1.3146,
+      "num_input_tokens_seen": 422369440,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5213904553080916,
+      "grad_norm": 0.39193689823150635,
+      "learning_rate": 2.4484075901863697e-05,
+      "loss": 1.3141,
+      "num_input_tokens_seen": 422783140,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5218750189282664,
+      "grad_norm": 0.447322279214859,
+      "learning_rate": 2.444484835229316e-05,
+      "loss": 1.2537,
+      "num_input_tokens_seen": 423177384,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5223595825484412,
+      "grad_norm": 0.3933646082878113,
+      "learning_rate": 2.4405622170178483e-05,
+      "loss": 1.3232,
+      "num_input_tokens_seen": 423569888,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5228441461686161,
+      "grad_norm": 0.4285268187522888,
+      "learning_rate": 2.436639745214201e-05,
+      "loss": 1.2985,
+      "num_input_tokens_seen": 423969632,
+      "step": 10790
+    },
+    {
+      "epoch": 0.5233287097887909,
+      "grad_norm": 0.40538379549980164,
+      "learning_rate": 2.432717429480254e-05,
+      "loss": 1.3513,
+      "num_input_tokens_seen": 424379412,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5238132734089657,
+      "grad_norm": 0.4205954670906067,
+      "learning_rate": 2.4287952794774972e-05,
+      "loss": 1.2476,
+      "num_input_tokens_seen": 424732692,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5242978370291405,
+      "grad_norm": 0.4187707304954529,
+      "learning_rate": 2.424873304867018e-05,
+      "loss": 1.2784,
+      "num_input_tokens_seen": 425099744,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5247824006493153,
+      "grad_norm": 0.4088019132614136,
+      "learning_rate": 2.420951515309466e-05,
+      "loss": 1.3619,
+      "num_input_tokens_seen": 425533784,
+      "step": 10830
+    },
+    {
+      "epoch": 0.52526696426949,
+      "grad_norm": 0.4790429472923279,
+      "learning_rate": 2.4170299204650402e-05,
+      "loss": 1.3588,
+      "num_input_tokens_seen": 425907244,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5257515278896648,
+      "grad_norm": 0.39465224742889404,
+      "learning_rate": 2.4131085299934552e-05,
+      "loss": 1.2817,
+      "num_input_tokens_seen": 426303280,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5262360915098396,
+      "grad_norm": 0.3853622376918793,
+      "learning_rate": 2.4091873535539263e-05,
+      "loss": 1.3173,
+      "num_input_tokens_seen": 426708880,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5267206551300144,
+      "grad_norm": 0.47276604175567627,
+      "learning_rate": 2.40526640080514e-05,
+      "loss": 1.267,
+      "num_input_tokens_seen": 427072384,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5272052187501893,
+      "grad_norm": 0.4207751452922821,
+      "learning_rate": 2.40134568140523e-05,
+      "loss": 1.3232,
+      "num_input_tokens_seen": 427483780,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5276897823703641,
+      "grad_norm": 0.4074230194091797,
+      "learning_rate": 2.3974252050117578e-05,
+      "loss": 1.3059,
+      "num_input_tokens_seen": 427878960,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5281743459905389,
+      "grad_norm": 0.4160650968551636,
+      "learning_rate": 2.3935049812816853e-05,
+      "loss": 1.284,
+      "num_input_tokens_seen": 428262904,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5286589096107137,
+      "grad_norm": 0.4167976975440979,
+      "learning_rate": 2.3895850198713532e-05,
+      "loss": 1.3627,
+      "num_input_tokens_seen": 428669280,
+      "step": 10910
+    },
+    {
+      "epoch": 0.5291434732308885,
+      "grad_norm": 0.3857592046260834,
+      "learning_rate": 2.3856653304364528e-05,
+      "loss": 1.2677,
+      "num_input_tokens_seen": 429050584,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5296280368510633,
+      "grad_norm": 0.39841026067733765,
+      "learning_rate": 2.38174592263201e-05,
+      "loss": 1.2863,
+      "num_input_tokens_seen": 429432948,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5301126004712381,
+      "grad_norm": 0.37593045830726624,
+      "learning_rate": 2.377826806112352e-05,
+      "loss": 1.319,
+      "num_input_tokens_seen": 429843780,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5305971640914129,
+      "grad_norm": 0.39701202511787415,
+      "learning_rate": 2.3739079905310925e-05,
+      "loss": 1.3137,
+      "num_input_tokens_seen": 430264816,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5310817277115877,
+      "grad_norm": 0.4029529094696045,
+      "learning_rate": 2.3699894855411025e-05,
+      "loss": 1.3228,
+      "num_input_tokens_seen": 430690612,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5315662913317626,
+      "grad_norm": 0.4053628742694855,
+      "learning_rate": 2.366071300794489e-05,
+      "loss": 1.3084,
+      "num_input_tokens_seen": 431088820,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5320508549519374,
+      "grad_norm": 0.4208217263221741,
+      "learning_rate": 2.362153445942567e-05,
+      "loss": 1.2865,
+      "num_input_tokens_seen": 431510088,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5325354185721122,
+      "grad_norm": 0.38513317704200745,
+      "learning_rate": 2.3582359306358425e-05,
+      "loss": 1.2924,
+      "num_input_tokens_seen": 431912904,
+      "step": 10990
+    },
+    {
+      "epoch": 0.533019982192287,
+      "grad_norm": 0.3804323077201843,
+      "learning_rate": 2.354318764523984e-05,
+      "loss": 1.3148,
+      "num_input_tokens_seen": 432316308,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5335045458124618,
+      "grad_norm": 0.4341334104537964,
+      "learning_rate": 2.3504019572557978e-05,
+      "loss": 1.3025,
+      "num_input_tokens_seen": 432706392,
+      "step": 11010
+    },
+    {
+      "epoch": 0.5339891094326366,
+      "grad_norm": 0.3772577941417694,
+      "learning_rate": 2.3464855184792103e-05,
+      "loss": 1.3058,
+      "num_input_tokens_seen": 433114000,
+      "step": 11020
+    },
+    {
+      "epoch": 0.5344736730528113,
+      "grad_norm": 0.37960031628608704,
+      "learning_rate": 2.3425694578412357e-05,
+      "loss": 1.2901,
+      "num_input_tokens_seen": 433515624,
+      "step": 11030
+    },
+    {
+      "epoch": 0.5349582366729861,
+      "grad_norm": 0.3838542401790619,
+      "learning_rate": 2.338653784987961e-05,
+      "loss": 1.3182,
+      "num_input_tokens_seen": 433899616,
+      "step": 11040
+    },
+    {
+      "epoch": 0.5354428002931609,
+      "grad_norm": 0.40158811211586,
+      "learning_rate": 2.3347385095645143e-05,
+      "loss": 1.3337,
+      "num_input_tokens_seen": 434295476,
+      "step": 11050
+    },
+    {
+      "epoch": 0.5359273639133358,
+      "grad_norm": 0.41163861751556396,
+      "learning_rate": 2.3308236412150488e-05,
+      "loss": 1.2498,
+      "num_input_tokens_seen": 434692404,
+      "step": 11060
+    },
+    {
+      "epoch": 0.5364119275335106,
+      "grad_norm": 0.4120139181613922,
+      "learning_rate": 2.3269091895827096e-05,
+      "loss": 1.325,
+      "num_input_tokens_seen": 435084432,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5368964911536854,
+      "grad_norm": 0.4136423170566559,
+      "learning_rate": 2.3229951643096215e-05,
+      "loss": 1.3232,
+      "num_input_tokens_seen": 435481608,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5373810547738602,
+      "grad_norm": 0.3973861038684845,
+      "learning_rate": 2.3190815750368534e-05,
+      "loss": 1.2942,
+      "num_input_tokens_seen": 435879740,
+      "step": 11090
+    },
+    {
+      "epoch": 0.537865618394035,
+      "grad_norm": 0.3832756280899048,
+      "learning_rate": 2.3151684314044042e-05,
+      "loss": 1.291,
+      "num_input_tokens_seen": 436231608,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5383501820142098,
+      "grad_norm": 0.446608304977417,
+      "learning_rate": 2.3112557430511734e-05,
+      "loss": 1.2808,
+      "num_input_tokens_seen": 436633972,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5388347456343846,
+      "grad_norm": 0.41678836941719055,
+      "learning_rate": 2.3073435196149392e-05,
+      "loss": 1.3131,
+      "num_input_tokens_seen": 437031292,
+      "step": 11120
+    },
+    {
+      "epoch": 0.5393193092545594,
+      "grad_norm": 0.3977677822113037,
+      "learning_rate": 2.3034317707323364e-05,
+      "loss": 1.2948,
+      "num_input_tokens_seen": 437405436,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5398038728747342,
+      "grad_norm": 0.4163575768470764,
+      "learning_rate": 2.2995205060388265e-05,
+      "loss": 1.3057,
+      "num_input_tokens_seen": 437781780,
+      "step": 11140
+    },
+    {
+      "epoch": 0.5402884364949091,
+      "grad_norm": 0.38908636569976807,
+      "learning_rate": 2.295609735168684e-05,
+      "loss": 1.323,
+      "num_input_tokens_seen": 438167672,
+      "step": 11150
+    },
+    {
+      "epoch": 0.5407730001150839,
+      "grad_norm": 0.3993242681026459,
+      "learning_rate": 2.2916994677549614e-05,
+      "loss": 1.2775,
+      "num_input_tokens_seen": 438551092,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5412575637352587,
+      "grad_norm": 0.43453851342201233,
+      "learning_rate": 2.2877897134294755e-05,
+      "loss": 1.3369,
+      "num_input_tokens_seen": 438925056,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5417421273554335,
+      "grad_norm": 0.38773950934410095,
+      "learning_rate": 2.2838804818227766e-05,
+      "loss": 1.3376,
+      "num_input_tokens_seen": 439315640,
+      "step": 11180
+    },
+    {
+      "epoch": 0.5422266909756083,
+      "grad_norm": 0.4047930836677551,
+      "learning_rate": 2.2799717825641297e-05,
+      "loss": 1.2998,
+      "num_input_tokens_seen": 439714968,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5427112545957831,
+      "grad_norm": 0.385567307472229,
+      "learning_rate": 2.2760636252814858e-05,
+      "loss": 1.3205,
+      "num_input_tokens_seen": 440096324,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5431958182159579,
+      "grad_norm": 0.3837738335132599,
+      "learning_rate": 2.2721560196014635e-05,
+      "loss": 1.2782,
+      "num_input_tokens_seen": 440489640,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5436803818361327,
+      "grad_norm": 0.4107000231742859,
+      "learning_rate": 2.26824897514932e-05,
+      "loss": 1.356,
+      "num_input_tokens_seen": 440874496,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5441649454563076,
+      "grad_norm": 0.4507005214691162,
+      "learning_rate": 2.2643425015489335e-05,
+      "loss": 1.2991,
+      "num_input_tokens_seen": 441274860,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5446495090764824,
+      "grad_norm": 0.40049469470977783,
+      "learning_rate": 2.260436608422772e-05,
+      "loss": 1.3207,
+      "num_input_tokens_seen": 441672140,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5451340726966571,
+      "grad_norm": 0.4210105836391449,
+      "learning_rate": 2.2565313053918764e-05,
+      "loss": 1.3396,
+      "num_input_tokens_seen": 442056044,
+      "step": 11250
+    },
+    {
+      "epoch": 0.5456186363168319,
+      "grad_norm": 0.40691545605659485,
+      "learning_rate": 2.252626602075835e-05,
+      "loss": 1.3008,
+      "num_input_tokens_seen": 442444840,
+      "step": 11260
+    },
+    {
+      "epoch": 0.5461031999370067,
+      "grad_norm": 0.377591073513031,
+      "learning_rate": 2.2487225080927553e-05,
+      "loss": 1.2803,
+      "num_input_tokens_seen": 442832056,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5465877635571815,
+      "grad_norm": 0.36699971556663513,
+      "learning_rate": 2.244819033059248e-05,
+      "loss": 1.3336,
+      "num_input_tokens_seen": 443227492,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5470723271773563,
+      "grad_norm": 0.3749854862689972,
+      "learning_rate": 2.2409161865903952e-05,
+      "loss": 1.31,
+      "num_input_tokens_seen": 443640864,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5475568907975311,
+      "grad_norm": 0.5490464568138123,
+      "learning_rate": 2.2370139782997342e-05,
+      "loss": 1.3327,
+      "num_input_tokens_seen": 444016308,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5480414544177059,
+      "grad_norm": 0.4179551601409912,
+      "learning_rate": 2.2331124177992274e-05,
+      "loss": 1.3102,
+      "num_input_tokens_seen": 444404972,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5485260180378808,
+      "grad_norm": 0.3857383728027344,
+      "learning_rate": 2.2292115146992438e-05,
+      "loss": 1.2987,
+      "num_input_tokens_seen": 444771056,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5490105816580556,
+      "grad_norm": 0.40112438797950745,
+      "learning_rate": 2.2253112786085313e-05,
+      "loss": 1.2942,
+      "num_input_tokens_seen": 445173356,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5494951452782304,
+      "grad_norm": 0.4082011580467224,
+      "learning_rate": 2.2214117191341972e-05,
+      "loss": 1.2856,
+      "num_input_tokens_seen": 445519032,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5499797088984052,
+      "grad_norm": 0.423109769821167,
+      "learning_rate": 2.2175128458816792e-05,
+      "loss": 1.3118,
+      "num_input_tokens_seen": 445906772,
+      "step": 11350
+    },
+    {
+      "epoch": 0.55046427251858,
+      "grad_norm": 0.42687487602233887,
+      "learning_rate": 2.213614668454728e-05,
+      "loss": 1.2962,
+      "num_input_tokens_seen": 446254804,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5509488361387548,
+      "grad_norm": 0.3661687672138214,
+      "learning_rate": 2.2097171964553757e-05,
+      "loss": 1.3353,
+      "num_input_tokens_seen": 446614404,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5514333997589296,
+      "grad_norm": 0.39668744802474976,
+      "learning_rate": 2.2058204394839217e-05,
+      "loss": 1.3309,
+      "num_input_tokens_seen": 447010280,
+      "step": 11380
+    },
+    {
+      "epoch": 0.5519179633791044,
+      "grad_norm": 0.3846248984336853,
+      "learning_rate": 2.201924407138902e-05,
+      "loss": 1.2841,
+      "num_input_tokens_seen": 447381468,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5524025269992792,
+      "grad_norm": 0.39855462312698364,
+      "learning_rate": 2.1980291090170664e-05,
+      "loss": 1.3553,
+      "num_input_tokens_seen": 447788436,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5528870906194541,
+      "grad_norm": 0.40765848755836487,
+      "learning_rate": 2.19413455471336e-05,
+      "loss": 1.2543,
+      "num_input_tokens_seen": 448189856,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5533716542396289,
+      "grad_norm": 0.3901714086532593,
+      "learning_rate": 2.1902407538208897e-05,
+      "loss": 1.3356,
+      "num_input_tokens_seen": 448564344,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5538562178598037,
+      "grad_norm": 0.3917088806629181,
+      "learning_rate": 2.1863477159309132e-05,
+      "loss": 1.3467,
+      "num_input_tokens_seen": 448945696,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5543407814799785,
+      "grad_norm": 0.4396210312843323,
+      "learning_rate": 2.182455450632803e-05,
+      "loss": 1.3022,
+      "num_input_tokens_seen": 449313644,
+      "step": 11440
+    },
+    {
+      "epoch": 0.5548253451001532,
+      "grad_norm": 0.41703107953071594,
+      "learning_rate": 2.178563967514034e-05,
+      "loss": 1.2391,
+      "num_input_tokens_seen": 449716576,
+      "step": 11450
+    },
+    {
+      "epoch": 0.555309908720328,
+      "grad_norm": 0.405318945646286,
+      "learning_rate": 2.1746732761601486e-05,
+      "loss": 1.2382,
+      "num_input_tokens_seen": 450092104,
+      "step": 11460
+    },
+    {
+      "epoch": 0.5557944723405028,
+      "grad_norm": 0.37074992060661316,
+      "learning_rate": 2.1707833861547442e-05,
+      "loss": 1.2947,
+      "num_input_tokens_seen": 450487336,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5562790359606776,
+      "grad_norm": 0.4145805835723877,
+      "learning_rate": 2.1668943070794407e-05,
+      "loss": 1.3364,
+      "num_input_tokens_seen": 450861056,
+      "step": 11480
+    },
+    {
+      "epoch": 0.5567635995808524,
+      "grad_norm": 0.42270269989967346,
+      "learning_rate": 2.163006048513863e-05,
+      "loss": 1.354,
+      "num_input_tokens_seen": 451258688,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5572481632010273,
+      "grad_norm": 0.3963187336921692,
+      "learning_rate": 2.1591186200356122e-05,
+      "loss": 1.281,
+      "num_input_tokens_seen": 451641956,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5577327268212021,
+      "grad_norm": 0.4037591814994812,
+      "learning_rate": 2.1552320312202485e-05,
+      "loss": 1.2815,
+      "num_input_tokens_seen": 452047412,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5582172904413769,
+      "grad_norm": 0.3662770390510559,
+      "learning_rate": 2.1513462916412592e-05,
+      "loss": 1.3455,
+      "num_input_tokens_seen": 452448840,
+      "step": 11520
+    },
+    {
+      "epoch": 0.5587018540615517,
+      "grad_norm": 0.3583228290081024,
+      "learning_rate": 2.147461410870043e-05,
+      "loss": 1.2794,
+      "num_input_tokens_seen": 452843044,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5591864176817265,
+      "grad_norm": 0.37632209062576294,
+      "learning_rate": 2.143577398475883e-05,
+      "loss": 1.2674,
+      "num_input_tokens_seen": 453264900,
+      "step": 11540
+    },
+    {
+      "epoch": 0.5596709813019013,
+      "grad_norm": 0.3996264636516571,
+      "learning_rate": 2.139694264025922e-05,
+      "loss": 1.2744,
+      "num_input_tokens_seen": 453667984,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5601555449220761,
+      "grad_norm": 0.3832734525203705,
+      "learning_rate": 2.135812017085142e-05,
+      "loss": 1.2997,
+      "num_input_tokens_seen": 454063124,
+      "step": 11560
+    },
+    {
+      "epoch": 0.5606401085422509,
+      "grad_norm": 0.4112671911716461,
+      "learning_rate": 2.1319306672163355e-05,
+      "loss": 1.326,
+      "num_input_tokens_seen": 454468736,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5611246721624257,
+      "grad_norm": 0.40387192368507385,
+      "learning_rate": 2.1280502239800905e-05,
+      "loss": 1.2733,
+      "num_input_tokens_seen": 454868724,
+      "step": 11580
+    },
+    {
+      "epoch": 0.5616092357826006,
+      "grad_norm": 0.4105452001094818,
+      "learning_rate": 2.1241706969347554e-05,
+      "loss": 1.2681,
+      "num_input_tokens_seen": 455262064,
+      "step": 11590
+    },
+    {
+      "epoch": 0.5620937994027754,
+      "grad_norm": 0.41394761204719543,
+      "learning_rate": 2.1202920956364282e-05,
+      "loss": 1.2333,
+      "num_input_tokens_seen": 455641940,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5625783630229502,
+      "grad_norm": 0.47459670901298523,
+      "learning_rate": 2.116414429638922e-05,
+      "loss": 1.3239,
+      "num_input_tokens_seen": 456036468,
+      "step": 11610
+    },
+    {
+      "epoch": 0.563062926643125,
+      "grad_norm": 0.38530948758125305,
+      "learning_rate": 2.112537708493749e-05,
+      "loss": 1.3167,
+      "num_input_tokens_seen": 456420516,
+      "step": 11620
+    },
+    {
+      "epoch": 0.5635474902632998,
+      "grad_norm": 0.3913155794143677,
+      "learning_rate": 2.108661941750091e-05,
+      "loss": 1.3043,
+      "num_input_tokens_seen": 456824708,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5640320538834745,
+      "grad_norm": 0.44334185123443604,
+      "learning_rate": 2.1047871389547826e-05,
+      "loss": 1.3133,
+      "num_input_tokens_seen": 457192180,
+      "step": 11640
+    },
+    {
+      "epoch": 0.5645166175036493,
+      "grad_norm": 0.41720449924468994,
+      "learning_rate": 2.1009133096522805e-05,
+      "loss": 1.2314,
+      "num_input_tokens_seen": 457554472,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5650011811238241,
+      "grad_norm": 0.3976454734802246,
+      "learning_rate": 2.0970404633846453e-05,
+      "loss": 1.3181,
+      "num_input_tokens_seen": 457940752,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5654857447439989,
+      "grad_norm": 0.3785874843597412,
+      "learning_rate": 2.0931686096915172e-05,
+      "loss": 1.2951,
+      "num_input_tokens_seen": 458329516,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5659703083641738,
+      "grad_norm": 0.42193588614463806,
+      "learning_rate": 2.0892977581100884e-05,
+      "loss": 1.2895,
+      "num_input_tokens_seen": 458728512,
+      "step": 11680
+    },
+    {
+      "epoch": 0.5664548719843486,
+      "grad_norm": 0.4394080936908722,
+      "learning_rate": 2.085427918175086e-05,
+      "loss": 1.287,
+      "num_input_tokens_seen": 459138172,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5669394356045234,
+      "grad_norm": 0.39886605739593506,
+      "learning_rate": 2.0815590994187416e-05,
+      "loss": 1.3094,
+      "num_input_tokens_seen": 459549264,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5674239992246982,
+      "grad_norm": 0.37033554911613464,
+      "learning_rate": 2.0776913113707766e-05,
+      "loss": 1.3196,
+      "num_input_tokens_seen": 459931536,
+      "step": 11710
+    },
+    {
+      "epoch": 0.567908562844873,
+      "grad_norm": 0.38384774327278137,
+      "learning_rate": 2.0738245635583675e-05,
+      "loss": 1.2916,
+      "num_input_tokens_seen": 460349548,
+      "step": 11720
+    },
+    {
+      "epoch": 0.5683931264650478,
+      "grad_norm": 0.41543838381767273,
+      "learning_rate": 2.0699588655061337e-05,
+      "loss": 1.34,
+      "num_input_tokens_seen": 460751796,
+      "step": 11730
+    },
+    {
+      "epoch": 0.5688776900852226,
+      "grad_norm": 0.38962802290916443,
+      "learning_rate": 2.066094226736104e-05,
+      "loss": 1.2622,
+      "num_input_tokens_seen": 461166532,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5693622537053974,
+      "grad_norm": 0.3820241391658783,
+      "learning_rate": 2.0622306567677026e-05,
+      "loss": 1.3371,
+      "num_input_tokens_seen": 461544700,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5698468173255722,
+      "grad_norm": 0.37688279151916504,
+      "learning_rate": 2.0583681651177177e-05,
+      "loss": 1.3028,
+      "num_input_tokens_seen": 461923852,
+      "step": 11760
+    },
+    {
+      "epoch": 0.5703313809457471,
+      "grad_norm": 0.41758424043655396,
+      "learning_rate": 2.0545067613002844e-05,
+      "loss": 1.303,
+      "num_input_tokens_seen": 462284468,
+      "step": 11770
+    },
+    {
+      "epoch": 0.5708159445659219,
+      "grad_norm": 0.41535744071006775,
+      "learning_rate": 2.050646454826854e-05,
+      "loss": 1.2806,
+      "num_input_tokens_seen": 462662724,
+      "step": 11780
+    },
+    {
+      "epoch": 0.5713005081860967,
+      "grad_norm": 0.41372743248939514,
+      "learning_rate": 2.0467872552061785e-05,
+      "loss": 1.2763,
+      "num_input_tokens_seen": 463065836,
+      "step": 11790
+    },
+    {
+      "epoch": 0.5717850718062715,
+      "grad_norm": 0.4078621566295624,
+      "learning_rate": 2.042929171944283e-05,
+      "loss": 1.2739,
+      "num_input_tokens_seen": 463465784,
+      "step": 11800
+    },
+    {
+      "epoch": 0.5722696354264463,
+      "grad_norm": 0.4270489811897278,
+      "learning_rate": 2.03907221454444e-05,
+      "loss": 1.2816,
+      "num_input_tokens_seen": 463873784,
+      "step": 11810
+    },
+    {
+      "epoch": 0.5727541990466211,
+      "grad_norm": 0.40874189138412476,
+      "learning_rate": 2.0352163925071526e-05,
+      "loss": 1.3118,
+      "num_input_tokens_seen": 464264876,
+      "step": 11820
+    },
+    {
+      "epoch": 0.5732387626667959,
+      "grad_norm": 0.4195123016834259,
+      "learning_rate": 2.031361715330124e-05,
+      "loss": 1.28,
+      "num_input_tokens_seen": 464655612,
+      "step": 11830
+    },
+    {
+      "epoch": 0.5737233262869706,
+      "grad_norm": 0.42843663692474365,
+      "learning_rate": 2.0275081925082408e-05,
+      "loss": 1.2797,
+      "num_input_tokens_seen": 465021840,
+      "step": 11840
+    },
+    {
+      "epoch": 0.5742078899071454,
+      "grad_norm": 0.4026115834712982,
+      "learning_rate": 2.0236558335335418e-05,
+      "loss": 1.2662,
+      "num_input_tokens_seen": 465412176,
+      "step": 11850
+    },
+    {
+      "epoch": 0.5746924535273203,
+      "grad_norm": 0.4160013794898987,
+      "learning_rate": 2.0198046478952034e-05,
+      "loss": 1.3044,
+      "num_input_tokens_seen": 465798436,
+      "step": 11860
+    },
+    {
+      "epoch": 0.5751770171474951,
+      "grad_norm": 0.3887726068496704,
+      "learning_rate": 2.0159546450795076e-05,
+      "loss": 1.271,
+      "num_input_tokens_seen": 466194580,
+      "step": 11870
+    },
+    {
+      "epoch": 0.5756615807676699,
+      "grad_norm": 0.4644850492477417,
+      "learning_rate": 2.012105834569827e-05,
+      "loss": 1.3343,
+      "num_input_tokens_seen": 466583580,
+      "step": 11880
+    },
+    {
+      "epoch": 0.5761461443878447,
+      "grad_norm": 0.37905943393707275,
+      "learning_rate": 2.008258225846594e-05,
+      "loss": 1.2825,
+      "num_input_tokens_seen": 466995580,
+      "step": 11890
+    },
+    {
+      "epoch": 0.5766307080080195,
+      "grad_norm": 0.4146198034286499,
+      "learning_rate": 2.0044118283872842e-05,
+      "loss": 1.2915,
+      "num_input_tokens_seen": 467394060,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5771152716281943,
+      "grad_norm": 0.40112000703811646,
+      "learning_rate": 2.0005666516663844e-05,
+      "loss": 1.2813,
+      "num_input_tokens_seen": 467762492,
+      "step": 11910
+    },
+    {
+      "epoch": 0.5775998352483691,
+      "grad_norm": 0.3847469091415405,
+      "learning_rate": 1.9967227051553798e-05,
+      "loss": 1.3034,
+      "num_input_tokens_seen": 468158852,
+      "step": 11920
+    },
+    {
+      "epoch": 0.5780843988685439,
+      "grad_norm": 0.39969122409820557,
+      "learning_rate": 1.992879998322723e-05,
+      "loss": 1.2962,
+      "num_input_tokens_seen": 468564816,
+      "step": 11930
+    },
+    {
+      "epoch": 0.5785689624887188,
+      "grad_norm": 0.42271485924720764,
+      "learning_rate": 1.9890385406338118e-05,
+      "loss": 1.2821,
+      "num_input_tokens_seen": 468965016,
+      "step": 11940
+    },
+    {
+      "epoch": 0.5790535261088936,
+      "grad_norm": 0.4109669029712677,
+      "learning_rate": 1.9851983415509704e-05,
+      "loss": 1.2964,
+      "num_input_tokens_seen": 469374000,
+      "step": 11950
+    },
+    {
+      "epoch": 0.5795380897290684,
+      "grad_norm": 0.38713210821151733,
+      "learning_rate": 1.981359410533418e-05,
+      "loss": 1.2827,
+      "num_input_tokens_seen": 469791140,
+      "step": 11960
+    },
+    {
+      "epoch": 0.5800226533492432,
+      "grad_norm": 0.39154934883117676,
+      "learning_rate": 1.9775217570372556e-05,
+      "loss": 1.2617,
+      "num_input_tokens_seen": 470165196,
+      "step": 11970
+    },
+    {
+      "epoch": 0.580507216969418,
+      "grad_norm": 0.4089168310165405,
+      "learning_rate": 1.9736853905154334e-05,
+      "loss": 1.2968,
+      "num_input_tokens_seen": 470550712,
+      "step": 11980
+    },
+    {
+      "epoch": 0.5809917805895928,
+      "grad_norm": 0.45818933844566345,
+      "learning_rate": 1.9698503204177342e-05,
+      "loss": 1.2931,
+      "num_input_tokens_seen": 470936976,
+      "step": 11990
+    },
+    {
+      "epoch": 0.5814763442097676,
+      "grad_norm": 0.42065727710723877,
+      "learning_rate": 1.9660165561907447e-05,
+      "loss": 1.3308,
+      "num_input_tokens_seen": 471314876,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5814763442097676,
+      "eval_loss": 1.4008795022964478,
+      "eval_runtime": 3.8344,
+      "eval_samples_per_second": 39.12,
+      "eval_steps_per_second": 4.955,
+      "num_input_tokens_seen": 471314876,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5819609078299424,
+      "grad_norm": 0.38812288641929626,
+      "learning_rate": 1.9621841072778387e-05,
+      "loss": 1.2928,
+      "num_input_tokens_seen": 471727100,
+      "step": 12010
+    },
+    {
+      "epoch": 0.5824454714501172,
+      "grad_norm": 0.402675062417984,
+      "learning_rate": 1.9583529831191448e-05,
+      "loss": 1.3312,
+      "num_input_tokens_seen": 472107148,
+      "step": 12020
+    },
+    {
+      "epoch": 0.5829300350702921,
+      "grad_norm": 0.4156324863433838,
+      "learning_rate": 1.954523193151534e-05,
+      "loss": 1.3157,
+      "num_input_tokens_seen": 472518180,
+      "step": 12030
+    },
+    {
+      "epoch": 0.5834145986904669,
+      "grad_norm": 0.40975773334503174,
+      "learning_rate": 1.9506947468085866e-05,
+      "loss": 1.2565,
+      "num_input_tokens_seen": 472894628,
+      "step": 12040
+    },
+    {
+      "epoch": 0.5838991623106417,
+      "grad_norm": 0.4518705904483795,
+      "learning_rate": 1.9468676535205767e-05,
+      "loss": 1.2686,
+      "num_input_tokens_seen": 473270452,
+      "step": 12050
+    },
+    {
+      "epoch": 0.5843837259308164,
+      "grad_norm": 0.4540398120880127,
+      "learning_rate": 1.9430419227144443e-05,
+      "loss": 1.2771,
+      "num_input_tokens_seen": 473644072,
+      "step": 12060
+    },
+    {
+      "epoch": 0.5848682895509912,
+      "grad_norm": 0.40955179929733276,
+      "learning_rate": 1.939217563813771e-05,
+      "loss": 1.3421,
+      "num_input_tokens_seen": 474010688,
+      "step": 12070
+    },
+    {
+      "epoch": 0.585352853171166,
+      "grad_norm": 0.4388551115989685,
+      "learning_rate": 1.935394586238763e-05,
+      "loss": 1.3129,
+      "num_input_tokens_seen": 474390028,
+      "step": 12080
+    },
+    {
+      "epoch": 0.5858374167913408,
+      "grad_norm": 0.41205498576164246,
+      "learning_rate": 1.93157299940622e-05,
+      "loss": 1.2662,
+      "num_input_tokens_seen": 474801152,
+      "step": 12090
+    },
+    {
+      "epoch": 0.5863219804115156,
+      "grad_norm": 0.41944679617881775,
+      "learning_rate": 1.92775281272952e-05,
+      "loss": 1.2786,
+      "num_input_tokens_seen": 475187368,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5868065440316904,
+      "grad_norm": 0.4028654098510742,
+      "learning_rate": 1.9239340356185892e-05,
+      "loss": 1.2792,
+      "num_input_tokens_seen": 475599252,
+      "step": 12110
+    },
+    {
+      "epoch": 0.5872911076518653,
+      "grad_norm": 0.44524598121643066,
+      "learning_rate": 1.9201166774798833e-05,
+      "loss": 1.3175,
+      "num_input_tokens_seen": 475971476,
+      "step": 12120
+    },
+    {
+      "epoch": 0.5877756712720401,
+      "grad_norm": 0.3849859833717346,
+      "learning_rate": 1.9163007477163616e-05,
+      "loss": 1.2327,
+      "num_input_tokens_seen": 476379776,
+      "step": 12130
+    },
+    {
+      "epoch": 0.5882602348922149,
+      "grad_norm": 0.4216983914375305,
+      "learning_rate": 1.9124862557274668e-05,
+      "loss": 1.2934,
+      "num_input_tokens_seen": 476772296,
+      "step": 12140
+    },
+    {
+      "epoch": 0.5887447985123897,
+      "grad_norm": 0.4107081890106201,
+      "learning_rate": 1.9086732109090965e-05,
+      "loss": 1.3092,
+      "num_input_tokens_seen": 477178020,
+      "step": 12150
+    },
+    {
+      "epoch": 0.5892293621325645,
+      "grad_norm": 0.43871021270751953,
+      "learning_rate": 1.904861622653589e-05,
+      "loss": 1.3208,
+      "num_input_tokens_seen": 477570016,
+      "step": 12160
+    },
+    {
+      "epoch": 0.5897139257527393,
+      "grad_norm": 0.4565945565700531,
+      "learning_rate": 1.9010515003496892e-05,
+      "loss": 1.2592,
+      "num_input_tokens_seen": 477957740,
+      "step": 12170
+    },
+    {
+      "epoch": 0.5901984893729141,
+      "grad_norm": 0.4425964653491974,
+      "learning_rate": 1.8972428533825345e-05,
+      "loss": 1.2855,
+      "num_input_tokens_seen": 478348188,
+      "step": 12180
+    },
+    {
+      "epoch": 0.5906830529930889,
+      "grad_norm": 0.41607752442359924,
+      "learning_rate": 1.8934356911336283e-05,
+      "loss": 1.3222,
+      "num_input_tokens_seen": 478741832,
+      "step": 12190
+    },
+    {
+      "epoch": 0.5911676166132637,
+      "grad_norm": 0.43811753392219543,
+      "learning_rate": 1.8896300229808144e-05,
+      "loss": 1.2914,
+      "num_input_tokens_seen": 479136864,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5916521802334386,
+      "grad_norm": 0.4176853895187378,
+      "learning_rate": 1.8858258582982597e-05,
+      "loss": 1.3503,
+      "num_input_tokens_seen": 479527104,
+      "step": 12210
+    },
+    {
+      "epoch": 0.5921367438536134,
+      "grad_norm": 0.3926706612110138,
+      "learning_rate": 1.8820232064564233e-05,
+      "loss": 1.286,
+      "num_input_tokens_seen": 479884732,
+      "step": 12220
+    },
+    {
+      "epoch": 0.5926213074737882,
+      "grad_norm": 0.38026413321495056,
+      "learning_rate": 1.878222076822043e-05,
+      "loss": 1.2862,
+      "num_input_tokens_seen": 480292660,
+      "step": 12230
+    },
+    {
+      "epoch": 0.593105871093963,
+      "grad_norm": 0.4443627595901489,
+      "learning_rate": 1.8744224787581024e-05,
+      "loss": 1.2863,
+      "num_input_tokens_seen": 480669036,
+      "step": 12240
+    },
+    {
+      "epoch": 0.5935904347141377,
+      "grad_norm": 0.3894060552120209,
+      "learning_rate": 1.870624421623816e-05,
+      "loss": 1.315,
+      "num_input_tokens_seen": 481033564,
+      "step": 12250
+    },
+    {
+      "epoch": 0.5940749983343125,
+      "grad_norm": 0.4227178692817688,
+      "learning_rate": 1.8668279147746e-05,
+      "loss": 1.3146,
+      "num_input_tokens_seen": 481449056,
+      "step": 12260
+    },
+    {
+      "epoch": 0.5945595619544873,
+      "grad_norm": 0.38117271661758423,
+      "learning_rate": 1.863032967562055e-05,
+      "loss": 1.2849,
+      "num_input_tokens_seen": 481832856,
+      "step": 12270
+    },
+    {
+      "epoch": 0.5950441255746621,
+      "grad_norm": 0.446122944355011,
+      "learning_rate": 1.859239589333936e-05,
+      "loss": 1.3231,
+      "num_input_tokens_seen": 482219828,
+      "step": 12280
+    },
+    {
+      "epoch": 0.5955286891948369,
+      "grad_norm": 0.4982196092605591,
+      "learning_rate": 1.855447789434137e-05,
+      "loss": 1.2839,
+      "num_input_tokens_seen": 482624804,
+      "step": 12290
+    },
+    {
+      "epoch": 0.5960132528150118,
+      "grad_norm": 0.44610318541526794,
+      "learning_rate": 1.851657577202661e-05,
+      "loss": 1.2712,
+      "num_input_tokens_seen": 483017968,
+      "step": 12300
+    },
+    {
+      "epoch": 0.5964978164351866,
+      "grad_norm": 0.4101884067058563,
+      "learning_rate": 1.8478689619756026e-05,
+      "loss": 1.3168,
+      "num_input_tokens_seen": 483408892,
+      "step": 12310
+    },
+    {
+      "epoch": 0.5969823800553614,
+      "grad_norm": 0.38893336057662964,
+      "learning_rate": 1.8440819530851225e-05,
+      "loss": 1.2848,
+      "num_input_tokens_seen": 483774232,
+      "step": 12320
+    },
+    {
+      "epoch": 0.5974669436755362,
+      "grad_norm": 0.4360596835613251,
+      "learning_rate": 1.8402965598594227e-05,
+      "loss": 1.3122,
+      "num_input_tokens_seen": 484172492,
+      "step": 12330
+    },
+    {
+      "epoch": 0.597951507295711,
+      "grad_norm": 0.3694493770599365,
+      "learning_rate": 1.8365127916227288e-05,
+      "loss": 1.2786,
+      "num_input_tokens_seen": 484548300,
+      "step": 12340
+    },
+    {
+      "epoch": 0.5984360709158858,
+      "grad_norm": 0.4072704315185547,
+      "learning_rate": 1.8327306576952592e-05,
+      "loss": 1.3197,
+      "num_input_tokens_seen": 484966920,
+      "step": 12350
+    },
+    {
+      "epoch": 0.5989206345360606,
+      "grad_norm": 0.43515607714653015,
+      "learning_rate": 1.828950167393211e-05,
+      "loss": 1.2959,
+      "num_input_tokens_seen": 485363496,
+      "step": 12360
+    },
+    {
+      "epoch": 0.5994051981562354,
+      "grad_norm": 0.3934329152107239,
+      "learning_rate": 1.8251713300287294e-05,
+      "loss": 1.2764,
+      "num_input_tokens_seen": 485781532,
+      "step": 12370
+    },
+    {
+      "epoch": 0.5998897617764102,
+      "grad_norm": 0.43624651432037354,
+      "learning_rate": 1.821394154909891e-05,
+      "loss": 1.2791,
+      "num_input_tokens_seen": 486165704,
+      "step": 12380
+    },
+    {
+      "epoch": 0.6003743253965851,
+      "grad_norm": 0.4212031960487366,
+      "learning_rate": 1.817618651340675e-05,
+      "loss": 1.2857,
+      "num_input_tokens_seen": 486559644,
+      "step": 12390
+    },
+    {
+      "epoch": 0.6008588890167599,
+      "grad_norm": 0.4342302680015564,
+      "learning_rate": 1.813844828620946e-05,
+      "loss": 1.265,
+      "num_input_tokens_seen": 486973628,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6013434526369347,
+      "grad_norm": 0.35909196734428406,
+      "learning_rate": 1.8100726960464254e-05,
+      "loss": 1.2953,
+      "num_input_tokens_seen": 487389200,
+      "step": 12410
+    },
+    {
+      "epoch": 0.6018280162571095,
+      "grad_norm": 0.4191993772983551,
+      "learning_rate": 1.8063022629086752e-05,
+      "loss": 1.2548,
+      "num_input_tokens_seen": 487788196,
+      "step": 12420
+    },
+    {
+      "epoch": 0.6023125798772843,
+      "grad_norm": 0.4022323191165924,
+      "learning_rate": 1.8025335384950665e-05,
+      "loss": 1.2569,
+      "num_input_tokens_seen": 488152244,
+      "step": 12430
+    },
+    {
+      "epoch": 0.602797143497459,
+      "grad_norm": 0.3971206545829773,
+      "learning_rate": 1.7987665320887666e-05,
+      "loss": 1.3111,
+      "num_input_tokens_seen": 488555012,
+      "step": 12440
+    },
+    {
+      "epoch": 0.6032817071176338,
+      "grad_norm": 0.40955471992492676,
+      "learning_rate": 1.795001252968706e-05,
+      "loss": 1.2905,
+      "num_input_tokens_seen": 488963324,
+      "step": 12450
+    },
+    {
+      "epoch": 0.6037662707378086,
+      "grad_norm": 0.4047541320323944,
+      "learning_rate": 1.7912377104095645e-05,
+      "loss": 1.3335,
+      "num_input_tokens_seen": 489371688,
+      "step": 12460
+    },
+    {
+      "epoch": 0.6042508343579834,
+      "grad_norm": 0.4085901379585266,
+      "learning_rate": 1.787475913681743e-05,
+      "loss": 1.2776,
+      "num_input_tokens_seen": 489791296,
+      "step": 12470
+    },
+    {
+      "epoch": 0.6047353979781583,
+      "grad_norm": 0.40488049387931824,
+      "learning_rate": 1.783715872051341e-05,
+      "loss": 1.3007,
+      "num_input_tokens_seen": 490158224,
+      "step": 12480
+    },
+    {
+      "epoch": 0.6052199615983331,
+      "grad_norm": 0.37421244382858276,
+      "learning_rate": 1.7799575947801374e-05,
+      "loss": 1.3284,
+      "num_input_tokens_seen": 490589156,
+      "step": 12490
+    },
+    {
+      "epoch": 0.6057045252185079,
+      "grad_norm": 0.44858112931251526,
+      "learning_rate": 1.776201091125561e-05,
+      "loss": 1.2499,
+      "num_input_tokens_seen": 491003784,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6061890888386827,
+      "grad_norm": 0.3810955286026001,
+      "learning_rate": 1.7724463703406766e-05,
+      "loss": 1.2935,
+      "num_input_tokens_seen": 491417140,
+      "step": 12510
+    },
+    {
+      "epoch": 0.6066736524588575,
+      "grad_norm": 0.4152781665325165,
+      "learning_rate": 1.768693441674153e-05,
+      "loss": 1.3209,
+      "num_input_tokens_seen": 491809688,
+      "step": 12520
+    },
+    {
+      "epoch": 0.6071582160790323,
+      "grad_norm": 0.4103911221027374,
+      "learning_rate": 1.764942314370248e-05,
+      "loss": 1.2924,
+      "num_input_tokens_seen": 492210364,
+      "step": 12530
+    },
+    {
+      "epoch": 0.6076427796992071,
+      "grad_norm": 0.4096033275127411,
+      "learning_rate": 1.761192997668781e-05,
+      "loss": 1.25,
+      "num_input_tokens_seen": 492627760,
+      "step": 12540
+    },
+    {
+      "epoch": 0.6081273433193819,
+      "grad_norm": 0.42093944549560547,
+      "learning_rate": 1.7574455008051115e-05,
+      "loss": 1.2955,
+      "num_input_tokens_seen": 493015580,
+      "step": 12550
+    },
+    {
+      "epoch": 0.6086119069395567,
+      "grad_norm": 0.43841052055358887,
+      "learning_rate": 1.753699833010115e-05,
+      "loss": 1.2778,
+      "num_input_tokens_seen": 493413392,
+      "step": 12560
+    },
+    {
+      "epoch": 0.6090964705597316,
+      "grad_norm": 0.3942921459674835,
+      "learning_rate": 1.7499560035101653e-05,
+      "loss": 1.2609,
+      "num_input_tokens_seen": 493804560,
+      "step": 12570
+    },
+    {
+      "epoch": 0.6095810341799064,
+      "grad_norm": 0.42653924226760864,
+      "learning_rate": 1.746214021527103e-05,
+      "loss": 1.2611,
+      "num_input_tokens_seen": 494185728,
+      "step": 12580
+    },
+    {
+      "epoch": 0.6100655978000812,
+      "grad_norm": 0.3946724236011505,
+      "learning_rate": 1.7424738962782222e-05,
+      "loss": 1.2609,
+      "num_input_tokens_seen": 494597784,
+      "step": 12590
+    },
+    {
+      "epoch": 0.610550161420256,
+      "grad_norm": 0.37868139147758484,
+      "learning_rate": 1.7387356369762426e-05,
+      "loss": 1.2714,
+      "num_input_tokens_seen": 494995584,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6110347250404308,
+      "grad_norm": 0.42459574341773987,
+      "learning_rate": 1.734999252829285e-05,
+      "loss": 1.2983,
+      "num_input_tokens_seen": 495396436,
+      "step": 12610
+    },
+    {
+      "epoch": 0.6115192886606056,
+      "grad_norm": 0.3998323976993561,
+      "learning_rate": 1.7312647530408548e-05,
+      "loss": 1.2819,
+      "num_input_tokens_seen": 495790568,
+      "step": 12620
+    },
+    {
+      "epoch": 0.6120038522807804,
+      "grad_norm": 0.4067113995552063,
+      "learning_rate": 1.7275321468098133e-05,
+      "loss": 1.2594,
+      "num_input_tokens_seen": 496182596,
+      "step": 12630
+    },
+    {
+      "epoch": 0.6124884159009552,
+      "grad_norm": 0.41096609830856323,
+      "learning_rate": 1.72380144333036e-05,
+      "loss": 1.2807,
+      "num_input_tokens_seen": 496584560,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6129729795211301,
+      "grad_norm": 0.448887437582016,
+      "learning_rate": 1.720072651792004e-05,
+      "loss": 1.2719,
+      "num_input_tokens_seen": 496979328,
+      "step": 12650
+    },
+    {
+      "epoch": 0.6134575431413049,
+      "grad_norm": 0.4053841829299927,
+      "learning_rate": 1.716345781379549e-05,
+      "loss": 1.3734,
+      "num_input_tokens_seen": 497342904,
+      "step": 12660
+    },
+    {
+      "epoch": 0.6139421067614796,
+      "grad_norm": 0.4214945137500763,
+      "learning_rate": 1.7126208412730628e-05,
+      "loss": 1.3149,
+      "num_input_tokens_seen": 497758472,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6144266703816544,
+      "grad_norm": 0.42576828598976135,
+      "learning_rate": 1.708897840647861e-05,
+      "loss": 1.2727,
+      "num_input_tokens_seen": 498155548,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6149112340018292,
+      "grad_norm": 0.3854653537273407,
+      "learning_rate": 1.7051767886744808e-05,
+      "loss": 1.2532,
+      "num_input_tokens_seen": 498540656,
+      "step": 12690
+    },
+    {
+      "epoch": 0.615395797622004,
+      "grad_norm": 0.42444708943367004,
+      "learning_rate": 1.701457694518661e-05,
+      "loss": 1.2709,
+      "num_input_tokens_seen": 498973360,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6158803612421788,
+      "grad_norm": 0.3879097104072571,
+      "learning_rate": 1.697740567341314e-05,
+      "loss": 1.3207,
+      "num_input_tokens_seen": 499404028,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6163649248623536,
+      "grad_norm": 0.3972528874874115,
+      "learning_rate": 1.694025416298511e-05,
+      "loss": 1.3065,
+      "num_input_tokens_seen": 499780220,
+      "step": 12720
+    },
+    {
+      "epoch": 0.6168494884825284,
+      "grad_norm": 0.4193166196346283,
+      "learning_rate": 1.6903122505414552e-05,
+      "loss": 1.3359,
+      "num_input_tokens_seen": 500185720,
+      "step": 12730
+    },
+    {
+      "epoch": 0.6173340521027033,
+      "grad_norm": 0.40607649087905884,
+      "learning_rate": 1.686601079216457e-05,
+      "loss": 1.2789,
+      "num_input_tokens_seen": 500562168,
+      "step": 12740
+    },
+    {
+      "epoch": 0.6178186157228781,
+      "grad_norm": 0.4222082197666168,
+      "learning_rate": 1.682891911464917e-05,
+      "loss": 1.2436,
+      "num_input_tokens_seen": 500990044,
+      "step": 12750
+    },
+    {
+      "epoch": 0.6183031793430529,
+      "grad_norm": 0.40841108560562134,
+      "learning_rate": 1.6791847564232982e-05,
+      "loss": 1.2466,
+      "num_input_tokens_seen": 501376276,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6187877429632277,
+      "grad_norm": 0.4501042664051056,
+      "learning_rate": 1.6754796232231084e-05,
+      "loss": 1.3059,
+      "num_input_tokens_seen": 501734004,
+      "step": 12770
+    },
+    {
+      "epoch": 0.6192723065834025,
+      "grad_norm": 0.3940059244632721,
+      "learning_rate": 1.6717765209908722e-05,
+      "loss": 1.3457,
+      "num_input_tokens_seen": 502130812,
+      "step": 12780
+    },
+    {
+      "epoch": 0.6197568702035773,
+      "grad_norm": 0.372949481010437,
+      "learning_rate": 1.668075458848115e-05,
+      "loss": 1.2881,
+      "num_input_tokens_seen": 502544872,
+      "step": 12790
+    },
+    {
+      "epoch": 0.6202414338237521,
+      "grad_norm": 0.4232177138328552,
+      "learning_rate": 1.6643764459113324e-05,
+      "loss": 1.2777,
+      "num_input_tokens_seen": 502909916,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6207259974439269,
+      "grad_norm": 0.40586578845977783,
+      "learning_rate": 1.6606794912919776e-05,
+      "loss": 1.2462,
+      "num_input_tokens_seen": 503313708,
+      "step": 12810
+    },
+    {
+      "epoch": 0.6212105610641017,
+      "grad_norm": 0.40091219544410706,
+      "learning_rate": 1.6569846040964293e-05,
+      "loss": 1.305,
+      "num_input_tokens_seen": 503697412,
+      "step": 12820
+    },
+    {
+      "epoch": 0.6216951246842766,
+      "grad_norm": 0.42310020327568054,
+      "learning_rate": 1.653291793425978e-05,
+      "loss": 1.2631,
+      "num_input_tokens_seen": 504107344,
+      "step": 12830
+    },
+    {
+      "epoch": 0.6221796883044514,
+      "grad_norm": 0.42501911520957947,
+      "learning_rate": 1.6496010683767936e-05,
+      "loss": 1.275,
+      "num_input_tokens_seen": 504491792,
+      "step": 12840
+    },
+    {
+      "epoch": 0.6226642519246262,
+      "grad_norm": 0.42510515451431274,
+      "learning_rate": 1.6459124380399144e-05,
+      "loss": 1.2764,
+      "num_input_tokens_seen": 504881504,
+      "step": 12850
+    },
+    {
+      "epoch": 0.623148815544801,
+      "grad_norm": 0.4044088125228882,
+      "learning_rate": 1.6422259115012165e-05,
+      "loss": 1.2925,
+      "num_input_tokens_seen": 505293712,
+      "step": 12860
+    },
+    {
+      "epoch": 0.6236333791649757,
+      "grad_norm": 0.4045158326625824,
+      "learning_rate": 1.638541497841392e-05,
+      "loss": 1.2921,
+      "num_input_tokens_seen": 505655600,
+      "step": 12870
+    },
+    {
+      "epoch": 0.6241179427851505,
+      "grad_norm": 0.42825374007225037,
+      "learning_rate": 1.6348592061359334e-05,
+      "loss": 1.2857,
+      "num_input_tokens_seen": 506040624,
+      "step": 12880
+    },
+    {
+      "epoch": 0.6246025064053253,
+      "grad_norm": 0.39818504452705383,
+      "learning_rate": 1.6311790454551e-05,
+      "loss": 1.2945,
+      "num_input_tokens_seen": 506464380,
+      "step": 12890
+    },
+    {
+      "epoch": 0.6250870700255001,
+      "grad_norm": 0.46967610716819763,
+      "learning_rate": 1.6275010248639085e-05,
+      "loss": 1.2733,
+      "num_input_tokens_seen": 506879284,
+      "step": 12900
+    },
+    {
+      "epoch": 0.6255716336456749,
+      "grad_norm": 0.46247559785842896,
+      "learning_rate": 1.6238251534220982e-05,
+      "loss": 1.2628,
+      "num_input_tokens_seen": 507255672,
+      "step": 12910
+    },
+    {
+      "epoch": 0.6260561972658498,
+      "grad_norm": 0.39192506670951843,
+      "learning_rate": 1.6201514401841204e-05,
+      "loss": 1.2749,
+      "num_input_tokens_seen": 507642376,
+      "step": 12920
+    },
+    {
+      "epoch": 0.6265407608860246,
+      "grad_norm": 0.41501137614250183,
+      "learning_rate": 1.6164798941991046e-05,
+      "loss": 1.2847,
+      "num_input_tokens_seen": 508025372,
+      "step": 12930
+    },
+    {
+      "epoch": 0.6270253245061994,
+      "grad_norm": 0.38921937346458435,
+      "learning_rate": 1.6128105245108464e-05,
+      "loss": 1.2737,
+      "num_input_tokens_seen": 508418732,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6275098881263742,
+      "grad_norm": 0.3864487111568451,
+      "learning_rate": 1.609143340157777e-05,
+      "loss": 1.2822,
+      "num_input_tokens_seen": 508802832,
+      "step": 12950
+    },
+    {
+      "epoch": 0.627994451746549,
+      "grad_norm": 0.39535877108573914,
+      "learning_rate": 1.6054783501729488e-05,
+      "loss": 1.3428,
+      "num_input_tokens_seen": 509204136,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6284790153667238,
+      "grad_norm": 0.47437265515327454,
+      "learning_rate": 1.6018155635840046e-05,
+      "loss": 1.261,
+      "num_input_tokens_seen": 509580876,
+      "step": 12970
+    },
+    {
+      "epoch": 0.6289635789868986,
+      "grad_norm": 0.40359655022621155,
+      "learning_rate": 1.5981549894131628e-05,
+      "loss": 1.2695,
+      "num_input_tokens_seen": 509961156,
+      "step": 12980
+    },
+    {
+      "epoch": 0.6294481426070734,
+      "grad_norm": 0.4515499770641327,
+      "learning_rate": 1.5944966366771928e-05,
+      "loss": 1.3089,
+      "num_input_tokens_seen": 510362116,
+      "step": 12990
+    },
+    {
+      "epoch": 0.6299327062272482,
+      "grad_norm": 0.384777307510376,
+      "learning_rate": 1.5908405143873878e-05,
+      "loss": 1.296,
+      "num_input_tokens_seen": 510738872,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6304172698474231,
+      "grad_norm": 0.36402514576911926,
+      "learning_rate": 1.587186631549552e-05,
+      "loss": 1.2881,
+      "num_input_tokens_seen": 511139464,
+      "step": 13010
+    },
+    {
+      "epoch": 0.6309018334675979,
+      "grad_norm": 0.36122068762779236,
+      "learning_rate": 1.5835349971639694e-05,
+      "loss": 1.2815,
+      "num_input_tokens_seen": 511548508,
+      "step": 13020
+    },
+    {
+      "epoch": 0.6313863970877727,
+      "grad_norm": 0.4500305652618408,
+      "learning_rate": 1.5798856202253885e-05,
+      "loss": 1.2844,
+      "num_input_tokens_seen": 511919064,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6318709607079475,
+      "grad_norm": 0.41978541016578674,
+      "learning_rate": 1.5762385097229952e-05,
+      "loss": 1.3016,
+      "num_input_tokens_seen": 512296380,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6323555243281223,
+      "grad_norm": 0.40535783767700195,
+      "learning_rate": 1.5725936746403952e-05,
+      "loss": 1.3093,
+      "num_input_tokens_seen": 512698204,
+      "step": 13050
+    },
+    {
+      "epoch": 0.632840087948297,
+      "grad_norm": 0.41118356585502625,
+      "learning_rate": 1.568951123955585e-05,
+      "loss": 1.3026,
+      "num_input_tokens_seen": 513094244,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6333246515684718,
+      "grad_norm": 0.4092055559158325,
+      "learning_rate": 1.565310866640939e-05,
+      "loss": 1.2814,
+      "num_input_tokens_seen": 513478972,
+      "step": 13070
+    },
+    {
+      "epoch": 0.6338092151886466,
+      "grad_norm": 0.42620420455932617,
+      "learning_rate": 1.561672911663179e-05,
+      "loss": 1.2906,
+      "num_input_tokens_seen": 513892816,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6342937788088214,
+      "grad_norm": 0.40342986583709717,
+      "learning_rate": 1.558037267983358e-05,
+      "loss": 1.2528,
+      "num_input_tokens_seen": 514268756,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6347783424289963,
+      "grad_norm": 0.44240349531173706,
+      "learning_rate": 1.554403944556834e-05,
+      "loss": 1.2711,
+      "num_input_tokens_seen": 514676816,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6352629060491711,
+      "grad_norm": 0.38063910603523254,
+      "learning_rate": 1.5507729503332503e-05,
+      "loss": 1.2958,
+      "num_input_tokens_seen": 515102476,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6357474696693459,
+      "grad_norm": 0.4031839072704315,
+      "learning_rate": 1.547144294256514e-05,
+      "loss": 1.3018,
+      "num_input_tokens_seen": 515484244,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6362320332895207,
+      "grad_norm": 0.4612565338611603,
+      "learning_rate": 1.5435179852647712e-05,
+      "loss": 1.2909,
+      "num_input_tokens_seen": 515868796,
+      "step": 13130
+    },
+    {
+      "epoch": 0.6367165969096955,
+      "grad_norm": 0.4232649803161621,
+      "learning_rate": 1.539894032290389e-05,
+      "loss": 1.2522,
+      "num_input_tokens_seen": 516277108,
+      "step": 13140
+    },
+    {
+      "epoch": 0.6372011605298703,
+      "grad_norm": 0.4205443859100342,
+      "learning_rate": 1.5362724442599276e-05,
+      "loss": 1.3396,
+      "num_input_tokens_seen": 516659132,
+      "step": 13150
+    },
+    {
+      "epoch": 0.6376857241500451,
+      "grad_norm": 0.3871660828590393,
+      "learning_rate": 1.532653230094125e-05,
+      "loss": 1.312,
+      "num_input_tokens_seen": 517083020,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6381702877702199,
+      "grad_norm": 0.42630356550216675,
+      "learning_rate": 1.529036398707869e-05,
+      "loss": 1.2414,
+      "num_input_tokens_seen": 517464668,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6386548513903947,
+      "grad_norm": 0.40976426005363464,
+      "learning_rate": 1.5254219590101816e-05,
+      "loss": 1.2985,
+      "num_input_tokens_seen": 517860704,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6391394150105696,
+      "grad_norm": 0.4328073561191559,
+      "learning_rate": 1.5218099199041902e-05,
+      "loss": 1.2917,
+      "num_input_tokens_seen": 518276916,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6396239786307444,
+      "grad_norm": 0.3879329264163971,
+      "learning_rate": 1.5182002902871123e-05,
+      "loss": 1.3094,
+      "num_input_tokens_seen": 518654656,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6401085422509192,
+      "grad_norm": 0.41468775272369385,
+      "learning_rate": 1.5145930790502267e-05,
+      "loss": 1.307,
+      "num_input_tokens_seen": 519033292,
+      "step": 13210
+    },
+    {
+      "epoch": 0.640593105871094,
+      "grad_norm": 0.40821224451065063,
+      "learning_rate": 1.5109882950788586e-05,
+      "loss": 1.2277,
+      "num_input_tokens_seen": 519406144,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6410776694912688,
+      "grad_norm": 0.38090160489082336,
+      "learning_rate": 1.5073859472523514e-05,
+      "loss": 1.2876,
+      "num_input_tokens_seen": 519814676,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6415622331114436,
+      "grad_norm": 0.39433416724205017,
+      "learning_rate": 1.50378604444405e-05,
+      "loss": 1.2361,
+      "num_input_tokens_seen": 520205984,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6420467967316184,
+      "grad_norm": 0.3543625771999359,
+      "learning_rate": 1.5001885955212758e-05,
+      "loss": 1.276,
+      "num_input_tokens_seen": 520602212,
+      "step": 13250
+    },
+    {
+      "epoch": 0.6425313603517931,
+      "grad_norm": 0.43784913420677185,
+      "learning_rate": 1.4965936093453054e-05,
+      "loss": 1.307,
+      "num_input_tokens_seen": 520984044,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6430159239719679,
+      "grad_norm": 0.38104525208473206,
+      "learning_rate": 1.4930010947713513e-05,
+      "loss": 1.287,
+      "num_input_tokens_seen": 521381332,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6435004875921428,
+      "grad_norm": 0.42758843302726746,
+      "learning_rate": 1.4894110606485334e-05,
+      "loss": 1.3144,
+      "num_input_tokens_seen": 521758556,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6439850512123176,
+      "grad_norm": 0.3898857831954956,
+      "learning_rate": 1.4858235158198675e-05,
+      "loss": 1.2667,
+      "num_input_tokens_seen": 522155024,
+      "step": 13290
+    },
+    {
+      "epoch": 0.6444696148324924,
+      "grad_norm": 0.406747967004776,
+      "learning_rate": 1.482238469122232e-05,
+      "loss": 1.277,
+      "num_input_tokens_seen": 522511408,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6449541784526672,
+      "grad_norm": 0.3949616849422455,
+      "learning_rate": 1.4786559293863566e-05,
+      "loss": 1.2483,
+      "num_input_tokens_seen": 522886168,
+      "step": 13310
+    },
+    {
+      "epoch": 0.645438742072842,
+      "grad_norm": 0.4430643916130066,
+      "learning_rate": 1.4750759054367923e-05,
+      "loss": 1.2268,
+      "num_input_tokens_seen": 523282444,
+      "step": 13320
+    },
+    {
+      "epoch": 0.6459233056930168,
+      "grad_norm": 0.4223615825176239,
+      "learning_rate": 1.4714984060918962e-05,
+      "loss": 1.272,
+      "num_input_tokens_seen": 523654248,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6464078693131916,
+      "grad_norm": 0.3976212739944458,
+      "learning_rate": 1.4679234401638043e-05,
+      "loss": 1.2618,
+      "num_input_tokens_seen": 524040568,
+      "step": 13340
+    },
+    {
+      "epoch": 0.6468924329333664,
+      "grad_norm": 0.40596914291381836,
+      "learning_rate": 1.464351016458414e-05,
+      "loss": 1.2804,
+      "num_input_tokens_seen": 524461668,
+      "step": 13350
+    },
+    {
+      "epoch": 0.6473769965535413,
+      "grad_norm": 0.3791384696960449,
+      "learning_rate": 1.460781143775359e-05,
+      "loss": 1.304,
+      "num_input_tokens_seen": 524855056,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6478615601737161,
+      "grad_norm": 0.3824685215950012,
+      "learning_rate": 1.457213830907992e-05,
+      "loss": 1.288,
+      "num_input_tokens_seen": 525251636,
+      "step": 13370
+    },
+    {
+      "epoch": 0.6483461237938909,
+      "grad_norm": 0.39863574504852295,
+      "learning_rate": 1.453649086643356e-05,
+      "loss": 1.2836,
+      "num_input_tokens_seen": 525642284,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6488306874140657,
+      "grad_norm": 0.4155414402484894,
+      "learning_rate": 1.4500869197621708e-05,
+      "loss": 1.2526,
+      "num_input_tokens_seen": 526045452,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6493152510342405,
+      "grad_norm": 0.41306737065315247,
+      "learning_rate": 1.446527339038808e-05,
+      "loss": 1.2497,
+      "num_input_tokens_seen": 526454148,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6497998146544153,
+      "grad_norm": 0.4045097529888153,
+      "learning_rate": 1.4429703532412642e-05,
+      "loss": 1.2635,
+      "num_input_tokens_seen": 526858372,
+      "step": 13410
+    },
+    {
+      "epoch": 0.6502843782745901,
+      "grad_norm": 0.39056453108787537,
+      "learning_rate": 1.4394159711311494e-05,
+      "loss": 1.3007,
+      "num_input_tokens_seen": 527262932,
+      "step": 13420
+    },
+    {
+      "epoch": 0.6507689418947649,
+      "grad_norm": 0.4115632474422455,
+      "learning_rate": 1.435864201463657e-05,
+      "loss": 1.2925,
+      "num_input_tokens_seen": 527689048,
+      "step": 13430
+    },
+    {
+      "epoch": 0.6512535055149397,
+      "grad_norm": 0.399703711271286,
+      "learning_rate": 1.4323150529875462e-05,
+      "loss": 1.2915,
+      "num_input_tokens_seen": 528076752,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6517380691351146,
+      "grad_norm": 0.39391404390335083,
+      "learning_rate": 1.4287685344451202e-05,
+      "loss": 1.2642,
+      "num_input_tokens_seen": 528478876,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6522226327552894,
+      "grad_norm": 0.41902291774749756,
+      "learning_rate": 1.4252246545722048e-05,
+      "loss": 1.2977,
+      "num_input_tokens_seen": 528884840,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6527071963754641,
+      "grad_norm": 0.42969298362731934,
+      "learning_rate": 1.4216834220981235e-05,
+      "loss": 1.2721,
+      "num_input_tokens_seen": 529259760,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6531917599956389,
+      "grad_norm": 0.41451066732406616,
+      "learning_rate": 1.4181448457456814e-05,
+      "loss": 1.2888,
+      "num_input_tokens_seen": 529676680,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6536763236158137,
+      "grad_norm": 0.4217407703399658,
+      "learning_rate": 1.4146089342311391e-05,
+      "loss": 1.2666,
+      "num_input_tokens_seen": 530073596,
+      "step": 13490
+    },
+    {
+      "epoch": 0.6541608872359885,
+      "grad_norm": 0.3723660111427307,
+      "learning_rate": 1.4110756962641952e-05,
+      "loss": 1.2463,
+      "num_input_tokens_seen": 530447064,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6546454508561633,
+      "grad_norm": 0.42004016041755676,
+      "learning_rate": 1.4075451405479598e-05,
+      "loss": 1.2802,
+      "num_input_tokens_seen": 530851496,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6551300144763381,
+      "grad_norm": 0.4187595546245575,
+      "learning_rate": 1.4040172757789388e-05,
+      "loss": 1.2933,
+      "num_input_tokens_seen": 531264604,
+      "step": 13520
+    },
+    {
+      "epoch": 0.6556145780965129,
+      "grad_norm": 0.4250360429286957,
+      "learning_rate": 1.4004921106470098e-05,
+      "loss": 1.2751,
+      "num_input_tokens_seen": 531679120,
+      "step": 13530
+    },
+    {
+      "epoch": 0.6560991417166878,
+      "grad_norm": 0.37126293778419495,
+      "learning_rate": 1.3969696538353977e-05,
+      "loss": 1.284,
+      "num_input_tokens_seen": 532051596,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6565837053368626,
+      "grad_norm": 0.3943636119365692,
+      "learning_rate": 1.3934499140206596e-05,
+      "loss": 1.2617,
+      "num_input_tokens_seen": 532444860,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6570682689570374,
+      "grad_norm": 0.4207961857318878,
+      "learning_rate": 1.3899328998726574e-05,
+      "loss": 1.2434,
+      "num_input_tokens_seen": 532852472,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6575528325772122,
+      "grad_norm": 0.41514161229133606,
+      "learning_rate": 1.3864186200545403e-05,
+      "loss": 1.2763,
+      "num_input_tokens_seen": 533252152,
+      "step": 13570
+    },
+    {
+      "epoch": 0.658037396197387,
+      "grad_norm": 0.422195702791214,
+      "learning_rate": 1.3829070832227234e-05,
+      "loss": 1.2649,
+      "num_input_tokens_seen": 533646968,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6585219598175618,
+      "grad_norm": 0.3905238211154938,
+      "learning_rate": 1.3793982980268644e-05,
+      "loss": 1.2159,
+      "num_input_tokens_seen": 534000320,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6590065234377366,
+      "grad_norm": 0.43043792247772217,
+      "learning_rate": 1.3758922731098406e-05,
+      "loss": 1.3498,
+      "num_input_tokens_seen": 534396632,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6594910870579114,
+      "grad_norm": 0.4008113741874695,
+      "learning_rate": 1.372389017107735e-05,
+      "loss": 1.2894,
+      "num_input_tokens_seen": 534782340,
+      "step": 13610
+    },
+    {
+      "epoch": 0.6599756506780862,
+      "grad_norm": 0.3774920403957367,
+      "learning_rate": 1.3688885386498052e-05,
+      "loss": 1.2679,
+      "num_input_tokens_seen": 535198784,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6604602142982611,
+      "grad_norm": 0.42692962288856506,
+      "learning_rate": 1.3653908463584717e-05,
+      "loss": 1.239,
+      "num_input_tokens_seen": 535613348,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6609447779184359,
+      "grad_norm": 0.4580807387828827,
+      "learning_rate": 1.3618959488492875e-05,
+      "loss": 1.2745,
+      "num_input_tokens_seen": 536005904,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6614293415386107,
+      "grad_norm": 0.4274024963378906,
+      "learning_rate": 1.3584038547309253e-05,
+      "loss": 1.2859,
+      "num_input_tokens_seen": 536409736,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6619139051587855,
+      "grad_norm": 0.43948325514793396,
+      "learning_rate": 1.3549145726051514e-05,
+      "loss": 1.2886,
+      "num_input_tokens_seen": 536786188,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6623984687789602,
+      "grad_norm": 0.39728376269340515,
+      "learning_rate": 1.3514281110668036e-05,
+      "loss": 1.2886,
+      "num_input_tokens_seen": 537166236,
+      "step": 13670
+    },
+    {
+      "epoch": 0.662883032399135,
+      "grad_norm": 0.39317986369132996,
+      "learning_rate": 1.3479444787037756e-05,
+      "loss": 1.3038,
+      "num_input_tokens_seen": 537555492,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6633675960193098,
+      "grad_norm": 0.40623709559440613,
+      "learning_rate": 1.3444636840969882e-05,
+      "loss": 1.292,
+      "num_input_tokens_seen": 537970244,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6638521596394846,
+      "grad_norm": 0.4472850263118744,
+      "learning_rate": 1.340985735820376e-05,
+      "loss": 1.2384,
+      "num_input_tokens_seen": 538370372,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6643367232596594,
+      "grad_norm": 0.3942728638648987,
+      "learning_rate": 1.3375106424408584e-05,
+      "loss": 1.2675,
+      "num_input_tokens_seen": 538759172,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6648212868798343,
+      "grad_norm": 0.4153252840042114,
+      "learning_rate": 1.3340384125183263e-05,
+      "loss": 1.3135,
+      "num_input_tokens_seen": 539150680,
+      "step": 13720
+    },
+    {
+      "epoch": 0.6653058505000091,
+      "grad_norm": 0.41260483860969543,
+      "learning_rate": 1.330569054605616e-05,
+      "loss": 1.2691,
+      "num_input_tokens_seen": 539546148,
+      "step": 13730
+    },
+    {
+      "epoch": 0.6657904141201839,
+      "grad_norm": 0.4087775647640228,
+      "learning_rate": 1.3271025772484897e-05,
+      "loss": 1.2561,
+      "num_input_tokens_seen": 539930372,
+      "step": 13740
+    },
+    {
+      "epoch": 0.6662749777403587,
+      "grad_norm": 0.40349823236465454,
+      "learning_rate": 1.3236389889856123e-05,
+      "loss": 1.2255,
+      "num_input_tokens_seen": 540338764,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6667595413605335,
+      "grad_norm": 0.3926162123680115,
+      "learning_rate": 1.3201782983485356e-05,
+      "loss": 1.2884,
+      "num_input_tokens_seen": 540728228,
+      "step": 13760
+    },
+    {
+      "epoch": 0.6672441049807083,
+      "grad_norm": 0.4218115508556366,
+      "learning_rate": 1.3167205138616703e-05,
+      "loss": 1.2595,
+      "num_input_tokens_seen": 541134348,
+      "step": 13770
+    },
+    {
+      "epoch": 0.6677286686008831,
+      "grad_norm": 0.4189454913139343,
+      "learning_rate": 1.3132656440422711e-05,
+      "loss": 1.3104,
+      "num_input_tokens_seen": 541508532,
+      "step": 13780
+    },
+    {
+      "epoch": 0.6682132322210579,
+      "grad_norm": 0.4163365364074707,
+      "learning_rate": 1.3098136974004136e-05,
+      "loss": 1.2894,
+      "num_input_tokens_seen": 541904872,
+      "step": 13790
+    },
+    {
+      "epoch": 0.6686977958412327,
+      "grad_norm": 0.39865246415138245,
+      "learning_rate": 1.30636468243897e-05,
+      "loss": 1.2838,
+      "num_input_tokens_seen": 542327136,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6691823594614076,
+      "grad_norm": 0.4374254047870636,
+      "learning_rate": 1.3029186076535948e-05,
+      "loss": 1.2899,
+      "num_input_tokens_seen": 542708876,
+      "step": 13810
+    },
+    {
+      "epoch": 0.6696669230815824,
+      "grad_norm": 0.39832600951194763,
+      "learning_rate": 1.2994754815326976e-05,
+      "loss": 1.3043,
+      "num_input_tokens_seen": 543083288,
+      "step": 13820
+    },
+    {
+      "epoch": 0.6701514867017572,
+      "grad_norm": 0.38928577303886414,
+      "learning_rate": 1.2960353125574264e-05,
+      "loss": 1.2409,
+      "num_input_tokens_seen": 543498476,
+      "step": 13830
+    },
+    {
+      "epoch": 0.670636050321932,
+      "grad_norm": 0.3987908363342285,
+      "learning_rate": 1.2925981092016434e-05,
+      "loss": 1.3112,
+      "num_input_tokens_seen": 543904852,
+      "step": 13840
+    },
+    {
+      "epoch": 0.6711206139421068,
+      "grad_norm": 0.43027713894844055,
+      "learning_rate": 1.2891638799319078e-05,
+      "loss": 1.2587,
+      "num_input_tokens_seen": 544297824,
+      "step": 13850
+    },
+    {
+      "epoch": 0.6716051775622816,
+      "grad_norm": 0.4249833822250366,
+      "learning_rate": 1.2857326332074516e-05,
+      "loss": 1.3004,
+      "num_input_tokens_seen": 544682872,
+      "step": 13860
+    },
+    {
+      "epoch": 0.6720897411824563,
+      "grad_norm": 0.42288899421691895,
+      "learning_rate": 1.2823043774801625e-05,
+      "loss": 1.3173,
+      "num_input_tokens_seen": 545074136,
+      "step": 13870
+    },
+    {
+      "epoch": 0.6725743048026311,
+      "grad_norm": 0.4400785267353058,
+      "learning_rate": 1.278879121194556e-05,
+      "loss": 1.225,
+      "num_input_tokens_seen": 545481540,
+      "step": 13880
+    },
+    {
+      "epoch": 0.6730588684228059,
+      "grad_norm": 0.42327094078063965,
+      "learning_rate": 1.275456872787765e-05,
+      "loss": 1.2532,
+      "num_input_tokens_seen": 545873512,
+      "step": 13890
+    },
+    {
+      "epoch": 0.6735434320429808,
+      "grad_norm": 0.39755240082740784,
+      "learning_rate": 1.2720376406895086e-05,
+      "loss": 1.2726,
+      "num_input_tokens_seen": 546300044,
+      "step": 13900
+    },
+    {
+      "epoch": 0.6740279956631556,
+      "grad_norm": 0.39591267704963684,
+      "learning_rate": 1.2686214333220787e-05,
+      "loss": 1.2871,
+      "num_input_tokens_seen": 546682376,
+      "step": 13910
+    },
+    {
+      "epoch": 0.6745125592833304,
+      "grad_norm": 0.3895485997200012,
+      "learning_rate": 1.2652082591003173e-05,
+      "loss": 1.2779,
+      "num_input_tokens_seen": 547085092,
+      "step": 13920
+    },
+    {
+      "epoch": 0.6749971229035052,
+      "grad_norm": 0.43440917134284973,
+      "learning_rate": 1.261798126431592e-05,
+      "loss": 1.2538,
+      "num_input_tokens_seen": 547501820,
+      "step": 13930
+    },
+    {
+      "epoch": 0.67548168652368,
+      "grad_norm": 0.39425063133239746,
+      "learning_rate": 1.2583910437157825e-05,
+      "loss": 1.3221,
+      "num_input_tokens_seen": 547901628,
+      "step": 13940
+    },
+    {
+      "epoch": 0.6759662501438548,
+      "grad_norm": 0.3926689922809601,
+      "learning_rate": 1.2549870193452513e-05,
+      "loss": 1.3095,
+      "num_input_tokens_seen": 548291540,
+      "step": 13950
+    },
+    {
+      "epoch": 0.6764508137640296,
+      "grad_norm": 0.4056353271007538,
+      "learning_rate": 1.2515860617048314e-05,
+      "loss": 1.2636,
+      "num_input_tokens_seen": 548701312,
+      "step": 13960
+    },
+    {
+      "epoch": 0.6769353773842044,
+      "grad_norm": 0.43241339921951294,
+      "learning_rate": 1.2481881791717996e-05,
+      "loss": 1.2654,
+      "num_input_tokens_seen": 549068240,
+      "step": 13970
+    },
+    {
+      "epoch": 0.6774199410043792,
+      "grad_norm": 0.3984728753566742,
+      "learning_rate": 1.2447933801158593e-05,
+      "loss": 1.213,
+      "num_input_tokens_seen": 549458460,
+      "step": 13980
+    },
+    {
+      "epoch": 0.6779045046245541,
+      "grad_norm": 0.39436209201812744,
+      "learning_rate": 1.2414016728991171e-05,
+      "loss": 1.2805,
+      "num_input_tokens_seen": 549845624,
+      "step": 13990
+    },
+    {
+      "epoch": 0.6783890682447289,
+      "grad_norm": 0.4029475450515747,
+      "learning_rate": 1.2380130658760653e-05,
+      "loss": 1.2622,
+      "num_input_tokens_seen": 550234352,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6783890682447289,
+      "eval_loss": 1.3831204175949097,
+      "eval_runtime": 3.6695,
+      "eval_samples_per_second": 40.878,
+      "eval_steps_per_second": 5.178,
+      "num_input_tokens_seen": 550234352,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6788736318649037,
+      "grad_norm": 0.3884779214859009,
+      "learning_rate": 1.2346275673935592e-05,
+      "loss": 1.2657,
+      "num_input_tokens_seen": 550605424,
+      "step": 14010
+    },
+    {
+      "epoch": 0.6793581954850785,
+      "grad_norm": 0.3887692391872406,
+      "learning_rate": 1.2312451857907983e-05,
+      "loss": 1.2375,
+      "num_input_tokens_seen": 551020804,
+      "step": 14020
+    },
+    {
+      "epoch": 0.6798427591052533,
+      "grad_norm": 0.44228610396385193,
+      "learning_rate": 1.2278659293993011e-05,
+      "loss": 1.2443,
+      "num_input_tokens_seen": 551415568,
+      "step": 14030
+    },
+    {
+      "epoch": 0.6803273227254281,
+      "grad_norm": 0.4105127155780792,
+      "learning_rate": 1.2244898065428918e-05,
+      "loss": 1.256,
+      "num_input_tokens_seen": 551790224,
+      "step": 14040
+    },
+    {
+      "epoch": 0.6808118863456029,
+      "grad_norm": 0.367942750453949,
+      "learning_rate": 1.2211168255376747e-05,
+      "loss": 1.313,
+      "num_input_tokens_seen": 552201712,
+      "step": 14050
+    },
+    {
+      "epoch": 0.6812964499657777,
+      "grad_norm": 0.43449220061302185,
+      "learning_rate": 1.217746994692014e-05,
+      "loss": 1.2926,
+      "num_input_tokens_seen": 552593300,
+      "step": 14060
+    },
+    {
+      "epoch": 0.6817810135859526,
+      "grad_norm": 0.3836462199687958,
+      "learning_rate": 1.2143803223065161e-05,
+      "loss": 1.3065,
+      "num_input_tokens_seen": 552995148,
+      "step": 14070
+    },
+    {
+      "epoch": 0.6822655772061273,
+      "grad_norm": 0.4054076373577118,
+      "learning_rate": 1.2110168166740057e-05,
+      "loss": 1.3149,
+      "num_input_tokens_seen": 553394712,
+      "step": 14080
+    },
+    {
+      "epoch": 0.6827501408263021,
+      "grad_norm": 0.39375001192092896,
+      "learning_rate": 1.2076564860795095e-05,
+      "loss": 1.2744,
+      "num_input_tokens_seen": 553807736,
+      "step": 14090
+    },
+    {
+      "epoch": 0.6832347044464769,
+      "grad_norm": 0.4356180727481842,
+      "learning_rate": 1.2042993388002302e-05,
+      "loss": 1.2847,
+      "num_input_tokens_seen": 554190868,
+      "step": 14100
+    },
+    {
+      "epoch": 0.6837192680666517,
+      "grad_norm": 0.4105941951274872,
+      "learning_rate": 1.2009453831055331e-05,
+      "loss": 1.2395,
+      "num_input_tokens_seen": 554599144,
+      "step": 14110
+    },
+    {
+      "epoch": 0.6842038316868265,
+      "grad_norm": 0.43507689237594604,
+      "learning_rate": 1.1975946272569177e-05,
+      "loss": 1.2725,
+      "num_input_tokens_seen": 554992164,
+      "step": 14120
+    },
+    {
+      "epoch": 0.6846883953070013,
+      "grad_norm": 0.4143732786178589,
+      "learning_rate": 1.194247079508006e-05,
+      "loss": 1.2748,
+      "num_input_tokens_seen": 555351352,
+      "step": 14130
+    },
+    {
+      "epoch": 0.6851729589271761,
+      "grad_norm": 0.4113134741783142,
+      "learning_rate": 1.1909027481045138e-05,
+      "loss": 1.2729,
+      "num_input_tokens_seen": 555746752,
+      "step": 14140
+    },
+    {
+      "epoch": 0.6856575225473509,
+      "grad_norm": 0.400814414024353,
+      "learning_rate": 1.1875616412842368e-05,
+      "loss": 1.2404,
+      "num_input_tokens_seen": 556135216,
+      "step": 14150
+    },
+    {
+      "epoch": 0.6861420861675258,
+      "grad_norm": 0.386210560798645,
+      "learning_rate": 1.1842237672770277e-05,
+      "loss": 1.2877,
+      "num_input_tokens_seen": 556532556,
+      "step": 14160
+    },
+    {
+      "epoch": 0.6866266497877006,
+      "grad_norm": 0.41966861486434937,
+      "learning_rate": 1.1808891343047754e-05,
+      "loss": 1.2242,
+      "num_input_tokens_seen": 556962628,
+      "step": 14170
+    },
+    {
+      "epoch": 0.6871112134078754,
+      "grad_norm": 0.4356541037559509,
+      "learning_rate": 1.1775577505813868e-05,
+      "loss": 1.2896,
+      "num_input_tokens_seen": 557340936,
+      "step": 14180
+    },
+    {
+      "epoch": 0.6875957770280502,
+      "grad_norm": 0.38502469658851624,
+      "learning_rate": 1.1742296243127621e-05,
+      "loss": 1.3201,
+      "num_input_tokens_seen": 557716816,
+      "step": 14190
+    },
+    {
+      "epoch": 0.688080340648225,
+      "grad_norm": 0.4457418620586395,
+      "learning_rate": 1.1709047636967812e-05,
+      "loss": 1.2624,
+      "num_input_tokens_seen": 558088256,
+      "step": 14200
+    },
+    {
+      "epoch": 0.6885649042683998,
+      "grad_norm": 0.43167582154273987,
+      "learning_rate": 1.1675831769232775e-05,
+      "loss": 1.274,
+      "num_input_tokens_seen": 558496660,
+      "step": 14210
+    },
+    {
+      "epoch": 0.6890494678885746,
+      "grad_norm": 0.42342841625213623,
+      "learning_rate": 1.1642648721740226e-05,
+      "loss": 1.2851,
+      "num_input_tokens_seen": 558846140,
+      "step": 14220
+    },
+    {
+      "epoch": 0.6895340315087494,
+      "grad_norm": 0.4272722005844116,
+      "learning_rate": 1.1609498576227008e-05,
+      "loss": 1.2644,
+      "num_input_tokens_seen": 559205772,
+      "step": 14230
+    },
+    {
+      "epoch": 0.6900185951289242,
+      "grad_norm": 0.4222122132778168,
+      "learning_rate": 1.1576381414348953e-05,
+      "loss": 1.287,
+      "num_input_tokens_seen": 559613824,
+      "step": 14240
+    },
+    {
+      "epoch": 0.6905031587490991,
+      "grad_norm": 0.5002584457397461,
+      "learning_rate": 1.1543297317680607e-05,
+      "loss": 1.3271,
+      "num_input_tokens_seen": 560013268,
+      "step": 14250
+    },
+    {
+      "epoch": 0.6909877223692739,
+      "grad_norm": 0.4087948501110077,
+      "learning_rate": 1.1510246367715122e-05,
+      "loss": 1.2673,
+      "num_input_tokens_seen": 560392376,
+      "step": 14260
+    },
+    {
+      "epoch": 0.6914722859894487,
+      "grad_norm": 0.41717860102653503,
+      "learning_rate": 1.1477228645863944e-05,
+      "loss": 1.2923,
+      "num_input_tokens_seen": 560761820,
+      "step": 14270
+    },
+    {
+      "epoch": 0.6919568496096234,
+      "grad_norm": 0.3851679563522339,
+      "learning_rate": 1.1444244233456717e-05,
+      "loss": 1.2749,
+      "num_input_tokens_seen": 561148148,
+      "step": 14280
+    },
+    {
+      "epoch": 0.6924414132297982,
+      "grad_norm": 0.41185683012008667,
+      "learning_rate": 1.1411293211741014e-05,
+      "loss": 1.234,
+      "num_input_tokens_seen": 561552512,
+      "step": 14290
+    },
+    {
+      "epoch": 0.692925976849973,
+      "grad_norm": 0.3968884348869324,
+      "learning_rate": 1.1378375661882181e-05,
+      "loss": 1.2464,
+      "num_input_tokens_seen": 561934928,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6934105404701478,
+      "grad_norm": 0.38210412859916687,
+      "learning_rate": 1.1345491664963078e-05,
+      "loss": 1.247,
+      "num_input_tokens_seen": 562352708,
+      "step": 14310
+    },
+    {
+      "epoch": 0.6938951040903226,
+      "grad_norm": 0.3991890847682953,
+      "learning_rate": 1.1312641301983954e-05,
+      "loss": 1.2252,
+      "num_input_tokens_seen": 562733284,
+      "step": 14320
+    },
+    {
+      "epoch": 0.6943796677104974,
+      "grad_norm": 0.3969036340713501,
+      "learning_rate": 1.1279824653862197e-05,
+      "loss": 1.2681,
+      "num_input_tokens_seen": 563140896,
+      "step": 14330
+    },
+    {
+      "epoch": 0.6948642313306723,
+      "grad_norm": 0.42541587352752686,
+      "learning_rate": 1.1247041801432137e-05,
+      "loss": 1.2644,
+      "num_input_tokens_seen": 563495812,
+      "step": 14340
+    },
+    {
+      "epoch": 0.6953487949508471,
+      "grad_norm": 0.3810971677303314,
+      "learning_rate": 1.1214292825444883e-05,
+      "loss": 1.2588,
+      "num_input_tokens_seen": 563921972,
+      "step": 14350
+    },
+    {
+      "epoch": 0.6958333585710219,
+      "grad_norm": 0.4099074602127075,
+      "learning_rate": 1.1181577806568064e-05,
+      "loss": 1.2685,
+      "num_input_tokens_seen": 564303812,
+      "step": 14360
+    },
+    {
+      "epoch": 0.6963179221911967,
+      "grad_norm": 0.42169204354286194,
+      "learning_rate": 1.1148896825385707e-05,
+      "loss": 1.2461,
+      "num_input_tokens_seen": 564712192,
+      "step": 14370
+    },
+    {
+      "epoch": 0.6968024858113715,
+      "grad_norm": 0.42555099725723267,
+      "learning_rate": 1.111624996239796e-05,
+      "loss": 1.274,
+      "num_input_tokens_seen": 565108896,
+      "step": 14380
+    },
+    {
+      "epoch": 0.6972870494315463,
+      "grad_norm": 0.4312445819377899,
+      "learning_rate": 1.108363729802096e-05,
+      "loss": 1.272,
+      "num_input_tokens_seen": 565492580,
+      "step": 14390
+    },
+    {
+      "epoch": 0.6977716130517211,
+      "grad_norm": 0.40941333770751953,
+      "learning_rate": 1.1051058912586579e-05,
+      "loss": 1.2707,
+      "num_input_tokens_seen": 565921508,
+      "step": 14400
+    },
+    {
+      "epoch": 0.6982561766718959,
+      "grad_norm": 0.4134999215602875,
+      "learning_rate": 1.1018514886342279e-05,
+      "loss": 1.2187,
+      "num_input_tokens_seen": 566308100,
+      "step": 14410
+    },
+    {
+      "epoch": 0.6987407402920707,
+      "grad_norm": 0.3979948163032532,
+      "learning_rate": 1.0986005299450858e-05,
+      "loss": 1.2768,
+      "num_input_tokens_seen": 566705680,
+      "step": 14420
+    },
+    {
+      "epoch": 0.6992253039122456,
+      "grad_norm": 0.41298606991767883,
+      "learning_rate": 1.0953530231990311e-05,
+      "loss": 1.2341,
+      "num_input_tokens_seen": 567123900,
+      "step": 14430
+    },
+    {
+      "epoch": 0.6997098675324204,
+      "grad_norm": 0.42366859316825867,
+      "learning_rate": 1.0921089763953594e-05,
+      "loss": 1.2959,
+      "num_input_tokens_seen": 567534336,
+      "step": 14440
+    },
+    {
+      "epoch": 0.7001944311525952,
+      "grad_norm": 0.39607560634613037,
+      "learning_rate": 1.0888683975248431e-05,
+      "loss": 1.2823,
+      "num_input_tokens_seen": 567938980,
+      "step": 14450
+    },
+    {
+      "epoch": 0.70067899477277,
+      "grad_norm": 0.38363564014434814,
+      "learning_rate": 1.0856312945697142e-05,
+      "loss": 1.2062,
+      "num_input_tokens_seen": 568304904,
+      "step": 14460
+    },
+    {
+      "epoch": 0.7011635583929448,
+      "grad_norm": 0.4327394366264343,
+      "learning_rate": 1.0823976755036393e-05,
+      "loss": 1.2881,
+      "num_input_tokens_seen": 568703588,
+      "step": 14470
+    },
+    {
+      "epoch": 0.7016481220131195,
+      "grad_norm": 0.40510445833206177,
+      "learning_rate": 1.079167548291708e-05,
+      "loss": 1.2547,
+      "num_input_tokens_seen": 569104248,
+      "step": 14480
+    },
+    {
+      "epoch": 0.7021326856332943,
+      "grad_norm": 0.41055554151535034,
+      "learning_rate": 1.075940920890404e-05,
+      "loss": 1.2536,
+      "num_input_tokens_seen": 569485804,
+      "step": 14490
+    },
+    {
+      "epoch": 0.7026172492534691,
+      "grad_norm": 0.4175165295600891,
+      "learning_rate": 1.0727178012475944e-05,
+      "loss": 1.2334,
+      "num_input_tokens_seen": 569887012,
+      "step": 14500
+    },
+    {
+      "epoch": 0.7031018128736439,
+      "grad_norm": 0.43152111768722534,
+      "learning_rate": 1.0694981973025022e-05,
+      "loss": 1.2328,
+      "num_input_tokens_seen": 570274412,
+      "step": 14510
+    },
+    {
+      "epoch": 0.7035863764938188,
+      "grad_norm": 0.4261281490325928,
+      "learning_rate": 1.0662821169856948e-05,
+      "loss": 1.2766,
+      "num_input_tokens_seen": 570667476,
+      "step": 14520
+    },
+    {
+      "epoch": 0.7040709401139936,
+      "grad_norm": 0.4019312858581543,
+      "learning_rate": 1.0630695682190554e-05,
+      "loss": 1.2272,
+      "num_input_tokens_seen": 571069868,
+      "step": 14530
+    },
+    {
+      "epoch": 0.7045555037341684,
+      "grad_norm": 0.4646441638469696,
+      "learning_rate": 1.0598605589157726e-05,
+      "loss": 1.2188,
+      "num_input_tokens_seen": 571471600,
+      "step": 14540
+    },
+    {
+      "epoch": 0.7050400673543432,
+      "grad_norm": 0.40755367279052734,
+      "learning_rate": 1.0566550969803127e-05,
+      "loss": 1.286,
+      "num_input_tokens_seen": 571874376,
+      "step": 14550
+    },
+    {
+      "epoch": 0.705524630974518,
+      "grad_norm": 0.4092569947242737,
+      "learning_rate": 1.0534531903084065e-05,
+      "loss": 1.2766,
+      "num_input_tokens_seen": 572267796,
+      "step": 14560
+    },
+    {
+      "epoch": 0.7060091945946928,
+      "grad_norm": 0.443362295627594,
+      "learning_rate": 1.0502548467870284e-05,
+      "loss": 1.2842,
+      "num_input_tokens_seen": 572646180,
+      "step": 14570
+    },
+    {
+      "epoch": 0.7064937582148676,
+      "grad_norm": 0.39347755908966064,
+      "learning_rate": 1.0470600742943726e-05,
+      "loss": 1.2541,
+      "num_input_tokens_seen": 573058952,
+      "step": 14580
+    },
+    {
+      "epoch": 0.7069783218350424,
+      "grad_norm": 0.4157586395740509,
+      "learning_rate": 1.0438688806998395e-05,
+      "loss": 1.2842,
+      "num_input_tokens_seen": 573460344,
+      "step": 14590
+    },
+    {
+      "epoch": 0.7074628854552172,
+      "grad_norm": 0.45576024055480957,
+      "learning_rate": 1.0406812738640134e-05,
+      "loss": 1.2606,
+      "num_input_tokens_seen": 573848724,
+      "step": 14600
+    },
+    {
+      "epoch": 0.7079474490753921,
+      "grad_norm": 0.41932404041290283,
+      "learning_rate": 1.037497261638645e-05,
+      "loss": 1.2402,
+      "num_input_tokens_seen": 574233032,
+      "step": 14610
+    },
+    {
+      "epoch": 0.7084320126955669,
+      "grad_norm": 0.4327371120452881,
+      "learning_rate": 1.0343168518666272e-05,
+      "loss": 1.2565,
+      "num_input_tokens_seen": 574627952,
+      "step": 14620
+    },
+    {
+      "epoch": 0.7089165763157417,
+      "grad_norm": 0.4439123272895813,
+      "learning_rate": 1.0311400523819831e-05,
+      "loss": 1.2774,
+      "num_input_tokens_seen": 575040240,
+      "step": 14630
+    },
+    {
+      "epoch": 0.7094011399359165,
+      "grad_norm": 0.4237552583217621,
+      "learning_rate": 1.0279668710098401e-05,
+      "loss": 1.2613,
+      "num_input_tokens_seen": 575434676,
+      "step": 14640
+    },
+    {
+      "epoch": 0.7098857035560913,
+      "grad_norm": 0.4437905251979828,
+      "learning_rate": 1.0247973155664156e-05,
+      "loss": 1.2636,
+      "num_input_tokens_seen": 575812584,
+      "step": 14650
+    },
+    {
+      "epoch": 0.7103702671762661,
+      "grad_norm": 0.3982216715812683,
+      "learning_rate": 1.0216313938589936e-05,
+      "loss": 1.2758,
+      "num_input_tokens_seen": 576228776,
+      "step": 14660
+    },
+    {
+      "epoch": 0.7108548307964409,
+      "grad_norm": 0.36928027868270874,
+      "learning_rate": 1.0184691136859096e-05,
+      "loss": 1.2293,
+      "num_input_tokens_seen": 576617456,
+      "step": 14670
+    },
+    {
+      "epoch": 0.7113393944166156,
+      "grad_norm": 0.3991413712501526,
+      "learning_rate": 1.0153104828365261e-05,
+      "loss": 1.2731,
+      "num_input_tokens_seen": 576986340,
+      "step": 14680
+    },
+    {
+      "epoch": 0.7118239580367904,
+      "grad_norm": 0.4070897698402405,
+      "learning_rate": 1.0121555090912207e-05,
+      "loss": 1.2772,
+      "num_input_tokens_seen": 577384544,
+      "step": 14690
+    },
+    {
+      "epoch": 0.7123085216569653,
+      "grad_norm": 0.3921777904033661,
+      "learning_rate": 1.0090042002213587e-05,
+      "loss": 1.2332,
+      "num_input_tokens_seen": 577772956,
+      "step": 14700
+    },
+    {
+      "epoch": 0.7127930852771401,
+      "grad_norm": 0.37934672832489014,
+      "learning_rate": 1.0058565639892808e-05,
+      "loss": 1.2527,
+      "num_input_tokens_seen": 578188080,
+      "step": 14710
+    },
+    {
+      "epoch": 0.7132776488973149,
+      "grad_norm": 0.3901313841342926,
+      "learning_rate": 1.0027126081482801e-05,
+      "loss": 1.2523,
+      "num_input_tokens_seen": 578567732,
+      "step": 14720
+    },
+    {
+      "epoch": 0.7137622125174897,
+      "grad_norm": 0.4196103513240814,
+      "learning_rate": 9.995723404425845e-06,
+      "loss": 1.2586,
+      "num_input_tokens_seen": 578986928,
+      "step": 14730
+    },
+    {
+      "epoch": 0.7142467761376645,
+      "grad_norm": 0.40894681215286255,
+      "learning_rate": 9.964357686073378e-06,
+      "loss": 1.2532,
+      "num_input_tokens_seen": 579387500,
+      "step": 14740
+    },
+    {
+      "epoch": 0.7147313397578393,
+      "grad_norm": 0.38982832431793213,
+      "learning_rate": 9.933029003685778e-06,
+      "loss": 1.2751,
+      "num_input_tokens_seen": 579749804,
+      "step": 14750
+    },
+    {
+      "epoch": 0.7152159033780141,
+      "grad_norm": 0.3959210515022278,
+      "learning_rate": 9.901737434432226e-06,
+      "loss": 1.2404,
+      "num_input_tokens_seen": 580156476,
+      "step": 14760
+    },
+    {
+      "epoch": 0.7157004669981889,
+      "grad_norm": 0.4251919984817505,
+      "learning_rate": 9.870483055390456e-06,
+      "loss": 1.2346,
+      "num_input_tokens_seen": 580522596,
+      "step": 14770
+    },
+    {
+      "epoch": 0.7161850306183638,
+      "grad_norm": 0.48525333404541016,
+      "learning_rate": 9.839265943546627e-06,
+      "loss": 1.2809,
+      "num_input_tokens_seen": 580900272,
+      "step": 14780
+    },
+    {
+      "epoch": 0.7166695942385386,
+      "grad_norm": 0.3704177439212799,
+      "learning_rate": 9.808086175795061e-06,
+      "loss": 1.2318,
+      "num_input_tokens_seen": 581280140,
+      "step": 14790
+    },
+    {
+      "epoch": 0.7171541578587134,
+      "grad_norm": 0.4149393141269684,
+      "learning_rate": 9.77694382893814e-06,
+      "loss": 1.3142,
+      "num_input_tokens_seen": 581666396,
+      "step": 14800
+    },
+    {
+      "epoch": 0.7176387214788882,
+      "grad_norm": 0.4752131402492523,
+      "learning_rate": 9.745838979686026e-06,
+      "loss": 1.2692,
+      "num_input_tokens_seen": 582040104,
+      "step": 14810
+    },
+    {
+      "epoch": 0.718123285099063,
+      "grad_norm": 0.41852232813835144,
+      "learning_rate": 9.714771704656553e-06,
+      "loss": 1.2908,
+      "num_input_tokens_seen": 582431284,
+      "step": 14820
+    },
+    {
+      "epoch": 0.7186078487192378,
+      "grad_norm": 0.3744739592075348,
+      "learning_rate": 9.683742080374968e-06,
+      "loss": 1.2438,
+      "num_input_tokens_seen": 582835184,
+      "step": 14830
+    },
+    {
+      "epoch": 0.7190924123394126,
+      "grad_norm": 0.3988076448440552,
+      "learning_rate": 9.652750183273806e-06,
+      "loss": 1.2978,
+      "num_input_tokens_seen": 583230036,
+      "step": 14840
+    },
+    {
+      "epoch": 0.7195769759595874,
+      "grad_norm": 0.41156208515167236,
+      "learning_rate": 9.621796089692667e-06,
+      "loss": 1.2709,
+      "num_input_tokens_seen": 583626572,
+      "step": 14850
+    },
+    {
+      "epoch": 0.7200615395797622,
+      "grad_norm": 0.437142014503479,
+      "learning_rate": 9.59087987587801e-06,
+      "loss": 1.2503,
+      "num_input_tokens_seen": 584037012,
+      "step": 14860
+    },
+    {
+      "epoch": 0.7205461031999371,
+      "grad_norm": 0.42297014594078064,
+      "learning_rate": 9.560001617983005e-06,
+      "loss": 1.2808,
+      "num_input_tokens_seen": 584438176,
+      "step": 14870
+    },
+    {
+      "epoch": 0.7210306668201119,
+      "grad_norm": 0.4088895618915558,
+      "learning_rate": 9.529161392067336e-06,
+      "loss": 1.2203,
+      "num_input_tokens_seen": 584823036,
+      "step": 14880
+    },
+    {
+      "epoch": 0.7215152304402866,
+      "grad_norm": 0.4062732458114624,
+      "learning_rate": 9.498359274097002e-06,
+      "loss": 1.2212,
+      "num_input_tokens_seen": 585208164,
+      "step": 14890
+    },
+    {
+      "epoch": 0.7219997940604614,
+      "grad_norm": 0.4162246882915497,
+      "learning_rate": 9.467595339944116e-06,
+      "loss": 1.3,
+      "num_input_tokens_seen": 585566620,
+      "step": 14900
+    },
+    {
+      "epoch": 0.7224843576806362,
+      "grad_norm": 0.38531064987182617,
+      "learning_rate": 9.436869665386763e-06,
+      "loss": 1.2206,
+      "num_input_tokens_seen": 585980372,
+      "step": 14910
+    },
+    {
+      "epoch": 0.722968921300811,
+      "grad_norm": 0.42929065227508545,
+      "learning_rate": 9.40618232610876e-06,
+      "loss": 1.2744,
+      "num_input_tokens_seen": 586389084,
+      "step": 14920
+    },
+    {
+      "epoch": 0.7234534849209858,
+      "grad_norm": 0.4108678102493286,
+      "learning_rate": 9.375533397699523e-06,
+      "loss": 1.252,
+      "num_input_tokens_seen": 586790192,
+      "step": 14930
+    },
+    {
+      "epoch": 0.7239380485411606,
+      "grad_norm": 0.43513932824134827,
+      "learning_rate": 9.344922955653826e-06,
+      "loss": 1.2669,
+      "num_input_tokens_seen": 587176208,
+      "step": 14940
+    },
+    {
+      "epoch": 0.7244226121613354,
+      "grad_norm": 0.4044954776763916,
+      "learning_rate": 9.314351075371674e-06,
+      "loss": 1.2949,
+      "num_input_tokens_seen": 587570128,
+      "step": 14950
+    },
+    {
+      "epoch": 0.7249071757815103,
+      "grad_norm": 0.4217306077480316,
+      "learning_rate": 9.283817832158053e-06,
+      "loss": 1.3133,
+      "num_input_tokens_seen": 587951740,
+      "step": 14960
+    },
+    {
+      "epoch": 0.7253917394016851,
+      "grad_norm": 0.4031387269496918,
+      "learning_rate": 9.253323301222802e-06,
+      "loss": 1.2583,
+      "num_input_tokens_seen": 588343360,
+      "step": 14970
+    },
+    {
+      "epoch": 0.7258763030218599,
+      "grad_norm": 0.4108414649963379,
+      "learning_rate": 9.222867557680403e-06,
+      "loss": 1.2722,
+      "num_input_tokens_seen": 588756676,
+      "step": 14980
+    },
+    {
+      "epoch": 0.7263608666420347,
+      "grad_norm": 0.4133433699607849,
+      "learning_rate": 9.192450676549774e-06,
+      "loss": 1.2386,
+      "num_input_tokens_seen": 589152220,
+      "step": 14990
+    },
+    {
+      "epoch": 0.7268454302622095,
+      "grad_norm": 0.3888295590877533,
+      "learning_rate": 9.162072732754132e-06,
+      "loss": 1.2964,
+      "num_input_tokens_seen": 589548180,
+      "step": 15000
+    },
+    {
+      "epoch": 0.7273299938823843,
+      "grad_norm": 0.40823793411254883,
+      "learning_rate": 9.131733801120771e-06,
+      "loss": 1.249,
+      "num_input_tokens_seen": 589932416,
+      "step": 15010
+    },
+    {
+      "epoch": 0.7278145575025591,
+      "grad_norm": 0.41177472472190857,
+      "learning_rate": 9.1014339563809e-06,
+      "loss": 1.2636,
+      "num_input_tokens_seen": 590315064,
+      "step": 15020
+    },
+    {
+      "epoch": 0.7282991211227339,
+      "grad_norm": 0.4307875633239746,
+      "learning_rate": 9.071173273169428e-06,
+      "loss": 1.2832,
+      "num_input_tokens_seen": 590718608,
+      "step": 15030
+    },
+    {
+      "epoch": 0.7287836847429087,
+      "grad_norm": 0.3676307499408722,
+      "learning_rate": 9.040951826024824e-06,
+      "loss": 1.2616,
+      "num_input_tokens_seen": 591105428,
+      "step": 15040
+    },
+    {
+      "epoch": 0.7292682483630836,
+      "grad_norm": 0.4145369529724121,
+      "learning_rate": 9.010769689388885e-06,
+      "loss": 1.2932,
+      "num_input_tokens_seen": 591484840,
+      "step": 15050
+    },
+    {
+      "epoch": 0.7297528119832584,
+      "grad_norm": 0.38058140873908997,
+      "learning_rate": 8.980626937606612e-06,
+      "loss": 1.2451,
+      "num_input_tokens_seen": 591878184,
+      "step": 15060
+    },
+    {
+      "epoch": 0.7302373756034332,
+      "grad_norm": 0.41027316451072693,
+      "learning_rate": 8.950523644925954e-06,
+      "loss": 1.2548,
+      "num_input_tokens_seen": 592242680,
+      "step": 15070
+    },
+    {
+      "epoch": 0.730721939223608,
+      "grad_norm": 0.4113955497741699,
+      "learning_rate": 8.920459885497703e-06,
+      "loss": 1.2306,
+      "num_input_tokens_seen": 592616904,
+      "step": 15080
+    },
+    {
+      "epoch": 0.7312065028437827,
+      "grad_norm": 0.4587390422821045,
+      "learning_rate": 8.890435733375232e-06,
+      "loss": 1.2836,
+      "num_input_tokens_seen": 592980040,
+      "step": 15090
+    },
+    {
+      "epoch": 0.7316910664639575,
+      "grad_norm": 0.4060867726802826,
+      "learning_rate": 8.860451262514386e-06,
+      "loss": 1.2048,
+      "num_input_tokens_seen": 593376740,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7321756300841323,
+      "grad_norm": 0.44127029180526733,
+      "learning_rate": 8.830506546773257e-06,
+      "loss": 1.2557,
+      "num_input_tokens_seen": 593757812,
+      "step": 15110
+    },
+    {
+      "epoch": 0.7326601937043071,
+      "grad_norm": 0.423213392496109,
+      "learning_rate": 8.800601659911998e-06,
+      "loss": 1.2426,
+      "num_input_tokens_seen": 594122040,
+      "step": 15120
+    },
+    {
+      "epoch": 0.7331447573244819,
+      "grad_norm": 0.41493481397628784,
+      "learning_rate": 8.770736675592678e-06,
+      "loss": 1.2721,
+      "num_input_tokens_seen": 594520000,
+      "step": 15130
+    },
+    {
+      "epoch": 0.7336293209446568,
+      "grad_norm": 0.46575504541397095,
+      "learning_rate": 8.740911667379053e-06,
+      "loss": 1.2268,
+      "num_input_tokens_seen": 594873824,
+      "step": 15140
+    },
+    {
+      "epoch": 0.7341138845648316,
+      "grad_norm": 0.4391193091869354,
+      "learning_rate": 8.711126708736426e-06,
+      "loss": 1.313,
+      "num_input_tokens_seen": 595233156,
+      "step": 15150
+    },
+    {
+      "epoch": 0.7345984481850064,
+      "grad_norm": 0.41552045941352844,
+      "learning_rate": 8.681381873031447e-06,
+      "loss": 1.2626,
+      "num_input_tokens_seen": 595644240,
+      "step": 15160
+    },
+    {
+      "epoch": 0.7350830118051812,
+      "grad_norm": 0.3735989034175873,
+      "learning_rate": 8.651677233531943e-06,
+      "loss": 1.2961,
+      "num_input_tokens_seen": 596014316,
+      "step": 15170
+    },
+    {
+      "epoch": 0.735567575425356,
+      "grad_norm": 0.45374736189842224,
+      "learning_rate": 8.6220128634067e-06,
+      "loss": 1.2512,
+      "num_input_tokens_seen": 596420640,
+      "step": 15180
+    },
+    {
+      "epoch": 0.7360521390455308,
+      "grad_norm": 0.4039424657821655,
+      "learning_rate": 8.592388835725352e-06,
+      "loss": 1.2505,
+      "num_input_tokens_seen": 596817584,
+      "step": 15190
+    },
+    {
+      "epoch": 0.7365367026657056,
+      "grad_norm": 0.4076104164123535,
+      "learning_rate": 8.56280522345812e-06,
+      "loss": 1.2252,
+      "num_input_tokens_seen": 597205152,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7370212662858804,
+      "grad_norm": 0.4360331594944,
+      "learning_rate": 8.533262099475708e-06,
+      "loss": 1.2516,
+      "num_input_tokens_seen": 597588540,
+      "step": 15210
+    },
+    {
+      "epoch": 0.7375058299060552,
+      "grad_norm": 0.39837440848350525,
+      "learning_rate": 8.503759536549066e-06,
+      "loss": 1.2656,
+      "num_input_tokens_seen": 597989068,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7379903935262301,
+      "grad_norm": 0.45183074474334717,
+      "learning_rate": 8.474297607349252e-06,
+      "loss": 1.3073,
+      "num_input_tokens_seen": 598396680,
+      "step": 15230
+    },
+    {
+      "epoch": 0.7384749571464049,
+      "grad_norm": 0.46322163939476013,
+      "learning_rate": 8.44487638444721e-06,
+      "loss": 1.281,
+      "num_input_tokens_seen": 598789612,
+      "step": 15240
+    },
+    {
+      "epoch": 0.7389595207665797,
+      "grad_norm": 0.38120391964912415,
+      "learning_rate": 8.415495940313637e-06,
+      "loss": 1.2602,
+      "num_input_tokens_seen": 599183232,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7394440843867545,
+      "grad_norm": 0.3967757225036621,
+      "learning_rate": 8.386156347318785e-06,
+      "loss": 1.264,
+      "num_input_tokens_seen": 599587496,
+      "step": 15260
+    },
+    {
+      "epoch": 0.7399286480069293,
+      "grad_norm": 0.44284841418266296,
+      "learning_rate": 8.356857677732258e-06,
+      "loss": 1.2691,
+      "num_input_tokens_seen": 599970304,
+      "step": 15270
+    },
+    {
+      "epoch": 0.740413211627104,
+      "grad_norm": 0.42405474185943604,
+      "learning_rate": 8.327600003722887e-06,
+      "loss": 1.3021,
+      "num_input_tokens_seen": 600356376,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7408977752472788,
+      "grad_norm": 0.3977855145931244,
+      "learning_rate": 8.298383397358494e-06,
+      "loss": 1.2675,
+      "num_input_tokens_seen": 600728604,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7413823388674536,
+      "grad_norm": 0.41853830218315125,
+      "learning_rate": 8.269207930605757e-06,
+      "loss": 1.2954,
+      "num_input_tokens_seen": 601127264,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7418669024876284,
+      "grad_norm": 0.4441795349121094,
+      "learning_rate": 8.240073675330023e-06,
+      "loss": 1.2563,
+      "num_input_tokens_seen": 601500628,
+      "step": 15310
+    },
+    {
+      "epoch": 0.7423514661078033,
+      "grad_norm": 0.37602853775024414,
+      "learning_rate": 8.210980703295126e-06,
+      "loss": 1.2418,
+      "num_input_tokens_seen": 601875864,
+      "step": 15320
+    },
+    {
+      "epoch": 0.7428360297279781,
+      "grad_norm": 0.4239734411239624,
+      "learning_rate": 8.181929086163186e-06,
+      "loss": 1.2742,
+      "num_input_tokens_seen": 602271004,
+      "step": 15330
+    },
+    {
+      "epoch": 0.7433205933481529,
+      "grad_norm": 0.39539673924446106,
+      "learning_rate": 8.1529188954945e-06,
+      "loss": 1.2483,
+      "num_input_tokens_seen": 602647156,
+      "step": 15340
+    },
+    {
+      "epoch": 0.7438051569683277,
+      "grad_norm": 0.3857654631137848,
+      "learning_rate": 8.123950202747274e-06,
+      "loss": 1.2448,
+      "num_input_tokens_seen": 603043708,
+      "step": 15350
+    },
+    {
+      "epoch": 0.7442897205885025,
+      "grad_norm": 0.41120001673698425,
+      "learning_rate": 8.095023079277541e-06,
+      "loss": 1.2468,
+      "num_input_tokens_seen": 603427072,
+      "step": 15360
+    },
+    {
+      "epoch": 0.7447742842086773,
+      "grad_norm": 0.38883382081985474,
+      "learning_rate": 8.066137596338908e-06,
+      "loss": 1.2563,
+      "num_input_tokens_seen": 603820012,
+      "step": 15370
+    },
+    {
+      "epoch": 0.7452588478288521,
+      "grad_norm": 0.3978869915008545,
+      "learning_rate": 8.037293825082423e-06,
+      "loss": 1.2982,
+      "num_input_tokens_seen": 604199356,
+      "step": 15380
+    },
+    {
+      "epoch": 0.7457434114490269,
+      "grad_norm": 0.39663106203079224,
+      "learning_rate": 8.008491836556408e-06,
+      "loss": 1.2548,
+      "num_input_tokens_seen": 604594724,
+      "step": 15390
+    },
+    {
+      "epoch": 0.7462279750692017,
+      "grad_norm": 0.3799634873867035,
+      "learning_rate": 7.979731701706231e-06,
+      "loss": 1.2978,
+      "num_input_tokens_seen": 604995016,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7467125386893766,
+      "grad_norm": 0.424040824174881,
+      "learning_rate": 7.951013491374193e-06,
+      "loss": 1.2551,
+      "num_input_tokens_seen": 605385864,
+      "step": 15410
+    },
+    {
+      "epoch": 0.7471971023095514,
+      "grad_norm": 0.47143998742103577,
+      "learning_rate": 7.922337276299305e-06,
+      "loss": 1.3032,
+      "num_input_tokens_seen": 605778132,
+      "step": 15420
+    },
+    {
+      "epoch": 0.7476816659297262,
+      "grad_norm": 0.439042866230011,
+      "learning_rate": 7.89370312711715e-06,
+      "loss": 1.2905,
+      "num_input_tokens_seen": 606175808,
+      "step": 15430
+    },
+    {
+      "epoch": 0.748166229549901,
+      "grad_norm": 0.42106932401657104,
+      "learning_rate": 7.86511111435969e-06,
+      "loss": 1.251,
+      "num_input_tokens_seen": 606587468,
+      "step": 15440
+    },
+    {
+      "epoch": 0.7486507931700758,
+      "grad_norm": 0.434646874666214,
+      "learning_rate": 7.836561308455109e-06,
+      "loss": 1.2324,
+      "num_input_tokens_seen": 606989572,
+      "step": 15450
+    },
+    {
+      "epoch": 0.7491353567902506,
+      "grad_norm": 0.3873103857040405,
+      "learning_rate": 7.80805377972759e-06,
+      "loss": 1.249,
+      "num_input_tokens_seen": 607386504,
+      "step": 15460
+    },
+    {
+      "epoch": 0.7496199204104254,
+      "grad_norm": 0.42348745465278625,
+      "learning_rate": 7.779588598397222e-06,
+      "loss": 1.1877,
+      "num_input_tokens_seen": 607763740,
+      "step": 15470
+    },
+    {
+      "epoch": 0.7501044840306001,
+      "grad_norm": 0.46129289269447327,
+      "learning_rate": 7.751165834579744e-06,
+      "loss": 1.2666,
+      "num_input_tokens_seen": 608138336,
+      "step": 15480
+    },
+    {
+      "epoch": 0.750589047650775,
+      "grad_norm": 0.4692579507827759,
+      "learning_rate": 7.722785558286447e-06,
+      "loss": 1.2385,
+      "num_input_tokens_seen": 608572348,
+      "step": 15490
+    },
+    {
+      "epoch": 0.7510736112709498,
+      "grad_norm": 0.4252379536628723,
+      "learning_rate": 7.694447839423936e-06,
+      "loss": 1.238,
+      "num_input_tokens_seen": 608980572,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7515581748911246,
+      "grad_norm": 0.37597569823265076,
+      "learning_rate": 7.666152747794006e-06,
+      "loss": 1.2893,
+      "num_input_tokens_seen": 609382708,
+      "step": 15510
+    },
+    {
+      "epoch": 0.7520427385112994,
+      "grad_norm": 0.39012569189071655,
+      "learning_rate": 7.63790035309346e-06,
+      "loss": 1.3176,
+      "num_input_tokens_seen": 609759336,
+      "step": 15520
+    },
+    {
+      "epoch": 0.7525273021314742,
+      "grad_norm": 0.398588091135025,
+      "learning_rate": 7.609690724913901e-06,
+      "loss": 1.2815,
+      "num_input_tokens_seen": 610153132,
+      "step": 15530
+    },
+    {
+      "epoch": 0.753011865751649,
+      "grad_norm": 0.4179481565952301,
+      "learning_rate": 7.581523932741619e-06,
+      "loss": 1.2689,
+      "num_input_tokens_seen": 610529384,
+      "step": 15540
+    },
+    {
+      "epoch": 0.7534964293718238,
+      "grad_norm": 0.3886168599128723,
+      "learning_rate": 7.553400045957362e-06,
+      "loss": 1.2504,
+      "num_input_tokens_seen": 610922988,
+      "step": 15550
+    },
+    {
+      "epoch": 0.7539809929919986,
+      "grad_norm": 0.3878910541534424,
+      "learning_rate": 7.525319133836223e-06,
+      "loss": 1.2506,
+      "num_input_tokens_seen": 611272416,
+      "step": 15560
+    },
+    {
+      "epoch": 0.7544655566121734,
+      "grad_norm": 0.4470667541027069,
+      "learning_rate": 7.497281265547406e-06,
+      "loss": 1.241,
+      "num_input_tokens_seen": 611675884,
+      "step": 15570
+    },
+    {
+      "epoch": 0.7549501202323483,
+      "grad_norm": 0.440876841545105,
+      "learning_rate": 7.469286510154116e-06,
+      "loss": 1.2208,
+      "num_input_tokens_seen": 612084480,
+      "step": 15580
+    },
+    {
+      "epoch": 0.7554346838525231,
+      "grad_norm": 0.4724419414997101,
+      "learning_rate": 7.441334936613353e-06,
+      "loss": 1.2509,
+      "num_input_tokens_seen": 612451128,
+      "step": 15590
+    },
+    {
+      "epoch": 0.7559192474726979,
+      "grad_norm": 0.4163588881492615,
+      "learning_rate": 7.413426613775759e-06,
+      "loss": 1.2477,
+      "num_input_tokens_seen": 612831956,
+      "step": 15600
+    },
+    {
+      "epoch": 0.7564038110928727,
+      "grad_norm": 0.42442935705184937,
+      "learning_rate": 7.385561610385414e-06,
+      "loss": 1.2763,
+      "num_input_tokens_seen": 613196848,
+      "step": 15610
+    },
+    {
+      "epoch": 0.7568883747130475,
+      "grad_norm": 0.4455620050430298,
+      "learning_rate": 7.357739995079724e-06,
+      "loss": 1.2434,
+      "num_input_tokens_seen": 613631540,
+      "step": 15620
+    },
+    {
+      "epoch": 0.7573729383332223,
+      "grad_norm": 0.3932070732116699,
+      "learning_rate": 7.329961836389198e-06,
+      "loss": 1.2842,
+      "num_input_tokens_seen": 614045632,
+      "step": 15630
+    },
+    {
+      "epoch": 0.7578575019533971,
+      "grad_norm": 0.40935105085372925,
+      "learning_rate": 7.302227202737316e-06,
+      "loss": 1.2179,
+      "num_input_tokens_seen": 614417140,
+      "step": 15640
+    },
+    {
+      "epoch": 0.7583420655735719,
+      "grad_norm": 0.3958241939544678,
+      "learning_rate": 7.274536162440351e-06,
+      "loss": 1.2848,
+      "num_input_tokens_seen": 614833672,
+      "step": 15650
+    },
+    {
+      "epoch": 0.7588266291937467,
+      "grad_norm": 0.43931275606155396,
+      "learning_rate": 7.246888783707173e-06,
+      "loss": 1.2335,
+      "num_input_tokens_seen": 615222276,
+      "step": 15660
+    },
+    {
+      "epoch": 0.7593111928139216,
+      "grad_norm": 0.4984539747238159,
+      "learning_rate": 7.219285134639134e-06,
+      "loss": 1.2827,
+      "num_input_tokens_seen": 615624440,
+      "step": 15670
+    },
+    {
+      "epoch": 0.7597957564340964,
+      "grad_norm": 0.40793275833129883,
+      "learning_rate": 7.191725283229839e-06,
+      "loss": 1.2562,
+      "num_input_tokens_seen": 616014768,
+      "step": 15680
+    },
+    {
+      "epoch": 0.7602803200542712,
+      "grad_norm": 0.41619017720222473,
+      "learning_rate": 7.164209297365043e-06,
+      "loss": 1.2754,
+      "num_input_tokens_seen": 616391320,
+      "step": 15690
+    },
+    {
+      "epoch": 0.760764883674446,
+      "grad_norm": 0.43882516026496887,
+      "learning_rate": 7.136737244822422e-06,
+      "loss": 1.2868,
+      "num_input_tokens_seen": 616766344,
+      "step": 15700
+    },
+    {
+      "epoch": 0.7612494472946207,
+      "grad_norm": 0.3904348611831665,
+      "learning_rate": 7.109309193271454e-06,
+      "loss": 1.2305,
+      "num_input_tokens_seen": 617152204,
+      "step": 15710
+    },
+    {
+      "epoch": 0.7617340109147955,
+      "grad_norm": 0.41161251068115234,
+      "learning_rate": 7.081925210273227e-06,
+      "loss": 1.2739,
+      "num_input_tokens_seen": 617555064,
+      "step": 15720
+    },
+    {
+      "epoch": 0.7622185745349703,
+      "grad_norm": 0.44097068905830383,
+      "learning_rate": 7.054585363280286e-06,
+      "loss": 1.2619,
+      "num_input_tokens_seen": 617932060,
+      "step": 15730
+    },
+    {
+      "epoch": 0.7627031381551451,
+      "grad_norm": 0.46235036849975586,
+      "learning_rate": 7.027289719636437e-06,
+      "loss": 1.2797,
+      "num_input_tokens_seen": 618306600,
+      "step": 15740
+    },
+    {
+      "epoch": 0.7631877017753199,
+      "grad_norm": 0.421164870262146,
+      "learning_rate": 7.0000383465766345e-06,
+      "loss": 1.2303,
+      "num_input_tokens_seen": 618670828,
+      "step": 15750
+    },
+    {
+      "epoch": 0.7636722653954948,
+      "grad_norm": 0.3979032039642334,
+      "learning_rate": 6.972831311226758e-06,
+      "loss": 1.2491,
+      "num_input_tokens_seen": 619060076,
+      "step": 15760
+    },
+    {
+      "epoch": 0.7641568290156696,
+      "grad_norm": 0.41532114148139954,
+      "learning_rate": 6.945668680603487e-06,
+      "loss": 1.2922,
+      "num_input_tokens_seen": 619436564,
+      "step": 15770
+    },
+    {
+      "epoch": 0.7646413926358444,
+      "grad_norm": 0.41047605872154236,
+      "learning_rate": 6.918550521614137e-06,
+      "loss": 1.2638,
+      "num_input_tokens_seen": 619817996,
+      "step": 15780
+    },
+    {
+      "epoch": 0.7651259562560192,
+      "grad_norm": 0.39262655377388,
+      "learning_rate": 6.891476901056445e-06,
+      "loss": 1.2943,
+      "num_input_tokens_seen": 620243736,
+      "step": 15790
+    },
+    {
+      "epoch": 0.765610519876194,
+      "grad_norm": 0.4346751272678375,
+      "learning_rate": 6.864447885618477e-06,
+      "loss": 1.2818,
+      "num_input_tokens_seen": 620638616,
+      "step": 15800
+    },
+    {
+      "epoch": 0.7660950834963688,
+      "grad_norm": 0.39605090022087097,
+      "learning_rate": 6.837463541878394e-06,
+      "loss": 1.2827,
+      "num_input_tokens_seen": 621007792,
+      "step": 15810
+    },
+    {
+      "epoch": 0.7665796471165436,
+      "grad_norm": 0.45163053274154663,
+      "learning_rate": 6.810523936304356e-06,
+      "loss": 1.2719,
+      "num_input_tokens_seen": 621383392,
+      "step": 15820
+    },
+    {
+      "epoch": 0.7670642107367184,
+      "grad_norm": 0.38712289929389954,
+      "learning_rate": 6.783629135254288e-06,
+      "loss": 1.2749,
+      "num_input_tokens_seen": 621788080,
+      "step": 15830
+    },
+    {
+      "epoch": 0.7675487743568932,
+      "grad_norm": 0.417074590921402,
+      "learning_rate": 6.756779204975785e-06,
+      "loss": 1.2551,
+      "num_input_tokens_seen": 622190428,
+      "step": 15840
+    },
+    {
+      "epoch": 0.7680333379770681,
+      "grad_norm": 0.4212455749511719,
+      "learning_rate": 6.729974211605888e-06,
+      "loss": 1.2964,
+      "num_input_tokens_seen": 622563676,
+      "step": 15850
+    },
+    {
+      "epoch": 0.7685179015972429,
+      "grad_norm": 0.43890535831451416,
+      "learning_rate": 6.703214221170961e-06,
+      "loss": 1.2982,
+      "num_input_tokens_seen": 622940944,
+      "step": 15860
+    },
+    {
+      "epoch": 0.7690024652174177,
+      "grad_norm": 0.4360447824001312,
+      "learning_rate": 6.676499299586525e-06,
+      "loss": 1.2862,
+      "num_input_tokens_seen": 623355512,
+      "step": 15870
+    },
+    {
+      "epoch": 0.7694870288375925,
+      "grad_norm": 0.41077935695648193,
+      "learning_rate": 6.649829512657082e-06,
+      "loss": 1.269,
+      "num_input_tokens_seen": 623743220,
+      "step": 15880
+    },
+    {
+      "epoch": 0.7699715924577673,
+      "grad_norm": 0.40932127833366394,
+      "learning_rate": 6.623204926075938e-06,
+      "loss": 1.2807,
+      "num_input_tokens_seen": 624121016,
+      "step": 15890
+    },
+    {
+      "epoch": 0.770456156077942,
+      "grad_norm": 0.42621883749961853,
+      "learning_rate": 6.596625605425083e-06,
+      "loss": 1.2542,
+      "num_input_tokens_seen": 624541652,
+      "step": 15900
+    },
+    {
+      "epoch": 0.7709407196981168,
+      "grad_norm": 0.4213699400424957,
+      "learning_rate": 6.570091616175014e-06,
+      "loss": 1.2228,
+      "num_input_tokens_seen": 624945824,
+      "step": 15910
+    },
+    {
+      "epoch": 0.7714252833182916,
+      "grad_norm": 0.4320572018623352,
+      "learning_rate": 6.543603023684536e-06,
+      "loss": 1.2724,
+      "num_input_tokens_seen": 625344196,
+      "step": 15920
+    },
+    {
+      "epoch": 0.7719098469384664,
+      "grad_norm": 0.403472363948822,
+      "learning_rate": 6.5171598932006665e-06,
+      "loss": 1.3135,
+      "num_input_tokens_seen": 625739548,
+      "step": 15930
+    },
+    {
+      "epoch": 0.7723944105586413,
+      "grad_norm": 0.4023309051990509,
+      "learning_rate": 6.49076228985841e-06,
+      "loss": 1.261,
+      "num_input_tokens_seen": 626125928,
+      "step": 15940
+    },
+    {
+      "epoch": 0.7728789741788161,
+      "grad_norm": 0.39641690254211426,
+      "learning_rate": 6.464410278680658e-06,
+      "loss": 1.2736,
+      "num_input_tokens_seen": 626532060,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7733635377989909,
+      "grad_norm": 0.39902058243751526,
+      "learning_rate": 6.4381039245779675e-06,
+      "loss": 1.2552,
+      "num_input_tokens_seen": 626924212,
+      "step": 15960
+    },
+    {
+      "epoch": 0.7738481014191657,
+      "grad_norm": 0.4693625271320343,
+      "learning_rate": 6.411843292348465e-06,
+      "loss": 1.31,
+      "num_input_tokens_seen": 627344236,
+      "step": 15970
+    },
+    {
+      "epoch": 0.7743326650393405,
+      "grad_norm": 0.4080311357975006,
+      "learning_rate": 6.385628446677624e-06,
+      "loss": 1.2568,
+      "num_input_tokens_seen": 627754136,
+      "step": 15980
+    },
+    {
+      "epoch": 0.7748172286595153,
+      "grad_norm": 0.4102632403373718,
+      "learning_rate": 6.359459452138161e-06,
+      "loss": 1.3095,
+      "num_input_tokens_seen": 628131720,
+      "step": 15990
+    },
+    {
+      "epoch": 0.7753017922796901,
+      "grad_norm": 0.4116896986961365,
+      "learning_rate": 6.33333637318983e-06,
+      "loss": 1.2585,
+      "num_input_tokens_seen": 628560668,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7753017922796901,
+      "eval_loss": 1.3683557510375977,
+      "eval_runtime": 3.6064,
+      "eval_samples_per_second": 41.592,
+      "eval_steps_per_second": 5.268,
+      "num_input_tokens_seen": 628560668,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7757863558998649,
+      "grad_norm": 0.3886633813381195,
+      "learning_rate": 6.3072592741793e-06,
+      "loss": 1.2845,
+      "num_input_tokens_seen": 628949040,
+      "step": 16010
+    },
+    {
+      "epoch": 0.7762709195200397,
+      "grad_norm": 0.4156300127506256,
+      "learning_rate": 6.28122821933998e-06,
+      "loss": 1.3016,
+      "num_input_tokens_seen": 629334480,
+      "step": 16020
+    },
+    {
+      "epoch": 0.7767554831402146,
+      "grad_norm": 0.40398070216178894,
+      "learning_rate": 6.255243272791858e-06,
+      "loss": 1.2423,
+      "num_input_tokens_seen": 629704276,
+      "step": 16030
+    },
+    {
+      "epoch": 0.7772400467603894,
+      "grad_norm": 0.3824484944343567,
+      "learning_rate": 6.2293044985413555e-06,
+      "loss": 1.2772,
+      "num_input_tokens_seen": 630086976,
+      "step": 16040
+    },
+    {
+      "epoch": 0.7777246103805642,
+      "grad_norm": 0.39248552918434143,
+      "learning_rate": 6.203411960481145e-06,
+      "loss": 1.254,
+      "num_input_tokens_seen": 630477336,
+      "step": 16050
+    },
+    {
+      "epoch": 0.778209174000739,
+      "grad_norm": 0.37206169962882996,
+      "learning_rate": 6.17756572239003e-06,
+      "loss": 1.2717,
+      "num_input_tokens_seen": 630871612,
+      "step": 16060
+    },
+    {
+      "epoch": 0.7786937376209138,
+      "grad_norm": 0.40856611728668213,
+      "learning_rate": 6.151765847932747e-06,
+      "loss": 1.2494,
+      "num_input_tokens_seen": 631248400,
+      "step": 16070
+    },
+    {
+      "epoch": 0.7791783012410886,
+      "grad_norm": 0.40056514739990234,
+      "learning_rate": 6.126012400659856e-06,
+      "loss": 1.2542,
+      "num_input_tokens_seen": 631641360,
+      "step": 16080
+    },
+    {
+      "epoch": 0.7796628648612633,
+      "grad_norm": 0.4258357882499695,
+      "learning_rate": 6.1003054440075205e-06,
+      "loss": 1.2451,
+      "num_input_tokens_seen": 632006744,
+      "step": 16090
+    },
+    {
+      "epoch": 0.7801474284814381,
+      "grad_norm": 0.39988163113594055,
+      "learning_rate": 6.074645041297425e-06,
+      "loss": 1.2772,
+      "num_input_tokens_seen": 632371660,
+      "step": 16100
+    },
+    {
+      "epoch": 0.7806319921016129,
+      "grad_norm": 0.44167232513427734,
+      "learning_rate": 6.049031255736548e-06,
+      "loss": 1.2338,
+      "num_input_tokens_seen": 632740436,
+      "step": 16110
+    },
+    {
+      "epoch": 0.7811165557217878,
+      "grad_norm": 0.3898947536945343,
+      "learning_rate": 6.023464150417077e-06,
+      "loss": 1.2255,
+      "num_input_tokens_seen": 633086408,
+      "step": 16120
+    },
+    {
+      "epoch": 0.7816011193419626,
+      "grad_norm": 0.43151912093162537,
+      "learning_rate": 5.997943788316179e-06,
+      "loss": 1.2518,
+      "num_input_tokens_seen": 633479808,
+      "step": 16130
+    },
+    {
+      "epoch": 0.7820856829621374,
+      "grad_norm": 0.39515140652656555,
+      "learning_rate": 5.972470232295907e-06,
+      "loss": 1.2406,
+      "num_input_tokens_seen": 633863396,
+      "step": 16140
+    },
+    {
+      "epoch": 0.7825702465823122,
+      "grad_norm": 0.4262303411960602,
+      "learning_rate": 5.947043545103012e-06,
+      "loss": 1.2579,
+      "num_input_tokens_seen": 634250168,
+      "step": 16150
+    },
+    {
+      "epoch": 0.783054810202487,
+      "grad_norm": 0.40577808022499084,
+      "learning_rate": 5.921663789368806e-06,
+      "loss": 1.3019,
+      "num_input_tokens_seen": 634607024,
+      "step": 16160
+    },
+    {
+      "epoch": 0.7835393738226618,
+      "grad_norm": 0.42164546251296997,
+      "learning_rate": 5.896331027608978e-06,
+      "loss": 1.2479,
+      "num_input_tokens_seen": 635005404,
+      "step": 16170
+    },
+    {
+      "epoch": 0.7840239374428366,
+      "grad_norm": 0.4001612663269043,
+      "learning_rate": 5.871045322223481e-06,
+      "loss": 1.2902,
+      "num_input_tokens_seen": 635356840,
+      "step": 16180
+    },
+    {
+      "epoch": 0.7845085010630114,
+      "grad_norm": 0.40439048409461975,
+      "learning_rate": 5.845806735496362e-06,
+      "loss": 1.2674,
+      "num_input_tokens_seen": 635733620,
+      "step": 16190
+    },
+    {
+      "epoch": 0.7849930646831863,
+      "grad_norm": 0.37555184960365295,
+      "learning_rate": 5.820615329595575e-06,
+      "loss": 1.2414,
+      "num_input_tokens_seen": 636132896,
+      "step": 16200
+    },
+    {
+      "epoch": 0.7854776283033611,
+      "grad_norm": 0.4048125147819519,
+      "learning_rate": 5.795471166572894e-06,
+      "loss": 1.2734,
+      "num_input_tokens_seen": 636522108,
+      "step": 16210
+    },
+    {
+      "epoch": 0.7859621919235359,
+      "grad_norm": 0.3880002200603485,
+      "learning_rate": 5.770374308363693e-06,
+      "loss": 1.2619,
+      "num_input_tokens_seen": 636887204,
+      "step": 16220
+    },
+    {
+      "epoch": 0.7864467555437107,
+      "grad_norm": 0.4102485477924347,
+      "learning_rate": 5.745324816786854e-06,
+      "loss": 1.235,
+      "num_input_tokens_seen": 637306444,
+      "step": 16230
+    },
+    {
+      "epoch": 0.7869313191638855,
+      "grad_norm": 0.42181381583213806,
+      "learning_rate": 5.720322753544549e-06,
+      "loss": 1.2281,
+      "num_input_tokens_seen": 637683960,
+      "step": 16240
+    },
+    {
+      "epoch": 0.7874158827840603,
+      "grad_norm": 0.40080034732818604,
+      "learning_rate": 5.695368180222163e-06,
+      "loss": 1.1679,
+      "num_input_tokens_seen": 638089180,
+      "step": 16250
+    },
+    {
+      "epoch": 0.7879004464042351,
+      "grad_norm": 0.39268258213996887,
+      "learning_rate": 5.670461158288071e-06,
+      "loss": 1.2564,
+      "num_input_tokens_seen": 638480792,
+      "step": 16260
+    },
+    {
+      "epoch": 0.7883850100244099,
+      "grad_norm": 0.3921567499637604,
+      "learning_rate": 5.6456017490935405e-06,
+      "loss": 1.2426,
+      "num_input_tokens_seen": 638897224,
+      "step": 16270
+    },
+    {
+      "epoch": 0.7888695736445847,
+      "grad_norm": 0.3720284402370453,
+      "learning_rate": 5.620790013872543e-06,
+      "loss": 1.2322,
+      "num_input_tokens_seen": 639288872,
+      "step": 16280
+    },
+    {
+      "epoch": 0.7893541372647596,
+      "grad_norm": 0.41209957003593445,
+      "learning_rate": 5.596026013741631e-06,
+      "loss": 1.3017,
+      "num_input_tokens_seen": 639672792,
+      "step": 16290
+    },
+    {
+      "epoch": 0.7898387008849344,
+      "grad_norm": 0.43729230761528015,
+      "learning_rate": 5.571309809699771e-06,
+      "loss": 1.3003,
+      "num_input_tokens_seen": 640077948,
+      "step": 16300
+    },
+    {
+      "epoch": 0.7903232645051091,
+      "grad_norm": 0.3936404883861542,
+      "learning_rate": 5.546641462628194e-06,
+      "loss": 1.3,
+      "num_input_tokens_seen": 640466812,
+      "step": 16310
+    },
+    {
+      "epoch": 0.7908078281252839,
+      "grad_norm": 0.396609365940094,
+      "learning_rate": 5.522021033290265e-06,
+      "loss": 1.2776,
+      "num_input_tokens_seen": 640869644,
+      "step": 16320
+    },
+    {
+      "epoch": 0.7912923917454587,
+      "grad_norm": 0.39924660325050354,
+      "learning_rate": 5.4974485823312885e-06,
+      "loss": 1.2602,
+      "num_input_tokens_seen": 641267224,
+      "step": 16330
+    },
+    {
+      "epoch": 0.7917769553656335,
+      "grad_norm": 0.4168340563774109,
+      "learning_rate": 5.472924170278418e-06,
+      "loss": 1.2327,
+      "num_input_tokens_seen": 641650820,
+      "step": 16340
+    },
+    {
+      "epoch": 0.7922615189858083,
+      "grad_norm": 0.3994021415710449,
+      "learning_rate": 5.448447857540453e-06,
+      "loss": 1.257,
+      "num_input_tokens_seen": 642044792,
+      "step": 16350
+    },
+    {
+      "epoch": 0.7927460826059831,
+      "grad_norm": 0.40958625078201294,
+      "learning_rate": 5.424019704407735e-06,
+      "loss": 1.277,
+      "num_input_tokens_seen": 642438212,
+      "step": 16360
+    },
+    {
+      "epoch": 0.7932306462261579,
+      "grad_norm": 0.4128318428993225,
+      "learning_rate": 5.3996397710519565e-06,
+      "loss": 1.2549,
+      "num_input_tokens_seen": 642855308,
+      "step": 16370
+    },
+    {
+      "epoch": 0.7937152098463328,
+      "grad_norm": 0.38638776540756226,
+      "learning_rate": 5.37530811752606e-06,
+      "loss": 1.2651,
+      "num_input_tokens_seen": 643244872,
+      "step": 16380
+    },
+    {
+      "epoch": 0.7941997734665076,
+      "grad_norm": 0.41983383893966675,
+      "learning_rate": 5.351024803764035e-06,
+      "loss": 1.2892,
+      "num_input_tokens_seen": 643614276,
+      "step": 16390
+    },
+    {
+      "epoch": 0.7946843370866824,
+      "grad_norm": 0.41671621799468994,
+      "learning_rate": 5.32678988958083e-06,
+      "loss": 1.2707,
+      "num_input_tokens_seen": 644000564,
+      "step": 16400
+    },
+    {
+      "epoch": 0.7951689007068572,
+      "grad_norm": 0.4639870822429657,
+      "learning_rate": 5.302603434672149e-06,
+      "loss": 1.2279,
+      "num_input_tokens_seen": 644383116,
+      "step": 16410
+    },
+    {
+      "epoch": 0.795653464327032,
+      "grad_norm": 0.4280005693435669,
+      "learning_rate": 5.278465498614349e-06,
+      "loss": 1.2616,
+      "num_input_tokens_seen": 644805824,
+      "step": 16420
+    },
+    {
+      "epoch": 0.7961380279472068,
+      "grad_norm": 0.45124003291130066,
+      "learning_rate": 5.254376140864273e-06,
+      "loss": 1.2564,
+      "num_input_tokens_seen": 645207656,
+      "step": 16430
+    },
+    {
+      "epoch": 0.7966225915673816,
+      "grad_norm": 0.42907872796058655,
+      "learning_rate": 5.230335420759089e-06,
+      "loss": 1.2791,
+      "num_input_tokens_seen": 645619040,
+      "step": 16440
+    },
+    {
+      "epoch": 0.7971071551875564,
+      "grad_norm": 0.4425942599773407,
+      "learning_rate": 5.206343397516178e-06,
+      "loss": 1.2267,
+      "num_input_tokens_seen": 646010824,
+      "step": 16450
+    },
+    {
+      "epoch": 0.7975917188077312,
+      "grad_norm": 0.41618916392326355,
+      "learning_rate": 5.182400130232962e-06,
+      "loss": 1.2694,
+      "num_input_tokens_seen": 646440088,
+      "step": 16460
+    },
+    {
+      "epoch": 0.7980762824279061,
+      "grad_norm": 0.3770136833190918,
+      "learning_rate": 5.1585056778867766e-06,
+      "loss": 1.2545,
+      "num_input_tokens_seen": 646840628,
+      "step": 16470
+    },
+    {
+      "epoch": 0.7985608460480809,
+      "grad_norm": 0.39917248487472534,
+      "learning_rate": 5.134660099334699e-06,
+      "loss": 1.2714,
+      "num_input_tokens_seen": 647216352,
+      "step": 16480
+    },
+    {
+      "epoch": 0.7990454096682557,
+      "grad_norm": 0.41502079367637634,
+      "learning_rate": 5.110863453313436e-06,
+      "loss": 1.2083,
+      "num_input_tokens_seen": 647613200,
+      "step": 16490
+    },
+    {
+      "epoch": 0.7995299732884305,
+      "grad_norm": 0.43602079153060913,
+      "learning_rate": 5.087115798439146e-06,
+      "loss": 1.2545,
+      "num_input_tokens_seen": 647990596,
+      "step": 16500
+    },
+    {
+      "epoch": 0.8000145369086052,
+      "grad_norm": 0.3962024748325348,
+      "learning_rate": 5.063417193207337e-06,
+      "loss": 1.2527,
+      "num_input_tokens_seen": 648402756,
+      "step": 16510
+    },
+    {
+      "epoch": 0.80049910052878,
+      "grad_norm": 0.4318599998950958,
+      "learning_rate": 5.039767695992664e-06,
+      "loss": 1.2438,
+      "num_input_tokens_seen": 648791928,
+      "step": 16520
+    },
+    {
+      "epoch": 0.8009836641489548,
+      "grad_norm": 0.41063135862350464,
+      "learning_rate": 5.016167365048857e-06,
+      "loss": 1.2647,
+      "num_input_tokens_seen": 649194520,
+      "step": 16530
+    },
+    {
+      "epoch": 0.8014682277691296,
+      "grad_norm": 0.419166624546051,
+      "learning_rate": 4.992616258508501e-06,
+      "loss": 1.2353,
+      "num_input_tokens_seen": 649580868,
+      "step": 16540
+    },
+    {
+      "epoch": 0.8019527913893044,
+      "grad_norm": 0.4164583683013916,
+      "learning_rate": 4.969114434382966e-06,
+      "loss": 1.2328,
+      "num_input_tokens_seen": 649985136,
+      "step": 16550
+    },
+    {
+      "epoch": 0.8024373550094793,
+      "grad_norm": 0.41067010164260864,
+      "learning_rate": 4.945661950562195e-06,
+      "loss": 1.2201,
+      "num_input_tokens_seen": 650381492,
+      "step": 16560
+    },
+    {
+      "epoch": 0.8029219186296541,
+      "grad_norm": 0.38203445076942444,
+      "learning_rate": 4.922258864814619e-06,
+      "loss": 1.242,
+      "num_input_tokens_seen": 650781812,
+      "step": 16570
+    },
+    {
+      "epoch": 0.8034064822498289,
+      "grad_norm": 0.3730851709842682,
+      "learning_rate": 4.8989052347869876e-06,
+      "loss": 1.2403,
+      "num_input_tokens_seen": 651179148,
+      "step": 16580
+    },
+    {
+      "epoch": 0.8038910458700037,
+      "grad_norm": 0.38860711455345154,
+      "learning_rate": 4.875601118004228e-06,
+      "loss": 1.2151,
+      "num_input_tokens_seen": 651589328,
+      "step": 16590
+    },
+    {
+      "epoch": 0.8043756094901785,
+      "grad_norm": 0.437386691570282,
+      "learning_rate": 4.852346571869307e-06,
+      "loss": 1.2417,
+      "num_input_tokens_seen": 651973144,
+      "step": 16600
+    },
+    {
+      "epoch": 0.8048601731103533,
+      "grad_norm": 0.3950743079185486,
+      "learning_rate": 4.8291416536630805e-06,
+      "loss": 1.2422,
+      "num_input_tokens_seen": 652374852,
+      "step": 16610
+    },
+    {
+      "epoch": 0.8053447367305281,
+      "grad_norm": 0.4481166899204254,
+      "learning_rate": 4.805986420544173e-06,
+      "loss": 1.2728,
+      "num_input_tokens_seen": 652767160,
+      "step": 16620
+    },
+    {
+      "epoch": 0.8058293003507029,
+      "grad_norm": 0.4067986011505127,
+      "learning_rate": 4.782880929548808e-06,
+      "loss": 1.2141,
+      "num_input_tokens_seen": 653149252,
+      "step": 16630
+    },
+    {
+      "epoch": 0.8063138639708777,
+      "grad_norm": 0.44263729453086853,
+      "learning_rate": 4.7598252375907035e-06,
+      "loss": 1.2469,
+      "num_input_tokens_seen": 653534468,
+      "step": 16640
+    },
+    {
+      "epoch": 0.8067984275910526,
+      "grad_norm": 0.4220990836620331,
+      "learning_rate": 4.736819401460893e-06,
+      "loss": 1.2297,
+      "num_input_tokens_seen": 653904180,
+      "step": 16650
+    },
+    {
+      "epoch": 0.8072829912112274,
+      "grad_norm": 0.4025983512401581,
+      "learning_rate": 4.713863477827626e-06,
+      "loss": 1.2693,
+      "num_input_tokens_seen": 654309572,
+      "step": 16660
+    },
+    {
+      "epoch": 0.8077675548314022,
+      "grad_norm": 0.44055184721946716,
+      "learning_rate": 4.690957523236178e-06,
+      "loss": 1.2899,
+      "num_input_tokens_seen": 654738392,
+      "step": 16670
+    },
+    {
+      "epoch": 0.808252118451577,
+      "grad_norm": 0.43497493863105774,
+      "learning_rate": 4.66810159410877e-06,
+      "loss": 1.2421,
+      "num_input_tokens_seen": 655115200,
+      "step": 16680
+    },
+    {
+      "epoch": 0.8087366820717518,
+      "grad_norm": 0.38774144649505615,
+      "learning_rate": 4.645295746744374e-06,
+      "loss": 1.2939,
+      "num_input_tokens_seen": 655499648,
+      "step": 16690
+    },
+    {
+      "epoch": 0.8092212456919265,
+      "grad_norm": 0.41730353236198425,
+      "learning_rate": 4.622540037318618e-06,
+      "loss": 1.2312,
+      "num_input_tokens_seen": 655885096,
+      "step": 16700
+    },
+    {
+      "epoch": 0.8097058093121013,
+      "grad_norm": 0.4214309751987457,
+      "learning_rate": 4.5998345218836304e-06,
+      "loss": 1.2738,
+      "num_input_tokens_seen": 656267208,
+      "step": 16710
+    },
+    {
+      "epoch": 0.8101903729322761,
+      "grad_norm": 0.40911537408828735,
+      "learning_rate": 4.577179256367886e-06,
+      "loss": 1.2121,
+      "num_input_tokens_seen": 656652956,
+      "step": 16720
+    },
+    {
+      "epoch": 0.8106749365524509,
+      "grad_norm": 0.4146646559238434,
+      "learning_rate": 4.554574296576092e-06,
+      "loss": 1.2153,
+      "num_input_tokens_seen": 657038508,
+      "step": 16730
+    },
+    {
+      "epoch": 0.8111595001726258,
+      "grad_norm": 0.430595338344574,
+      "learning_rate": 4.532019698189044e-06,
+      "loss": 1.2613,
+      "num_input_tokens_seen": 657434028,
+      "step": 16740
+    },
+    {
+      "epoch": 0.8116440637928006,
+      "grad_norm": 0.3903404772281647,
+      "learning_rate": 4.509515516763493e-06,
+      "loss": 1.2283,
+      "num_input_tokens_seen": 657845788,
+      "step": 16750
+    },
+    {
+      "epoch": 0.8121286274129754,
+      "grad_norm": 0.4230201840400696,
+      "learning_rate": 4.487061807731982e-06,
+      "loss": 1.2466,
+      "num_input_tokens_seen": 658242124,
+      "step": 16760
+    },
+    {
+      "epoch": 0.8126131910331502,
+      "grad_norm": 0.39636996388435364,
+      "learning_rate": 4.464658626402751e-06,
+      "loss": 1.2527,
+      "num_input_tokens_seen": 658637400,
+      "step": 16770
+    },
+    {
+      "epoch": 0.813097754653325,
+      "grad_norm": 0.4210509657859802,
+      "learning_rate": 4.442306027959564e-06,
+      "loss": 1.2837,
+      "num_input_tokens_seen": 659032724,
+      "step": 16780
+    },
+    {
+      "epoch": 0.8135823182734998,
+      "grad_norm": 0.4173252582550049,
+      "learning_rate": 4.4200040674616e-06,
+      "loss": 1.2855,
+      "num_input_tokens_seen": 659452924,
+      "step": 16790
+    },
+    {
+      "epoch": 0.8140668818936746,
+      "grad_norm": 0.43643826246261597,
+      "learning_rate": 4.397752799843294e-06,
+      "loss": 1.2185,
+      "num_input_tokens_seen": 659860564,
+      "step": 16800
+    },
+    {
+      "epoch": 0.8145514455138494,
+      "grad_norm": 0.4406135380268097,
+      "learning_rate": 4.375552279914233e-06,
+      "loss": 1.3007,
+      "num_input_tokens_seen": 660247376,
+      "step": 16810
+    },
+    {
+      "epoch": 0.8150360091340242,
+      "grad_norm": 0.41418373584747314,
+      "learning_rate": 4.353402562358977e-06,
+      "loss": 1.2831,
+      "num_input_tokens_seen": 660644512,
+      "step": 16820
+    },
+    {
+      "epoch": 0.8155205727541991,
+      "grad_norm": 0.3968205749988556,
+      "learning_rate": 4.331303701736969e-06,
+      "loss": 1.2532,
+      "num_input_tokens_seen": 661014408,
+      "step": 16830
+    },
+    {
+      "epoch": 0.8160051363743739,
+      "grad_norm": 0.4220852255821228,
+      "learning_rate": 4.309255752482378e-06,
+      "loss": 1.2457,
+      "num_input_tokens_seen": 661407992,
+      "step": 16840
+    },
+    {
+      "epoch": 0.8164896999945487,
+      "grad_norm": 0.418720543384552,
+      "learning_rate": 4.2872587689039484e-06,
+      "loss": 1.2173,
+      "num_input_tokens_seen": 661784464,
+      "step": 16850
+    },
+    {
+      "epoch": 0.8169742636147235,
+      "grad_norm": 0.38268375396728516,
+      "learning_rate": 4.265312805184909e-06,
+      "loss": 1.237,
+      "num_input_tokens_seen": 662147388,
+      "step": 16860
+    },
+    {
+      "epoch": 0.8174588272348983,
+      "grad_norm": 0.40058982372283936,
+      "learning_rate": 4.24341791538281e-06,
+      "loss": 1.2639,
+      "num_input_tokens_seen": 662530408,
+      "step": 16870
+    },
+    {
+      "epoch": 0.8179433908550731,
+      "grad_norm": 0.4105454981327057,
+      "learning_rate": 4.221574153429392e-06,
+      "loss": 1.2431,
+      "num_input_tokens_seen": 662921336,
+      "step": 16880
+    },
+    {
+      "epoch": 0.8184279544752479,
+      "grad_norm": 0.41436707973480225,
+      "learning_rate": 4.1997815731304515e-06,
+      "loss": 1.2628,
+      "num_input_tokens_seen": 663310688,
+      "step": 16890
+    },
+    {
+      "epoch": 0.8189125180954226,
+      "grad_norm": 0.39193400740623474,
+      "learning_rate": 4.178040228165725e-06,
+      "loss": 1.2385,
+      "num_input_tokens_seen": 663687484,
+      "step": 16900
+    },
+    {
+      "epoch": 0.8193970817155974,
+      "grad_norm": 0.4802253246307373,
+      "learning_rate": 4.156350172088736e-06,
+      "loss": 1.2784,
+      "num_input_tokens_seen": 664078036,
+      "step": 16910
+    },
+    {
+      "epoch": 0.8198816453357723,
+      "grad_norm": 0.4015842080116272,
+      "learning_rate": 4.134711458326681e-06,
+      "loss": 1.2329,
+      "num_input_tokens_seen": 664483676,
+      "step": 16920
+    },
+    {
+      "epoch": 0.8203662089559471,
+      "grad_norm": 0.40182778239250183,
+      "learning_rate": 4.11312414018028e-06,
+      "loss": 1.2525,
+      "num_input_tokens_seen": 664851144,
+      "step": 16930
+    },
+    {
+      "epoch": 0.8208507725761219,
+      "grad_norm": 0.4445132315158844,
+      "learning_rate": 4.091588270823671e-06,
+      "loss": 1.2506,
+      "num_input_tokens_seen": 665274412,
+      "step": 16940
+    },
+    {
+      "epoch": 0.8213353361962967,
+      "grad_norm": 0.43962743878364563,
+      "learning_rate": 4.070103903304237e-06,
+      "loss": 1.299,
+      "num_input_tokens_seen": 665663080,
+      "step": 16950
+    },
+    {
+      "epoch": 0.8218198998164715,
+      "grad_norm": 0.43457552790641785,
+      "learning_rate": 4.048671090542522e-06,
+      "loss": 1.2759,
+      "num_input_tokens_seen": 666049536,
+      "step": 16960
+    },
+    {
+      "epoch": 0.8223044634366463,
+      "grad_norm": 0.44402769207954407,
+      "learning_rate": 4.0272898853320835e-06,
+      "loss": 1.2648,
+      "num_input_tokens_seen": 666424352,
+      "step": 16970
+    },
+    {
+      "epoch": 0.8227890270568211,
+      "grad_norm": 0.4204462468624115,
+      "learning_rate": 4.005960340339335e-06,
+      "loss": 1.2761,
+      "num_input_tokens_seen": 666812808,
+      "step": 16980
+    },
+    {
+      "epoch": 0.8232735906769959,
+      "grad_norm": 0.4030398428440094,
+      "learning_rate": 3.984682508103466e-06,
+      "loss": 1.2865,
+      "num_input_tokens_seen": 667198772,
+      "step": 16990
+    },
+    {
+      "epoch": 0.8237581542971708,
+      "grad_norm": 0.3986146152019501,
+      "learning_rate": 3.963456441036259e-06,
+      "loss": 1.247,
+      "num_input_tokens_seen": 667609040,
+      "step": 17000
+    },
+    {
+      "epoch": 0.8242427179173456,
+      "grad_norm": 0.4464845061302185,
+      "learning_rate": 3.942282191422017e-06,
+      "loss": 1.2972,
+      "num_input_tokens_seen": 668009280,
+      "step": 17010
+    },
+    {
+      "epoch": 0.8247272815375204,
+      "grad_norm": 0.37678518891334534,
+      "learning_rate": 3.9211598114173855e-06,
+      "loss": 1.2387,
+      "num_input_tokens_seen": 668416796,
+      "step": 17020
+    },
+    {
+      "epoch": 0.8252118451576952,
+      "grad_norm": 0.38661307096481323,
+      "learning_rate": 3.900089353051259e-06,
+      "loss": 1.2197,
+      "num_input_tokens_seen": 668795620,
+      "step": 17030
+    },
+    {
+      "epoch": 0.82569640877787,
+      "grad_norm": 0.3967253267765045,
+      "learning_rate": 3.879070868224616e-06,
+      "loss": 1.2319,
+      "num_input_tokens_seen": 669174880,
+      "step": 17040
+    },
+    {
+      "epoch": 0.8261809723980448,
+      "grad_norm": 0.45880743861198425,
+      "learning_rate": 3.858104408710445e-06,
+      "loss": 1.2629,
+      "num_input_tokens_seen": 669581256,
+      "step": 17050
+    },
+    {
+      "epoch": 0.8266655360182196,
+      "grad_norm": 0.40062230825424194,
+      "learning_rate": 3.837190026153548e-06,
+      "loss": 1.3047,
+      "num_input_tokens_seen": 669998664,
+      "step": 17060
+    },
+    {
+      "epoch": 0.8271500996383944,
+      "grad_norm": 0.42311957478523254,
+      "learning_rate": 3.816327772070483e-06,
+      "loss": 1.2558,
+      "num_input_tokens_seen": 670389116,
+      "step": 17070
+    },
+    {
+      "epoch": 0.8276346632585692,
+      "grad_norm": 0.41024044156074524,
+      "learning_rate": 3.7955176978493822e-06,
+      "loss": 1.2791,
+      "num_input_tokens_seen": 670805700,
+      "step": 17080
+    },
+    {
+      "epoch": 0.8281192268787441,
+      "grad_norm": 0.4341394305229187,
+      "learning_rate": 3.7747598547498682e-06,
+      "loss": 1.229,
+      "num_input_tokens_seen": 671163656,
+      "step": 17090
+    },
+    {
+      "epoch": 0.8286037904989189,
+      "grad_norm": 0.42642325162887573,
+      "learning_rate": 3.754054293902884e-06,
+      "loss": 1.2721,
+      "num_input_tokens_seen": 671556364,
+      "step": 17100
+    },
+    {
+      "epoch": 0.8290883541190937,
+      "grad_norm": 0.41226011514663696,
+      "learning_rate": 3.7334010663106044e-06,
+      "loss": 1.2968,
+      "num_input_tokens_seen": 671957808,
+      "step": 17110
+    },
+    {
+      "epoch": 0.8295729177392684,
+      "grad_norm": 0.4139423966407776,
+      "learning_rate": 3.712800222846302e-06,
+      "loss": 1.2627,
+      "num_input_tokens_seen": 672364408,
+      "step": 17120
+    },
+    {
+      "epoch": 0.8300574813594432,
+      "grad_norm": 0.4178198575973511,
+      "learning_rate": 3.6922518142541994e-06,
+      "loss": 1.2778,
+      "num_input_tokens_seen": 672770620,
+      "step": 17130
+    },
+    {
+      "epoch": 0.830542044979618,
+      "grad_norm": 0.43971824645996094,
+      "learning_rate": 3.6717558911493784e-06,
+      "loss": 1.1902,
+      "num_input_tokens_seen": 673162284,
+      "step": 17140
+    },
+    {
+      "epoch": 0.8310266085997928,
+      "grad_norm": 0.40751203894615173,
+      "learning_rate": 3.6513125040176205e-06,
+      "loss": 1.2592,
+      "num_input_tokens_seen": 673552816,
+      "step": 17150
+    },
+    {
+      "epoch": 0.8315111722199676,
+      "grad_norm": 0.3929533064365387,
+      "learning_rate": 3.6309217032153093e-06,
+      "loss": 1.2793,
+      "num_input_tokens_seen": 673963224,
+      "step": 17160
+    },
+    {
+      "epoch": 0.8319957358401424,
+      "grad_norm": 0.42938390374183655,
+      "learning_rate": 3.610583538969306e-06,
+      "loss": 1.2622,
+      "num_input_tokens_seen": 674355136,
+      "step": 17170
+    },
+    {
+      "epoch": 0.8324802994603173,
+      "grad_norm": 0.42667803168296814,
+      "learning_rate": 3.590298061376804e-06,
+      "loss": 1.3012,
+      "num_input_tokens_seen": 674750972,
+      "step": 17180
+    },
+    {
+      "epoch": 0.8329648630804921,
+      "grad_norm": 0.4376852214336395,
+      "learning_rate": 3.5700653204052167e-06,
+      "loss": 1.3022,
+      "num_input_tokens_seen": 675185324,
+      "step": 17190
+    },
+    {
+      "epoch": 0.8334494267006669,
+      "grad_norm": 0.40134289860725403,
+      "learning_rate": 3.5498853658920695e-06,
+      "loss": 1.2425,
+      "num_input_tokens_seen": 675605452,
+      "step": 17200
+    },
+    {
+      "epoch": 0.8339339903208417,
+      "grad_norm": 0.4355108439922333,
+      "learning_rate": 3.5297582475448483e-06,
+      "loss": 1.2516,
+      "num_input_tokens_seen": 676001164,
+      "step": 17210
+    },
+    {
+      "epoch": 0.8344185539410165,
+      "grad_norm": 0.3797595798969269,
+      "learning_rate": 3.509684014940906e-06,
+      "loss": 1.2325,
+      "num_input_tokens_seen": 676377212,
+      "step": 17220
+    },
+    {
+      "epoch": 0.8349031175611913,
+      "grad_norm": 0.42660579085350037,
+      "learning_rate": 3.489662717527312e-06,
+      "loss": 1.2337,
+      "num_input_tokens_seen": 676770320,
+      "step": 17230
+    },
+    {
+      "epoch": 0.8353876811813661,
+      "grad_norm": 0.39556610584259033,
+      "learning_rate": 3.4696944046207574e-06,
+      "loss": 1.3193,
+      "num_input_tokens_seen": 677180796,
+      "step": 17240
+    },
+    {
+      "epoch": 0.8358722448015409,
+      "grad_norm": 0.3952700197696686,
+      "learning_rate": 3.449779125407426e-06,
+      "loss": 1.2574,
+      "num_input_tokens_seen": 677593456,
+      "step": 17250
+    },
+    {
+      "epoch": 0.8363568084217157,
+      "grad_norm": 0.40727055072784424,
+      "learning_rate": 3.4299169289428446e-06,
+      "loss": 1.2599,
+      "num_input_tokens_seen": 677958328,
+      "step": 17260
+    },
+    {
+      "epoch": 0.8368413720418906,
+      "grad_norm": 0.4267754852771759,
+      "learning_rate": 3.4101078641518124e-06,
+      "loss": 1.2312,
+      "num_input_tokens_seen": 678375748,
+      "step": 17270
+    },
+    {
+      "epoch": 0.8373259356620654,
+      "grad_norm": 0.3855496644973755,
+      "learning_rate": 3.3903519798282353e-06,
+      "loss": 1.2167,
+      "num_input_tokens_seen": 678754312,
+      "step": 17280
+    },
+    {
+      "epoch": 0.8378104992822402,
+      "grad_norm": 0.39305436611175537,
+      "learning_rate": 3.3706493246350335e-06,
+      "loss": 1.2569,
+      "num_input_tokens_seen": 679131624,
+      "step": 17290
+    },
+    {
+      "epoch": 0.838295062902415,
+      "grad_norm": 0.395431786775589,
+      "learning_rate": 3.3509999471040136e-06,
+      "loss": 1.2482,
+      "num_input_tokens_seen": 679524344,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8387796265225897,
+      "grad_norm": 0.4162997901439667,
+      "learning_rate": 3.3314038956357514e-06,
+      "loss": 1.2221,
+      "num_input_tokens_seen": 679924936,
+      "step": 17310
+    },
+    {
+      "epoch": 0.8392641901427645,
+      "grad_norm": 0.42025381326675415,
+      "learning_rate": 3.3118612184994485e-06,
+      "loss": 1.2243,
+      "num_input_tokens_seen": 680326700,
+      "step": 17320
+    },
+    {
+      "epoch": 0.8397487537629393,
+      "grad_norm": 0.46221596002578735,
+      "learning_rate": 3.292371963832863e-06,
+      "loss": 1.2707,
+      "num_input_tokens_seen": 680740540,
+      "step": 17330
+    },
+    {
+      "epoch": 0.8402333173831141,
+      "grad_norm": 0.40451374650001526,
+      "learning_rate": 3.2729361796421387e-06,
+      "loss": 1.2341,
+      "num_input_tokens_seen": 681129484,
+      "step": 17340
+    },
+    {
+      "epoch": 0.8407178810032889,
+      "grad_norm": 0.43298593163490295,
+      "learning_rate": 3.253553913801727e-06,
+      "loss": 1.305,
+      "num_input_tokens_seen": 681512368,
+      "step": 17350
+    },
+    {
+      "epoch": 0.8412024446234638,
+      "grad_norm": 0.4344673156738281,
+      "learning_rate": 3.2342252140542373e-06,
+      "loss": 1.3109,
+      "num_input_tokens_seen": 681938304,
+      "step": 17360
+    },
+    {
+      "epoch": 0.8416870082436386,
+      "grad_norm": 0.39718765020370483,
+      "learning_rate": 3.2149501280103466e-06,
+      "loss": 1.2778,
+      "num_input_tokens_seen": 682342636,
+      "step": 17370
+    },
+    {
+      "epoch": 0.8421715718638134,
+      "grad_norm": 0.4262661933898926,
+      "learning_rate": 3.195728703148673e-06,
+      "loss": 1.2626,
+      "num_input_tokens_seen": 682748112,
+      "step": 17380
+    },
+    {
+      "epoch": 0.8426561354839882,
+      "grad_norm": 0.4599827826023102,
+      "learning_rate": 3.1765609868156325e-06,
+      "loss": 1.285,
+      "num_input_tokens_seen": 683133464,
+      "step": 17390
+    },
+    {
+      "epoch": 0.843140699104163,
+      "grad_norm": 0.3915078639984131,
+      "learning_rate": 3.1574470262253795e-06,
+      "loss": 1.2881,
+      "num_input_tokens_seen": 683506720,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8436252627243378,
+      "grad_norm": 0.4697626233100891,
+      "learning_rate": 3.138386868459622e-06,
+      "loss": 1.2346,
+      "num_input_tokens_seen": 683894420,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8441098263445126,
+      "grad_norm": 0.4377289116382599,
+      "learning_rate": 3.119380560467572e-06,
+      "loss": 1.2772,
+      "num_input_tokens_seen": 684275120,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8445943899646874,
+      "grad_norm": 0.4258815050125122,
+      "learning_rate": 3.1004281490657703e-06,
+      "loss": 1.2482,
+      "num_input_tokens_seen": 684658984,
+      "step": 17430
+    },
+    {
+      "epoch": 0.8450789535848622,
+      "grad_norm": 0.4437004625797272,
+      "learning_rate": 3.0815296809380167e-06,
+      "loss": 1.2674,
+      "num_input_tokens_seen": 685068552,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8455635172050371,
+      "grad_norm": 0.3915651738643646,
+      "learning_rate": 3.0626852026352347e-06,
+      "loss": 1.2626,
+      "num_input_tokens_seen": 685445348,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8460480808252119,
+      "grad_norm": 0.4123501777648926,
+      "learning_rate": 3.043894760575358e-06,
+      "loss": 1.2912,
+      "num_input_tokens_seen": 685845820,
+      "step": 17460
+    },
+    {
+      "epoch": 0.8465326444453867,
+      "grad_norm": 0.4246242344379425,
+      "learning_rate": 3.0251584010432127e-06,
+      "loss": 1.2455,
+      "num_input_tokens_seen": 686224024,
+      "step": 17470
+    },
+    {
+      "epoch": 0.8470172080655615,
+      "grad_norm": 0.43629172444343567,
+      "learning_rate": 3.00647617019042e-06,
+      "loss": 1.2333,
+      "num_input_tokens_seen": 686627324,
+      "step": 17480
+    },
+    {
+      "epoch": 0.8475017716857363,
+      "grad_norm": 0.3956005871295929,
+      "learning_rate": 2.9878481140352495e-06,
+      "loss": 1.2989,
+      "num_input_tokens_seen": 687019208,
+      "step": 17490
+    },
+    {
+      "epoch": 0.847986335305911,
+      "grad_norm": 0.434689462184906,
+      "learning_rate": 2.96927427846255e-06,
+      "loss": 1.2314,
+      "num_input_tokens_seen": 687411440,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8484708989260858,
+      "grad_norm": 0.3737858831882477,
+      "learning_rate": 2.9507547092236075e-06,
+      "loss": 1.2527,
+      "num_input_tokens_seen": 687806800,
+      "step": 17510
+    },
+    {
+      "epoch": 0.8489554625462606,
+      "grad_norm": 0.4112042784690857,
+      "learning_rate": 2.9322894519360237e-06,
+      "loss": 1.249,
+      "num_input_tokens_seen": 688225164,
+      "step": 17520
+    },
+    {
+      "epoch": 0.8494400261664354,
+      "grad_norm": 0.41528812050819397,
+      "learning_rate": 2.913878552083646e-06,
+      "loss": 1.3127,
+      "num_input_tokens_seen": 688617932,
+      "step": 17530
+    },
+    {
+      "epoch": 0.8499245897866103,
+      "grad_norm": 0.43437185883522034,
+      "learning_rate": 2.895522055016395e-06,
+      "loss": 1.2433,
+      "num_input_tokens_seen": 689007680,
+      "step": 17540
+    },
+    {
+      "epoch": 0.8504091534067851,
+      "grad_norm": 0.39245957136154175,
+      "learning_rate": 2.8772200059502153e-06,
+      "loss": 1.2776,
+      "num_input_tokens_seen": 689376876,
+      "step": 17550
+    },
+    {
+      "epoch": 0.8508937170269599,
+      "grad_norm": 0.44739076495170593,
+      "learning_rate": 2.8589724499669122e-06,
+      "loss": 1.2425,
+      "num_input_tokens_seen": 689774952,
+      "step": 17560
+    },
+    {
+      "epoch": 0.8513782806471347,
+      "grad_norm": 0.400081604719162,
+      "learning_rate": 2.840779432014079e-06,
+      "loss": 1.253,
+      "num_input_tokens_seen": 690191364,
+      "step": 17570
+    },
+    {
+      "epoch": 0.8518628442673095,
+      "grad_norm": 0.43295279145240784,
+      "learning_rate": 2.8226409969049627e-06,
+      "loss": 1.24,
+      "num_input_tokens_seen": 690593376,
+      "step": 17580
+    },
+    {
+      "epoch": 0.8523474078874843,
+      "grad_norm": 0.4127790927886963,
+      "learning_rate": 2.804557189318366e-06,
+      "loss": 1.2606,
+      "num_input_tokens_seen": 691002516,
+      "step": 17590
+    },
+    {
+      "epoch": 0.8528319715076591,
+      "grad_norm": 0.40065881609916687,
+      "learning_rate": 2.7865280537985233e-06,
+      "loss": 1.2413,
+      "num_input_tokens_seen": 691404960,
+      "step": 17600
+    },
+    {
+      "epoch": 0.8533165351278339,
+      "grad_norm": 0.45817670226097107,
+      "learning_rate": 2.768553634755011e-06,
+      "loss": 1.2612,
+      "num_input_tokens_seen": 691797660,
+      "step": 17610
+    },
+    {
+      "epoch": 0.8538010987480087,
+      "grad_norm": 0.36435696482658386,
+      "learning_rate": 2.750633976462616e-06,
+      "loss": 1.2247,
+      "num_input_tokens_seen": 692207740,
+      "step": 17620
+    },
+    {
+      "epoch": 0.8542856623681836,
+      "grad_norm": 0.41439151763916016,
+      "learning_rate": 2.732769123061249e-06,
+      "loss": 1.271,
+      "num_input_tokens_seen": 692585708,
+      "step": 17630
+    },
+    {
+      "epoch": 0.8547702259883584,
+      "grad_norm": 0.4093399941921234,
+      "learning_rate": 2.714959118555821e-06,
+      "loss": 1.2442,
+      "num_input_tokens_seen": 692977756,
+      "step": 17640
+    },
+    {
+      "epoch": 0.8552547896085332,
+      "grad_norm": 0.44113224744796753,
+      "learning_rate": 2.697204006816131e-06,
+      "loss": 1.2601,
+      "num_input_tokens_seen": 693350796,
+      "step": 17650
+    },
+    {
+      "epoch": 0.855739353228708,
+      "grad_norm": 0.40980538725852966,
+      "learning_rate": 2.6795038315767824e-06,
+      "loss": 1.2668,
+      "num_input_tokens_seen": 693759588,
+      "step": 17660
+    },
+    {
+      "epoch": 0.8562239168488828,
+      "grad_norm": 0.42793384194374084,
+      "learning_rate": 2.661858636437034e-06,
+      "loss": 1.2285,
+      "num_input_tokens_seen": 694174680,
+      "step": 17670
+    },
+    {
+      "epoch": 0.8567084804690576,
+      "grad_norm": 0.4959481954574585,
+      "learning_rate": 2.644268464860741e-06,
+      "loss": 1.2491,
+      "num_input_tokens_seen": 694561320,
+      "step": 17680
+    },
+    {
+      "epoch": 0.8571930440892324,
+      "grad_norm": 0.40092337131500244,
+      "learning_rate": 2.6267333601762088e-06,
+      "loss": 1.2301,
+      "num_input_tokens_seen": 694959912,
+      "step": 17690
+    },
+    {
+      "epoch": 0.8576776077094072,
+      "grad_norm": 0.40096816420555115,
+      "learning_rate": 2.6092533655761144e-06,
+      "loss": 1.2886,
+      "num_input_tokens_seen": 695328628,
+      "step": 17700
+    },
+    {
+      "epoch": 0.8581621713295821,
+      "grad_norm": 0.4054926335811615,
+      "learning_rate": 2.591828524117365e-06,
+      "loss": 1.2309,
+      "num_input_tokens_seen": 695748200,
+      "step": 17710
+    },
+    {
+      "epoch": 0.8586467349497569,
+      "grad_norm": 0.38719359040260315,
+      "learning_rate": 2.5744588787210366e-06,
+      "loss": 1.2367,
+      "num_input_tokens_seen": 696133424,
+      "step": 17720
+    },
+    {
+      "epoch": 0.8591312985699316,
+      "grad_norm": 0.386705219745636,
+      "learning_rate": 2.557144472172235e-06,
+      "loss": 1.2824,
+      "num_input_tokens_seen": 696536372,
+      "step": 17730
+    },
+    {
+      "epoch": 0.8596158621901064,
+      "grad_norm": 0.4401761293411255,
+      "learning_rate": 2.5398853471200105e-06,
+      "loss": 1.2132,
+      "num_input_tokens_seen": 696913248,
+      "step": 17740
+    },
+    {
+      "epoch": 0.8601004258102812,
+      "grad_norm": 0.43798384070396423,
+      "learning_rate": 2.522681546077224e-06,
+      "loss": 1.2504,
+      "num_input_tokens_seen": 697311808,
+      "step": 17750
+    },
+    {
+      "epoch": 0.860584989430456,
+      "grad_norm": 0.4158650040626526,
+      "learning_rate": 2.5055331114204798e-06,
+      "loss": 1.2485,
+      "num_input_tokens_seen": 697710452,
+      "step": 17760
+    },
+    {
+      "epoch": 0.8610695530506308,
+      "grad_norm": 0.4461188018321991,
+      "learning_rate": 2.4884400853900034e-06,
+      "loss": 1.2431,
+      "num_input_tokens_seen": 698116700,
+      "step": 17770
+    },
+    {
+      "epoch": 0.8615541166708056,
+      "grad_norm": 0.36897483468055725,
+      "learning_rate": 2.4714025100895155e-06,
+      "loss": 1.2314,
+      "num_input_tokens_seen": 698502984,
+      "step": 17780
+    },
+    {
+      "epoch": 0.8620386802909804,
+      "grad_norm": 0.38578104972839355,
+      "learning_rate": 2.4544204274861785e-06,
+      "loss": 1.2852,
+      "num_input_tokens_seen": 698869344,
+      "step": 17790
+    },
+    {
+      "epoch": 0.8625232439111553,
+      "grad_norm": 0.3806360363960266,
+      "learning_rate": 2.4374938794104407e-06,
+      "loss": 1.2656,
+      "num_input_tokens_seen": 699268952,
+      "step": 17800
+    },
+    {
+      "epoch": 0.8630078075313301,
+      "grad_norm": 0.43705320358276367,
+      "learning_rate": 2.420622907555975e-06,
+      "loss": 1.2623,
+      "num_input_tokens_seen": 699655576,
+      "step": 17810
+    },
+    {
+      "epoch": 0.8634923711515049,
+      "grad_norm": 0.41822245717048645,
+      "learning_rate": 2.403807553479548e-06,
+      "loss": 1.2524,
+      "num_input_tokens_seen": 700027896,
+      "step": 17820
+    },
+    {
+      "epoch": 0.8639769347716797,
+      "grad_norm": 0.4171805679798126,
+      "learning_rate": 2.3870478586009325e-06,
+      "loss": 1.2786,
+      "num_input_tokens_seen": 700417852,
+      "step": 17830
+    },
+    {
+      "epoch": 0.8644614983918545,
+      "grad_norm": 0.4310646057128906,
+      "learning_rate": 2.3703438642027927e-06,
+      "loss": 1.2672,
+      "num_input_tokens_seen": 700843676,
+      "step": 17840
+    },
+    {
+      "epoch": 0.8649460620120293,
+      "grad_norm": 0.41659244894981384,
+      "learning_rate": 2.353695611430609e-06,
+      "loss": 1.2511,
+      "num_input_tokens_seen": 701217408,
+      "step": 17850
+    },
+    {
+      "epoch": 0.8654306256322041,
+      "grad_norm": 0.3900809586048126,
+      "learning_rate": 2.337103141292535e-06,
+      "loss": 1.2216,
+      "num_input_tokens_seen": 701584684,
+      "step": 17860
+    },
+    {
+      "epoch": 0.8659151892523789,
+      "grad_norm": 0.4181850552558899,
+      "learning_rate": 2.3205664946593348e-06,
+      "loss": 1.3174,
+      "num_input_tokens_seen": 701960304,
+      "step": 17870
+    },
+    {
+      "epoch": 0.8663997528725537,
+      "grad_norm": 0.4225742220878601,
+      "learning_rate": 2.3040857122642674e-06,
+      "loss": 1.229,
+      "num_input_tokens_seen": 702347808,
+      "step": 17880
+    },
+    {
+      "epoch": 0.8668843164927286,
+      "grad_norm": 0.3922450542449951,
+      "learning_rate": 2.2876608347029816e-06,
+      "loss": 1.2615,
+      "num_input_tokens_seen": 702743944,
+      "step": 17890
+    },
+    {
+      "epoch": 0.8673688801129034,
+      "grad_norm": 0.3811093866825104,
+      "learning_rate": 2.2712919024334257e-06,
+      "loss": 1.2735,
+      "num_input_tokens_seen": 703123092,
+      "step": 17900
+    },
+    {
+      "epoch": 0.8678534437330782,
+      "grad_norm": 0.45579883456230164,
+      "learning_rate": 2.2549789557757327e-06,
+      "loss": 1.2724,
+      "num_input_tokens_seen": 703523388,
+      "step": 17910
+    },
+    {
+      "epoch": 0.868338007353253,
+      "grad_norm": 0.44050538539886475,
+      "learning_rate": 2.238722034912144e-06,
+      "loss": 1.2293,
+      "num_input_tokens_seen": 703903936,
+      "step": 17920
+    },
+    {
+      "epoch": 0.8688225709734277,
+      "grad_norm": 0.3877395987510681,
+      "learning_rate": 2.222521179886888e-06,
+      "loss": 1.2697,
+      "num_input_tokens_seen": 704304692,
+      "step": 17930
+    },
+    {
+      "epoch": 0.8693071345936025,
+      "grad_norm": 0.427232027053833,
+      "learning_rate": 2.206376430606097e-06,
+      "loss": 1.2374,
+      "num_input_tokens_seen": 704700272,
+      "step": 17940
+    },
+    {
+      "epoch": 0.8697916982137773,
+      "grad_norm": 0.39587920904159546,
+      "learning_rate": 2.1902878268376975e-06,
+      "loss": 1.2735,
+      "num_input_tokens_seen": 705078452,
+      "step": 17950
+    },
+    {
+      "epoch": 0.8702762618339521,
+      "grad_norm": 0.37119200825691223,
+      "learning_rate": 2.174255408211326e-06,
+      "loss": 1.3054,
+      "num_input_tokens_seen": 705480336,
+      "step": 17960
+    },
+    {
+      "epoch": 0.8707608254541269,
+      "grad_norm": 0.4567846357822418,
+      "learning_rate": 2.1582792142182117e-06,
+      "loss": 1.2335,
+      "num_input_tokens_seen": 705869500,
+      "step": 17970
+    },
+    {
+      "epoch": 0.8712453890743018,
+      "grad_norm": 0.39730075001716614,
+      "learning_rate": 2.1423592842111066e-06,
+      "loss": 1.2125,
+      "num_input_tokens_seen": 706254284,
+      "step": 17980
+    },
+    {
+      "epoch": 0.8717299526944766,
+      "grad_norm": 0.41465237736701965,
+      "learning_rate": 2.1264956574041513e-06,
+      "loss": 1.2287,
+      "num_input_tokens_seen": 706671516,
+      "step": 17990
+    },
+    {
+      "epoch": 0.8722145163146514,
+      "grad_norm": 0.3785446584224701,
+      "learning_rate": 2.1106883728728155e-06,
+      "loss": 1.2477,
+      "num_input_tokens_seen": 707047904,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8722145163146514,
+      "eval_loss": 1.3607689142227173,
+      "eval_runtime": 3.6634,
+      "eval_samples_per_second": 40.946,
+      "eval_steps_per_second": 5.186,
+      "num_input_tokens_seen": 707047904,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8726990799348262,
+      "grad_norm": 0.3779754042625427,
+      "learning_rate": 2.094937469553787e-06,
+      "loss": 1.2501,
+      "num_input_tokens_seen": 707448824,
+      "step": 18010
+    },
+    {
+      "epoch": 0.873183643555001,
+      "grad_norm": 0.40831905603408813,
+      "learning_rate": 2.079242986244867e-06,
+      "loss": 1.3242,
+      "num_input_tokens_seen": 707830972,
+      "step": 18020
+    },
+    {
+      "epoch": 0.8736682071751758,
+      "grad_norm": 0.4000668525695801,
+      "learning_rate": 2.063604961604884e-06,
+      "loss": 1.2698,
+      "num_input_tokens_seen": 708228628,
+      "step": 18030
+    },
+    {
+      "epoch": 0.8741527707953506,
+      "grad_norm": 0.4057333469390869,
+      "learning_rate": 2.0480234341535952e-06,
+      "loss": 1.3053,
+      "num_input_tokens_seen": 708602260,
+      "step": 18040
+    },
+    {
+      "epoch": 0.8746373344155254,
+      "grad_norm": 0.44601795077323914,
+      "learning_rate": 2.0324984422716046e-06,
+      "loss": 1.2743,
+      "num_input_tokens_seen": 709010156,
+      "step": 18050
+    },
+    {
+      "epoch": 0.8751218980357002,
+      "grad_norm": 0.4290080964565277,
+      "learning_rate": 2.0170300242002365e-06,
+      "loss": 1.2339,
+      "num_input_tokens_seen": 709382372,
+      "step": 18060
+    },
+    {
+      "epoch": 0.8756064616558751,
+      "grad_norm": 0.40737321972846985,
+      "learning_rate": 2.001618218041487e-06,
+      "loss": 1.2312,
+      "num_input_tokens_seen": 709780940,
+      "step": 18070
+    },
+    {
+      "epoch": 0.8760910252760499,
+      "grad_norm": 0.42168939113616943,
+      "learning_rate": 1.9862630617578816e-06,
+      "loss": 1.2517,
+      "num_input_tokens_seen": 710158796,
+      "step": 18080
+    },
+    {
+      "epoch": 0.8765755888962247,
+      "grad_norm": 0.4381299912929535,
+      "learning_rate": 1.9709645931724225e-06,
+      "loss": 1.2813,
+      "num_input_tokens_seen": 710561412,
+      "step": 18090
+    },
+    {
+      "epoch": 0.8770601525163995,
+      "grad_norm": 0.44368723034858704,
+      "learning_rate": 1.955722849968464e-06,
+      "loss": 1.3046,
+      "num_input_tokens_seen": 710960164,
+      "step": 18100
+    },
+    {
+      "epoch": 0.8775447161365743,
+      "grad_norm": 0.41541314125061035,
+      "learning_rate": 1.9405378696896487e-06,
+      "loss": 1.2637,
+      "num_input_tokens_seen": 711347332,
+      "step": 18110
+    },
+    {
+      "epoch": 0.878029279756749,
+      "grad_norm": 0.3992341160774231,
+      "learning_rate": 1.925409689739785e-06,
+      "loss": 1.2553,
+      "num_input_tokens_seen": 711716796,
+      "step": 18120
+    },
+    {
+      "epoch": 0.8785138433769238,
+      "grad_norm": 0.47702735662460327,
+      "learning_rate": 1.910338347382787e-06,
+      "loss": 1.2407,
+      "num_input_tokens_seen": 712071888,
+      "step": 18130
+    },
+    {
+      "epoch": 0.8789984069970986,
+      "grad_norm": 0.38431301712989807,
+      "learning_rate": 1.8953238797425442e-06,
+      "loss": 1.235,
+      "num_input_tokens_seen": 712445676,
+      "step": 18140
+    },
+    {
+      "epoch": 0.8794829706172734,
+      "grad_norm": 0.4030941128730774,
+      "learning_rate": 1.8803663238028707e-06,
+      "loss": 1.2388,
+      "num_input_tokens_seen": 712841240,
+      "step": 18150
+    },
+    {
+      "epoch": 0.8799675342374483,
+      "grad_norm": 0.3907192349433899,
+      "learning_rate": 1.8654657164073884e-06,
+      "loss": 1.2495,
+      "num_input_tokens_seen": 713199460,
+      "step": 18160
+    },
+    {
+      "epoch": 0.8804520978576231,
+      "grad_norm": 0.4238632917404175,
+      "learning_rate": 1.850622094259441e-06,
+      "loss": 1.329,
+      "num_input_tokens_seen": 713586584,
+      "step": 18170
+    },
+    {
+      "epoch": 0.8809366614777979,
+      "grad_norm": 0.43450093269348145,
+      "learning_rate": 1.835835493922014e-06,
+      "loss": 1.2552,
+      "num_input_tokens_seen": 713936208,
+      "step": 18180
+    },
+    {
+      "epoch": 0.8814212250979727,
+      "grad_norm": 0.36312198638916016,
+      "learning_rate": 1.821105951817617e-06,
+      "loss": 1.2135,
+      "num_input_tokens_seen": 714345664,
+      "step": 18190
+    },
+    {
+      "epoch": 0.8819057887181475,
+      "grad_norm": 0.3843260407447815,
+      "learning_rate": 1.8064335042282387e-06,
+      "loss": 1.2077,
+      "num_input_tokens_seen": 714749968,
+      "step": 18200
+    },
+    {
+      "epoch": 0.8823903523383223,
+      "grad_norm": 0.43907880783081055,
+      "learning_rate": 1.791818187295205e-06,
+      "loss": 1.2502,
+      "num_input_tokens_seen": 715139224,
+      "step": 18210
+    },
+    {
+      "epoch": 0.8828749159584971,
+      "grad_norm": 0.42311954498291016,
+      "learning_rate": 1.7772600370191433e-06,
+      "loss": 1.2607,
+      "num_input_tokens_seen": 715522824,
+      "step": 18220
+    },
+    {
+      "epoch": 0.8833594795786719,
+      "grad_norm": 0.41037461161613464,
+      "learning_rate": 1.762759089259844e-06,
+      "loss": 1.2311,
+      "num_input_tokens_seen": 715920216,
+      "step": 18230
+    },
+    {
+      "epoch": 0.8838440431988467,
+      "grad_norm": 0.4087775945663452,
+      "learning_rate": 1.7483153797362123e-06,
+      "loss": 1.2339,
+      "num_input_tokens_seen": 716298856,
+      "step": 18240
+    },
+    {
+      "epoch": 0.8843286068190216,
+      "grad_norm": 0.43544670939445496,
+      "learning_rate": 1.733928944026153e-06,
+      "loss": 1.258,
+      "num_input_tokens_seen": 716698420,
+      "step": 18250
+    },
+    {
+      "epoch": 0.8848131704391964,
+      "grad_norm": 0.4397813081741333,
+      "learning_rate": 1.7195998175665057e-06,
+      "loss": 1.2538,
+      "num_input_tokens_seen": 717109980,
+      "step": 18260
+    },
+    {
+      "epoch": 0.8852977340593712,
+      "grad_norm": 0.4167066812515259,
+      "learning_rate": 1.7053280356529283e-06,
+      "loss": 1.2635,
+      "num_input_tokens_seen": 717495908,
+      "step": 18270
+    },
+    {
+      "epoch": 0.885782297679546,
+      "grad_norm": 0.3873845040798187,
+      "learning_rate": 1.691113633439842e-06,
+      "loss": 1.2667,
+      "num_input_tokens_seen": 717871192,
+      "step": 18280
+    },
+    {
+      "epoch": 0.8862668612997208,
+      "grad_norm": 0.41001996397972107,
+      "learning_rate": 1.6769566459403224e-06,
+      "loss": 1.2516,
+      "num_input_tokens_seen": 718259268,
+      "step": 18290
+    },
+    {
+      "epoch": 0.8867514249198956,
+      "grad_norm": 0.454142689704895,
+      "learning_rate": 1.6628571080260196e-06,
+      "loss": 1.2868,
+      "num_input_tokens_seen": 718662496,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8872359885400704,
+      "grad_norm": 0.433021605014801,
+      "learning_rate": 1.6488150544270776e-06,
+      "loss": 1.2646,
+      "num_input_tokens_seen": 719043384,
+      "step": 18310
+    },
+    {
+      "epoch": 0.8877205521602451,
+      "grad_norm": 0.3922811448574066,
+      "learning_rate": 1.6348305197320417e-06,
+      "loss": 1.3003,
+      "num_input_tokens_seen": 719419832,
+      "step": 18320
+    },
+    {
+      "epoch": 0.8882051157804199,
+      "grad_norm": 0.41497138142585754,
+      "learning_rate": 1.6209035383877803e-06,
+      "loss": 1.2812,
+      "num_input_tokens_seen": 719822572,
+      "step": 18330
+    },
+    {
+      "epoch": 0.8886896794005948,
+      "grad_norm": 0.3779890239238739,
+      "learning_rate": 1.6070341446993875e-06,
+      "loss": 1.1811,
+      "num_input_tokens_seen": 720208548,
+      "step": 18340
+    },
+    {
+      "epoch": 0.8891742430207696,
+      "grad_norm": 0.4199633300304413,
+      "learning_rate": 1.5932223728301138e-06,
+      "loss": 1.2312,
+      "num_input_tokens_seen": 720585180,
+      "step": 18350
+    },
+    {
+      "epoch": 0.8896588066409444,
+      "grad_norm": 0.40874430537223816,
+      "learning_rate": 1.579468256801267e-06,
+      "loss": 1.1845,
+      "num_input_tokens_seen": 720970764,
+      "step": 18360
+    },
+    {
+      "epoch": 0.8901433702611192,
+      "grad_norm": 0.37856462597846985,
+      "learning_rate": 1.5657718304921492e-06,
+      "loss": 1.2665,
+      "num_input_tokens_seen": 721361356,
+      "step": 18370
+    },
+    {
+      "epoch": 0.890627933881294,
+      "grad_norm": 0.37885743379592896,
+      "learning_rate": 1.5521331276399488e-06,
+      "loss": 1.2479,
+      "num_input_tokens_seen": 721724800,
+      "step": 18380
+    },
+    {
+      "epoch": 0.8911124975014688,
+      "grad_norm": 0.40030255913734436,
+      "learning_rate": 1.538552181839678e-06,
+      "loss": 1.2585,
+      "num_input_tokens_seen": 722116248,
+      "step": 18390
+    },
+    {
+      "epoch": 0.8915970611216436,
+      "grad_norm": 0.43325039744377136,
+      "learning_rate": 1.525029026544067e-06,
+      "loss": 1.2446,
+      "num_input_tokens_seen": 722489540,
+      "step": 18400
+    },
+    {
+      "epoch": 0.8920816247418184,
+      "grad_norm": 0.3845292627811432,
+      "learning_rate": 1.511563695063517e-06,
+      "loss": 1.258,
+      "num_input_tokens_seen": 722864464,
+      "step": 18410
+    },
+    {
+      "epoch": 0.8925661883619933,
+      "grad_norm": 0.4011618196964264,
+      "learning_rate": 1.4981562205659772e-06,
+      "loss": 1.1975,
+      "num_input_tokens_seen": 723261460,
+      "step": 18420
+    },
+    {
+      "epoch": 0.8930507519821681,
+      "grad_norm": 0.4230853021144867,
+      "learning_rate": 1.4848066360768935e-06,
+      "loss": 1.2705,
+      "num_input_tokens_seen": 723627044,
+      "step": 18430
+    },
+    {
+      "epoch": 0.8935353156023429,
+      "grad_norm": 0.41304829716682434,
+      "learning_rate": 1.4715149744791156e-06,
+      "loss": 1.2206,
+      "num_input_tokens_seen": 724019484,
+      "step": 18440
+    },
+    {
+      "epoch": 0.8940198792225177,
+      "grad_norm": 0.40190738439559937,
+      "learning_rate": 1.458281268512815e-06,
+      "loss": 1.2595,
+      "num_input_tokens_seen": 724395804,
+      "step": 18450
+    },
+    {
+      "epoch": 0.8945044428426925,
+      "grad_norm": 0.40311187505722046,
+      "learning_rate": 1.445105550775408e-06,
+      "loss": 1.2737,
+      "num_input_tokens_seen": 724797948,
+      "step": 18460
+    },
+    {
+      "epoch": 0.8949890064628673,
+      "grad_norm": 0.42915475368499756,
+      "learning_rate": 1.431987853721467e-06,
+      "loss": 1.2391,
+      "num_input_tokens_seen": 725183560,
+      "step": 18470
+    },
+    {
+      "epoch": 0.8954735700830421,
+      "grad_norm": 0.406745046377182,
+      "learning_rate": 1.4189282096626593e-06,
+      "loss": 1.2801,
+      "num_input_tokens_seen": 725560960,
+      "step": 18480
+    },
+    {
+      "epoch": 0.8959581337032169,
+      "grad_norm": 0.43271973729133606,
+      "learning_rate": 1.405926650767639e-06,
+      "loss": 1.2748,
+      "num_input_tokens_seen": 725936388,
+      "step": 18490
+    },
+    {
+      "epoch": 0.8964426973233917,
+      "grad_norm": 0.41202259063720703,
+      "learning_rate": 1.3929832090620043e-06,
+      "loss": 1.2548,
+      "num_input_tokens_seen": 726331564,
+      "step": 18500
+    },
+    {
+      "epoch": 0.8969272609435666,
+      "grad_norm": 0.4043997526168823,
+      "learning_rate": 1.3800979164281775e-06,
+      "loss": 1.2473,
+      "num_input_tokens_seen": 726700276,
+      "step": 18510
+    },
+    {
+      "epoch": 0.8974118245637414,
+      "grad_norm": 0.38937658071517944,
+      "learning_rate": 1.3672708046053668e-06,
+      "loss": 1.2406,
+      "num_input_tokens_seen": 727097284,
+      "step": 18520
+    },
+    {
+      "epoch": 0.8978963881839161,
+      "grad_norm": 0.4417135715484619,
+      "learning_rate": 1.3545019051894537e-06,
+      "loss": 1.2141,
+      "num_input_tokens_seen": 727502228,
+      "step": 18530
+    },
+    {
+      "epoch": 0.8983809518040909,
+      "grad_norm": 0.393096923828125,
+      "learning_rate": 1.34179124963294e-06,
+      "loss": 1.2459,
+      "num_input_tokens_seen": 727852368,
+      "step": 18540
+    },
+    {
+      "epoch": 0.8988655154242657,
+      "grad_norm": 0.42874211072921753,
+      "learning_rate": 1.3291388692448503e-06,
+      "loss": 1.2397,
+      "num_input_tokens_seen": 728254728,
+      "step": 18550
+    },
+    {
+      "epoch": 0.8993500790444405,
+      "grad_norm": 0.40823593735694885,
+      "learning_rate": 1.3165447951906773e-06,
+      "loss": 1.2552,
+      "num_input_tokens_seen": 728648556,
+      "step": 18560
+    },
+    {
+      "epoch": 0.8998346426646153,
+      "grad_norm": 0.37802520394325256,
+      "learning_rate": 1.3040090584922921e-06,
+      "loss": 1.2287,
+      "num_input_tokens_seen": 729028520,
+      "step": 18570
+    },
+    {
+      "epoch": 0.9003192062847901,
+      "grad_norm": 0.39629054069519043,
+      "learning_rate": 1.29153169002785e-06,
+      "loss": 1.2521,
+      "num_input_tokens_seen": 729403340,
+      "step": 18580
+    },
+    {
+      "epoch": 0.9008037699049649,
+      "grad_norm": 0.38545021414756775,
+      "learning_rate": 1.2791127205317583e-06,
+      "loss": 1.2657,
+      "num_input_tokens_seen": 729803816,
+      "step": 18590
+    },
+    {
+      "epoch": 0.9012883335251398,
+      "grad_norm": 0.457553893327713,
+      "learning_rate": 1.2667521805945577e-06,
+      "loss": 1.2545,
+      "num_input_tokens_seen": 730207652,
+      "step": 18600
+    },
+    {
+      "epoch": 0.9017728971453146,
+      "grad_norm": 0.42532676458358765,
+      "learning_rate": 1.2544501006628768e-06,
+      "loss": 1.2469,
+      "num_input_tokens_seen": 730598644,
+      "step": 18610
+    },
+    {
+      "epoch": 0.9022574607654894,
+      "grad_norm": 0.4223385155200958,
+      "learning_rate": 1.2422065110393317e-06,
+      "loss": 1.2694,
+      "num_input_tokens_seen": 730997112,
+      "step": 18620
+    },
+    {
+      "epoch": 0.9027420243856642,
+      "grad_norm": 0.404644638299942,
+      "learning_rate": 1.2300214418824756e-06,
+      "loss": 1.2535,
+      "num_input_tokens_seen": 731392988,
+      "step": 18630
+    },
+    {
+      "epoch": 0.903226588005839,
+      "grad_norm": 0.39279234409332275,
+      "learning_rate": 1.2178949232067056e-06,
+      "loss": 1.2253,
+      "num_input_tokens_seen": 731765160,
+      "step": 18640
+    },
+    {
+      "epoch": 0.9037111516260138,
+      "grad_norm": 0.41039055585861206,
+      "learning_rate": 1.2058269848822052e-06,
+      "loss": 1.2272,
+      "num_input_tokens_seen": 732135724,
+      "step": 18650
+    },
+    {
+      "epoch": 0.9041957152461886,
+      "grad_norm": 0.38321247696876526,
+      "learning_rate": 1.1938176566348518e-06,
+      "loss": 1.2478,
+      "num_input_tokens_seen": 732558172,
+      "step": 18660
+    },
+    {
+      "epoch": 0.9046802788663634,
+      "grad_norm": 0.47206875681877136,
+      "learning_rate": 1.1818669680461636e-06,
+      "loss": 1.2375,
+      "num_input_tokens_seen": 732961984,
+      "step": 18670
+    },
+    {
+      "epoch": 0.9051648424865382,
+      "grad_norm": 0.3947557210922241,
+      "learning_rate": 1.169974948553207e-06,
+      "loss": 1.2745,
+      "num_input_tokens_seen": 733342248,
+      "step": 18680
+    },
+    {
+      "epoch": 0.9056494061067131,
+      "grad_norm": 0.3839742839336395,
+      "learning_rate": 1.1581416274485447e-06,
+      "loss": 1.2215,
+      "num_input_tokens_seen": 733722272,
+      "step": 18690
+    },
+    {
+      "epoch": 0.9061339697268879,
+      "grad_norm": 0.4032100737094879,
+      "learning_rate": 1.146367033880147e-06,
+      "loss": 1.2589,
+      "num_input_tokens_seen": 734138572,
+      "step": 18700
+    },
+    {
+      "epoch": 0.9066185333470627,
+      "grad_norm": 0.42585238814353943,
+      "learning_rate": 1.1346511968513218e-06,
+      "loss": 1.2102,
+      "num_input_tokens_seen": 734513304,
+      "step": 18710
+    },
+    {
+      "epoch": 0.9071030969672375,
+      "grad_norm": 0.4173504412174225,
+      "learning_rate": 1.122994145220657e-06,
+      "loss": 1.2864,
+      "num_input_tokens_seen": 734919300,
+      "step": 18720
+    },
+    {
+      "epoch": 0.9075876605874122,
+      "grad_norm": 0.4611232876777649,
+      "learning_rate": 1.1113959077019315e-06,
+      "loss": 1.2136,
+      "num_input_tokens_seen": 735332452,
+      "step": 18730
+    },
+    {
+      "epoch": 0.908072224207587,
+      "grad_norm": 0.410653680562973,
+      "learning_rate": 1.0998565128640615e-06,
+      "loss": 1.2745,
+      "num_input_tokens_seen": 735714372,
+      "step": 18740
+    },
+    {
+      "epoch": 0.9085567878277618,
+      "grad_norm": 0.39335983991622925,
+      "learning_rate": 1.0883759891310047e-06,
+      "loss": 1.2148,
+      "num_input_tokens_seen": 736147668,
+      "step": 18750
+    },
+    {
+      "epoch": 0.9090413514479366,
+      "grad_norm": 0.4009684920310974,
+      "learning_rate": 1.0769543647817293e-06,
+      "loss": 1.255,
+      "num_input_tokens_seen": 736533020,
+      "step": 18760
+    },
+    {
+      "epoch": 0.9095259150681114,
+      "grad_norm": 0.39060232043266296,
+      "learning_rate": 1.0655916679501026e-06,
+      "loss": 1.28,
+      "num_input_tokens_seen": 736930600,
+      "step": 18770
+    },
+    {
+      "epoch": 0.9100104786882863,
+      "grad_norm": 0.3796159327030182,
+      "learning_rate": 1.0542879266248501e-06,
+      "loss": 1.295,
+      "num_input_tokens_seen": 737329028,
+      "step": 18780
+    },
+    {
+      "epoch": 0.9104950423084611,
+      "grad_norm": 0.41085562109947205,
+      "learning_rate": 1.0430431686494768e-06,
+      "loss": 1.2471,
+      "num_input_tokens_seen": 737729856,
+      "step": 18790
+    },
+    {
+      "epoch": 0.9109796059286359,
+      "grad_norm": 0.38016316294670105,
+      "learning_rate": 1.031857421722196e-06,
+      "loss": 1.2601,
+      "num_input_tokens_seen": 738134952,
+      "step": 18800
+    },
+    {
+      "epoch": 0.9114641695488107,
+      "grad_norm": 0.42966845631599426,
+      "learning_rate": 1.0207307133958676e-06,
+      "loss": 1.2398,
+      "num_input_tokens_seen": 738559280,
+      "step": 18810
+    },
+    {
+      "epoch": 0.9119487331689855,
+      "grad_norm": 0.39582687616348267,
+      "learning_rate": 1.0096630710779264e-06,
+      "loss": 1.2218,
+      "num_input_tokens_seen": 738942248,
+      "step": 18820
+    },
+    {
+      "epoch": 0.9124332967891603,
+      "grad_norm": 0.3957637846469879,
+      "learning_rate": 9.98654522030318e-07,
+      "loss": 1.2582,
+      "num_input_tokens_seen": 739314768,
+      "step": 18830
+    },
+    {
+      "epoch": 0.9129178604093351,
+      "grad_norm": 0.4404972493648529,
+      "learning_rate": 9.877050933694176e-07,
+      "loss": 1.2526,
+      "num_input_tokens_seen": 739693344,
+      "step": 18840
+    },
+    {
+      "epoch": 0.9134024240295099,
+      "grad_norm": 0.4136838912963867,
+      "learning_rate": 9.768148120659903e-07,
+      "loss": 1.2503,
+      "num_input_tokens_seen": 740095948,
+      "step": 18850
+    },
+    {
+      "epoch": 0.9138869876496847,
+      "grad_norm": 0.4641477167606354,
+      "learning_rate": 9.65983704945092e-07,
+      "loss": 1.2321,
+      "num_input_tokens_seen": 740509384,
+      "step": 18860
+    },
+    {
+      "epoch": 0.9143715512698596,
+      "grad_norm": 0.37187913060188293,
+      "learning_rate": 9.55211798686037e-07,
+      "loss": 1.2635,
+      "num_input_tokens_seen": 740899344,
+      "step": 18870
+    },
+    {
+      "epoch": 0.9148561148900344,
+      "grad_norm": 0.4150676131248474,
+      "learning_rate": 9.444991198223008e-07,
+      "loss": 1.2356,
+      "num_input_tokens_seen": 741314140,
+      "step": 18880
+    },
+    {
+      "epoch": 0.9153406785102092,
+      "grad_norm": 0.413786917924881,
+      "learning_rate": 9.338456947414837e-07,
+      "loss": 1.2555,
+      "num_input_tokens_seen": 741697844,
+      "step": 18890
+    },
+    {
+      "epoch": 0.915825242130384,
+      "grad_norm": 0.49498802423477173,
+      "learning_rate": 9.23251549685214e-07,
+      "loss": 1.2668,
+      "num_input_tokens_seen": 742079296,
+      "step": 18900
+    },
+    {
+      "epoch": 0.9163098057505588,
+      "grad_norm": 0.39548417925834656,
+      "learning_rate": 9.127167107491174e-07,
+      "loss": 1.2235,
+      "num_input_tokens_seen": 742477492,
+      "step": 18910
+    },
+    {
+      "epoch": 0.9167943693707336,
+      "grad_norm": 0.3953016698360443,
+      "learning_rate": 9.022412038827227e-07,
+      "loss": 1.2328,
+      "num_input_tokens_seen": 742879824,
+      "step": 18920
+    },
+    {
+      "epoch": 0.9172789329909083,
+      "grad_norm": 0.422567754983902,
+      "learning_rate": 8.918250548894225e-07,
+      "loss": 1.2207,
+      "num_input_tokens_seen": 743283828,
+      "step": 18930
+    },
+    {
+      "epoch": 0.9177634966110831,
+      "grad_norm": 0.4394066333770752,
+      "learning_rate": 8.814682894263904e-07,
+      "loss": 1.2525,
+      "num_input_tokens_seen": 743671012,
+      "step": 18940
+    },
+    {
+      "epoch": 0.9182480602312579,
+      "grad_norm": 0.3766193389892578,
+      "learning_rate": 8.711709330045309e-07,
+      "loss": 1.2441,
+      "num_input_tokens_seen": 744042668,
+      "step": 18950
+    },
+    {
+      "epoch": 0.9187326238514328,
+      "grad_norm": 0.4037720561027527,
+      "learning_rate": 8.609330109884045e-07,
+      "loss": 1.3088,
+      "num_input_tokens_seen": 744414936,
+      "step": 18960
+    },
+    {
+      "epoch": 0.9192171874716076,
+      "grad_norm": 0.38858434557914734,
+      "learning_rate": 8.507545485961804e-07,
+      "loss": 1.2561,
+      "num_input_tokens_seen": 744849212,
+      "step": 18970
+    },
+    {
+      "epoch": 0.9197017510917824,
+      "grad_norm": 0.44780433177948,
+      "learning_rate": 8.406355708995672e-07,
+      "loss": 1.2863,
+      "num_input_tokens_seen": 745242736,
+      "step": 18980
+    },
+    {
+      "epoch": 0.9201863147119572,
+      "grad_norm": 0.4300279915332794,
+      "learning_rate": 8.305761028237353e-07,
+      "loss": 1.2229,
+      "num_input_tokens_seen": 745670052,
+      "step": 18990
+    },
+    {
+      "epoch": 0.920670878332132,
+      "grad_norm": 0.4177449345588684,
+      "learning_rate": 8.205761691472913e-07,
+      "loss": 1.2178,
+      "num_input_tokens_seen": 746084620,
+      "step": 19000
+    },
+    {
+      "epoch": 0.9211554419523068,
+      "grad_norm": 0.38727110624313354,
+      "learning_rate": 8.106357945021765e-07,
+      "loss": 1.2613,
+      "num_input_tokens_seen": 746466352,
+      "step": 19010
+    },
+    {
+      "epoch": 0.9216400055724816,
+      "grad_norm": 0.4461565315723419,
+      "learning_rate": 8.007550033736405e-07,
+      "loss": 1.2423,
+      "num_input_tokens_seen": 746860664,
+      "step": 19020
+    },
+    {
+      "epoch": 0.9221245691926564,
+      "grad_norm": 0.4176614582538605,
+      "learning_rate": 7.909338201001564e-07,
+      "loss": 1.2664,
+      "num_input_tokens_seen": 747238188,
+      "step": 19030
+    },
+    {
+      "epoch": 0.9226091328128312,
+      "grad_norm": 0.4129350781440735,
+      "learning_rate": 7.811722688733786e-07,
+      "loss": 1.2481,
+      "num_input_tokens_seen": 747640208,
+      "step": 19040
+    },
+    {
+      "epoch": 0.9230936964330061,
+      "grad_norm": 0.40543922781944275,
+      "learning_rate": 7.714703737380674e-07,
+      "loss": 1.2575,
+      "num_input_tokens_seen": 748033864,
+      "step": 19050
+    },
+    {
+      "epoch": 0.9235782600531809,
+      "grad_norm": 0.4024973511695862,
+      "learning_rate": 7.618281585920456e-07,
+      "loss": 1.2477,
+      "num_input_tokens_seen": 748427680,
+      "step": 19060
+    },
+    {
+      "epoch": 0.9240628236733557,
+      "grad_norm": 0.3884994387626648,
+      "learning_rate": 7.522456471861172e-07,
+      "loss": 1.2325,
+      "num_input_tokens_seen": 748812864,
+      "step": 19070
+    },
+    {
+      "epoch": 0.9245473872935305,
+      "grad_norm": 0.38652002811431885,
+      "learning_rate": 7.427228631240457e-07,
+      "loss": 1.2592,
+      "num_input_tokens_seen": 749169748,
+      "step": 19080
+    },
+    {
+      "epoch": 0.9250319509137053,
+      "grad_norm": 0.4229782521724701,
+      "learning_rate": 7.33259829862451e-07,
+      "loss": 1.2767,
+      "num_input_tokens_seen": 749552652,
+      "step": 19090
+    },
+    {
+      "epoch": 0.9255165145338801,
+      "grad_norm": 0.3931877315044403,
+      "learning_rate": 7.238565707107875e-07,
+      "loss": 1.2596,
+      "num_input_tokens_seen": 749938204,
+      "step": 19100
+    },
+    {
+      "epoch": 0.9260010781540549,
+      "grad_norm": 0.42200735211372375,
+      "learning_rate": 7.145131088312745e-07,
+      "loss": 1.2536,
+      "num_input_tokens_seen": 750322924,
+      "step": 19110
+    },
+    {
+      "epoch": 0.9264856417742297,
+      "grad_norm": 0.40737777948379517,
+      "learning_rate": 7.052294672388271e-07,
+      "loss": 1.2143,
+      "num_input_tokens_seen": 750735180,
+      "step": 19120
+    },
+    {
+      "epoch": 0.9269702053944046,
+      "grad_norm": 0.41996297240257263,
+      "learning_rate": 6.960056688010197e-07,
+      "loss": 1.2552,
+      "num_input_tokens_seen": 751134712,
+      "step": 19130
+    },
+    {
+      "epoch": 0.9274547690145793,
+      "grad_norm": 0.40312203764915466,
+      "learning_rate": 6.868417362380114e-07,
+      "loss": 1.2409,
+      "num_input_tokens_seen": 751509996,
+      "step": 19140
+    },
+    {
+      "epoch": 0.9279393326347541,
+      "grad_norm": 0.3921646773815155,
+      "learning_rate": 6.777376921225125e-07,
+      "loss": 1.2459,
+      "num_input_tokens_seen": 751937076,
+      "step": 19150
+    },
+    {
+      "epoch": 0.9284238962549289,
+      "grad_norm": 0.4335014522075653,
+      "learning_rate": 6.686935588797072e-07,
+      "loss": 1.2383,
+      "num_input_tokens_seen": 752333820,
+      "step": 19160
+    },
+    {
+      "epoch": 0.9289084598751037,
+      "grad_norm": 0.42503783106803894,
+      "learning_rate": 6.597093587872055e-07,
+      "loss": 1.2457,
+      "num_input_tokens_seen": 752732964,
+      "step": 19170
+    },
+    {
+      "epoch": 0.9293930234952785,
+      "grad_norm": 0.4474683701992035,
+      "learning_rate": 6.507851139749888e-07,
+      "loss": 1.2576,
+      "num_input_tokens_seen": 753114736,
+      "step": 19180
+    },
+    {
+      "epoch": 0.9298775871154533,
+      "grad_norm": 0.40870001912117004,
+      "learning_rate": 6.419208464253618e-07,
+      "loss": 1.2785,
+      "num_input_tokens_seen": 753496580,
+      "step": 19190
+    },
+    {
+      "epoch": 0.9303621507356281,
+      "grad_norm": 0.4168875217437744,
+      "learning_rate": 6.331165779728865e-07,
+      "loss": 1.2859,
+      "num_input_tokens_seen": 753875144,
+      "step": 19200
+    },
+    {
+      "epoch": 0.9308467143558029,
+      "grad_norm": 0.3911028504371643,
+      "learning_rate": 6.243723303043403e-07,
+      "loss": 1.2791,
+      "num_input_tokens_seen": 754260416,
+      "step": 19210
+    },
+    {
+      "epoch": 0.9313312779759778,
+      "grad_norm": 0.4180554151535034,
+      "learning_rate": 6.156881249586493e-07,
+      "loss": 1.2427,
+      "num_input_tokens_seen": 754653980,
+      "step": 19220
+    },
+    {
+      "epoch": 0.9318158415961526,
+      "grad_norm": 0.4038565158843994,
+      "learning_rate": 6.070639833268471e-07,
+      "loss": 1.2551,
+      "num_input_tokens_seen": 755023452,
+      "step": 19230
+    },
+    {
+      "epoch": 0.9323004052163274,
+      "grad_norm": 0.433095246553421,
+      "learning_rate": 5.984999266520214e-07,
+      "loss": 1.2713,
+      "num_input_tokens_seen": 755440128,
+      "step": 19240
+    },
+    {
+      "epoch": 0.9327849688365022,
+      "grad_norm": 0.4306170642375946,
+      "learning_rate": 5.899959760292478e-07,
+      "loss": 1.2995,
+      "num_input_tokens_seen": 755840776,
+      "step": 19250
+    },
+    {
+      "epoch": 0.933269532456677,
+      "grad_norm": 0.4218454658985138,
+      "learning_rate": 5.815521524055623e-07,
+      "loss": 1.2378,
+      "num_input_tokens_seen": 756263932,
+      "step": 19260
+    },
+    {
+      "epoch": 0.9337540960768518,
+      "grad_norm": 0.3932402729988098,
+      "learning_rate": 5.731684765798772e-07,
+      "loss": 1.2506,
+      "num_input_tokens_seen": 756648804,
+      "step": 19270
+    },
+    {
+      "epoch": 0.9342386596970266,
+      "grad_norm": 0.4214298129081726,
+      "learning_rate": 5.648449692029656e-07,
+      "loss": 1.2264,
+      "num_input_tokens_seen": 757063844,
+      "step": 19280
+    },
+    {
+      "epoch": 0.9347232233172014,
+      "grad_norm": 0.4255639612674713,
+      "learning_rate": 5.565816507773797e-07,
+      "loss": 1.2217,
+      "num_input_tokens_seen": 757437036,
+      "step": 19290
+    },
+    {
+      "epoch": 0.9352077869373762,
+      "grad_norm": 0.3967702090740204,
+      "learning_rate": 5.483785416574239e-07,
+      "loss": 1.3114,
+      "num_input_tokens_seen": 757815468,
+      "step": 19300
+    },
+    {
+      "epoch": 0.9356923505575511,
+      "grad_norm": 0.4092276096343994,
+      "learning_rate": 5.402356620490878e-07,
+      "loss": 1.2753,
+      "num_input_tokens_seen": 758208344,
+      "step": 19310
+    },
+    {
+      "epoch": 0.9361769141777259,
+      "grad_norm": 0.40975773334503174,
+      "learning_rate": 5.321530320100076e-07,
+      "loss": 1.2785,
+      "num_input_tokens_seen": 758618456,
+      "step": 19320
+    },
+    {
+      "epoch": 0.9366614777979007,
+      "grad_norm": 0.44307002425193787,
+      "learning_rate": 5.241306714494021e-07,
+      "loss": 1.2863,
+      "num_input_tokens_seen": 759009464,
+      "step": 19330
+    },
+    {
+      "epoch": 0.9371460414180754,
+      "grad_norm": 0.39563581347465515,
+      "learning_rate": 5.161686001280503e-07,
+      "loss": 1.2356,
+      "num_input_tokens_seen": 759418492,
+      "step": 19340
+    },
+    {
+      "epoch": 0.9376306050382502,
+      "grad_norm": 0.4311593472957611,
+      "learning_rate": 5.082668376582111e-07,
+      "loss": 1.2691,
+      "num_input_tokens_seen": 759767128,
+      "step": 19350
+    },
+    {
+      "epoch": 0.938115168658425,
+      "grad_norm": 0.3757501542568207,
+      "learning_rate": 5.00425403503596e-07,
+      "loss": 1.2662,
+      "num_input_tokens_seen": 760168828,
+      "step": 19360
+    },
+    {
+      "epoch": 0.9385997322785998,
+      "grad_norm": 0.4447803795337677,
+      "learning_rate": 4.926443169793154e-07,
+      "loss": 1.3027,
+      "num_input_tokens_seen": 760588676,
+      "step": 19370
+    },
+    {
+      "epoch": 0.9390842958987746,
+      "grad_norm": 0.4607948958873749,
+      "learning_rate": 4.849235972518295e-07,
+      "loss": 1.2566,
+      "num_input_tokens_seen": 760956864,
+      "step": 19380
+    },
+    {
+      "epoch": 0.9395688595189494,
+      "grad_norm": 0.3940238058567047,
+      "learning_rate": 4.772632633389063e-07,
+      "loss": 1.2854,
+      "num_input_tokens_seen": 761318912,
+      "step": 19390
+    },
+    {
+      "epoch": 0.9400534231391243,
+      "grad_norm": 0.4033295810222626,
+      "learning_rate": 4.6966333410956023e-07,
+      "loss": 1.2041,
+      "num_input_tokens_seen": 761704264,
+      "step": 19400
+    },
+    {
+      "epoch": 0.9405379867592991,
+      "grad_norm": 0.4064282178878784,
+      "learning_rate": 4.621238282840279e-07,
+      "loss": 1.24,
+      "num_input_tokens_seen": 762085876,
+      "step": 19410
+    },
+    {
+      "epoch": 0.9410225503794739,
+      "grad_norm": 0.39992567896842957,
+      "learning_rate": 4.546447644337065e-07,
+      "loss": 1.2397,
+      "num_input_tokens_seen": 762474368,
+      "step": 19420
+    },
+    {
+      "epoch": 0.9415071139996487,
+      "grad_norm": 0.4709777235984802,
+      "learning_rate": 4.4722616098110684e-07,
+      "loss": 1.2858,
+      "num_input_tokens_seen": 762874704,
+      "step": 19430
+    },
+    {
+      "epoch": 0.9419916776198235,
+      "grad_norm": 0.4245252311229706,
+      "learning_rate": 4.3986803619981973e-07,
+      "loss": 1.2707,
+      "num_input_tokens_seen": 763268244,
+      "step": 19440
+    },
+    {
+      "epoch": 0.9424762412399983,
+      "grad_norm": 0.44352298974990845,
+      "learning_rate": 4.325704082144666e-07,
+      "loss": 1.2645,
+      "num_input_tokens_seen": 763643580,
+      "step": 19450
+    },
+    {
+      "epoch": 0.9429608048601731,
+      "grad_norm": 0.39565107226371765,
+      "learning_rate": 4.2533329500063776e-07,
+      "loss": 1.2278,
+      "num_input_tokens_seen": 764054380,
+      "step": 19460
+    },
+    {
+      "epoch": 0.9434453684803479,
+      "grad_norm": 0.3986281454563141,
+      "learning_rate": 4.181567143848819e-07,
+      "loss": 1.2794,
+      "num_input_tokens_seen": 764458228,
+      "step": 19470
+    },
+    {
+      "epoch": 0.9439299321005227,
+      "grad_norm": 0.4145820140838623,
+      "learning_rate": 4.1104068404462514e-07,
+      "loss": 1.242,
+      "num_input_tokens_seen": 764816404,
+      "step": 19480
+    },
+    {
+      "epoch": 0.9444144957206976,
+      "grad_norm": 0.42509499192237854,
+      "learning_rate": 4.039852215081602e-07,
+      "loss": 1.2862,
+      "num_input_tokens_seen": 765190032,
+      "step": 19490
+    },
+    {
+      "epoch": 0.9448990593408724,
+      "grad_norm": 0.38184764981269836,
+      "learning_rate": 3.969903441545825e-07,
+      "loss": 1.2225,
+      "num_input_tokens_seen": 765565508,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9453836229610472,
+      "grad_norm": 0.460821270942688,
+      "learning_rate": 3.9005606921375117e-07,
+      "loss": 1.2652,
+      "num_input_tokens_seen": 765974548,
+      "step": 19510
+    },
+    {
+      "epoch": 0.945868186581222,
+      "grad_norm": 0.4286702573299408,
+      "learning_rate": 3.831824137662504e-07,
+      "loss": 1.2459,
+      "num_input_tokens_seen": 766399956,
+      "step": 19520
+    },
+    {
+      "epoch": 0.9463527502013968,
+      "grad_norm": 0.38720691204071045,
+      "learning_rate": 3.7636939474334775e-07,
+      "loss": 1.2141,
+      "num_input_tokens_seen": 766793544,
+      "step": 19530
+    },
+    {
+      "epoch": 0.9468373138215715,
+      "grad_norm": 0.3998976945877075,
+      "learning_rate": 3.696170289269524e-07,
+      "loss": 1.2611,
+      "num_input_tokens_seen": 767186876,
+      "step": 19540
+    },
+    {
+      "epoch": 0.9473218774417463,
+      "grad_norm": 0.4127962291240692,
+      "learning_rate": 3.6292533294955966e-07,
+      "loss": 1.2499,
+      "num_input_tokens_seen": 767600456,
+      "step": 19550
+    },
+    {
+      "epoch": 0.9478064410619211,
+      "grad_norm": 0.4086759388446808,
+      "learning_rate": 3.5629432329424006e-07,
+      "loss": 1.2433,
+      "num_input_tokens_seen": 768032396,
+      "step": 19560
+    },
+    {
+      "epoch": 0.9482910046820959,
+      "grad_norm": 0.3765683174133301,
+      "learning_rate": 3.49724016294567e-07,
+      "loss": 1.2663,
+      "num_input_tokens_seen": 768403312,
+      "step": 19570
+    },
+    {
+      "epoch": 0.9487755683022708,
+      "grad_norm": 0.47519636154174805,
+      "learning_rate": 3.432144281345973e-07,
+      "loss": 1.2692,
+      "num_input_tokens_seen": 768763468,
+      "step": 19580
+    },
+    {
+      "epoch": 0.9492601319224456,
+      "grad_norm": 0.41748225688934326,
+      "learning_rate": 3.3676557484881855e-07,
+      "loss": 1.2243,
+      "num_input_tokens_seen": 769110720,
+      "step": 19590
+    },
+    {
+      "epoch": 0.9497446955426204,
+      "grad_norm": 0.4386734962463379,
+      "learning_rate": 3.30377472322127e-07,
+      "loss": 1.2672,
+      "num_input_tokens_seen": 769514668,
+      "step": 19600
+    },
+    {
+      "epoch": 0.9502292591627952,
+      "grad_norm": 0.42562198638916016,
+      "learning_rate": 3.2405013628976077e-07,
+      "loss": 1.2504,
+      "num_input_tokens_seen": 769901384,
+      "step": 19610
+    },
+    {
+      "epoch": 0.95071382278297,
+      "grad_norm": 0.4268267750740051,
+      "learning_rate": 3.1778358233729154e-07,
+      "loss": 1.2657,
+      "num_input_tokens_seen": 770308688,
+      "step": 19620
+    },
+    {
+      "epoch": 0.9511983864031448,
+      "grad_norm": 0.4298213720321655,
+      "learning_rate": 3.1157782590056637e-07,
+      "loss": 1.1705,
+      "num_input_tokens_seen": 770709140,
+      "step": 19630
+    },
+    {
+      "epoch": 0.9516829500233196,
+      "grad_norm": 0.4501436650753021,
+      "learning_rate": 3.054328822656688e-07,
+      "loss": 1.2964,
+      "num_input_tokens_seen": 771100080,
+      "step": 19640
+    },
+    {
+      "epoch": 0.9521675136434944,
+      "grad_norm": 0.43033567070961,
+      "learning_rate": 2.9934876656890207e-07,
+      "loss": 1.2118,
+      "num_input_tokens_seen": 771491372,
+      "step": 19650
+    },
+    {
+      "epoch": 0.9526520772636692,
+      "grad_norm": 0.4044438302516937,
+      "learning_rate": 2.933254937967228e-07,
+      "loss": 1.2532,
+      "num_input_tokens_seen": 771854668,
+      "step": 19660
+    },
+    {
+      "epoch": 0.9531366408838441,
+      "grad_norm": 0.38256967067718506,
+      "learning_rate": 2.8736307878572957e-07,
+      "loss": 1.2422,
+      "num_input_tokens_seen": 772233856,
+      "step": 19670
+    },
+    {
+      "epoch": 0.9536212045040189,
+      "grad_norm": 0.4290720224380493,
+      "learning_rate": 2.8146153622260475e-07,
+      "loss": 1.2342,
+      "num_input_tokens_seen": 772629696,
+      "step": 19680
+    },
+    {
+      "epoch": 0.9541057681241937,
+      "grad_norm": 0.3771139681339264,
+      "learning_rate": 2.7562088064410354e-07,
+      "loss": 1.2594,
+      "num_input_tokens_seen": 773032252,
+      "step": 19690
+    },
+    {
+      "epoch": 0.9545903317443685,
+      "grad_norm": 0.4023270606994629,
+      "learning_rate": 2.6984112643698434e-07,
+      "loss": 1.2354,
+      "num_input_tokens_seen": 773437064,
+      "step": 19700
+    },
+    {
+      "epoch": 0.9550748953645433,
+      "grad_norm": 0.42096689343452454,
+      "learning_rate": 2.641222878380117e-07,
+      "loss": 1.2892,
+      "num_input_tokens_seen": 773816632,
+      "step": 19710
+    },
+    {
+      "epoch": 0.9555594589847181,
+      "grad_norm": 0.3777530789375305,
+      "learning_rate": 2.5846437893388977e-07,
+      "loss": 1.2892,
+      "num_input_tokens_seen": 774228364,
+      "step": 19720
+    },
+    {
+      "epoch": 0.9560440226048929,
+      "grad_norm": 0.43396878242492676,
+      "learning_rate": 2.52867413661248e-07,
+      "loss": 1.2451,
+      "num_input_tokens_seen": 774601636,
+      "step": 19730
+    },
+    {
+      "epoch": 0.9565285862250676,
+      "grad_norm": 0.39147117733955383,
+      "learning_rate": 2.4733140580658897e-07,
+      "loss": 1.2891,
+      "num_input_tokens_seen": 774994136,
+      "step": 19740
+    },
+    {
+      "epoch": 0.9570131498452424,
+      "grad_norm": 0.43016576766967773,
+      "learning_rate": 2.4185636900627417e-07,
+      "loss": 1.2161,
+      "num_input_tokens_seen": 775389680,
+      "step": 19750
+    },
+    {
+      "epoch": 0.9574977134654173,
+      "grad_norm": 0.4033438265323639,
+      "learning_rate": 2.3644231674647688e-07,
+      "loss": 1.2664,
+      "num_input_tokens_seen": 775778568,
+      "step": 19760
+    },
+    {
+      "epoch": 0.9579822770855921,
+      "grad_norm": 0.4349776804447174,
+      "learning_rate": 2.3108926236314887e-07,
+      "loss": 1.2115,
+      "num_input_tokens_seen": 776157176,
+      "step": 19770
+    },
+    {
+      "epoch": 0.9584668407057669,
+      "grad_norm": 0.4078141748905182,
+      "learning_rate": 2.2579721904199824e-07,
+      "loss": 1.2701,
+      "num_input_tokens_seen": 776547680,
+      "step": 19780
+    },
+    {
+      "epoch": 0.9589514043259417,
+      "grad_norm": 0.4067167639732361,
+      "learning_rate": 2.2056619981844495e-07,
+      "loss": 1.2334,
+      "num_input_tokens_seen": 776966012,
+      "step": 19790
+    },
+    {
+      "epoch": 0.9594359679461165,
+      "grad_norm": 0.41261714696884155,
+      "learning_rate": 2.153962175775931e-07,
+      "loss": 1.2612,
+      "num_input_tokens_seen": 777377344,
+      "step": 19800
+    },
+    {
+      "epoch": 0.9599205315662913,
+      "grad_norm": 0.3794651925563812,
+      "learning_rate": 2.1028728505420313e-07,
+      "loss": 1.2746,
+      "num_input_tokens_seen": 777740524,
+      "step": 19810
+    },
+    {
+      "epoch": 0.9604050951864661,
+      "grad_norm": 0.3992972671985626,
+      "learning_rate": 2.0523941483265586e-07,
+      "loss": 1.2302,
+      "num_input_tokens_seen": 778140684,
+      "step": 19820
+    },
+    {
+      "epoch": 0.9608896588066409,
+      "grad_norm": 0.4393966495990753,
+      "learning_rate": 2.0025261934691897e-07,
+      "loss": 1.2686,
+      "num_input_tokens_seen": 778528860,
+      "step": 19830
+    },
+    {
+      "epoch": 0.9613742224268158,
+      "grad_norm": 0.4146067798137665,
+      "learning_rate": 1.9532691088053056e-07,
+      "loss": 1.2656,
+      "num_input_tokens_seen": 778900016,
+      "step": 19840
+    },
+    {
+      "epoch": 0.9618587860469906,
+      "grad_norm": 0.43496647477149963,
+      "learning_rate": 1.904623015665463e-07,
+      "loss": 1.2421,
+      "num_input_tokens_seen": 779282552,
+      "step": 19850
+    },
+    {
+      "epoch": 0.9623433496671654,
+      "grad_norm": 0.43942153453826904,
+      "learning_rate": 1.8565880338752838e-07,
+      "loss": 1.2166,
+      "num_input_tokens_seen": 779634656,
+      "step": 19860
+    },
+    {
+      "epoch": 0.9628279132873402,
+      "grad_norm": 0.41181692481040955,
+      "learning_rate": 1.8091642817550935e-07,
+      "loss": 1.2809,
+      "num_input_tokens_seen": 780011584,
+      "step": 19870
+    },
+    {
+      "epoch": 0.963312476907515,
+      "grad_norm": 0.44306066632270813,
+      "learning_rate": 1.762351876119589e-07,
+      "loss": 1.2406,
+      "num_input_tokens_seen": 780398180,
+      "step": 19880
+    },
+    {
+      "epoch": 0.9637970405276898,
+      "grad_norm": 0.38546454906463623,
+      "learning_rate": 1.7161509322776437e-07,
+      "loss": 1.2775,
+      "num_input_tokens_seen": 780787352,
+      "step": 19890
+    },
+    {
+      "epoch": 0.9642816041478646,
+      "grad_norm": 0.425279825925827,
+      "learning_rate": 1.6705615640319472e-07,
+      "loss": 1.2662,
+      "num_input_tokens_seen": 781181080,
+      "step": 19900
+    },
+    {
+      "epoch": 0.9647661677680394,
+      "grad_norm": 0.4292610287666321,
+      "learning_rate": 1.625583883678755e-07,
+      "loss": 1.2033,
+      "num_input_tokens_seen": 781590820,
+      "step": 19910
+    },
+    {
+      "epoch": 0.9652507313882142,
+      "grad_norm": 0.44324782490730286,
+      "learning_rate": 1.5812180020075563e-07,
+      "loss": 1.2729,
+      "num_input_tokens_seen": 782000476,
+      "step": 19920
+    },
+    {
+      "epoch": 0.9657352950083891,
+      "grad_norm": 0.4017219543457031,
+      "learning_rate": 1.537464028300961e-07,
+      "loss": 1.2573,
+      "num_input_tokens_seen": 782384008,
+      "step": 19930
+    },
+    {
+      "epoch": 0.9662198586285639,
+      "grad_norm": 0.4076748490333557,
+      "learning_rate": 1.4943220703342031e-07,
+      "loss": 1.2051,
+      "num_input_tokens_seen": 782770540,
+      "step": 19940
+    },
+    {
+      "epoch": 0.9667044222487386,
+      "grad_norm": 0.4016420841217041,
+      "learning_rate": 1.4517922343750546e-07,
+      "loss": 1.2357,
+      "num_input_tokens_seen": 783161236,
+      "step": 19950
+    },
+    {
+      "epoch": 0.9671889858689134,
+      "grad_norm": 0.4644124507904053,
+      "learning_rate": 1.4098746251834938e-07,
+      "loss": 1.3148,
+      "num_input_tokens_seen": 783549976,
+      "step": 19960
+    },
+    {
+      "epoch": 0.9676735494890882,
+      "grad_norm": 0.4020436108112335,
+      "learning_rate": 1.3685693460114835e-07,
+      "loss": 1.2423,
+      "num_input_tokens_seen": 783956176,
+      "step": 19970
+    },
+    {
+      "epoch": 0.968158113109263,
+      "grad_norm": 0.42544135451316833,
+      "learning_rate": 1.3278764986025816e-07,
+      "loss": 1.2435,
+      "num_input_tokens_seen": 784370068,
+      "step": 19980
+    },
+    {
+      "epoch": 0.9686426767294378,
+      "grad_norm": 0.40704938769340515,
+      "learning_rate": 1.2877961831919416e-07,
+      "loss": 1.1991,
+      "num_input_tokens_seen": 784774032,
+      "step": 19990
+    },
+    {
+      "epoch": 0.9691272403496126,
+      "grad_norm": 0.45113271474838257,
+      "learning_rate": 1.248328498505813e-07,
+      "loss": 1.216,
+      "num_input_tokens_seen": 785148304,
+      "step": 20000
+    },
+    {
+      "epoch": 0.9691272403496126,
+      "eval_loss": 1.3588515520095825,
+      "eval_runtime": 4.2882,
+      "eval_samples_per_second": 34.98,
+      "eval_steps_per_second": 4.431,
+      "num_input_tokens_seen": 785148304,
+      "step": 20000
+    },
+    {
+      "epoch": 0.9696118039697874,
+      "grad_norm": 0.4246367812156677,
+      "learning_rate": 1.2094735417614579e-07,
+      "loss": 1.3037,
+      "num_input_tokens_seen": 785532360,
+      "step": 20010
+    },
+    {
+      "epoch": 0.9700963675899623,
+      "grad_norm": 0.40048691630363464,
+      "learning_rate": 1.1712314086668452e-07,
+      "loss": 1.2873,
+      "num_input_tokens_seen": 785948652,
+      "step": 20020
+    },
+    {
+      "epoch": 0.9705809312101371,
+      "grad_norm": 0.39899271726608276,
+      "learning_rate": 1.13360219342043e-07,
+      "loss": 1.2443,
+      "num_input_tokens_seen": 786340648,
+      "step": 20030
+    },
+    {
+      "epoch": 0.9710654948303119,
+      "grad_norm": 0.41089197993278503,
+      "learning_rate": 1.0965859887109575e-07,
+      "loss": 1.2696,
+      "num_input_tokens_seen": 786729636,
+      "step": 20040
+    },
+    {
+      "epoch": 0.9715500584504867,
+      "grad_norm": 0.3835326135158539,
+      "learning_rate": 1.0601828857171037e-07,
+      "loss": 1.2485,
+      "num_input_tokens_seen": 787122652,
+      "step": 20050
+    },
+    {
+      "epoch": 0.9720346220706615,
+      "grad_norm": 0.43811744451522827,
+      "learning_rate": 1.0243929741074465e-07,
+      "loss": 1.2809,
+      "num_input_tokens_seen": 787529572,
+      "step": 20060
+    },
+    {
+      "epoch": 0.9725191856908363,
+      "grad_norm": 0.4033362865447998,
+      "learning_rate": 9.892163420400779e-08,
+      "loss": 1.2607,
+      "num_input_tokens_seen": 787926640,
+      "step": 20070
+    },
+    {
+      "epoch": 0.9730037493110111,
+      "grad_norm": 0.4612255096435547,
+      "learning_rate": 9.546530761624928e-08,
+      "loss": 1.2536,
+      "num_input_tokens_seen": 788343440,
+      "step": 20080
+    },
+    {
+      "epoch": 0.9734883129311859,
+      "grad_norm": 0.4130823314189911,
+      "learning_rate": 9.20703261611311e-08,
+      "loss": 1.2822,
+      "num_input_tokens_seen": 788759176,
+      "step": 20090
+    },
+    {
+      "epoch": 0.9739728765513607,
+      "grad_norm": 0.46311095356941223,
+      "learning_rate": 8.873669820121111e-08,
+      "loss": 1.2476,
+      "num_input_tokens_seen": 789162040,
+      "step": 20100
+    },
+    {
+      "epoch": 0.9744574401715356,
+      "grad_norm": 0.3985438942909241,
+      "learning_rate": 8.546443194791808e-08,
+      "loss": 1.2453,
+      "num_input_tokens_seen": 789567068,
+      "step": 20110
+    },
+    {
+      "epoch": 0.9749420037917104,
+      "grad_norm": 0.40944841504096985,
+      "learning_rate": 8.22535354615378e-08,
+      "loss": 1.2376,
+      "num_input_tokens_seen": 789989960,
+      "step": 20120
+    },
+    {
+      "epoch": 0.9754265674118852,
+      "grad_norm": 0.41747674345970154,
+      "learning_rate": 7.910401665118528e-08,
+      "loss": 1.2289,
+      "num_input_tokens_seen": 790375840,
+      "step": 20130
+    },
+    {
+      "epoch": 0.97591113103206,
+      "grad_norm": 0.41349682211875916,
+      "learning_rate": 7.601588327479092e-08,
+      "loss": 1.2703,
+      "num_input_tokens_seen": 790758920,
+      "step": 20140
+    },
+    {
+      "epoch": 0.9763956946522347,
+      "grad_norm": 0.4096575677394867,
+      "learning_rate": 7.298914293907833e-08,
+      "loss": 1.2,
+      "num_input_tokens_seen": 791138556,
+      "step": 20150
+    },
+    {
+      "epoch": 0.9768802582724095,
+      "grad_norm": 0.39834490418434143,
+      "learning_rate": 7.002380309955314e-08,
+      "loss": 1.2784,
+      "num_input_tokens_seen": 791537516,
+      "step": 20160
+    },
+    {
+      "epoch": 0.9773648218925843,
+      "grad_norm": 0.41982364654541016,
+      "learning_rate": 6.711987106046979e-08,
+      "loss": 1.2283,
+      "num_input_tokens_seen": 791914936,
+      "step": 20170
+    },
+    {
+      "epoch": 0.9778493855127591,
+      "grad_norm": 0.4055969715118408,
+      "learning_rate": 6.427735397483148e-08,
+      "loss": 1.243,
+      "num_input_tokens_seen": 792312344,
+      "step": 20180
+    },
+    {
+      "epoch": 0.9783339491329339,
+      "grad_norm": 0.490959495306015,
+      "learning_rate": 6.149625884435407e-08,
+      "loss": 1.2525,
+      "num_input_tokens_seen": 792710660,
+      "step": 20190
+    },
+    {
+      "epoch": 0.9788185127531088,
+      "grad_norm": 0.4319050908088684,
+      "learning_rate": 5.877659251946332e-08,
+      "loss": 1.2896,
+      "num_input_tokens_seen": 793093572,
+      "step": 20200
+    },
+    {
+      "epoch": 0.9793030763732836,
+      "grad_norm": 0.42025476694107056,
+      "learning_rate": 5.611836169927276e-08,
+      "loss": 1.2736,
+      "num_input_tokens_seen": 793485640,
+      "step": 20210
+    },
+    {
+      "epoch": 0.9797876399934584,
+      "grad_norm": 0.47515881061553955,
+      "learning_rate": 5.35215729315669e-08,
+      "loss": 1.3133,
+      "num_input_tokens_seen": 793896728,
+      "step": 20220
+    },
+    {
+      "epoch": 0.9802722036136332,
+      "grad_norm": 0.41248735785484314,
+      "learning_rate": 5.0986232612787453e-08,
+      "loss": 1.2762,
+      "num_input_tokens_seen": 794295484,
+      "step": 20230
+    },
+    {
+      "epoch": 0.980756767233808,
+      "grad_norm": 0.4190135896205902,
+      "learning_rate": 4.851234698800833e-08,
+      "loss": 1.2217,
+      "num_input_tokens_seen": 794691580,
+      "step": 20240
+    },
+    {
+      "epoch": 0.9812413308539828,
+      "grad_norm": 0.41553497314453125,
+      "learning_rate": 4.609992215093839e-08,
+      "loss": 1.2819,
+      "num_input_tokens_seen": 795077668,
+      "step": 20250
+    },
+    {
+      "epoch": 0.9817258944741576,
+      "grad_norm": 0.42695561051368713,
+      "learning_rate": 4.374896404388818e-08,
+      "loss": 1.252,
+      "num_input_tokens_seen": 795454924,
+      "step": 20260
+    },
+    {
+      "epoch": 0.9822104580943324,
+      "grad_norm": 0.3926050066947937,
+      "learning_rate": 4.145947845776155e-08,
+      "loss": 1.2739,
+      "num_input_tokens_seen": 795863124,
+      "step": 20270
+    },
+    {
+      "epoch": 0.9826950217145072,
+      "grad_norm": 0.4227527678012848,
+      "learning_rate": 3.923147103204738e-08,
+      "loss": 1.2595,
+      "num_input_tokens_seen": 796247640,
+      "step": 20280
+    },
+    {
+      "epoch": 0.9831795853346821,
+      "grad_norm": 0.40944164991378784,
+      "learning_rate": 3.7064947254797365e-08,
+      "loss": 1.3064,
+      "num_input_tokens_seen": 796646028,
+      "step": 20290
+    },
+    {
+      "epoch": 0.9836641489548569,
+      "grad_norm": 0.39912155270576477,
+      "learning_rate": 3.4959912462620426e-08,
+      "loss": 1.1981,
+      "num_input_tokens_seen": 797037764,
+      "step": 20300
+    },
+    {
+      "epoch": 0.9841487125750317,
+      "grad_norm": 0.4131697118282318,
+      "learning_rate": 3.2916371840660544e-08,
+      "loss": 1.2223,
+      "num_input_tokens_seen": 797429064,
+      "step": 20310
+    },
+    {
+      "epoch": 0.9846332761952065,
+      "grad_norm": 0.4304356873035431,
+      "learning_rate": 3.093433042259119e-08,
+      "loss": 1.2225,
+      "num_input_tokens_seen": 797810616,
+      "step": 20320
+    },
+    {
+      "epoch": 0.9851178398153813,
+      "grad_norm": 0.4309372007846832,
+      "learning_rate": 2.9013793090598708e-08,
+      "loss": 1.2713,
+      "num_input_tokens_seen": 798207544,
+      "step": 20330
+    },
+    {
+      "epoch": 0.985602403435556,
+      "grad_norm": 0.4649990200996399,
+      "learning_rate": 2.715476457537114e-08,
+      "loss": 1.2487,
+      "num_input_tokens_seen": 798598064,
+      "step": 20340
+    },
+    {
+      "epoch": 0.9860869670557308,
+      "grad_norm": 0.4426388740539551,
+      "learning_rate": 2.535724945608997e-08,
+      "loss": 1.2796,
+      "num_input_tokens_seen": 798970944,
+      "step": 20350
+    },
+    {
+      "epoch": 0.9865715306759056,
+      "grad_norm": 0.4075920879840851,
+      "learning_rate": 2.3621252160413443e-08,
+      "loss": 1.2605,
+      "num_input_tokens_seen": 799356380,
+      "step": 20360
+    },
+    {
+      "epoch": 0.9870560942960804,
+      "grad_norm": 0.40336674451828003,
+      "learning_rate": 2.194677696447378e-08,
+      "loss": 1.2305,
+      "num_input_tokens_seen": 799753192,
+      "step": 20370
+    },
+    {
+      "epoch": 0.9875406579162553,
+      "grad_norm": 0.4300762712955475,
+      "learning_rate": 2.0333827992852217e-08,
+      "loss": 1.22,
+      "num_input_tokens_seen": 800120392,
+      "step": 20380
+    },
+    {
+      "epoch": 0.9880252215364301,
+      "grad_norm": 0.4360959827899933,
+      "learning_rate": 1.878240921858454e-08,
+      "loss": 1.2784,
+      "num_input_tokens_seen": 800495740,
+      "step": 20390
+    },
+    {
+      "epoch": 0.9885097851566049,
+      "grad_norm": 0.4353114664554596,
+      "learning_rate": 1.7292524463144445e-08,
+      "loss": 1.2539,
+      "num_input_tokens_seen": 800902016,
+      "step": 20400
+    },
+    {
+      "epoch": 0.9889943487767797,
+      "grad_norm": 0.43556949496269226,
+      "learning_rate": 1.5864177396432422e-08,
+      "loss": 1.227,
+      "num_input_tokens_seen": 801289824,
+      "step": 20410
+    },
+    {
+      "epoch": 0.9894789123969545,
+      "grad_norm": 0.4041503071784973,
+      "learning_rate": 1.4497371536770221e-08,
+      "loss": 1.2591,
+      "num_input_tokens_seen": 801688324,
+      "step": 20420
+    },
+    {
+      "epoch": 0.9899634760171293,
+      "grad_norm": 0.42014312744140625,
+      "learning_rate": 1.3192110250886957e-08,
+      "loss": 1.1945,
+      "num_input_tokens_seen": 802094388,
+      "step": 20430
+    },
+    {
+      "epoch": 0.9904480396373041,
+      "grad_norm": 0.44136643409729004,
+      "learning_rate": 1.1948396753919123e-08,
+      "loss": 1.274,
+      "num_input_tokens_seen": 802477196,
+      "step": 20440
+    },
+    {
+      "epoch": 0.9909326032574789,
+      "grad_norm": 0.4420247972011566,
+      "learning_rate": 1.0766234109393925e-08,
+      "loss": 1.2007,
+      "num_input_tokens_seen": 802877584,
+      "step": 20450
+    },
+    {
+      "epoch": 0.9914171668776537,
+      "grad_norm": 0.4464600086212158,
+      "learning_rate": 9.645625229232069e-09,
+      "loss": 1.219,
+      "num_input_tokens_seen": 803263948,
+      "step": 20460
+    },
+    {
+      "epoch": 0.9919017304978286,
+      "grad_norm": 0.4290953278541565,
+      "learning_rate": 8.586572873725551e-09,
+      "loss": 1.2345,
+      "num_input_tokens_seen": 803671476,
+      "step": 20470
+    },
+    {
+      "epoch": 0.9923862941180034,
+      "grad_norm": 0.40474194288253784,
+      "learning_rate": 7.589079651543207e-09,
+      "loss": 1.2742,
+      "num_input_tokens_seen": 804084040,
+      "step": 20480
+    },
+    {
+      "epoch": 0.9928708577381782,
+      "grad_norm": 0.41104060411453247,
+      "learning_rate": 6.653148019727939e-09,
+      "loss": 1.2496,
+      "num_input_tokens_seen": 804500276,
+      "step": 20490
+    },
+    {
+      "epoch": 0.993355421358353,
+      "grad_norm": 0.3800486922264099,
+      "learning_rate": 5.778780283671736e-09,
+      "loss": 1.2768,
+      "num_input_tokens_seen": 804875212,
+      "step": 20500
+    },
+    {
+      "epoch": 0.9938399849785278,
+      "grad_norm": 0.421836256980896,
+      "learning_rate": 4.96597859712955e-09,
+      "loss": 1.2106,
+      "num_input_tokens_seen": 805248472,
+      "step": 20510
+    },
+    {
+      "epoch": 0.9943245485987026,
+      "grad_norm": 0.4047603905200958,
+      "learning_rate": 4.2147449622026435e-09,
+      "loss": 1.2057,
+      "num_input_tokens_seen": 805647944,
+      "step": 20520
+    },
+    {
+      "epoch": 0.9948091122188774,
+      "grad_norm": 0.3970736265182495,
+      "learning_rate": 3.525081229338589e-09,
+      "loss": 1.2238,
+      "num_input_tokens_seen": 806023600,
+      "step": 20530
+    },
+    {
+      "epoch": 0.9952936758390521,
+      "grad_norm": 0.4089398682117462,
+      "learning_rate": 2.8969890973257198e-09,
+      "loss": 1.2181,
+      "num_input_tokens_seen": 806426272,
+      "step": 20540
+    },
+    {
+      "epoch": 0.995778239459227,
+      "grad_norm": 0.39923080801963806,
+      "learning_rate": 2.3304701132847996e-09,
+      "loss": 1.2999,
+      "num_input_tokens_seen": 806826016,
+      "step": 20550
+    },
+    {
+      "epoch": 0.9962628030794018,
+      "grad_norm": 0.3768922686576843,
+      "learning_rate": 1.8255256726745772e-09,
+      "loss": 1.2206,
+      "num_input_tokens_seen": 807236380,
+      "step": 20560
+    },
+    {
+      "epoch": 0.9967473666995766,
+      "grad_norm": 0.39100906252861023,
+      "learning_rate": 1.3821570192806832e-09,
+      "loss": 1.266,
+      "num_input_tokens_seen": 807644712,
+      "step": 20570
+    },
+    {
+      "epoch": 0.9972319303197514,
+      "grad_norm": 0.40794339776039124,
+      "learning_rate": 1.0003652452128532e-09,
+      "loss": 1.2559,
+      "num_input_tokens_seen": 808025808,
+      "step": 20580
+    },
+    {
+      "epoch": 0.9977164939399262,
+      "grad_norm": 0.4562262296676636,
+      "learning_rate": 6.801512909021535e-10,
+      "loss": 1.2459,
+      "num_input_tokens_seen": 808411128,
+      "step": 20590
+    },
+    {
+      "epoch": 0.998201057560101,
+      "grad_norm": 0.41534632444381714,
+      "learning_rate": 4.21515945106532e-10,
+      "loss": 1.2734,
+      "num_input_tokens_seen": 808793980,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9986856211802758,
+      "grad_norm": 0.4286879897117615,
+      "learning_rate": 2.2445984489971593e-10,
+      "loss": 1.2695,
+      "num_input_tokens_seen": 809188808,
+      "step": 20610
+    },
+    {
+      "epoch": 0.9991701848004506,
+      "grad_norm": 0.4052944779396057,
+      "learning_rate": 8.898347567121157e-11,
+      "loss": 1.2466,
+      "num_input_tokens_seen": 809544636,
+      "step": 20620
+    },
+    {
+      "epoch": 0.9996547484206254,
+      "grad_norm": 0.4112201929092407,
+      "learning_rate": 1.5087171129080092e-11,
+      "loss": 1.231,
+      "num_input_tokens_seen": 809914964,
+      "step": 20630
+    },
+    {
+      "epoch": 0.9999939429547479,
+      "num_input_tokens_seen": 810173896,
+      "step": 20637,
+      "total_flos": 4.073925813718745e+18,
+      "train_loss": 1.2524831550218467,
+      "train_runtime": 185903.2881,
+      "train_samples_per_second": 14.209,
+      "train_steps_per_second": 0.111
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 20637,
+  "num_input_tokens_seen": 810173896,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.073925813718745e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}