diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5519 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998921074234783,
+  "eval_steps": 400,
+  "global_step": 6661,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015011141081271255,
+      "grad_norm": 0.8901742100715637,
+      "learning_rate": 1.1976047904191619e-06,
+      "loss": 2.5666,
+      "num_input_tokens_seen": 5242880,
+      "step": 10
+    },
+    {
+      "epoch": 0.003002228216254251,
+      "grad_norm": 0.8234608769416809,
+      "learning_rate": 2.3952095808383237e-06,
+      "loss": 2.5448,
+      "num_input_tokens_seen": 10485760,
+      "step": 20
+    },
+    {
+      "epoch": 0.004503342324381376,
+      "grad_norm": 0.6512044668197632,
+      "learning_rate": 3.592814371257485e-06,
+      "loss": 2.4883,
+      "num_input_tokens_seen": 15728640,
+      "step": 30
+    },
+    {
+      "epoch": 0.006004456432508502,
+      "grad_norm": 0.6285302639007568,
+      "learning_rate": 4.7904191616766475e-06,
+      "loss": 2.4675,
+      "num_input_tokens_seen": 20971520,
+      "step": 40
+    },
+    {
+      "epoch": 0.007505570540635628,
+      "grad_norm": 0.5729206204414368,
+      "learning_rate": 5.9880239520958085e-06,
+      "loss": 2.4332,
+      "num_input_tokens_seen": 26214400,
+      "step": 50
+    },
+    {
+      "epoch": 0.009006684648762753,
+      "grad_norm": 0.54844731092453,
+      "learning_rate": 7.18562874251497e-06,
+      "loss": 2.4374,
+      "num_input_tokens_seen": 31457280,
+      "step": 60
+    },
+    {
+      "epoch": 0.01050779875688988,
+      "grad_norm": 0.5901165008544922,
+      "learning_rate": 8.383233532934131e-06,
+      "loss": 2.4057,
+      "num_input_tokens_seen": 36700160,
+      "step": 70
+    },
+    {
+      "epoch": 0.012008912865017004,
+      "grad_norm": 0.6103231906890869,
+      "learning_rate": 9.580838323353295e-06,
+      "loss": 2.3765,
+      "num_input_tokens_seen": 41943040,
+      "step": 80
+    },
+    {
+      "epoch": 0.013510026973144131,
+      "grad_norm": 0.502941906452179,
+      "learning_rate": 1.0778443113772457e-05,
+      "loss": 2.3877,
+      "num_input_tokens_seen": 47185920,
+      "step": 90
+    },
+    {
+      "epoch": 0.015011141081271256,
+      "grad_norm": 0.6203625202178955,
+      "learning_rate": 1.1976047904191617e-05,
+      "loss": 2.3622,
+      "num_input_tokens_seen": 52428800,
+      "step": 100
+    },
+    {
+      "epoch": 0.016512255189398382,
+      "grad_norm": 0.658467173576355,
+      "learning_rate": 1.3173652694610779e-05,
+      "loss": 2.3567,
+      "num_input_tokens_seen": 57671680,
+      "step": 110
+    },
+    {
+      "epoch": 0.018013369297525506,
+      "grad_norm": 0.5568761229515076,
+      "learning_rate": 1.437125748502994e-05,
+      "loss": 2.3679,
+      "num_input_tokens_seen": 62914560,
+      "step": 120
+    },
+    {
+      "epoch": 0.019514483405652632,
+      "grad_norm": 0.6238291263580322,
+      "learning_rate": 1.5568862275449103e-05,
+      "loss": 2.3829,
+      "num_input_tokens_seen": 68157440,
+      "step": 130
+    },
+    {
+      "epoch": 0.02101559751377976,
+      "grad_norm": 0.6000885367393494,
+      "learning_rate": 1.6766467065868263e-05,
+      "loss": 2.3472,
+      "num_input_tokens_seen": 73400320,
+      "step": 140
+    },
+    {
+      "epoch": 0.022516711621906885,
+      "grad_norm": 0.6424698829650879,
+      "learning_rate": 1.7964071856287426e-05,
+      "loss": 2.3285,
+      "num_input_tokens_seen": 78643200,
+      "step": 150
+    },
+    {
+      "epoch": 0.02401782573003401,
+      "grad_norm": 0.6143007874488831,
+      "learning_rate": 1.916167664670659e-05,
+      "loss": 2.3151,
+      "num_input_tokens_seen": 83886080,
+      "step": 160
+    },
+    {
+      "epoch": 0.025518939838161135,
+      "grad_norm": 0.6206329464912415,
+      "learning_rate": 2.035928143712575e-05,
+      "loss": 2.3368,
+      "num_input_tokens_seen": 89128960,
+      "step": 170
+    },
+    {
+      "epoch": 0.027020053946288262,
+      "grad_norm": 0.6878198385238647,
+      "learning_rate": 2.1556886227544914e-05,
+      "loss": 2.3367,
+      "num_input_tokens_seen": 94371840,
+      "step": 180
+    },
+    {
+      "epoch": 0.028521168054415385,
+      "grad_norm": 0.8266871571540833,
+      "learning_rate": 2.2754491017964074e-05,
+      "loss": 2.3286,
+      "num_input_tokens_seen": 99614720,
+      "step": 190
+    },
+    {
+      "epoch": 0.03002228216254251,
+      "grad_norm": 0.5854138135910034,
+      "learning_rate": 2.3952095808383234e-05,
+      "loss": 2.303,
+      "num_input_tokens_seen": 104857600,
+      "step": 200
+    },
+    {
+      "epoch": 0.031523396270669635,
+      "grad_norm": 0.7110793590545654,
+      "learning_rate": 2.5149700598802394e-05,
+      "loss": 2.2985,
+      "num_input_tokens_seen": 110100480,
+      "step": 210
+    },
+    {
+      "epoch": 0.033024510378796765,
+      "grad_norm": 0.6561000943183899,
+      "learning_rate": 2.6347305389221558e-05,
+      "loss": 2.3154,
+      "num_input_tokens_seen": 115343360,
+      "step": 220
+    },
+    {
+      "epoch": 0.03452562448692389,
+      "grad_norm": 0.7813587188720703,
+      "learning_rate": 2.754491017964072e-05,
+      "loss": 2.3104,
+      "num_input_tokens_seen": 120586240,
+      "step": 230
+    },
+    {
+      "epoch": 0.03602673859505101,
+      "grad_norm": 0.7437533140182495,
+      "learning_rate": 2.874251497005988e-05,
+      "loss": 2.3098,
+      "num_input_tokens_seen": 125829120,
+      "step": 240
+    },
+    {
+      "epoch": 0.03752785270317814,
+      "grad_norm": 0.8770821690559387,
+      "learning_rate": 2.9940119760479045e-05,
+      "loss": 2.3061,
+      "num_input_tokens_seen": 131072000,
+      "step": 250
+    },
+    {
+      "epoch": 0.039028966811305264,
+      "grad_norm": 0.6238060593605042,
+      "learning_rate": 3.1137724550898205e-05,
+      "loss": 2.3052,
+      "num_input_tokens_seen": 136314880,
+      "step": 260
+    },
+    {
+      "epoch": 0.040530080919432394,
+      "grad_norm": 0.678520679473877,
+      "learning_rate": 3.233532934131737e-05,
+      "loss": 2.2934,
+      "num_input_tokens_seen": 141557760,
+      "step": 270
+    },
+    {
+      "epoch": 0.04203119502755952,
+      "grad_norm": 0.7275491952896118,
+      "learning_rate": 3.3532934131736525e-05,
+      "loss": 2.2938,
+      "num_input_tokens_seen": 146800640,
+      "step": 280
+    },
+    {
+      "epoch": 0.04353230913568664,
+      "grad_norm": 0.6745551824569702,
+      "learning_rate": 3.473053892215569e-05,
+      "loss": 2.3137,
+      "num_input_tokens_seen": 152043520,
+      "step": 290
+    },
+    {
+      "epoch": 0.04503342324381377,
+      "grad_norm": 0.8107547163963318,
+      "learning_rate": 3.592814371257485e-05,
+      "loss": 2.2833,
+      "num_input_tokens_seen": 157286400,
+      "step": 300
+    },
+    {
+      "epoch": 0.046534537351940894,
+      "grad_norm": 0.9518006443977356,
+      "learning_rate": 3.7125748502994016e-05,
+      "loss": 2.2883,
+      "num_input_tokens_seen": 162529280,
+      "step": 310
+    },
+    {
+      "epoch": 0.04803565146006802,
+      "grad_norm": 0.7946741580963135,
+      "learning_rate": 3.832335329341318e-05,
+      "loss": 2.292,
+      "num_input_tokens_seen": 167772160,
+      "step": 320
+    },
+    {
+      "epoch": 0.04953676556819515,
+      "grad_norm": 0.8819884657859802,
+      "learning_rate": 3.9520958083832336e-05,
+      "loss": 2.2874,
+      "num_input_tokens_seen": 173015040,
+      "step": 330
+    },
+    {
+      "epoch": 0.05103787967632227,
+      "grad_norm": 0.8569954633712769,
+      "learning_rate": 3.9645487837613415e-05,
+      "loss": 2.2701,
+      "num_input_tokens_seen": 178257920,
+      "step": 340
+    },
+    {
+      "epoch": 0.05253899378444939,
+      "grad_norm": 0.6152970194816589,
+      "learning_rate": 3.907501942235145e-05,
+      "loss": 2.2995,
+      "num_input_tokens_seen": 183500800,
+      "step": 350
+    },
+    {
+      "epoch": 0.054040107892576524,
+      "grad_norm": 0.6690587997436523,
+      "learning_rate": 3.852848873813304e-05,
+      "loss": 2.2663,
+      "num_input_tokens_seen": 188743680,
+      "step": 360
+    },
+    {
+      "epoch": 0.05554122200070365,
+      "grad_norm": 0.7796323895454407,
+      "learning_rate": 3.800426718572961e-05,
+      "loss": 2.2597,
+      "num_input_tokens_seen": 193986560,
+      "step": 370
+    },
+    {
+      "epoch": 0.05704233610883077,
+      "grad_norm": 0.9035688638687134,
+      "learning_rate": 3.7500877182723136e-05,
+      "loss": 2.2369,
+      "num_input_tokens_seen": 199229440,
+      "step": 380
+    },
+    {
+      "epoch": 0.0585434502169579,
+      "grad_norm": 0.7932724356651306,
+      "learning_rate": 3.701697462322401e-05,
+      "loss": 2.2514,
+      "num_input_tokens_seen": 204472320,
+      "step": 390
+    },
+    {
+      "epoch": 0.06004456432508502,
+      "grad_norm": 0.6233640909194946,
+      "learning_rate": 3.6551333764994134e-05,
+      "loss": 2.2572,
+      "num_input_tokens_seen": 209715200,
+      "step": 400
+    },
+    {
+      "epoch": 0.06004456432508502,
+      "eval_accuracy": 0.5491493691493692,
+      "eval_loss": 2.2461655139923096,
+      "eval_runtime": 95.7686,
+      "eval_samples_per_second": 3.133,
+      "eval_steps_per_second": 0.783,
+      "num_input_tokens_seen": 209715200,
+      "step": 400
+    },
+    {
+      "epoch": 0.061545678433212146,
+      "grad_norm": 0.5706452131271362,
+      "learning_rate": 3.610283415670218e-05,
+      "loss": 2.2674,
+      "num_input_tokens_seen": 214958080,
+      "step": 410
+    },
+    {
+      "epoch": 0.06304679254133927,
+      "grad_norm": 0.6556965112686157,
+      "learning_rate": 3.567044928762396e-05,
+      "loss": 2.2898,
+      "num_input_tokens_seen": 220200960,
+      "step": 420
+    },
+    {
+      "epoch": 0.0645479066494664,
+      "grad_norm": 0.7172439098358154,
+      "learning_rate": 3.5253236697846886e-05,
+      "loss": 2.2644,
+      "num_input_tokens_seen": 225443840,
+      "step": 430
+    },
+    {
+      "epoch": 0.06604902075759353,
+      "grad_norm": 0.6364426016807556,
+      "learning_rate": 3.485032933195115e-05,
+      "loss": 2.2551,
+      "num_input_tokens_seen": 230686720,
+      "step": 440
+    },
+    {
+      "epoch": 0.06755013486572065,
+      "grad_norm": 0.7587452530860901,
+      "learning_rate": 3.446092795552023e-05,
+      "loss": 2.2576,
+      "num_input_tokens_seen": 235929600,
+      "step": 450
+    },
+    {
+      "epoch": 0.06905124897384778,
+      "grad_norm": 0.6086819767951965,
+      "learning_rate": 3.408429448345356e-05,
+      "loss": 2.2768,
+      "num_input_tokens_seen": 241172480,
+      "step": 460
+    },
+    {
+      "epoch": 0.0705523630819749,
+      "grad_norm": 0.681690514087677,
+      "learning_rate": 3.371974609328701e-05,
+      "loss": 2.2744,
+      "num_input_tokens_seen": 246415360,
+      "step": 470
+    },
+    {
+      "epoch": 0.07205347719010202,
+      "grad_norm": 0.8943272233009338,
+      "learning_rate": 3.336665001664587e-05,
+      "loss": 2.2536,
+      "num_input_tokens_seen": 251658240,
+      "step": 480
+    },
+    {
+      "epoch": 0.07355459129822915,
+      "grad_norm": 1.1301355361938477,
+      "learning_rate": 3.3024418918399747e-05,
+      "loss": 2.2329,
+      "num_input_tokens_seen": 256901120,
+      "step": 490
+    },
+    {
+      "epoch": 0.07505570540635628,
+      "grad_norm": 0.7338290214538574,
+      "learning_rate": 3.269250678672408e-05,
+      "loss": 2.2592,
+      "num_input_tokens_seen": 262144000,
+      "step": 500
+    },
+    {
+      "epoch": 0.07655681951448341,
+      "grad_norm": 0.7819264531135559,
+      "learning_rate": 3.23704052686231e-05,
+      "loss": 2.2453,
+      "num_input_tokens_seen": 267386880,
+      "step": 510
+    },
+    {
+      "epoch": 0.07805793362261053,
+      "grad_norm": 0.6750678420066833,
+      "learning_rate": 3.205764039495589e-05,
+      "loss": 2.2514,
+      "num_input_tokens_seen": 272629760,
+      "step": 520
+    },
+    {
+      "epoch": 0.07955904773073766,
+      "grad_norm": 0.5943534970283508,
+      "learning_rate": 3.175376964696401e-05,
+      "loss": 2.2346,
+      "num_input_tokens_seen": 277872640,
+      "step": 530
+    },
+    {
+      "epoch": 0.08106016183886479,
+      "grad_norm": 0.7100771069526672,
+      "learning_rate": 3.145837932299803e-05,
+      "loss": 2.2316,
+      "num_input_tokens_seen": 283115520,
+      "step": 540
+    },
+    {
+      "epoch": 0.0825612759469919,
+      "grad_norm": 0.6773257851600647,
+      "learning_rate": 3.117108216979904e-05,
+      "loss": 2.2343,
+      "num_input_tokens_seen": 288358400,
+      "step": 550
+    },
+    {
+      "epoch": 0.08406239005511904,
+      "grad_norm": 0.8613666296005249,
+      "learning_rate": 3.089151524748689e-05,
+      "loss": 2.253,
+      "num_input_tokens_seen": 293601280,
+      "step": 560
+    },
+    {
+      "epoch": 0.08556350416324617,
+      "grad_norm": 0.5922672748565674,
+      "learning_rate": 3.0619338001484006e-05,
+      "loss": 2.2032,
+      "num_input_tokens_seen": 298844160,
+      "step": 570
+    },
+    {
+      "epoch": 0.08706461827137328,
+      "grad_norm": 0.8399538397789001,
+      "learning_rate": 3.0354230518081458e-05,
+      "loss": 2.2445,
+      "num_input_tokens_seen": 304087040,
+      "step": 580
+    },
+    {
+      "epoch": 0.08856573237950041,
+      "grad_norm": 0.8287783265113831,
+      "learning_rate": 3.0095891943326868e-05,
+      "loss": 2.2473,
+      "num_input_tokens_seen": 309329920,
+      "step": 590
+    },
+    {
+      "epoch": 0.09006684648762754,
+      "grad_norm": 0.6455395221710205,
+      "learning_rate": 2.9844039047465857e-05,
+      "loss": 2.2276,
+      "num_input_tokens_seen": 314572800,
+      "step": 600
+    },
+    {
+      "epoch": 0.09156796059575466,
+      "grad_norm": 0.6312438249588013,
+      "learning_rate": 2.959840491936165e-05,
+      "loss": 2.2383,
+      "num_input_tokens_seen": 319815680,
+      "step": 610
+    },
+    {
+      "epoch": 0.09306907470388179,
+      "grad_norm": 0.5632859468460083,
+      "learning_rate": 2.9358737777209838e-05,
+      "loss": 2.2342,
+      "num_input_tokens_seen": 325058560,
+      "step": 620
+    },
+    {
+      "epoch": 0.09457018881200892,
+      "grad_norm": 0.6137943863868713,
+      "learning_rate": 2.9124799883500802e-05,
+      "loss": 2.2058,
+      "num_input_tokens_seen": 330301440,
+      "step": 630
+    },
+    {
+      "epoch": 0.09607130292013603,
+      "grad_norm": 0.7079706788063049,
+      "learning_rate": 2.8896366553599782e-05,
+      "loss": 2.2235,
+      "num_input_tokens_seen": 335544320,
+      "step": 640
+    },
+    {
+      "epoch": 0.09757241702826316,
+      "grad_norm": 0.6246356964111328,
+      "learning_rate": 2.8673225248545833e-05,
+      "loss": 2.2363,
+      "num_input_tokens_seen": 340787200,
+      "step": 650
+    },
+    {
+      "epoch": 0.0990735311363903,
+      "grad_norm": 0.7458660006523132,
+      "learning_rate": 2.8455174743743358e-05,
+      "loss": 2.2199,
+      "num_input_tokens_seen": 346030080,
+      "step": 660
+    },
+    {
+      "epoch": 0.10057464524451741,
+      "grad_norm": 0.6742898225784302,
+      "learning_rate": 2.8242024366155264e-05,
+      "loss": 2.2293,
+      "num_input_tokens_seen": 351272960,
+      "step": 670
+    },
+    {
+      "epoch": 0.10207575935264454,
+      "grad_norm": 0.5833361148834229,
+      "learning_rate": 2.8033593293425236e-05,
+      "loss": 2.2346,
+      "num_input_tokens_seen": 356515840,
+      "step": 680
+    },
+    {
+      "epoch": 0.10357687346077167,
+      "grad_norm": 0.6706376075744629,
+      "learning_rate": 2.7829709909073584e-05,
+      "loss": 2.2353,
+      "num_input_tokens_seen": 361758720,
+      "step": 690
+    },
+    {
+      "epoch": 0.10507798756889879,
+      "grad_norm": 0.7505273222923279,
+      "learning_rate": 2.763021120854076e-05,
+      "loss": 2.2463,
+      "num_input_tokens_seen": 367001600,
+      "step": 700
+    },
+    {
+      "epoch": 0.10657910167702592,
+      "grad_norm": 0.6324784755706787,
+      "learning_rate": 2.743494225140684e-05,
+      "loss": 2.2208,
+      "num_input_tokens_seen": 372244480,
+      "step": 710
+    },
+    {
+      "epoch": 0.10808021578515305,
+      "grad_norm": 0.6054283380508423,
+      "learning_rate": 2.7243755655603398e-05,
+      "loss": 2.214,
+      "num_input_tokens_seen": 377487360,
+      "step": 720
+    },
+    {
+      "epoch": 0.10958132989328016,
+      "grad_norm": 0.6269551515579224,
+      "learning_rate": 2.7056511129865727e-05,
+      "loss": 2.2426,
+      "num_input_tokens_seen": 382730240,
+      "step": 730
+    },
+    {
+      "epoch": 0.1110824440014073,
+      "grad_norm": 0.8448464870452881,
+      "learning_rate": 2.6873075041054796e-05,
+      "loss": 2.2214,
+      "num_input_tokens_seen": 387973120,
+      "step": 740
+    },
+    {
+      "epoch": 0.11258355810953442,
+      "grad_norm": 0.7521204948425293,
+      "learning_rate": 2.669332001331669e-05,
+      "loss": 2.2033,
+      "num_input_tokens_seen": 393216000,
+      "step": 750
+    },
+    {
+      "epoch": 0.11408467221766154,
+      "grad_norm": 0.5531004071235657,
+      "learning_rate": 2.6517124556347397e-05,
+      "loss": 2.2177,
+      "num_input_tokens_seen": 398458880,
+      "step": 760
+    },
+    {
+      "epoch": 0.11558578632578867,
+      "grad_norm": 0.5678554773330688,
+      "learning_rate": 2.6344372720297863e-05,
+      "loss": 2.2527,
+      "num_input_tokens_seen": 403701760,
+      "step": 770
+    },
+    {
+      "epoch": 0.1170869004339158,
+      "grad_norm": 0.5761403441429138,
+      "learning_rate": 2.6174953775092044e-05,
+      "loss": 2.1891,
+      "num_input_tokens_seen": 408944640,
+      "step": 780
+    },
+    {
+      "epoch": 0.11858801454204292,
+      "grad_norm": 0.614530622959137,
+      "learning_rate": 2.6008761912142832e-05,
+      "loss": 2.2389,
+      "num_input_tokens_seen": 414187520,
+      "step": 790
+    },
+    {
+      "epoch": 0.12008912865017005,
+      "grad_norm": 0.8077418208122253,
+      "learning_rate": 2.584569596664017e-05,
+      "loss": 2.2173,
+      "num_input_tokens_seen": 419430400,
+      "step": 800
+    },
+    {
+      "epoch": 0.12008912865017005,
+      "eval_accuracy": 0.5564314204314205,
+      "eval_loss": 2.1939218044281006,
+      "eval_runtime": 94.2566,
+      "eval_samples_per_second": 3.183,
+      "eval_steps_per_second": 0.796,
+      "num_input_tokens_seen": 419430400,
+      "step": 800
+    },
+    {
+      "epoch": 0.12159024275829718,
+      "grad_norm": 0.5838737487792969,
+      "learning_rate": 2.5685659158755362e-05,
+      "loss": 2.2407,
+      "num_input_tokens_seen": 424673280,
+      "step": 810
+    },
+    {
+      "epoch": 0.12309135686642429,
+      "grad_norm": 0.5616093873977661,
+      "learning_rate": 2.5528558852257423e-05,
+      "loss": 2.2387,
+      "num_input_tokens_seen": 429916160,
+      "step": 820
+    },
+    {
+      "epoch": 0.12459247097455142,
+      "grad_norm": 0.5916436910629272,
+      "learning_rate": 2.537430632917375e-05,
+      "loss": 2.1901,
+      "num_input_tokens_seen": 435159040,
+      "step": 830
+    },
+    {
+      "epoch": 0.12609358508267854,
+      "grad_norm": 0.7512922286987305,
+      "learning_rate": 2.5222816579249755e-05,
+      "loss": 2.2354,
+      "num_input_tokens_seen": 440401920,
+      "step": 840
+    },
+    {
+      "epoch": 0.12759469919080568,
+      "grad_norm": 0.6344349384307861,
+      "learning_rate": 2.507400810307242e-05,
+      "loss": 2.18,
+      "num_input_tokens_seen": 445644800,
+      "step": 850
+    },
+    {
+      "epoch": 0.1290958132989328,
+      "grad_norm": 0.6193192005157471,
+      "learning_rate": 2.4927802727821985e-05,
+      "loss": 2.2427,
+      "num_input_tokens_seen": 450887680,
+      "step": 860
+    },
+    {
+      "epoch": 0.13059692740705992,
+      "grad_norm": 0.5937076807022095,
+      "learning_rate": 2.4784125434705546e-05,
+      "loss": 2.2096,
+      "num_input_tokens_seen": 456130560,
+      "step": 870
+    },
+    {
+      "epoch": 0.13209804151518706,
+      "grad_norm": 0.6233764886856079,
+      "learning_rate": 2.4642904197207104e-05,
+      "loss": 2.2124,
+      "num_input_tokens_seen": 461373440,
+      "step": 880
+    },
+    {
+      "epoch": 0.13359915562331418,
+      "grad_norm": 0.7363754510879517,
+      "learning_rate": 2.4504069829361966e-05,
+      "loss": 2.1925,
+      "num_input_tokens_seen": 466616320,
+      "step": 890
+    },
+    {
+      "epoch": 0.1351002697314413,
+      "grad_norm": 0.5889034867286682,
+      "learning_rate": 2.4367555843329425e-05,
+      "loss": 2.2139,
+      "num_input_tokens_seen": 471859200,
+      "step": 900
+    },
+    {
+      "epoch": 0.13660138383956844,
+      "grad_norm": 0.5365704894065857,
+      "learning_rate": 2.4233298315597637e-05,
+      "loss": 2.2096,
+      "num_input_tokens_seen": 477102080,
+      "step": 910
+    },
+    {
+      "epoch": 0.13810249794769555,
+      "grad_norm": 0.5648247599601746,
+      "learning_rate": 2.4101235761209246e-05,
+      "loss": 2.1803,
+      "num_input_tokens_seen": 482344960,
+      "step": 920
+    },
+    {
+      "epoch": 0.13960361205582267,
+      "grad_norm": 0.6017926931381226,
+      "learning_rate": 2.3971309015445497e-05,
+      "loss": 2.1787,
+      "num_input_tokens_seen": 487587840,
+      "step": 930
+    },
+    {
+      "epoch": 0.1411047261639498,
+      "grad_norm": 0.6408767700195312,
+      "learning_rate": 2.384346112245184e-05,
+      "loss": 2.1998,
+      "num_input_tokens_seen": 492830720,
+      "step": 940
+    },
+    {
+      "epoch": 0.14260584027207693,
+      "grad_norm": 0.635210394859314,
+      "learning_rate": 2.3717637230328695e-05,
+      "loss": 2.2327,
+      "num_input_tokens_seen": 498073600,
+      "step": 950
+    },
+    {
+      "epoch": 0.14410695438020404,
+      "grad_norm": 0.6507686972618103,
+      "learning_rate": 2.3593784492248522e-05,
+      "loss": 2.2162,
+      "num_input_tokens_seen": 503316480,
+      "step": 960
+    },
+    {
+      "epoch": 0.1456080684883312,
+      "grad_norm": 0.6332815885543823,
+      "learning_rate": 2.3471851973194328e-05,
+      "loss": 2.2287,
+      "num_input_tokens_seen": 508559360,
+      "step": 970
+    },
+    {
+      "epoch": 0.1471091825964583,
+      "grad_norm": 0.6600320339202881,
+      "learning_rate": 2.3351790561945772e-05,
+      "loss": 2.1837,
+      "num_input_tokens_seen": 513802240,
+      "step": 980
+    },
+    {
+      "epoch": 0.14861029670458545,
+      "grad_norm": 0.5689282417297363,
+      "learning_rate": 2.3233552887967433e-05,
+      "loss": 2.1926,
+      "num_input_tokens_seen": 519045120,
+      "step": 990
+    },
+    {
+      "epoch": 0.15011141081271256,
+      "grad_norm": 0.762301504611969,
+      "learning_rate": 2.3117093242879823e-05,
+      "loss": 2.2092,
+      "num_input_tokens_seen": 524288000,
+      "step": 1000
+    },
+    {
+      "epoch": 0.15161252492083968,
+      "grad_norm": 0.5970498919487,
+      "learning_rate": 2.3002367506217465e-05,
+      "loss": 2.2208,
+      "num_input_tokens_seen": 529530880,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15311363902896682,
+      "grad_norm": 0.7378530502319336,
+      "learning_rate": 2.288933307520014e-05,
+      "loss": 2.193,
+      "num_input_tokens_seen": 534773760,
+      "step": 1020
+    },
+    {
+      "epoch": 0.15461475313709394,
+      "grad_norm": 0.6843045949935913,
+      "learning_rate": 2.2777948798263435e-05,
+      "loss": 2.201,
+      "num_input_tokens_seen": 540016640,
+      "step": 1030
+    },
+    {
+      "epoch": 0.15611586724522106,
+      "grad_norm": 0.6419305205345154,
+      "learning_rate": 2.2668174912113104e-05,
+      "loss": 2.2008,
+      "num_input_tokens_seen": 545259520,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1576169813533482,
+      "grad_norm": 0.6570454239845276,
+      "learning_rate": 2.2559972982084464e-05,
+      "loss": 2.1878,
+      "num_input_tokens_seen": 550502400,
+      "step": 1050
+    },
+    {
+      "epoch": 0.15911809546147532,
+      "grad_norm": 0.6204289197921753,
+      "learning_rate": 2.2453305845603817e-05,
+      "loss": 2.1965,
+      "num_input_tokens_seen": 555745280,
+      "step": 1060
+    },
+    {
+      "epoch": 0.16061920956960243,
+      "grad_norm": 0.7957982420921326,
+      "learning_rate": 2.234813755856288e-05,
+      "loss": 2.1886,
+      "num_input_tokens_seen": 560988160,
+      "step": 1070
+    },
+    {
+      "epoch": 0.16212032367772958,
+      "grad_norm": 0.6827828288078308,
+      "learning_rate": 2.2244433344430573e-05,
+      "loss": 2.1646,
+      "num_input_tokens_seen": 566231040,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1636214377858567,
+      "grad_norm": 0.8159109354019165,
+      "learning_rate": 2.2142159545938563e-05,
+      "loss": 2.1966,
+      "num_input_tokens_seen": 571473920,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1651225518939838,
+      "grad_norm": 0.6716334223747253,
+      "learning_rate": 2.2041283579187985e-05,
+      "loss": 2.173,
+      "num_input_tokens_seen": 576716800,
+      "step": 1100
+    },
+    {
+      "epoch": 0.16662366600211095,
+      "grad_norm": 0.5386361479759216,
+      "learning_rate": 2.1941773890035455e-05,
+      "loss": 2.1958,
+      "num_input_tokens_seen": 581959680,
+      "step": 1110
+    },
+    {
+      "epoch": 0.16812478011023807,
+      "grad_norm": 0.5818248391151428,
+      "learning_rate": 2.1843599912625603e-05,
+      "loss": 2.2007,
+      "num_input_tokens_seen": 587202560,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1696258942183652,
+      "grad_norm": 0.6868605613708496,
+      "learning_rate": 2.17467320299465e-05,
+      "loss": 2.2067,
+      "num_input_tokens_seen": 592445440,
+      "step": 1130
+    },
+    {
+      "epoch": 0.17112700832649233,
+      "grad_norm": 0.6540196537971497,
+      "learning_rate": 2.1651141536292296e-05,
+      "loss": 2.17,
+      "num_input_tokens_seen": 597688320,
+      "step": 1140
+    },
+    {
+      "epoch": 0.17262812243461945,
+      "grad_norm": 0.5799123644828796,
+      "learning_rate": 2.1556800601525106e-05,
+      "loss": 2.1932,
+      "num_input_tokens_seen": 602931200,
+      "step": 1150
+    },
+    {
+      "epoch": 0.17412923654274656,
+      "grad_norm": 0.7334414124488831,
+      "learning_rate": 2.1463682237035052e-05,
+      "loss": 2.1795,
+      "num_input_tokens_seen": 608174080,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1756303506508737,
+      "grad_norm": 0.5947369337081909,
+      "learning_rate": 2.137176026330393e-05,
+      "loss": 2.1544,
+      "num_input_tokens_seen": 613416960,
+      "step": 1170
+    },
+    {
+      "epoch": 0.17713146475900082,
+      "grad_norm": 0.6859150528907776,
+      "learning_rate": 2.128100927898401e-05,
+      "loss": 2.1821,
+      "num_input_tokens_seen": 618659840,
+      "step": 1180
+    },
+    {
+      "epoch": 0.17863257886712794,
+      "grad_norm": 0.6155475378036499,
+      "learning_rate": 2.1191404631408953e-05,
+      "loss": 2.1643,
+      "num_input_tokens_seen": 623902720,
+      "step": 1190
+    },
+    {
+      "epoch": 0.18013369297525508,
+      "grad_norm": 0.5512512922286987,
+      "learning_rate": 2.110292238845922e-05,
+      "loss": 2.1992,
+      "num_input_tokens_seen": 629145600,
+      "step": 1200
+    },
+    {
+      "epoch": 0.18013369297525508,
+      "eval_accuracy": 0.5604013024013024,
+      "eval_loss": 2.1688950061798096,
+      "eval_runtime": 95.4401,
+      "eval_samples_per_second": 3.143,
+      "eval_steps_per_second": 0.786,
+      "num_input_tokens_seen": 629145600,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1816348070833822,
+      "grad_norm": 0.6315280795097351,
+      "learning_rate": 2.1015539311708933e-05,
+      "loss": 2.1888,
+      "num_input_tokens_seen": 634388480,
+      "step": 1210
+    },
+    {
+      "epoch": 0.18313592119150932,
+      "grad_norm": 0.6135600209236145,
+      "learning_rate": 2.092923283078589e-05,
+      "loss": 2.2094,
+      "num_input_tokens_seen": 639631360,
+      "step": 1220
+    },
+    {
+      "epoch": 0.18463703529963646,
+      "grad_norm": 0.7772483229637146,
+      "learning_rate": 2.084398101888042e-05,
+      "loss": 2.1959,
+      "num_input_tokens_seen": 644874240,
+      "step": 1230
+    },
+    {
+      "epoch": 0.18613814940776358,
+      "grad_norm": 0.7412639856338501,
+      "learning_rate": 2.0759762569342743e-05,
+      "loss": 2.1783,
+      "num_input_tokens_seen": 650117120,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1876392635158907,
+      "grad_norm": 0.8701518177986145,
+      "learning_rate": 2.0676556773312137e-05,
+      "loss": 2.1231,
+      "num_input_tokens_seen": 655360000,
+      "step": 1250
+    },
+    {
+      "epoch": 0.18914037762401784,
+      "grad_norm": 0.7169715762138367,
+      "learning_rate": 2.059434349832459e-05,
+      "loss": 2.1964,
+      "num_input_tokens_seen": 660602880,
+      "step": 1260
+    },
+    {
+      "epoch": 0.19064149173214495,
+      "grad_norm": 0.5910638570785522,
+      "learning_rate": 2.0513103167848672e-05,
+      "loss": 2.1915,
+      "num_input_tokens_seen": 665845760,
+      "step": 1270
+    },
+    {
+      "epoch": 0.19214260584027207,
+      "grad_norm": 0.5827318429946899,
+      "learning_rate": 2.0432816741702553e-05,
+      "loss": 2.2007,
+      "num_input_tokens_seen": 671088640,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1936437199483992,
+      "grad_norm": 0.6388292908668518,
+      "learning_rate": 2.0353465697307495e-05,
+      "loss": 2.1615,
+      "num_input_tokens_seen": 676331520,
+      "step": 1290
+    },
+    {
+      "epoch": 0.19514483405652633,
+      "grad_norm": 0.6320261359214783,
+      "learning_rate": 2.0275032011736088e-05,
+      "loss": 2.1983,
+      "num_input_tokens_seen": 681574400,
+      "step": 1300
+    },
+    {
+      "epoch": 0.19664594816465344,
+      "grad_norm": 0.6461578011512756,
+      "learning_rate": 2.019749814451557e-05,
+      "loss": 2.177,
+      "num_input_tokens_seen": 686817280,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1981470622727806,
+      "grad_norm": 0.7476032376289368,
+      "learning_rate": 2.0120847021149106e-05,
+      "loss": 2.1549,
+      "num_input_tokens_seen": 692060160,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1996481763809077,
+      "grad_norm": 0.6539933681488037,
+      "learning_rate": 2.004506201731977e-05,
+      "loss": 2.1761,
+      "num_input_tokens_seen": 697303040,
+      "step": 1330
+    },
+    {
+      "epoch": 0.20114929048903482,
+      "grad_norm": 0.5645061135292053,
+      "learning_rate": 1.9970126943744094e-05,
+      "loss": 2.1721,
+      "num_input_tokens_seen": 702545920,
+      "step": 1340
+    },
+    {
+      "epoch": 0.20265040459716196,
+      "grad_norm": 0.6610363125801086,
+      "learning_rate": 1.9896026031643902e-05,
+      "loss": 2.1748,
+      "num_input_tokens_seen": 707788800,
+      "step": 1350
+    },
+    {
+      "epoch": 0.20415151870528908,
+      "grad_norm": 0.6691358089447021,
+      "learning_rate": 1.9822743918806708e-05,
+      "loss": 2.1685,
+      "num_input_tokens_seen": 713031680,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2056526328134162,
+      "grad_norm": 0.6636555790901184,
+      "learning_rate": 1.9750265636206772e-05,
+      "loss": 2.1817,
+      "num_input_tokens_seen": 718274560,
+      "step": 1370
+    },
+    {
+      "epoch": 0.20715374692154334,
+      "grad_norm": 0.7346318960189819,
+      "learning_rate": 1.9678576595160388e-05,
+      "loss": 2.1833,
+      "num_input_tokens_seen": 723517440,
+      "step": 1380
+    },
+    {
+      "epoch": 0.20865486102967046,
+      "grad_norm": 0.5943858623504639,
+      "learning_rate": 1.960766257499033e-05,
+      "loss": 2.2494,
+      "num_input_tokens_seen": 728760320,
+      "step": 1390
+    },
+    {
+      "epoch": 0.21015597513779757,
+      "grad_norm": 0.7019902467727661,
+      "learning_rate": 1.9537509711175725e-05,
+      "loss": 2.2019,
+      "num_input_tokens_seen": 734003200,
+      "step": 1400
+    },
+    {
+      "epoch": 0.21165708924592472,
+      "grad_norm": 0.5532287955284119,
+      "learning_rate": 1.946810448396509e-05,
+      "loss": 2.1975,
+      "num_input_tokens_seen": 739246080,
+      "step": 1410
+    },
+    {
+      "epoch": 0.21315820335405183,
+      "grad_norm": 0.5484923720359802,
+      "learning_rate": 1.939943370743111e-05,
+      "loss": 2.1773,
+      "num_input_tokens_seen": 744488960,
+      "step": 1420
+    },
+    {
+      "epoch": 0.21465931746217895,
+      "grad_norm": 0.6313416957855225,
+      "learning_rate": 1.9331484518947162e-05,
+      "loss": 2.1919,
+      "num_input_tokens_seen": 749731840,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2161604315703061,
+      "grad_norm": 0.5967773795127869,
+      "learning_rate": 1.926424436906652e-05,
+      "loss": 2.1957,
+      "num_input_tokens_seen": 754974720,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2176615456784332,
+      "grad_norm": 0.5972282290458679,
+      "learning_rate": 1.919770101178605e-05,
+      "loss": 2.1647,
+      "num_input_tokens_seen": 760217600,
+      "step": 1450
+    },
+    {
+      "epoch": 0.21916265978656033,
+      "grad_norm": 0.5539584755897522,
+      "learning_rate": 1.9131842495177355e-05,
+      "loss": 2.1815,
+      "num_input_tokens_seen": 765460480,
+      "step": 1460
+    },
+    {
+      "epoch": 0.22066377389468747,
+      "grad_norm": 0.7532745003700256,
+      "learning_rate": 1.9066657152369064e-05,
+      "loss": 2.2143,
+      "num_input_tokens_seen": 770703360,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2221648880028146,
+      "grad_norm": 0.537327229976654,
+      "learning_rate": 1.9002133592864807e-05,
+      "loss": 2.1541,
+      "num_input_tokens_seen": 775946240,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2236660021109417,
+      "grad_norm": 0.7467483878135681,
+      "learning_rate": 1.893826069418217e-05,
+      "loss": 2.1638,
+      "num_input_tokens_seen": 781189120,
+      "step": 1490
+    },
+    {
+      "epoch": 0.22516711621906885,
+      "grad_norm": 0.5528019070625305,
+      "learning_rate": 1.8875027593798816e-05,
+      "loss": 2.1865,
+      "num_input_tokens_seen": 786432000,
+      "step": 1500
+    },
+    {
+      "epoch": 0.22666823032719596,
+      "grad_norm": 0.6948214173316956,
+      "learning_rate": 1.8812423681392357e-05,
+      "loss": 2.166,
+      "num_input_tokens_seen": 791674880,
+      "step": 1510
+    },
+    {
+      "epoch": 0.22816934443532308,
+      "grad_norm": 0.6632937788963318,
+      "learning_rate": 1.8750438591361568e-05,
+      "loss": 2.1864,
+      "num_input_tokens_seen": 796917760,
+      "step": 1520
+    },
+    {
+      "epoch": 0.22967045854345022,
+      "grad_norm": 0.5619977116584778,
+      "learning_rate": 1.8689062195616828e-05,
+      "loss": 2.192,
+      "num_input_tokens_seen": 802160640,
+      "step": 1530
+    },
+    {
+      "epoch": 0.23117157265157734,
+      "grad_norm": 0.6643936038017273,
+      "learning_rate": 1.8628284596628512e-05,
+      "loss": 2.1749,
+      "num_input_tokens_seen": 807403520,
+      "step": 1540
+    },
+    {
+      "epoch": 0.23267268675970446,
+      "grad_norm": 0.6980964541435242,
+      "learning_rate": 1.8568096120722424e-05,
+      "loss": 2.1662,
+      "num_input_tokens_seen": 812646400,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2341738008678316,
+      "grad_norm": 0.8099634647369385,
+      "learning_rate": 1.8508487311612005e-05,
+      "loss": 2.1778,
+      "num_input_tokens_seen": 817889280,
+      "step": 1560
+    },
+    {
+      "epoch": 0.23567491497595872,
+      "grad_norm": 0.5385667085647583,
+      "learning_rate": 1.844944892415748e-05,
+      "loss": 2.1536,
+      "num_input_tokens_seen": 823132160,
+      "step": 1570
+    },
+    {
+      "epoch": 0.23717602908408583,
+      "grad_norm": 0.70493483543396,
+      "learning_rate": 1.8390971918342592e-05,
+      "loss": 2.192,
+      "num_input_tokens_seen": 828375040,
+      "step": 1580
+    },
+    {
+      "epoch": 0.23867714319221298,
+      "grad_norm": 0.5577861666679382,
+      "learning_rate": 1.833304745346004e-05,
+      "loss": 2.1941,
+      "num_input_tokens_seen": 833617920,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2401782573003401,
+      "grad_norm": 0.5603963136672974,
+      "learning_rate": 1.8275666882497067e-05,
+      "loss": 2.1543,
+      "num_input_tokens_seen": 838860800,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2401782573003401,
+      "eval_accuracy": 0.5632421652421652,
+      "eval_loss": 2.1520919799804688,
+      "eval_runtime": 94.9668,
+      "eval_samples_per_second": 3.159,
+      "eval_steps_per_second": 0.79,
+      "num_input_tokens_seen": 838860800,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2416793714084672,
+      "grad_norm": 0.6568350791931152,
+      "learning_rate": 1.821882174671319e-05,
+      "loss": 2.1729,
+      "num_input_tokens_seen": 844103680,
+      "step": 1610
+    },
+    {
+      "epoch": 0.24318048551659435,
+      "grad_norm": 0.6658337712287903,
+      "learning_rate": 1.8162503770402265e-05,
+      "loss": 2.1954,
+      "num_input_tokens_seen": 849346560,
+      "step": 1620
+    },
+    {
+      "epoch": 0.24468159962472147,
+      "grad_norm": 0.6133268475532532,
+      "learning_rate": 1.8106704855831622e-05,
+      "loss": 2.1887,
+      "num_input_tokens_seen": 854589440,
+      "step": 1630
+    },
+    {
+      "epoch": 0.24618271373284858,
+      "grad_norm": 0.6844907402992249,
+      "learning_rate": 1.805141707835109e-05,
+      "loss": 2.1772,
+      "num_input_tokens_seen": 859832320,
+      "step": 1640
+    },
+    {
+      "epoch": 0.24768382784097573,
+      "grad_norm": 0.6765184998512268,
+      "learning_rate": 1.799663268166542e-05,
+      "loss": 2.1631,
+      "num_input_tokens_seen": 865075200,
+      "step": 1650
+    },
+    {
+      "epoch": 0.24918494194910285,
+      "grad_norm": 0.7173587083816528,
+      "learning_rate": 1.7942344073263488e-05,
+      "loss": 2.1347,
+      "num_input_tokens_seen": 870318080,
+      "step": 1660
+    },
+    {
+      "epoch": 0.25068605605723,
+      "grad_norm": 0.907135546207428,
+      "learning_rate": 1.788854381999832e-05,
+      "loss": 2.1809,
+      "num_input_tokens_seen": 875560960,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2521871701653571,
+      "grad_norm": 0.6652698516845703,
+      "learning_rate": 1.783522464381198e-05,
+      "loss": 2.1558,
+      "num_input_tokens_seen": 880803840,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2536882842734842,
+      "grad_norm": 0.5732602477073669,
+      "learning_rate": 1.7782379417599864e-05,
+      "loss": 2.1788,
+      "num_input_tokens_seen": 886046720,
+      "step": 1690
+    },
+    {
+      "epoch": 0.25518939838161137,
+      "grad_norm": 0.5569264888763428,
+      "learning_rate": 1.7730001161208948e-05,
+      "loss": 2.1375,
+      "num_input_tokens_seen": 891289600,
+      "step": 1700
+    },
+    {
+      "epoch": 0.25669051248973845,
+      "grad_norm": 0.6948941946029663,
+      "learning_rate": 1.767808303756493e-05,
+      "loss": 2.148,
+      "num_input_tokens_seen": 896532480,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2581916265978656,
+      "grad_norm": 0.5902568101882935,
+      "learning_rate": 1.7626618348923443e-05,
+      "loss": 2.1651,
+      "num_input_tokens_seen": 901775360,
+      "step": 1720
+    },
+    {
+      "epoch": 0.25969274070599274,
+      "grad_norm": 0.6689503192901611,
+      "learning_rate": 1.757560053324057e-05,
+      "loss": 2.1761,
+      "num_input_tokens_seen": 907018240,
+      "step": 1730
+    },
+    {
+      "epoch": 0.26119385481411983,
+      "grad_norm": 0.6513310074806213,
+      "learning_rate": 1.7525023160658283e-05,
+      "loss": 2.195,
+      "num_input_tokens_seen": 912261120,
+      "step": 1740
+    },
+    {
+      "epoch": 0.262694968922247,
+      "grad_norm": 0.5774445533752441,
+      "learning_rate": 1.7474879930100483e-05,
+      "loss": 2.1681,
+      "num_input_tokens_seen": 917504000,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2641960830303741,
+      "grad_norm": 0.583354115486145,
+      "learning_rate": 1.7425164665975574e-05,
+      "loss": 2.1619,
+      "num_input_tokens_seen": 922746880,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2656971971385012,
+      "grad_norm": 0.6275432109832764,
+      "learning_rate": 1.7375871314981654e-05,
+      "loss": 2.1819,
+      "num_input_tokens_seen": 927989760,
+      "step": 1770
+    },
+    {
+      "epoch": 0.26719831124662835,
+      "grad_norm": 0.6457045674324036,
+      "learning_rate": 1.732699394301053e-05,
+      "loss": 2.135,
+      "num_input_tokens_seen": 933232640,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2686994253547555,
+      "grad_norm": 0.6066982746124268,
+      "learning_rate": 1.7278526732147035e-05,
+      "loss": 2.1782,
+      "num_input_tokens_seen": 938475520,
+      "step": 1790
+    },
+    {
+      "epoch": 0.2702005394628826,
+      "grad_norm": 0.6707212328910828,
+      "learning_rate": 1.7230463977760115e-05,
+      "loss": 2.2126,
+      "num_input_tokens_seen": 943718400,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2717016535710097,
+      "grad_norm": 0.6359196305274963,
+      "learning_rate": 1.71828000856825e-05,
+      "loss": 2.1631,
+      "num_input_tokens_seen": 948961280,
+      "step": 1810
+    },
+    {
+      "epoch": 0.27320276767913687,
+      "grad_norm": 0.5916318297386169,
+      "learning_rate": 1.713552956947563e-05,
+      "loss": 2.1535,
+      "num_input_tokens_seen": 954204160,
+      "step": 1820
+    },
+    {
+      "epoch": 0.27470388178726396,
+      "grad_norm": 0.7036842107772827,
+      "learning_rate": 1.708864704777699e-05,
+      "loss": 2.153,
+      "num_input_tokens_seen": 959447040,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2762049958953911,
+      "grad_norm": 0.6982750296592712,
+      "learning_rate": 1.704214724172678e-05,
+      "loss": 2.1806,
+      "num_input_tokens_seen": 964689920,
+      "step": 1840
+    },
+    {
+      "epoch": 0.27770611000351825,
+      "grad_norm": 0.6548587083816528,
+      "learning_rate": 1.699602497247121e-05,
+      "loss": 2.1547,
+      "num_input_tokens_seen": 969932800,
+      "step": 1850
+    },
+    {
+      "epoch": 0.27920722411164534,
+      "grad_norm": 0.6462974548339844,
+      "learning_rate": 1.6950275158739732e-05,
+      "loss": 2.1697,
+      "num_input_tokens_seen": 975175680,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2807083382197725,
+      "grad_norm": 0.5359413623809814,
+      "learning_rate": 1.690489281449364e-05,
+      "loss": 2.158,
+      "num_input_tokens_seen": 980418560,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2822094523278996,
+      "grad_norm": 0.6209542155265808,
+      "learning_rate": 1.6859873046643506e-05,
+      "loss": 2.1345,
+      "num_input_tokens_seen": 985661440,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2837105664360267,
+      "grad_norm": 0.602796733379364,
+      "learning_rate": 1.681521105283317e-05,
+      "loss": 2.1774,
+      "num_input_tokens_seen": 990904320,
+      "step": 1890
+    },
+    {
+      "epoch": 0.28521168054415386,
+      "grad_norm": 0.6523065567016602,
+      "learning_rate": 1.6770902119287942e-05,
+      "loss": 2.2011,
+      "num_input_tokens_seen": 996147200,
+      "step": 1900
+    },
+    {
+      "epoch": 0.286712794652281,
+      "grad_norm": 0.5961717963218689,
+      "learning_rate": 1.672694161872488e-05,
+      "loss": 2.1722,
+      "num_input_tokens_seen": 1001390080,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2882139087604081,
+      "grad_norm": 0.5899094343185425,
+      "learning_rate": 1.6683325008322934e-05,
+      "loss": 2.1353,
+      "num_input_tokens_seen": 1006632960,
+      "step": 1920
+    },
+    {
+      "epoch": 0.28971502286853523,
+      "grad_norm": 0.590152382850647,
+      "learning_rate": 1.6640047827751115e-05,
+      "loss": 2.186,
+      "num_input_tokens_seen": 1011875840,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2912161369766624,
+      "grad_norm": 0.6230789422988892,
+      "learning_rate": 1.6597105697252553e-05,
+      "loss": 2.1719,
+      "num_input_tokens_seen": 1017118720,
+      "step": 1940
+    },
+    {
+      "epoch": 0.29271725108478946,
+      "grad_norm": 0.6433463096618652,
+      "learning_rate": 1.655449431578271e-05,
+      "loss": 2.1711,
+      "num_input_tokens_seen": 1022361600,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2942183651929166,
+      "grad_norm": 0.6224901080131531,
+      "learning_rate": 1.6512209459199873e-05,
+      "loss": 2.1473,
+      "num_input_tokens_seen": 1027604480,
+      "step": 1960
+    },
+    {
+      "epoch": 0.29571947930104375,
+      "grad_norm": 0.6015456914901733,
+      "learning_rate": 1.6470246978506274e-05,
+      "loss": 2.1788,
+      "num_input_tokens_seen": 1032847360,
+      "step": 1970
+    },
+    {
+      "epoch": 0.2972205934091709,
+      "grad_norm": 0.7380375266075134,
+      "learning_rate": 1.6428602798138068e-05,
+      "loss": 2.1566,
+      "num_input_tokens_seen": 1038090240,
+      "step": 1980
+    },
+    {
+      "epoch": 0.298721707517298,
+      "grad_norm": 0.5713395476341248,
+      "learning_rate": 1.638727291430271e-05,
+      "loss": 2.1596,
+      "num_input_tokens_seen": 1043333120,
+      "step": 1990
+    },
+    {
+      "epoch": 0.30022282162542513,
+      "grad_norm": 0.6106094717979431,
+      "learning_rate": 1.634625339336204e-05,
+      "loss": 2.1532,
+      "num_input_tokens_seen": 1048576000,
+      "step": 2000
+    },
+    {
+      "epoch": 0.30022282162542513,
+      "eval_accuracy": 0.565023199023199,
+      "eval_loss": 2.1400792598724365,
+      "eval_runtime": 95.5183,
+      "eval_samples_per_second": 3.141,
+      "eval_steps_per_second": 0.785,
+      "num_input_tokens_seen": 1048576000,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3017239357335523,
+      "grad_norm": 0.6882668733596802,
+      "learning_rate": 1.6305540370259715e-05,
+      "loss": 2.1681,
+      "num_input_tokens_seen": 1053818880,
+      "step": 2010
+    },
+    {
+      "epoch": 0.30322504984167936,
+      "grad_norm": 0.5827280282974243,
+      "learning_rate": 1.6265130046991463e-05,
+      "loss": 2.1699,
+      "num_input_tokens_seen": 1059061760,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3047261639498065,
+      "grad_norm": 0.5182669162750244,
+      "learning_rate": 1.6225018691116927e-05,
+      "loss": 2.1495,
+      "num_input_tokens_seen": 1064304640,
+      "step": 2030
+    },
+    {
+      "epoch": 0.30622727805793365,
+      "grad_norm": 0.8182618618011475,
+      "learning_rate": 1.618520263431155e-05,
+      "loss": 2.1837,
+      "num_input_tokens_seen": 1069547520,
+      "step": 2040
+    },
+    {
+      "epoch": 0.30772839216606074,
+      "grad_norm": 0.5096232891082764,
+      "learning_rate": 1.6145678270957475e-05,
+      "loss": 2.1489,
+      "num_input_tokens_seen": 1074790400,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3092295062741879,
+      "grad_norm": 0.6253472566604614,
+      "learning_rate": 1.6106442056772048e-05,
+      "loss": 2.1699,
+      "num_input_tokens_seen": 1080033280,
+      "step": 2060
+    },
+    {
+      "epoch": 0.310730620382315,
+      "grad_norm": 0.5389710068702698,
+      "learning_rate": 1.6067490507472832e-05,
+      "loss": 2.1715,
+      "num_input_tokens_seen": 1085276160,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3122317344904421,
+      "grad_norm": 0.6460773348808289,
+      "learning_rate": 1.6028820197477945e-05,
+      "loss": 2.1601,
+      "num_input_tokens_seen": 1090519040,
+      "step": 2080
+    },
+    {
+      "epoch": 0.31373284859856926,
+      "grad_norm": 0.7538013458251953,
+      "learning_rate": 1.5990427758640674e-05,
+      "loss": 2.1677,
+      "num_input_tokens_seen": 1095761920,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3152339627066964,
+      "grad_norm": 0.6462552547454834,
+      "learning_rate": 1.5952309879017224e-05,
+      "loss": 2.1805,
+      "num_input_tokens_seen": 1101004800,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3167350768148235,
+      "grad_norm": 0.6027917861938477,
+      "learning_rate": 1.5914463301666658e-05,
+      "loss": 2.1415,
+      "num_input_tokens_seen": 1106247680,
+      "step": 2110
+    },
+    {
+      "epoch": 0.31823619092295063,
+      "grad_norm": 0.6895632147789001,
+      "learning_rate": 1.5876884823482006e-05,
+      "loss": 2.1587,
+      "num_input_tokens_seen": 1111490560,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3197373050310778,
+      "grad_norm": 0.6652107834815979,
+      "learning_rate": 1.5839571294051578e-05,
+      "loss": 2.1444,
+      "num_input_tokens_seen": 1116733440,
+      "step": 2130
+    },
+    {
+      "epoch": 0.32123841913920487,
+      "grad_norm": 0.7466081380844116,
+      "learning_rate": 1.580251961454958e-05,
+      "loss": 2.1251,
+      "num_input_tokens_seen": 1121976320,
+      "step": 2140
+    },
+    {
+      "epoch": 0.322739533247332,
+      "grad_norm": 0.5937127470970154,
+      "learning_rate": 1.5765726736655174e-05,
+      "loss": 2.1489,
+      "num_input_tokens_seen": 1127219200,
+      "step": 2150
+    },
+    {
+      "epoch": 0.32424064735545916,
+      "grad_norm": 0.6133046746253967,
+      "learning_rate": 1.5729189661499015e-05,
+      "loss": 2.1507,
+      "num_input_tokens_seen": 1132462080,
+      "step": 2160
+    },
+    {
+      "epoch": 0.32574176146358624,
+      "grad_norm": 0.7476202845573425,
+      "learning_rate": 1.5692905438636622e-05,
+      "loss": 2.1691,
+      "num_input_tokens_seen": 1137704960,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3272428755717134,
+      "grad_norm": 0.7822461724281311,
+      "learning_rate": 1.5656871165047605e-05,
+      "loss": 2.1522,
+      "num_input_tokens_seen": 1142947840,
+      "step": 2180
+    },
+    {
+      "epoch": 0.32874398967984053,
+      "grad_norm": 0.6776607632637024,
+      "learning_rate": 1.5621083984160086e-05,
+      "loss": 2.142,
+      "num_input_tokens_seen": 1148190720,
+      "step": 2190
+    },
+    {
+      "epoch": 0.3302451037879676,
+      "grad_norm": 0.7508281469345093,
+      "learning_rate": 1.558554108489952e-05,
+      "loss": 2.1723,
+      "num_input_tokens_seen": 1153433600,
+      "step": 2200
+    },
+    {
+      "epoch": 0.33174621789609476,
+      "grad_norm": 0.7543405890464783,
+      "learning_rate": 1.5550239700761246e-05,
+      "loss": 2.1442,
+      "num_input_tokens_seen": 1158676480,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3332473320042219,
+      "grad_norm": 0.9116769433021545,
+      "learning_rate": 1.5515177108906e-05,
+      "loss": 2.1916,
+      "num_input_tokens_seen": 1163919360,
+      "step": 2220
+    },
+    {
+      "epoch": 0.334748446112349,
+      "grad_norm": 0.7372170686721802,
+      "learning_rate": 1.5480350629277787e-05,
+      "loss": 2.1546,
+      "num_input_tokens_seen": 1169162240,
+      "step": 2230
+    },
+    {
+      "epoch": 0.33624956022047614,
+      "grad_norm": 0.5467498302459717,
+      "learning_rate": 1.5445757623743444e-05,
+      "loss": 2.1919,
+      "num_input_tokens_seen": 1174405120,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3377506743286033,
+      "grad_norm": 0.5791755318641663,
+      "learning_rate": 1.5411395495253218e-05,
+      "loss": 2.1782,
+      "num_input_tokens_seen": 1179648000,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3392517884367304,
+      "grad_norm": 0.7308974862098694,
+      "learning_rate": 1.5377261687021863e-05,
+      "loss": 2.1682,
+      "num_input_tokens_seen": 1184890880,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3407529025448575,
+      "grad_norm": 0.6968861222267151,
+      "learning_rate": 1.5343353681729532e-05,
+      "loss": 2.1724,
+      "num_input_tokens_seen": 1190133760,
+      "step": 2270
+    },
+    {
+      "epoch": 0.34225401665298466,
+      "grad_norm": 0.6837549805641174,
+      "learning_rate": 1.5309669000742003e-05,
+      "loss": 2.1783,
+      "num_input_tokens_seen": 1195376640,
+      "step": 2280
+    },
+    {
+      "epoch": 0.34375513076111175,
+      "grad_norm": 0.6059392094612122,
+      "learning_rate": 1.5276205203349658e-05,
+      "loss": 2.1454,
+      "num_input_tokens_seen": 1200619520,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3452562448692389,
+      "grad_norm": 0.5880239605903625,
+      "learning_rate": 1.5242959886024648e-05,
+      "loss": 2.1428,
+      "num_input_tokens_seen": 1205862400,
+      "step": 2300
+    },
+    {
+      "epoch": 0.34675735897736604,
+      "grad_norm": 0.6651424169540405,
+      "learning_rate": 1.5209930681695804e-05,
+      "loss": 2.1518,
+      "num_input_tokens_seen": 1211105280,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3482584730854931,
+      "grad_norm": 0.5734732151031494,
+      "learning_rate": 1.5177115259040729e-05,
+      "loss": 2.1327,
+      "num_input_tokens_seen": 1216348160,
+      "step": 2320
+    },
+    {
+      "epoch": 0.34975958719362027,
+      "grad_norm": 0.6250523328781128,
+      "learning_rate": 1.514451132179463e-05,
+      "loss": 2.1459,
+      "num_input_tokens_seen": 1221591040,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3512607013017474,
+      "grad_norm": 0.6869091391563416,
+      "learning_rate": 1.5112116608075403e-05,
+      "loss": 2.1352,
+      "num_input_tokens_seen": 1226833920,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3527618154098745,
+      "grad_norm": 0.8655930161476135,
+      "learning_rate": 1.5079928889724547e-05,
+      "loss": 2.1581,
+      "num_input_tokens_seen": 1232076800,
+      "step": 2350
+    },
+    {
+      "epoch": 0.35426292951800165,
+      "grad_norm": 0.6587986946105957,
+      "learning_rate": 1.5047945971663434e-05,
+      "loss": 2.1783,
+      "num_input_tokens_seen": 1237319680,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3557640436261288,
+      "grad_norm": 0.6270188689231873,
+      "learning_rate": 1.501616569126455e-05,
+      "loss": 2.1665,
+      "num_input_tokens_seen": 1242562560,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3572651577342559,
+      "grad_norm": 0.6337898969650269,
+      "learning_rate": 1.4984585917737282e-05,
+      "loss": 2.177,
+      "num_input_tokens_seen": 1247805440,
+      "step": 2380
+    },
+    {
+      "epoch": 0.358766271842383,
+      "grad_norm": 0.6973663568496704,
+      "learning_rate": 1.4953204551527831e-05,
+      "loss": 2.1493,
+      "num_input_tokens_seen": 1253048320,
+      "step": 2390
+    },
+    {
+      "epoch": 0.36026738595051017,
+      "grad_norm": 0.5605215430259705,
+      "learning_rate": 1.4922019523732929e-05,
+      "loss": 2.1688,
+      "num_input_tokens_seen": 1258291200,
+      "step": 2400
+    },
+    {
+      "epoch": 0.36026738595051017,
+      "eval_accuracy": 0.5663190883190883,
+      "eval_loss": 2.1306827068328857,
+      "eval_runtime": 94.9906,
+      "eval_samples_per_second": 3.158,
+      "eval_steps_per_second": 0.79,
+      "num_input_tokens_seen": 1258291200,
+      "step": 2400
+    },
+    {
+      "epoch": 0.36176850005863725,
+      "grad_norm": 0.7090189456939697,
+      "learning_rate": 1.4891028795526912e-05,
+      "loss": 2.1446,
+      "num_input_tokens_seen": 1263534080,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3632696141667644,
+      "grad_norm": 0.7129746079444885,
+      "learning_rate": 1.4860230357601855e-05,
+      "loss": 2.178,
+      "num_input_tokens_seen": 1268776960,
+      "step": 2420
+    },
+    {
+      "epoch": 0.36477072827489154,
+      "grad_norm": 0.5609286427497864,
+      "learning_rate": 1.4829622229620384e-05,
+      "loss": 2.1846,
+      "num_input_tokens_seen": 1274019840,
+      "step": 2430
+    },
+    {
+      "epoch": 0.36627184238301863,
+      "grad_norm": 0.6806447505950928,
+      "learning_rate": 1.4799202459680824e-05,
+      "loss": 2.156,
+      "num_input_tokens_seen": 1279262720,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3677729564911458,
+      "grad_norm": 0.5787296295166016,
+      "learning_rate": 1.4768969123794386e-05,
+      "loss": 2.128,
+      "num_input_tokens_seen": 1284505600,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3692740705992729,
+      "grad_norm": 0.7626485824584961,
+      "learning_rate": 1.4738920325374027e-05,
+      "loss": 2.1607,
+      "num_input_tokens_seen": 1289748480,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3707751847074,
+      "grad_norm": 0.6390326619148254,
+      "learning_rate": 1.4709054194734743e-05,
+      "loss": 2.1821,
+      "num_input_tokens_seen": 1294991360,
+      "step": 2470
+    },
+    {
+      "epoch": 0.37227629881552715,
+      "grad_norm": 0.5435773730278015,
+      "learning_rate": 1.4679368888604919e-05,
+      "loss": 2.1762,
+      "num_input_tokens_seen": 1300234240,
+      "step": 2480
+    },
+    {
+      "epoch": 0.3737774129236543,
+      "grad_norm": 0.5910453200340271,
+      "learning_rate": 1.464986258964849e-05,
+      "loss": 2.1586,
+      "num_input_tokens_seen": 1305477120,
+      "step": 2490
+    },
+    {
+      "epoch": 0.3752785270317814,
+      "grad_norm": 0.5591862797737122,
+      "learning_rate": 1.4620533505997654e-05,
+      "loss": 2.1512,
+      "num_input_tokens_seen": 1310720000,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3767796411399085,
+      "grad_norm": 0.6790704727172852,
+      "learning_rate": 1.459137987079579e-05,
+      "loss": 2.1696,
+      "num_input_tokens_seen": 1315962880,
+      "step": 2510
+    },
+    {
+      "epoch": 0.37828075524803567,
+      "grad_norm": 0.649533212184906,
+      "learning_rate": 1.4562399941750401e-05,
+      "loss": 2.1357,
+      "num_input_tokens_seen": 1321205760,
+      "step": 2520
+    },
+    {
+      "epoch": 0.37978186935616276,
+      "grad_norm": 0.5689984560012817,
+      "learning_rate": 1.453359200069576e-05,
+      "loss": 2.1972,
+      "num_input_tokens_seen": 1326448640,
+      "step": 2530
+    },
+    {
+      "epoch": 0.3812829834642899,
+      "grad_norm": 0.7337691187858582,
+      "learning_rate": 1.4504954353165044e-05,
+      "loss": 2.1198,
+      "num_input_tokens_seen": 1331691520,
+      "step": 2540
+    },
+    {
+      "epoch": 0.38278409757241705,
+      "grad_norm": 0.550014853477478,
+      "learning_rate": 1.447648532797172e-05,
+      "loss": 2.1267,
+      "num_input_tokens_seen": 1336934400,
+      "step": 2550
+    },
+    {
+      "epoch": 0.38428521168054414,
+      "grad_norm": 0.6532608270645142,
+      "learning_rate": 1.4448183276799891e-05,
+      "loss": 2.1735,
+      "num_input_tokens_seen": 1342177280,
+      "step": 2560
+    },
+    {
+      "epoch": 0.3857863257886713,
+      "grad_norm": 0.5801606774330139,
+      "learning_rate": 1.4420046573803481e-05,
+      "loss": 2.113,
+      "num_input_tokens_seen": 1347420160,
+      "step": 2570
+    },
+    {
+      "epoch": 0.3872874398967984,
+      "grad_norm": 0.6582128405570984,
+      "learning_rate": 1.4392073615213913e-05,
+      "loss": 2.1431,
+      "num_input_tokens_seen": 1352663040,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3887885540049255,
+      "grad_norm": 0.6207413077354431,
+      "learning_rate": 1.436426281895616e-05,
+      "loss": 2.1683,
+      "num_input_tokens_seen": 1357905920,
+      "step": 2590
+    },
+    {
+      "epoch": 0.39028966811305266,
+      "grad_norm": 0.5679153800010681,
+      "learning_rate": 1.4336612624272917e-05,
+      "loss": 2.1194,
+      "num_input_tokens_seen": 1363148800,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3917907822211798,
+      "grad_norm": 0.6869505643844604,
+      "learning_rate": 1.4309121491356698e-05,
+      "loss": 2.1033,
+      "num_input_tokens_seen": 1368391680,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3932918963293069,
+      "grad_norm": 0.661869466304779,
+      "learning_rate": 1.4281787900989672e-05,
+      "loss": 2.161,
+      "num_input_tokens_seen": 1373634560,
+      "step": 2620
+    },
+    {
+      "epoch": 0.39479301043743403,
+      "grad_norm": 0.6266052722930908,
+      "learning_rate": 1.4254610354191023e-05,
+      "loss": 2.1838,
+      "num_input_tokens_seen": 1378877440,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3962941245455612,
+      "grad_norm": 0.6248527765274048,
+      "learning_rate": 1.4227587371871679e-05,
+      "loss": 2.1255,
+      "num_input_tokens_seen": 1384120320,
+      "step": 2640
+    },
+    {
+      "epoch": 0.39779523865368827,
+      "grad_norm": 0.7318665385246277,
+      "learning_rate": 1.4200717494496206e-05,
+      "loss": 2.1487,
+      "num_input_tokens_seen": 1389363200,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3992963527618154,
+      "grad_norm": 0.7174627780914307,
+      "learning_rate": 1.4173999281751702e-05,
+      "loss": 2.1375,
+      "num_input_tokens_seen": 1394606080,
+      "step": 2660
+    },
+    {
+      "epoch": 0.40079746686994255,
+      "grad_norm": 0.6472954154014587,
+      "learning_rate": 1.4147431312223518e-05,
+      "loss": 2.1546,
+      "num_input_tokens_seen": 1399848960,
+      "step": 2670
+    },
+    {
+      "epoch": 0.40229858097806964,
+      "grad_norm": 0.588234543800354,
+      "learning_rate": 1.4121012183077632e-05,
+      "loss": 2.1437,
+      "num_input_tokens_seen": 1405091840,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4037996950861968,
+      "grad_norm": 0.5872112512588501,
+      "learning_rate": 1.4094740509749542e-05,
+      "loss": 2.1672,
+      "num_input_tokens_seen": 1410334720,
+      "step": 2690
+    },
+    {
+      "epoch": 0.40530080919432393,
+      "grad_norm": 0.6200748682022095,
+      "learning_rate": 1.406861492563948e-05,
+      "loss": 2.1434,
+      "num_input_tokens_seen": 1415577600,
+      "step": 2700
+    },
+    {
+      "epoch": 0.406801923302451,
+      "grad_norm": 0.7556484937667847,
+      "learning_rate": 1.4042634081813838e-05,
+      "loss": 2.1603,
+      "num_input_tokens_seen": 1420820480,
+      "step": 2710
+    },
+    {
+      "epoch": 0.40830303741057816,
+      "grad_norm": 0.6454352736473083,
+      "learning_rate": 1.4016796646712618e-05,
+      "loss": 2.1783,
+      "num_input_tokens_seen": 1426063360,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4098041515187053,
+      "grad_norm": 0.5394317507743835,
+      "learning_rate": 1.3991101305862803e-05,
+      "loss": 2.1646,
+      "num_input_tokens_seen": 1431306240,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4113052656268324,
+      "grad_norm": 0.6699331998825073,
+      "learning_rate": 1.396554676159745e-05,
+      "loss": 2.1597,
+      "num_input_tokens_seen": 1436549120,
+      "step": 2740
+    },
+    {
+      "epoch": 0.41280637973495954,
+      "grad_norm": 0.6491063237190247,
+      "learning_rate": 1.3940131732780461e-05,
+      "loss": 2.1236,
+      "num_input_tokens_seen": 1441792000,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4143074938430867,
+      "grad_norm": 0.6439197063446045,
+      "learning_rate": 1.3914854954536792e-05,
+      "loss": 2.164,
+      "num_input_tokens_seen": 1447034880,
+      "step": 2760
+    },
+    {
+      "epoch": 0.41580860795121377,
+      "grad_norm": 0.7333915829658508,
+      "learning_rate": 1.3889715177988056e-05,
+      "loss": 2.1417,
+      "num_input_tokens_seen": 1452277760,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4173097220593409,
+      "grad_norm": 0.5554391145706177,
+      "learning_rate": 1.386471116999334e-05,
+      "loss": 2.1821,
+      "num_input_tokens_seen": 1457520640,
+      "step": 2780
+    },
+    {
+      "epoch": 0.41881083616746806,
+      "grad_norm": 0.6607913374900818,
+      "learning_rate": 1.3839841712895161e-05,
+      "loss": 2.1341,
+      "num_input_tokens_seen": 1462763520,
+      "step": 2790
+    },
+    {
+      "epoch": 0.42031195027559515,
+      "grad_norm": 0.6506746411323547,
+      "learning_rate": 1.381510560427038e-05,
+      "loss": 2.1443,
+      "num_input_tokens_seen": 1468006400,
+      "step": 2800
+    },
+    {
+      "epoch": 0.42031195027559515,
+      "eval_accuracy": 0.5676076516076516,
+      "eval_loss": 2.1226789951324463,
+      "eval_runtime": 94.7272,
+      "eval_samples_per_second": 3.167,
+      "eval_steps_per_second": 0.792,
+      "num_input_tokens_seen": 1468006400,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4218130643837223,
+      "grad_norm": 0.6999175548553467,
+      "learning_rate": 1.3790501656686045e-05,
+      "loss": 2.1222,
+      "num_input_tokens_seen": 1473249280,
+      "step": 2810
+    },
+    {
+      "epoch": 0.42331417849184944,
+      "grad_norm": 0.6637996435165405,
+      "learning_rate": 1.3766028697459948e-05,
+      "loss": 2.1752,
+      "num_input_tokens_seen": 1478492160,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4248152925999765,
+      "grad_norm": 0.675205409526825,
+      "learning_rate": 1.3741685568425882e-05,
+      "loss": 2.1379,
+      "num_input_tokens_seen": 1483735040,
+      "step": 2830
+    },
+    {
+      "epoch": 0.42631640670810367,
+      "grad_norm": 0.7251035571098328,
+      "learning_rate": 1.371747112570342e-05,
+      "loss": 2.1499,
+      "num_input_tokens_seen": 1488977920,
+      "step": 2840
+    },
+    {
+      "epoch": 0.4278175208162308,
+      "grad_norm": 0.8489077687263489,
+      "learning_rate": 1.3693384239472163e-05,
+      "loss": 2.1291,
+      "num_input_tokens_seen": 1494220800,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4293186349243579,
+      "grad_norm": 0.5617297887802124,
+      "learning_rate": 1.3669423793750302e-05,
+      "loss": 2.145,
+      "num_input_tokens_seen": 1499463680,
+      "step": 2860
+    },
+    {
+      "epoch": 0.43081974903248504,
+      "grad_norm": 0.5285384654998779,
+      "learning_rate": 1.3645588686177466e-05,
+      "loss": 2.1492,
+      "num_input_tokens_seen": 1504706560,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4323208631406122,
+      "grad_norm": 0.6254436373710632,
+      "learning_rate": 1.3621877827801699e-05,
+      "loss": 2.1196,
+      "num_input_tokens_seen": 1509949440,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4338219772487393,
+      "grad_norm": 0.5884484052658081,
+      "learning_rate": 1.3598290142870485e-05,
+      "loss": 2.1484,
+      "num_input_tokens_seen": 1515192320,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4353230913568664,
+      "grad_norm": 0.7654128074645996,
+      "learning_rate": 1.357482456862576e-05,
+      "loss": 2.1568,
+      "num_input_tokens_seen": 1520435200,
+      "step": 2900
+    },
+    {
+      "epoch": 0.43682420546499356,
+      "grad_norm": 0.839947521686554,
+      "learning_rate": 1.3551480055102794e-05,
+      "loss": 2.1651,
+      "num_input_tokens_seen": 1525678080,
+      "step": 2910
+    },
+    {
+      "epoch": 0.43832531957312065,
+      "grad_norm": 0.828296959400177,
+      "learning_rate": 1.3528255564932864e-05,
+      "loss": 2.1342,
+      "num_input_tokens_seen": 1530920960,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4398264336812478,
+      "grad_norm": 0.5922654271125793,
+      "learning_rate": 1.3505150073149634e-05,
+      "loss": 2.1632,
+      "num_input_tokens_seen": 1536163840,
+      "step": 2930
+    },
+    {
+      "epoch": 0.44132754778937494,
+      "grad_norm": 0.6538131833076477,
+      "learning_rate": 1.3482162566999154e-05,
+      "loss": 2.1633,
+      "num_input_tokens_seen": 1541406720,
+      "step": 2940
+    },
+    {
+      "epoch": 0.44282866189750203,
+      "grad_norm": 0.7411973476409912,
+      "learning_rate": 1.3459292045753423e-05,
+      "loss": 2.1349,
+      "num_input_tokens_seen": 1546649600,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4443297760056292,
+      "grad_norm": 0.6776240468025208,
+      "learning_rate": 1.3436537520527398e-05,
+      "loss": 2.1368,
+      "num_input_tokens_seen": 1551892480,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4458308901137563,
+      "grad_norm": 0.6425411105155945,
+      "learning_rate": 1.3413898014099404e-05,
+      "loss": 2.1358,
+      "num_input_tokens_seen": 1557135360,
+      "step": 2970
+    },
+    {
+      "epoch": 0.4473320042218834,
+      "grad_norm": 0.6736078858375549,
+      "learning_rate": 1.3391372560734868e-05,
+      "loss": 2.1154,
+      "num_input_tokens_seen": 1562378240,
+      "step": 2980
+    },
+    {
+      "epoch": 0.44883311833001055,
+      "grad_norm": 0.6204555630683899,
+      "learning_rate": 1.3368960206013277e-05,
+      "loss": 2.1744,
+      "num_input_tokens_seen": 1567621120,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4503342324381377,
+      "grad_norm": 0.6431263089179993,
+      "learning_rate": 1.3346660006658346e-05,
+      "loss": 2.1463,
+      "num_input_tokens_seen": 1572864000,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4518353465462648,
+      "grad_norm": 0.6702350378036499,
+      "learning_rate": 1.332447103037125e-05,
+      "loss": 2.1422,
+      "num_input_tokens_seen": 1578106880,
+      "step": 3010
+    },
+    {
+      "epoch": 0.4533364606543919,
+      "grad_norm": 0.5336691737174988,
+      "learning_rate": 1.330239235566693e-05,
+      "loss": 2.1515,
+      "num_input_tokens_seen": 1583349760,
+      "step": 3020
+    },
+    {
+      "epoch": 0.45483757476251907,
+      "grad_norm": 0.6110213398933411,
+      "learning_rate": 1.3280423071713355e-05,
+      "loss": 2.1564,
+      "num_input_tokens_seen": 1588592640,
+      "step": 3030
+    },
+    {
+      "epoch": 0.45633868887064616,
+      "grad_norm": 0.6357016563415527,
+      "learning_rate": 1.3258562278173699e-05,
+      "loss": 2.1591,
+      "num_input_tokens_seen": 1593835520,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4578398029787733,
+      "grad_norm": 0.609986424446106,
+      "learning_rate": 1.3236809085051366e-05,
+      "loss": 2.1862,
+      "num_input_tokens_seen": 1599078400,
+      "step": 3050
+    },
+    {
+      "epoch": 0.45934091708690045,
+      "grad_norm": 0.8125211000442505,
+      "learning_rate": 1.3215162612537805e-05,
+      "loss": 2.121,
+      "num_input_tokens_seen": 1604321280,
+      "step": 3060
+    },
+    {
+      "epoch": 0.46084203119502753,
+      "grad_norm": 0.5937681198120117,
+      "learning_rate": 1.3193621990863051e-05,
+      "loss": 2.1548,
+      "num_input_tokens_seen": 1609564160,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4623431453031547,
+      "grad_norm": 0.6638383865356445,
+      "learning_rate": 1.3172186360148932e-05,
+      "loss": 2.1251,
+      "num_input_tokens_seen": 1614807040,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4638442594112818,
+      "grad_norm": 0.6757727861404419,
+      "learning_rate": 1.3150854870264907e-05,
+      "loss": 2.1388,
+      "num_input_tokens_seen": 1620049920,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4653453735194089,
+      "grad_norm": 0.5711122155189514,
+      "learning_rate": 1.3129626680686454e-05,
+      "loss": 2.1498,
+      "num_input_tokens_seen": 1625292800,
+      "step": 3100
+    },
+    {
+      "epoch": 0.46684648762753606,
+      "grad_norm": 0.6421942114830017,
+      "learning_rate": 1.3108500960355947e-05,
+      "loss": 2.149,
+      "num_input_tokens_seen": 1630535680,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4683476017356632,
+      "grad_norm": 0.7078666090965271,
+      "learning_rate": 1.3087476887546022e-05,
+      "loss": 2.148,
+      "num_input_tokens_seen": 1635778560,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4698487158437903,
+      "grad_norm": 0.69465172290802,
+      "learning_rate": 1.3066553649725324e-05,
+      "loss": 2.1375,
+      "num_input_tokens_seen": 1641021440,
+      "step": 3130
+    },
+    {
+      "epoch": 0.47134982995191743,
+      "grad_norm": 0.6873762011528015,
+      "learning_rate": 1.304573044342661e-05,
+      "loss": 2.1458,
+      "num_input_tokens_seen": 1646264320,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4728509440600446,
+      "grad_norm": 0.6284067034721375,
+      "learning_rate": 1.302500647411715e-05,
+      "loss": 2.1442,
+      "num_input_tokens_seen": 1651507200,
+      "step": 3150
+    },
+    {
+      "epoch": 0.47435205816817166,
+      "grad_norm": 0.8749895095825195,
+      "learning_rate": 1.3004380956071416e-05,
+      "loss": 2.1304,
+      "num_input_tokens_seen": 1656750080,
+      "step": 3160
+    },
+    {
+      "epoch": 0.4758531722762988,
+      "grad_norm": 0.6920912265777588,
+      "learning_rate": 1.298385311224594e-05,
+      "loss": 2.1086,
+      "num_input_tokens_seen": 1661992960,
+      "step": 3170
+    },
+    {
+      "epoch": 0.47735428638442595,
+      "grad_norm": 0.6951104402542114,
+      "learning_rate": 1.296342217415636e-05,
+      "loss": 2.1271,
+      "num_input_tokens_seen": 1667235840,
+      "step": 3180
+    },
+    {
+      "epoch": 0.47885540049255304,
+      "grad_norm": 0.784321665763855,
+      "learning_rate": 1.2943087381756598e-05,
+      "loss": 2.1074,
+      "num_input_tokens_seen": 1672478720,
+      "step": 3190
+    },
+    {
+      "epoch": 0.4803565146006802,
+      "grad_norm": 0.6487578749656677,
+      "learning_rate": 1.2922847983320086e-05,
+      "loss": 2.1105,
+      "num_input_tokens_seen": 1677721600,
+      "step": 3200
+    },
+    {
+      "epoch": 0.4803565146006802,
+      "eval_accuracy": 0.5689149369149369,
+      "eval_loss": 2.115828514099121,
+      "eval_runtime": 94.253,
+      "eval_samples_per_second": 3.183,
+      "eval_steps_per_second": 0.796,
+      "num_input_tokens_seen": 1677721600,
+      "step": 3200
+    },
+    {
+      "epoch": 0.48185762870880733,
+      "grad_norm": 0.7569764852523804,
+      "learning_rate": 1.2902703235323062e-05,
+      "loss": 2.1195,
+      "num_input_tokens_seen": 1682964480,
+      "step": 3210
+    },
+    {
+      "epoch": 0.4833587428169344,
+      "grad_norm": 0.5782780051231384,
+      "learning_rate": 1.2882652402329833e-05,
+      "loss": 2.1186,
+      "num_input_tokens_seen": 1688207360,
+      "step": 3220
+    },
+    {
+      "epoch": 0.48485985692506156,
+      "grad_norm": 0.6713874936103821,
+      "learning_rate": 1.286269475688002e-05,
+      "loss": 2.1446,
+      "num_input_tokens_seen": 1693450240,
+      "step": 3230
+    },
+    {
+      "epoch": 0.4863609710331887,
+      "grad_norm": 0.6497519016265869,
+      "learning_rate": 1.2842829579377681e-05,
+      "loss": 2.1245,
+      "num_input_tokens_seen": 1698693120,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4878620851413158,
+      "grad_norm": 0.6722620129585266,
+      "learning_rate": 1.2823056157982359e-05,
+      "loss": 2.108,
+      "num_input_tokens_seen": 1703936000,
+      "step": 3250
+    },
+    {
+      "epoch": 0.48936319924944294,
+      "grad_norm": 0.6298756003379822,
+      "learning_rate": 1.2803373788501927e-05,
+      "loss": 2.1412,
+      "num_input_tokens_seen": 1709178880,
+      "step": 3260
+    },
+    {
+      "epoch": 0.4908643133575701,
+      "grad_norm": 0.6180531978607178,
+      "learning_rate": 1.2783781774287272e-05,
+      "loss": 2.16,
+      "num_input_tokens_seen": 1714421760,
+      "step": 3270
+    },
+    {
+      "epoch": 0.49236542746569717,
+      "grad_norm": 0.6927182674407959,
+      "learning_rate": 1.2764279426128711e-05,
+      "loss": 2.1604,
+      "num_input_tokens_seen": 1719664640,
+      "step": 3280
+    },
+    {
+      "epoch": 0.4938665415738243,
+      "grad_norm": 0.8382894396781921,
+      "learning_rate": 1.2744866062154176e-05,
+      "loss": 2.1509,
+      "num_input_tokens_seen": 1724907520,
+      "step": 3290
+    },
+    {
+      "epoch": 0.49536765568195146,
+      "grad_norm": 0.634811282157898,
+      "learning_rate": 1.2725541007729061e-05,
+      "loss": 2.1651,
+      "num_input_tokens_seen": 1730150400,
+      "step": 3300
+    },
+    {
+      "epoch": 0.4968687697900786,
+      "grad_norm": 0.5942947268486023,
+      "learning_rate": 1.2706303595357772e-05,
+      "loss": 2.158,
+      "num_input_tokens_seen": 1735393280,
+      "step": 3310
+    },
+    {
+      "epoch": 0.4983698838982057,
+      "grad_norm": 0.6313604712486267,
+      "learning_rate": 1.2687153164586875e-05,
+      "loss": 2.1394,
+      "num_input_tokens_seen": 1740636160,
+      "step": 3320
+    },
+    {
+      "epoch": 0.49987099800633283,
+      "grad_norm": 0.604307234287262,
+      "learning_rate": 1.2668089061909872e-05,
+      "loss": 2.1611,
+      "num_input_tokens_seen": 1745879040,
+      "step": 3330
+    },
+    {
+      "epoch": 0.50137211211446,
+      "grad_norm": 0.6554628610610962,
+      "learning_rate": 1.2649110640673518e-05,
+      "loss": 2.1072,
+      "num_input_tokens_seen": 1751121920,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5028732262225871,
+      "grad_norm": 0.7124990820884705,
+      "learning_rate": 1.2630217260985716e-05,
+      "loss": 2.1346,
+      "num_input_tokens_seen": 1756364800,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5043743403307142,
+      "grad_norm": 0.784233033657074,
+      "learning_rate": 1.2611408289624877e-05,
+      "loss": 2.1628,
+      "num_input_tokens_seen": 1761607680,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5058754544388413,
+      "grad_norm": 0.8534424901008606,
+      "learning_rate": 1.2592683099950795e-05,
+      "loss": 2.1577,
+      "num_input_tokens_seen": 1766850560,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5073765685469684,
+      "grad_norm": 0.7110067009925842,
+      "learning_rate": 1.2574041071816953e-05,
+      "loss": 2.1333,
+      "num_input_tokens_seen": 1772093440,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5088776826550956,
+      "grad_norm": 0.7778078317642212,
+      "learning_rate": 1.2555481591484266e-05,
+      "loss": 2.1136,
+      "num_input_tokens_seen": 1777336320,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5103787967632227,
+      "grad_norm": 0.7516434192657471,
+      "learning_rate": 1.253700405153621e-05,
+      "loss": 2.1456,
+      "num_input_tokens_seen": 1782579200,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5118799108713499,
+      "grad_norm": 0.8273611664772034,
+      "learning_rate": 1.2518607850795318e-05,
+      "loss": 2.1022,
+      "num_input_tokens_seen": 1787822080,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5133810249794769,
+      "grad_norm": 0.6412326097488403,
+      "learning_rate": 1.2500292394241044e-05,
+      "loss": 2.1175,
+      "num_input_tokens_seen": 1793064960,
+      "step": 3420
+    },
+    {
+      "epoch": 0.514882139087604,
+      "grad_norm": 0.6667389273643494,
+      "learning_rate": 1.2482057092928916e-05,
+      "loss": 2.1253,
+      "num_input_tokens_seen": 1798307840,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5163832531957312,
+      "grad_norm": 0.8034870624542236,
+      "learning_rate": 1.2463901363910993e-05,
+      "loss": 2.1233,
+      "num_input_tokens_seen": 1803550720,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5178843673038583,
+      "grad_norm": 0.646260678768158,
+      "learning_rate": 1.2445824630157606e-05,
+      "loss": 2.0914,
+      "num_input_tokens_seen": 1808793600,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5193854814119855,
+      "grad_norm": 0.7674567699432373,
+      "learning_rate": 1.2427826320480309e-05,
+      "loss": 2.0866,
+      "num_input_tokens_seen": 1814036480,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5208865955201126,
+      "grad_norm": 0.7650302648544312,
+      "learning_rate": 1.2409905869456093e-05,
+      "loss": 2.1287,
+      "num_input_tokens_seen": 1819279360,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5223877096282397,
+      "grad_norm": 0.6644052267074585,
+      "learning_rate": 1.2392062717352773e-05,
+      "loss": 2.122,
+      "num_input_tokens_seen": 1824522240,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5238888237363668,
+      "grad_norm": 0.8389644026756287,
+      "learning_rate": 1.2374296310055552e-05,
+      "loss": 2.1453,
+      "num_input_tokens_seen": 1829765120,
+      "step": 3490
+    },
+    {
+      "epoch": 0.525389937844494,
+      "grad_norm": 0.7249365448951721,
+      "learning_rate": 1.2356606098994752e-05,
+      "loss": 2.1418,
+      "num_input_tokens_seen": 1835008000,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5268910519526211,
+      "grad_norm": 0.7120906710624695,
+      "learning_rate": 1.2338991541074669e-05,
+      "loss": 2.1318,
+      "num_input_tokens_seen": 1840250880,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5283921660607482,
+      "grad_norm": 0.7992458343505859,
+      "learning_rate": 1.2321452098603552e-05,
+      "loss": 2.144,
+      "num_input_tokens_seen": 1845493760,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5298932801688754,
+      "grad_norm": 0.8479806780815125,
+      "learning_rate": 1.2303987239224644e-05,
+      "loss": 2.1429,
+      "num_input_tokens_seen": 1850736640,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5313943942770024,
+      "grad_norm": 0.670669436454773,
+      "learning_rate": 1.2286596435848341e-05,
+      "loss": 2.1395,
+      "num_input_tokens_seen": 1855979520,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5328955083851296,
+      "grad_norm": 0.7732242345809937,
+      "learning_rate": 1.2269279166585364e-05,
+      "loss": 2.1563,
+      "num_input_tokens_seen": 1861222400,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5343966224932567,
+      "grad_norm": 0.9772087931632996,
+      "learning_rate": 1.2252034914680983e-05,
+      "loss": 2.1375,
+      "num_input_tokens_seen": 1866465280,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5358977366013838,
+      "grad_norm": 0.5742955207824707,
+      "learning_rate": 1.2234863168450241e-05,
+      "loss": 2.1077,
+      "num_input_tokens_seen": 1871708160,
+      "step": 3570
+    },
+    {
+      "epoch": 0.537398850709511,
+      "grad_norm": 0.7332693934440613,
+      "learning_rate": 1.2217763421214203e-05,
+      "loss": 2.1338,
+      "num_input_tokens_seen": 1876951040,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5388999648176381,
+      "grad_norm": 0.6025305390357971,
+      "learning_rate": 1.2200735171237142e-05,
+      "loss": 2.1233,
+      "num_input_tokens_seen": 1882193920,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5404010789257652,
+      "grad_norm": 0.6781802773475647,
+      "learning_rate": 1.2183777921664712e-05,
+      "loss": 2.1045,
+      "num_input_tokens_seen": 1887436800,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5404010789257652,
+      "eval_accuracy": 0.56999592999593,
+      "eval_loss": 2.108959674835205,
+      "eval_runtime": 94.9286,
+      "eval_samples_per_second": 3.16,
+      "eval_steps_per_second": 0.79,
+      "num_input_tokens_seen": 1887436800,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5419021930338923,
+      "grad_norm": 0.6946921944618225,
+      "learning_rate": 1.2166891180463065e-05,
+      "loss": 2.1645,
+      "num_input_tokens_seen": 1892679680,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5434033071420195,
+      "grad_norm": 0.7196519374847412,
+      "learning_rate": 1.2150074460358885e-05,
+      "loss": 2.1177,
+      "num_input_tokens_seen": 1897922560,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5449044212501466,
+      "grad_norm": 0.6712445616722107,
+      "learning_rate": 1.2133327278780315e-05,
+      "loss": 2.118,
+      "num_input_tokens_seen": 1903165440,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5464055353582737,
+      "grad_norm": 0.6647999286651611,
+      "learning_rate": 1.2116649157798819e-05,
+      "loss": 2.1208,
+      "num_input_tokens_seen": 1908408320,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5479066494664009,
+      "grad_norm": 0.663698136806488,
+      "learning_rate": 1.2100039624071883e-05,
+      "loss": 2.1387,
+      "num_input_tokens_seen": 1913651200,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5494077635745279,
+      "grad_norm": 0.873149037361145,
+      "learning_rate": 1.2083498208786586e-05,
+      "loss": 2.1205,
+      "num_input_tokens_seen": 1918894080,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5509088776826551,
+      "grad_norm": 0.6879292726516724,
+      "learning_rate": 1.2067024447604041e-05,
+      "loss": 2.1089,
+      "num_input_tokens_seen": 1924136960,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5524099917907822,
+      "grad_norm": 0.7777056694030762,
+      "learning_rate": 1.2050617880604623e-05,
+      "loss": 2.1087,
+      "num_input_tokens_seen": 1929379840,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5539111058989094,
+      "grad_norm": 0.7917245030403137,
+      "learning_rate": 1.2034278052234061e-05,
+      "loss": 2.094,
+      "num_input_tokens_seen": 1934622720,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5554122200070365,
+      "grad_norm": 0.7892484664916992,
+      "learning_rate": 1.2018004511250296e-05,
+      "loss": 2.111,
+      "num_input_tokens_seen": 1939865600,
+      "step": 3700
+    },
+    {
+      "epoch": 0.5569133341151636,
+      "grad_norm": 0.8272449374198914,
+      "learning_rate": 1.2001796810671147e-05,
+      "loss": 2.1574,
+      "num_input_tokens_seen": 1945108480,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5584144482232907,
+      "grad_norm": 0.6054387092590332,
+      "learning_rate": 1.1985654507722748e-05,
+      "loss": 2.1226,
+      "num_input_tokens_seen": 1950351360,
+      "step": 3720
+    },
+    {
+      "epoch": 0.5599155623314178,
+      "grad_norm": 0.5401285886764526,
+      "learning_rate": 1.1969577163788735e-05,
+      "loss": 2.1093,
+      "num_input_tokens_seen": 1955594240,
+      "step": 3730
+    },
+    {
+      "epoch": 0.561416676439545,
+      "grad_norm": 0.6505089998245239,
+      "learning_rate": 1.1953564344360196e-05,
+      "loss": 2.1177,
+      "num_input_tokens_seen": 1960837120,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5629177905476721,
+      "grad_norm": 0.5896223783493042,
+      "learning_rate": 1.1937615618986343e-05,
+      "loss": 2.1246,
+      "num_input_tokens_seen": 1966080000,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5644189046557992,
+      "grad_norm": 0.9334360361099243,
+      "learning_rate": 1.192173056122592e-05,
+      "loss": 2.0941,
+      "num_input_tokens_seen": 1971322880,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5659200187639264,
+      "grad_norm": 0.8824881315231323,
+      "learning_rate": 1.1905908748599307e-05,
+      "loss": 2.135,
+      "num_input_tokens_seen": 1976565760,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5674211328720534,
+      "grad_norm": 0.9339269399642944,
+      "learning_rate": 1.1890149762541318e-05,
+      "loss": 2.1136,
+      "num_input_tokens_seen": 1981808640,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5689222469801806,
+      "grad_norm": 0.7937804460525513,
+      "learning_rate": 1.1874453188354716e-05,
+      "loss": 2.1191,
+      "num_input_tokens_seen": 1987051520,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5704233610883077,
+      "grad_norm": 0.7529720067977905,
+      "learning_rate": 1.1858818615164347e-05,
+      "loss": 2.1283,
+      "num_input_tokens_seen": 1992294400,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5719244751964349,
+      "grad_norm": 0.5535932779312134,
+      "learning_rate": 1.1843245635871996e-05,
+      "loss": 2.1311,
+      "num_input_tokens_seen": 1997537280,
+      "step": 3810
+    },
+    {
+      "epoch": 0.573425589304562,
+      "grad_norm": 0.6932185888290405,
+      "learning_rate": 1.1827733847111846e-05,
+      "loss": 2.1332,
+      "num_input_tokens_seen": 2002780160,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5749267034126891,
+      "grad_norm": 0.7156121730804443,
+      "learning_rate": 1.181228284920661e-05,
+      "loss": 2.1264,
+      "num_input_tokens_seen": 2008023040,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5764278175208162,
+      "grad_norm": 0.6193491220474243,
+      "learning_rate": 1.1796892246124261e-05,
+      "loss": 2.1376,
+      "num_input_tokens_seen": 2013265920,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5779289316289433,
+      "grad_norm": 0.6163820624351501,
+      "learning_rate": 1.1781561645435414e-05,
+      "loss": 2.0823,
+      "num_input_tokens_seen": 2018508800,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5794300457370705,
+      "grad_norm": 0.7165153622627258,
+      "learning_rate": 1.1766290658271293e-05,
+      "loss": 2.1194,
+      "num_input_tokens_seen": 2023751680,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5809311598451976,
+      "grad_norm": 0.6811047196388245,
+      "learning_rate": 1.1751078899282295e-05,
+      "loss": 2.1564,
+      "num_input_tokens_seen": 2028994560,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5824322739533248,
+      "grad_norm": 0.9540090560913086,
+      "learning_rate": 1.1735925986597164e-05,
+      "loss": 2.1241,
+      "num_input_tokens_seen": 2034237440,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5839333880614519,
+      "grad_norm": 0.6102780103683472,
+      "learning_rate": 1.1720831541782706e-05,
+      "loss": 2.0992,
+      "num_input_tokens_seen": 2039480320,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5854345021695789,
+      "grad_norm": 0.6293305158615112,
+      "learning_rate": 1.170579518980411e-05,
+      "loss": 2.1285,
+      "num_input_tokens_seen": 2044723200,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5869356162777061,
+      "grad_norm": 0.7680484056472778,
+      "learning_rate": 1.169081655898581e-05,
+      "loss": 2.0774,
+      "num_input_tokens_seen": 2049966080,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5884367303858332,
+      "grad_norm": 0.7843759059906006,
+      "learning_rate": 1.1675895280972886e-05,
+      "loss": 2.1271,
+      "num_input_tokens_seen": 2055208960,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5899378444939604,
+      "grad_norm": 0.8438634276390076,
+      "learning_rate": 1.1661030990693031e-05,
+      "loss": 2.1214,
+      "num_input_tokens_seen": 2060451840,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5914389586020875,
+      "grad_norm": 0.7281661033630371,
+      "learning_rate": 1.1646223326319031e-05,
+      "loss": 2.1074,
+      "num_input_tokens_seen": 2065694720,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5929400727102146,
+      "grad_norm": 0.7339427471160889,
+      "learning_rate": 1.163147192923176e-05,
+      "loss": 2.1571,
+      "num_input_tokens_seen": 2070937600,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5944411868183418,
+      "grad_norm": 0.583240807056427,
+      "learning_rate": 1.1616776443983717e-05,
+      "loss": 2.1111,
+      "num_input_tokens_seen": 2076180480,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5959423009264688,
+      "grad_norm": 0.5651365518569946,
+      "learning_rate": 1.1602136518263054e-05,
+      "loss": 2.0996,
+      "num_input_tokens_seen": 2081423360,
+      "step": 3970
+    },
+    {
+      "epoch": 0.597443415034596,
+      "grad_norm": 0.6529115438461304,
+      "learning_rate": 1.1587551802858083e-05,
+      "loss": 2.1478,
+      "num_input_tokens_seen": 2086666240,
+      "step": 3980
+    },
+    {
+      "epoch": 0.5989445291427231,
+      "grad_norm": 0.8366259932518005,
+      "learning_rate": 1.1573021951622309e-05,
+      "loss": 2.1437,
+      "num_input_tokens_seen": 2091909120,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6004456432508503,
+      "grad_norm": 0.5824588537216187,
+      "learning_rate": 1.1558546621439912e-05,
+      "loss": 2.1181,
+      "num_input_tokens_seen": 2097152000,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6004456432508503,
+      "eval_accuracy": 0.5708310948310948,
+      "eval_loss": 2.104544162750244,
+      "eval_runtime": 94.4692,
+      "eval_samples_per_second": 3.176,
+      "eval_steps_per_second": 0.794,
+      "num_input_tokens_seen": 2097152000,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6019467573589774,
+      "grad_norm": 0.6183150410652161,
+      "learning_rate": 1.1544125472191724e-05,
+      "loss": 2.1217,
+      "num_input_tokens_seen": 2102394880,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6034478714671045,
+      "grad_norm": 0.6359286904335022,
+      "learning_rate": 1.152975816672165e-05,
+      "loss": 2.0985,
+      "num_input_tokens_seen": 2107637760,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6049489855752316,
+      "grad_norm": 0.6601809859275818,
+      "learning_rate": 1.1515444370803586e-05,
+      "loss": 2.1115,
+      "num_input_tokens_seen": 2112880640,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6064500996833587,
+      "grad_norm": 0.6836331486701965,
+      "learning_rate": 1.1501183753108733e-05,
+      "loss": 2.1012,
+      "num_input_tokens_seen": 2118123520,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6079512137914859,
+      "grad_norm": 0.6830666661262512,
+      "learning_rate": 1.148697598517341e-05,
+      "loss": 2.1594,
+      "num_input_tokens_seen": 2123366400,
+      "step": 4050
+    },
+    {
+      "epoch": 0.609452327899613,
+      "grad_norm": 0.6801722049713135,
+      "learning_rate": 1.147282074136726e-05,
+      "loss": 2.1227,
+      "num_input_tokens_seen": 2128609280,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6109534420077402,
+      "grad_norm": 0.764336347579956,
+      "learning_rate": 1.1458717698861917e-05,
+      "loss": 2.1268,
+      "num_input_tokens_seen": 2133852160,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6124545561158673,
+      "grad_norm": 1.0238054990768433,
+      "learning_rate": 1.144466653760007e-05,
+      "loss": 2.1217,
+      "num_input_tokens_seen": 2139095040,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6139556702239943,
+      "grad_norm": 0.6901516914367676,
+      "learning_rate": 1.1430666940264972e-05,
+      "loss": 2.1304,
+      "num_input_tokens_seen": 2144337920,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6154567843321215,
+      "grad_norm": 0.6896687150001526,
+      "learning_rate": 1.141671859225032e-05,
+      "loss": 2.1239,
+      "num_input_tokens_seen": 2149580800,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6169578984402486,
+      "grad_norm": 0.8825215101242065,
+      "learning_rate": 1.1402821181630592e-05,
+      "loss": 2.1061,
+      "num_input_tokens_seen": 2154823680,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6184590125483758,
+      "grad_norm": 0.7571860551834106,
+      "learning_rate": 1.1388974399131718e-05,
+      "loss": 2.1222,
+      "num_input_tokens_seen": 2160066560,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6199601266565029,
+      "grad_norm": 0.6392367482185364,
+      "learning_rate": 1.1375177938102187e-05,
+      "loss": 2.156,
+      "num_input_tokens_seen": 2165309440,
+      "step": 4130
+    },
+    {
+      "epoch": 0.62146124076463,
+      "grad_norm": 0.7527381777763367,
+      "learning_rate": 1.136143149448452e-05,
+      "loss": 2.1233,
+      "num_input_tokens_seen": 2170552320,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6229623548727571,
+      "grad_norm": 0.7589773535728455,
+      "learning_rate": 1.1347734766787132e-05,
+      "loss": 2.1081,
+      "num_input_tokens_seen": 2175795200,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6244634689808842,
+      "grad_norm": 0.8642693161964417,
+      "learning_rate": 1.1334087456056552e-05,
+      "loss": 2.1131,
+      "num_input_tokens_seen": 2181038080,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6259645830890114,
+      "grad_norm": 0.7018624544143677,
+      "learning_rate": 1.1320489265850016e-05,
+      "loss": 2.1194,
+      "num_input_tokens_seen": 2186280960,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6274656971971385,
+      "grad_norm": 0.632854163646698,
+      "learning_rate": 1.1306939902208426e-05,
+      "loss": 2.1336,
+      "num_input_tokens_seen": 2191523840,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6289668113052657,
+      "grad_norm": 0.8238286972045898,
+      "learning_rate": 1.1293439073629661e-05,
+      "loss": 2.1341,
+      "num_input_tokens_seen": 2196766720,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6304679254133928,
+      "grad_norm": 0.8351567387580872,
+      "learning_rate": 1.1279986491042232e-05,
+      "loss": 2.0898,
+      "num_input_tokens_seen": 2202009600,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6319690395215198,
+      "grad_norm": 0.5938683748245239,
+      "learning_rate": 1.1266581867779284e-05,
+      "loss": 2.1127,
+      "num_input_tokens_seen": 2207252480,
+      "step": 4210
+    },
+    {
+      "epoch": 0.633470153629647,
+      "grad_norm": 0.7919456958770752,
+      "learning_rate": 1.1253224919552947e-05,
+      "loss": 2.1014,
+      "num_input_tokens_seen": 2212495360,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6349712677377741,
+      "grad_norm": 0.6826079487800598,
+      "learning_rate": 1.1239915364429005e-05,
+      "loss": 2.121,
+      "num_input_tokens_seen": 2217738240,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6364723818459013,
+      "grad_norm": 0.7500185966491699,
+      "learning_rate": 1.1226652922801909e-05,
+      "loss": 2.123,
+      "num_input_tokens_seen": 2222981120,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6379734959540284,
+      "grad_norm": 0.6904823780059814,
+      "learning_rate": 1.1213437317370097e-05,
+      "loss": 2.1258,
+      "num_input_tokens_seen": 2228224000,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6394746100621556,
+      "grad_norm": 0.6400703191757202,
+      "learning_rate": 1.1200268273111648e-05,
+      "loss": 2.1147,
+      "num_input_tokens_seen": 2233466880,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6409757241702826,
+      "grad_norm": 0.805190920829773,
+      "learning_rate": 1.1187145517260246e-05,
+      "loss": 2.1117,
+      "num_input_tokens_seen": 2238709760,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6424768382784097,
+      "grad_norm": 0.6888982653617859,
+      "learning_rate": 1.117406877928144e-05,
+      "loss": 2.1171,
+      "num_input_tokens_seen": 2243952640,
+      "step": 4280
+    },
+    {
+      "epoch": 0.6439779523865369,
+      "grad_norm": 0.7432284355163574,
+      "learning_rate": 1.1161037790849228e-05,
+      "loss": 2.1098,
+      "num_input_tokens_seen": 2249195520,
+      "step": 4290
+    },
+    {
+      "epoch": 0.645479066494664,
+      "grad_norm": 0.6752641201019287,
+      "learning_rate": 1.114805228582293e-05,
+      "loss": 2.1208,
+      "num_input_tokens_seen": 2254438400,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6469801806027912,
+      "grad_norm": 0.8564686179161072,
+      "learning_rate": 1.1135112000224364e-05,
+      "loss": 2.1125,
+      "num_input_tokens_seen": 2259681280,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6484812947109183,
+      "grad_norm": 0.9626083970069885,
+      "learning_rate": 1.1122216672215286e-05,
+      "loss": 2.1215,
+      "num_input_tokens_seen": 2264924160,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6499824088190453,
+      "grad_norm": 0.5604343414306641,
+      "learning_rate": 1.1109366042075172e-05,
+      "loss": 2.1667,
+      "num_input_tokens_seen": 2270167040,
+      "step": 4330
+    },
+    {
+      "epoch": 0.6514835229271725,
+      "grad_norm": 0.5957362651824951,
+      "learning_rate": 1.1096559852179205e-05,
+      "loss": 2.0962,
+      "num_input_tokens_seen": 2275409920,
+      "step": 4340
+    },
+    {
+      "epoch": 0.6529846370352996,
+      "grad_norm": 0.6427695751190186,
+      "learning_rate": 1.1083797846976627e-05,
+      "loss": 2.1203,
+      "num_input_tokens_seen": 2280652800,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6544857511434268,
+      "grad_norm": 0.6548131108283997,
+      "learning_rate": 1.1071079772969281e-05,
+      "loss": 2.1172,
+      "num_input_tokens_seen": 2285895680,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6559868652515539,
+      "grad_norm": 0.6767981052398682,
+      "learning_rate": 1.1058405378690493e-05,
+      "loss": 2.1267,
+      "num_input_tokens_seen": 2291138560,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6574879793596811,
+      "grad_norm": 0.7053226232528687,
+      "learning_rate": 1.1045774414684167e-05,
+      "loss": 2.1183,
+      "num_input_tokens_seen": 2296381440,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6589890934678081,
+      "grad_norm": 0.7884171605110168,
+      "learning_rate": 1.1033186633484174e-05,
+      "loss": 2.1103,
+      "num_input_tokens_seen": 2301624320,
+      "step": 4390
+    },
+    {
+      "epoch": 0.6604902075759352,
+      "grad_norm": 0.6177472472190857,
+      "learning_rate": 1.1020641789593993e-05,
+      "loss": 2.127,
+      "num_input_tokens_seen": 2306867200,
+      "step": 4400
+    },
+    {
+      "epoch": 0.6604902075759352,
+      "eval_accuracy": 0.5715555555555556,
+      "eval_loss": 2.099358081817627,
+      "eval_runtime": 93.9705,
+      "eval_samples_per_second": 3.192,
+      "eval_steps_per_second": 0.798,
+      "num_input_tokens_seen": 2306867200,
+      "step": 4400
+    },
+    {
+      "epoch": 0.6619913216840624,
+      "grad_norm": 0.6775453090667725,
+      "learning_rate": 1.1008139639466584e-05,
+      "loss": 2.1329,
+      "num_input_tokens_seen": 2312110080,
+      "step": 4410
+    },
+    {
+      "epoch": 0.6634924357921895,
+      "grad_norm": 0.6154870986938477,
+      "learning_rate": 1.0995679941484547e-05,
+      "loss": 2.0883,
+      "num_input_tokens_seen": 2317352960,
+      "step": 4420
+    },
+    {
+      "epoch": 0.6649935499003167,
+      "grad_norm": 0.7121848464012146,
+      "learning_rate": 1.0983262455940505e-05,
+      "loss": 2.1364,
+      "num_input_tokens_seen": 2322595840,
+      "step": 4430
+    },
+    {
+      "epoch": 0.6664946640084438,
+      "grad_norm": 0.7415539622306824,
+      "learning_rate": 1.0970886945017727e-05,
+      "loss": 2.1281,
+      "num_input_tokens_seen": 2327838720,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6679957781165708,
+      "grad_norm": 0.678644597530365,
+      "learning_rate": 1.0958553172771004e-05,
+      "loss": 2.0817,
+      "num_input_tokens_seen": 2333081600,
+      "step": 4450
+    },
+    {
+      "epoch": 0.669496892224698,
+      "grad_norm": 0.6753328442573547,
+      "learning_rate": 1.0946260905107762e-05,
+      "loss": 2.1169,
+      "num_input_tokens_seen": 2338324480,
+      "step": 4460
+    },
+    {
+      "epoch": 0.6709980063328251,
+      "grad_norm": 0.6162213087081909,
+      "learning_rate": 1.0934009909769385e-05,
+      "loss": 2.1551,
+      "num_input_tokens_seen": 2343567360,
+      "step": 4470
+    },
+    {
+      "epoch": 0.6724991204409523,
+      "grad_norm": 0.7211172580718994,
+      "learning_rate": 1.0921799956312801e-05,
+      "loss": 2.1153,
+      "num_input_tokens_seen": 2348810240,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6740002345490794,
+      "grad_norm": 0.8724449872970581,
+      "learning_rate": 1.090963081609226e-05,
+      "loss": 2.1171,
+      "num_input_tokens_seen": 2354053120,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6755013486572066,
+      "grad_norm": 0.8566068410873413,
+      "learning_rate": 1.0897502262241359e-05,
+      "loss": 2.1138,
+      "num_input_tokens_seen": 2359296000,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6770024627653336,
+      "grad_norm": 0.6851250529289246,
+      "learning_rate": 1.0885414069655284e-05,
+      "loss": 2.1129,
+      "num_input_tokens_seen": 2364538880,
+      "step": 4510
+    },
+    {
+      "epoch": 0.6785035768734607,
+      "grad_norm": 0.6527068614959717,
+      "learning_rate": 1.087336601497325e-05,
+      "loss": 2.1354,
+      "num_input_tokens_seen": 2369781760,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6800046909815879,
+      "grad_norm": 0.6223852038383484,
+      "learning_rate": 1.086135787656117e-05,
+      "loss": 2.1055,
+      "num_input_tokens_seen": 2375024640,
+      "step": 4530
+    },
+    {
+      "epoch": 0.681505805089715,
+      "grad_norm": 0.7803609371185303,
+      "learning_rate": 1.0849389434494532e-05,
+      "loss": 2.1206,
+      "num_input_tokens_seen": 2380267520,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6830069191978422,
+      "grad_norm": 0.8386520743370056,
+      "learning_rate": 1.0837460470541495e-05,
+      "loss": 2.1396,
+      "num_input_tokens_seen": 2385510400,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6845080333059693,
+      "grad_norm": 0.8144930601119995,
+      "learning_rate": 1.0825570768146148e-05,
+      "loss": 2.1149,
+      "num_input_tokens_seen": 2390753280,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6860091474140964,
+      "grad_norm": 0.8149564266204834,
+      "learning_rate": 1.081372011241202e-05,
+      "loss": 2.1056,
+      "num_input_tokens_seen": 2395996160,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6875102615222235,
+      "grad_norm": 0.5903097987174988,
+      "learning_rate": 1.0801908290085765e-05,
+      "loss": 2.123,
+      "num_input_tokens_seen": 2401239040,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6890113756303506,
+      "grad_norm": 0.6024293899536133,
+      "learning_rate": 1.0790135089541035e-05,
+      "loss": 2.1504,
+      "num_input_tokens_seen": 2406481920,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6905124897384778,
+      "grad_norm": 1.085522174835205,
+      "learning_rate": 1.0778400300762553e-05,
+      "loss": 2.1158,
+      "num_input_tokens_seen": 2411724800,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6920136038466049,
+      "grad_norm": 0.6899899840354919,
+      "learning_rate": 1.0766703715330396e-05,
+      "loss": 2.0817,
+      "num_input_tokens_seen": 2416967680,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6935147179547321,
+      "grad_norm": 0.732567310333252,
+      "learning_rate": 1.075504512640443e-05,
+      "loss": 2.1013,
+      "num_input_tokens_seen": 2422210560,
+      "step": 4620
+    },
+    {
+      "epoch": 0.6950158320628591,
+      "grad_norm": 0.7568468451499939,
+      "learning_rate": 1.0743424328708955e-05,
+      "loss": 2.1554,
+      "num_input_tokens_seen": 2427453440,
+      "step": 4630
+    },
+    {
+      "epoch": 0.6965169461709863,
+      "grad_norm": 0.7178658843040466,
+      "learning_rate": 1.0731841118517526e-05,
+      "loss": 2.0937,
+      "num_input_tokens_seen": 2432696320,
+      "step": 4640
+    },
+    {
+      "epoch": 0.6980180602791134,
+      "grad_norm": 0.8591102361679077,
+      "learning_rate": 1.072029529363794e-05,
+      "loss": 2.11,
+      "num_input_tokens_seen": 2437939200,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6995191743872405,
+      "grad_norm": 0.776842474937439,
+      "learning_rate": 1.0708786653397427e-05,
+      "loss": 2.1335,
+      "num_input_tokens_seen": 2443182080,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7010202884953677,
+      "grad_norm": 0.6280748248100281,
+      "learning_rate": 1.0697314998628e-05,
+      "loss": 2.0915,
+      "num_input_tokens_seen": 2448424960,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7025214026034948,
+      "grad_norm": 0.6168721318244934,
+      "learning_rate": 1.0685880131651965e-05,
+      "loss": 2.0978,
+      "num_input_tokens_seen": 2453667840,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7040225167116219,
+      "grad_norm": 0.7713338732719421,
+      "learning_rate": 1.067448185626763e-05,
+      "loss": 2.0992,
+      "num_input_tokens_seen": 2458910720,
+      "step": 4690
+    },
+    {
+      "epoch": 0.705523630819749,
+      "grad_norm": 0.9560806751251221,
+      "learning_rate": 1.0663119977735152e-05,
+      "loss": 2.1351,
+      "num_input_tokens_seen": 2464153600,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7070247449278761,
+      "grad_norm": 0.7966359257698059,
+      "learning_rate": 1.0651794302762573e-05,
+      "loss": 2.1285,
+      "num_input_tokens_seen": 2469396480,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7085258590360033,
+      "grad_norm": 0.7075307965278625,
+      "learning_rate": 1.0640504639492005e-05,
+      "loss": 2.1108,
+      "num_input_tokens_seen": 2474639360,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7100269731441304,
+      "grad_norm": 0.7893349528312683,
+      "learning_rate": 1.0629250797485977e-05,
+      "loss": 2.1133,
+      "num_input_tokens_seen": 2479882240,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7115280872522576,
+      "grad_norm": 0.6469416618347168,
+      "learning_rate": 1.0618032587713944e-05,
+      "loss": 2.1259,
+      "num_input_tokens_seen": 2485125120,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7130292013603846,
+      "grad_norm": 0.6949597001075745,
+      "learning_rate": 1.0606849822538959e-05,
+      "loss": 2.1272,
+      "num_input_tokens_seen": 2490368000,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7145303154685118,
+      "grad_norm": 0.6256146430969238,
+      "learning_rate": 1.0595702315704477e-05,
+      "loss": 2.1132,
+      "num_input_tokens_seen": 2495610880,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7160314295766389,
+      "grad_norm": 0.644726037979126,
+      "learning_rate": 1.0584589882321336e-05,
+      "loss": 2.1304,
+      "num_input_tokens_seen": 2500853760,
+      "step": 4770
+    },
+    {
+      "epoch": 0.717532543684766,
+      "grad_norm": 0.9808601140975952,
+      "learning_rate": 1.0573512338854876e-05,
+      "loss": 2.1061,
+      "num_input_tokens_seen": 2506096640,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7190336577928932,
+      "grad_norm": 0.829621434211731,
+      "learning_rate": 1.0562469503112205e-05,
+      "loss": 2.1288,
+      "num_input_tokens_seen": 2511339520,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7205347719010203,
+      "grad_norm": 1.08700692653656,
+      "learning_rate": 1.055146119422961e-05,
+      "loss": 2.1265,
+      "num_input_tokens_seen": 2516582400,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7205347719010203,
+      "eval_accuracy": 0.571921855921856,
+      "eval_loss": 2.095761299133301,
+      "eval_runtime": 95.3789,
+      "eval_samples_per_second": 3.145,
+      "eval_steps_per_second": 0.786,
+      "num_input_tokens_seen": 2516582400,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7220358860091474,
+      "grad_norm": 0.7580691576004028,
+      "learning_rate": 1.0540487232660126e-05,
+      "loss": 2.118,
+      "num_input_tokens_seen": 2521825280,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7235370001172745,
+      "grad_norm": 0.7149703502655029,
+      "learning_rate": 1.0529547440161225e-05,
+      "loss": 2.0822,
+      "num_input_tokens_seen": 2527068160,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7250381142254017,
+      "grad_norm": 0.6624520421028137,
+      "learning_rate": 1.0518641639782667e-05,
+      "loss": 2.1325,
+      "num_input_tokens_seen": 2532311040,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7265392283335288,
+      "grad_norm": 0.7207245826721191,
+      "learning_rate": 1.0507769655854467e-05,
+      "loss": 2.1231,
+      "num_input_tokens_seen": 2537553920,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7280403424416559,
+      "grad_norm": 0.6605533361434937,
+      "learning_rate": 1.0496931313975018e-05,
+      "loss": 2.0944,
+      "num_input_tokens_seen": 2542796800,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7295414565497831,
+      "grad_norm": 0.6779858469963074,
+      "learning_rate": 1.0486126440999343e-05,
+      "loss": 2.1355,
+      "num_input_tokens_seen": 2548039680,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7310425706579101,
+      "grad_norm": 0.9669274091720581,
+      "learning_rate": 1.047535486502747e-05,
+      "loss": 2.1198,
+      "num_input_tokens_seen": 2553282560,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7325436847660373,
+      "grad_norm": 0.6352229714393616,
+      "learning_rate": 1.0464616415392945e-05,
+      "loss": 2.0968,
+      "num_input_tokens_seen": 2558525440,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7340447988741644,
+      "grad_norm": 0.9474151134490967,
+      "learning_rate": 1.0453910922651489e-05,
+      "loss": 2.1213,
+      "num_input_tokens_seen": 2563768320,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7355459129822915,
+      "grad_norm": 0.6039556264877319,
+      "learning_rate": 1.0443238218569753e-05,
+      "loss": 2.1092,
+      "num_input_tokens_seen": 2569011200,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7370470270904187,
+      "grad_norm": 0.6552874445915222,
+      "learning_rate": 1.0432598136114214e-05,
+      "loss": 2.1292,
+      "num_input_tokens_seen": 2574254080,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7385481411985458,
+      "grad_norm": 0.6924867033958435,
+      "learning_rate": 1.042199050944021e-05,
+      "loss": 2.1396,
+      "num_input_tokens_seen": 2579496960,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7400492553066729,
+      "grad_norm": 0.7396571636199951,
+      "learning_rate": 1.0411415173881066e-05,
+      "loss": 2.1028,
+      "num_input_tokens_seen": 2584739840,
+      "step": 4930
+    },
+    {
+      "epoch": 0.7415503694148,
+      "grad_norm": 0.737847626209259,
+      "learning_rate": 1.0400871965937369e-05,
+      "loss": 2.0786,
+      "num_input_tokens_seen": 2589982720,
+      "step": 4940
+    },
+    {
+      "epoch": 0.7430514835229272,
+      "grad_norm": 0.7622154355049133,
+      "learning_rate": 1.0390360723266348e-05,
+      "loss": 2.1172,
+      "num_input_tokens_seen": 2595225600,
+      "step": 4950
+    },
+    {
+      "epoch": 0.7445525976310543,
+      "grad_norm": 0.9400896430015564,
+      "learning_rate": 1.0379881284671372e-05,
+      "loss": 2.1024,
+      "num_input_tokens_seen": 2600468480,
+      "step": 4960
+    },
+    {
+      "epoch": 0.7460537117391814,
+      "grad_norm": 0.8766177296638489,
+      "learning_rate": 1.0369433490091569e-05,
+      "loss": 2.1389,
+      "num_input_tokens_seen": 2605711360,
+      "step": 4970
+    },
+    {
+      "epoch": 0.7475548258473086,
+      "grad_norm": 0.9279113411903381,
+      "learning_rate": 1.0359017180591565e-05,
+      "loss": 2.1417,
+      "num_input_tokens_seen": 2610954240,
+      "step": 4980
+    },
+    {
+      "epoch": 0.7490559399554356,
+      "grad_norm": 0.6392689347267151,
+      "learning_rate": 1.0348632198351312e-05,
+      "loss": 2.1454,
+      "num_input_tokens_seen": 2616197120,
+      "step": 4990
+    },
+    {
+      "epoch": 0.7505570540635628,
+      "grad_norm": 0.6666054129600525,
+      "learning_rate": 1.0338278386656068e-05,
+      "loss": 2.0914,
+      "num_input_tokens_seen": 2621440000,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7520581681716899,
+      "grad_norm": 0.8447917103767395,
+      "learning_rate": 1.0327955589886445e-05,
+      "loss": 2.112,
+      "num_input_tokens_seen": 2626682880,
+      "step": 5010
+    },
+    {
+      "epoch": 0.753559282279817,
+      "grad_norm": 0.7507166862487793,
+      "learning_rate": 1.0317663653508592e-05,
+      "loss": 2.0921,
+      "num_input_tokens_seen": 2631925760,
+      "step": 5020
+    },
+    {
+      "epoch": 0.7550603963879442,
+      "grad_norm": 0.7831541299819946,
+      "learning_rate": 1.0307402424064481e-05,
+      "loss": 2.1418,
+      "num_input_tokens_seen": 2637168640,
+      "step": 5030
+    },
+    {
+      "epoch": 0.7565615104960713,
+      "grad_norm": 0.6328726410865784,
+      "learning_rate": 1.0297171749162294e-05,
+      "loss": 2.1253,
+      "num_input_tokens_seen": 2642411520,
+      "step": 5040
+    },
+    {
+      "epoch": 0.7580626246041984,
+      "grad_norm": 0.8324751257896423,
+      "learning_rate": 1.0286971477466913e-05,
+      "loss": 2.1222,
+      "num_input_tokens_seen": 2647654400,
+      "step": 5050
+    },
+    {
+      "epoch": 0.7595637387123255,
+      "grad_norm": 0.7374483942985535,
+      "learning_rate": 1.0276801458690536e-05,
+      "loss": 2.1117,
+      "num_input_tokens_seen": 2652897280,
+      "step": 5060
+    },
+    {
+      "epoch": 0.7610648528204527,
+      "grad_norm": 0.787074625492096,
+      "learning_rate": 1.0266661543583343e-05,
+      "loss": 2.0947,
+      "num_input_tokens_seen": 2658140160,
+      "step": 5070
+    },
+    {
+      "epoch": 0.7625659669285798,
+      "grad_norm": 0.6950966119766235,
+      "learning_rate": 1.0256551583924336e-05,
+      "loss": 2.1237,
+      "num_input_tokens_seen": 2663383040,
+      "step": 5080
+    },
+    {
+      "epoch": 0.764067081036707,
+      "grad_norm": 0.7359448671340942,
+      "learning_rate": 1.024647143251221e-05,
+      "loss": 2.1298,
+      "num_input_tokens_seen": 2668625920,
+      "step": 5090
+    },
+    {
+      "epoch": 0.7655681951448341,
+      "grad_norm": 0.7270290851593018,
+      "learning_rate": 1.0236420943156362e-05,
+      "loss": 2.1171,
+      "num_input_tokens_seen": 2673868800,
+      "step": 5100
+    },
+    {
+      "epoch": 0.7670693092529611,
+      "grad_norm": 0.9731076955795288,
+      "learning_rate": 1.022639997066799e-05,
+      "loss": 2.0728,
+      "num_input_tokens_seen": 2679111680,
+      "step": 5110
+    },
+    {
+      "epoch": 0.7685704233610883,
+      "grad_norm": 0.7202324867248535,
+      "learning_rate": 1.0216408370851276e-05,
+      "loss": 2.1083,
+      "num_input_tokens_seen": 2684354560,
+      "step": 5120
+    },
+    {
+      "epoch": 0.7700715374692154,
+      "grad_norm": 0.7496753334999084,
+      "learning_rate": 1.020644600049467e-05,
+      "loss": 2.105,
+      "num_input_tokens_seen": 2689597440,
+      "step": 5130
+    },
+    {
+      "epoch": 0.7715726515773426,
+      "grad_norm": 0.663766622543335,
+      "learning_rate": 1.0196512717362284e-05,
+      "loss": 2.1095,
+      "num_input_tokens_seen": 2694840320,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7730737656854697,
+      "grad_norm": 0.697460949420929,
+      "learning_rate": 1.0186608380185336e-05,
+      "loss": 2.115,
+      "num_input_tokens_seen": 2700083200,
+      "step": 5150
+    },
+    {
+      "epoch": 0.7745748797935968,
+      "grad_norm": 0.7035109400749207,
+      "learning_rate": 1.0176732848653748e-05,
+      "loss": 2.1035,
+      "num_input_tokens_seen": 2705326080,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7760759939017239,
+      "grad_norm": 0.7405036091804504,
+      "learning_rate": 1.0166885983407759e-05,
+      "loss": 2.1061,
+      "num_input_tokens_seen": 2710568960,
+      "step": 5170
+    },
+    {
+      "epoch": 0.777577108009851,
+      "grad_norm": 0.7046319842338562,
+      "learning_rate": 1.0157067646029694e-05,
+      "loss": 2.1098,
+      "num_input_tokens_seen": 2715811840,
+      "step": 5180
+    },
+    {
+      "epoch": 0.7790782221179782,
+      "grad_norm": 0.7800982594490051,
+      "learning_rate": 1.0147277699035774e-05,
+      "loss": 2.0871,
+      "num_input_tokens_seen": 2721054720,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7805793362261053,
+      "grad_norm": 0.6889191269874573,
+      "learning_rate": 1.0137516005868044e-05,
+      "loss": 2.0951,
+      "num_input_tokens_seen": 2726297600,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7805793362261053,
+      "eval_accuracy": 0.5728083028083029,
+      "eval_loss": 2.090898036956787,
+      "eval_runtime": 93.559,
+      "eval_samples_per_second": 3.207,
+      "eval_steps_per_second": 0.802,
+      "num_input_tokens_seen": 2726297600,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7820804503342325,
+      "grad_norm": 0.8273455500602722,
+      "learning_rate": 1.0127782430886362e-05,
+      "loss": 2.0874,
+      "num_input_tokens_seen": 2731540480,
+      "step": 5210
+    },
+    {
+      "epoch": 0.7835815644423596,
+      "grad_norm": 0.7020549178123474,
+      "learning_rate": 1.0118076839360487e-05,
+      "loss": 2.1437,
+      "num_input_tokens_seen": 2736783360,
+      "step": 5220
+    },
+    {
+      "epoch": 0.7850826785504866,
+      "grad_norm": 0.6244192719459534,
+      "learning_rate": 1.0108399097462245e-05,
+      "loss": 2.114,
+      "num_input_tokens_seen": 2742026240,
+      "step": 5230
+    },
+    {
+      "epoch": 0.7865837926586138,
+      "grad_norm": 0.6874685883522034,
+      "learning_rate": 1.0098749072257785e-05,
+      "loss": 2.0951,
+      "num_input_tokens_seen": 2747269120,
+      "step": 5240
+    },
+    {
+      "epoch": 0.7880849067667409,
+      "grad_norm": 0.9486780762672424,
+      "learning_rate": 1.0089126631699902e-05,
+      "loss": 2.1273,
+      "num_input_tokens_seen": 2752512000,
+      "step": 5250
+    },
+    {
+      "epoch": 0.7895860208748681,
+      "grad_norm": 0.8076638579368591,
+      "learning_rate": 1.0079531644620446e-05,
+      "loss": 2.0947,
+      "num_input_tokens_seen": 2757754880,
+      "step": 5260
+    },
+    {
+      "epoch": 0.7910871349829952,
+      "grad_norm": 0.7314202785491943,
+      "learning_rate": 1.0069963980722823e-05,
+      "loss": 2.1266,
+      "num_input_tokens_seen": 2762997760,
+      "step": 5270
+    },
+    {
+      "epoch": 0.7925882490911224,
+      "grad_norm": 0.6520930528640747,
+      "learning_rate": 1.0060423510574553e-05,
+      "loss": 2.1178,
+      "num_input_tokens_seen": 2768240640,
+      "step": 5280
+    },
+    {
+      "epoch": 0.7940893631992495,
+      "grad_norm": 0.7945283651351929,
+      "learning_rate": 1.0050910105599924e-05,
+      "loss": 2.1345,
+      "num_input_tokens_seen": 2773483520,
+      "step": 5290
+    },
+    {
+      "epoch": 0.7955904773073765,
+      "grad_norm": 0.9129828810691833,
+      "learning_rate": 1.0041423638072707e-05,
+      "loss": 2.1111,
+      "num_input_tokens_seen": 2778726400,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7970915914155037,
+      "grad_norm": 0.7378903031349182,
+      "learning_rate": 1.0031963981108954e-05,
+      "loss": 2.1075,
+      "num_input_tokens_seen": 2783969280,
+      "step": 5310
+    },
+    {
+      "epoch": 0.7985927055236308,
+      "grad_norm": 0.8121985197067261,
+      "learning_rate": 1.0022531008659884e-05,
+      "loss": 2.105,
+      "num_input_tokens_seen": 2789212160,
+      "step": 5320
+    },
+    {
+      "epoch": 0.800093819631758,
+      "grad_norm": 0.9001929759979248,
+      "learning_rate": 1.0013124595504802e-05,
+      "loss": 2.1389,
+      "num_input_tokens_seen": 2794455040,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8015949337398851,
+      "grad_norm": 0.6680558323860168,
+      "learning_rate": 1.0003744617244145e-05,
+      "loss": 2.1057,
+      "num_input_tokens_seen": 2799697920,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8030960478480123,
+      "grad_norm": 0.7327044606208801,
+      "learning_rate": 9.994390950292557e-06,
+      "loss": 2.1291,
+      "num_input_tokens_seen": 2804940800,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8045971619561393,
+      "grad_norm": 0.8424970507621765,
+      "learning_rate": 9.985063471872047e-06,
+      "loss": 2.1063,
+      "num_input_tokens_seen": 2810183680,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8060982760642664,
+      "grad_norm": 0.8721758127212524,
+      "learning_rate": 9.975762060005233e-06,
+      "loss": 2.1093,
+      "num_input_tokens_seen": 2815426560,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8075993901723936,
+      "grad_norm": 0.8354282379150391,
+      "learning_rate": 9.966486593508638e-06,
+      "loss": 2.1093,
+      "num_input_tokens_seen": 2820669440,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8091005042805207,
+      "grad_norm": 0.7554700970649719,
+      "learning_rate": 9.957236951986044e-06,
+      "loss": 2.1185,
+      "num_input_tokens_seen": 2825912320,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8106016183886479,
+      "grad_norm": 0.6626455783843994,
+      "learning_rate": 9.948013015821951e-06,
+      "loss": 2.0802,
+      "num_input_tokens_seen": 2831155200,
+      "step": 5400
+    },
+    {
+      "epoch": 0.812102732496775,
+      "grad_norm": 0.7216958403587341,
+      "learning_rate": 9.938814666175072e-06,
+      "loss": 2.0879,
+      "num_input_tokens_seen": 2836398080,
+      "step": 5410
+    },
+    {
+      "epoch": 0.813603846604902,
+      "grad_norm": 0.7840401530265808,
+      "learning_rate": 9.929641784971893e-06,
+      "loss": 2.1033,
+      "num_input_tokens_seen": 2841640960,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8151049607130292,
+      "grad_norm": 0.8764855265617371,
+      "learning_rate": 9.920494254900313e-06,
+      "loss": 2.0822,
+      "num_input_tokens_seen": 2846883840,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8166060748211563,
+      "grad_norm": 0.8843982219696045,
+      "learning_rate": 9.911371959403354e-06,
+      "loss": 2.1132,
+      "num_input_tokens_seen": 2852126720,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8181071889292835,
+      "grad_norm": 0.7158339023590088,
+      "learning_rate": 9.902274782672901e-06,
+      "loss": 2.1019,
+      "num_input_tokens_seen": 2857369600,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8196083030374106,
+      "grad_norm": 0.6980221271514893,
+      "learning_rate": 9.893202609643547e-06,
+      "loss": 2.0935,
+      "num_input_tokens_seen": 2862612480,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8211094171455378,
+      "grad_norm": 0.6455651521682739,
+      "learning_rate": 9.88415532598647e-06,
+      "loss": 2.1212,
+      "num_input_tokens_seen": 2867855360,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8226105312536648,
+      "grad_norm": 0.6872067451477051,
+      "learning_rate": 9.875132818103386e-06,
+      "loss": 2.0931,
+      "num_input_tokens_seen": 2873098240,
+      "step": 5480
+    },
+    {
+      "epoch": 0.8241116453617919,
+      "grad_norm": 0.8125963807106018,
+      "learning_rate": 9.86613497312055e-06,
+      "loss": 2.1199,
+      "num_input_tokens_seen": 2878341120,
+      "step": 5490
+    },
+    {
+      "epoch": 0.8256127594699191,
+      "grad_norm": 0.840162992477417,
+      "learning_rate": 9.857161678882842e-06,
+      "loss": 2.131,
+      "num_input_tokens_seen": 2883584000,
+      "step": 5500
+    },
+    {
+      "epoch": 0.8271138735780462,
+      "grad_norm": 0.774443507194519,
+      "learning_rate": 9.848212823947872e-06,
+      "loss": 2.0972,
+      "num_input_tokens_seen": 2888826880,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8286149876861734,
+      "grad_norm": 0.7709413766860962,
+      "learning_rate": 9.839288297580194e-06,
+      "loss": 2.1215,
+      "num_input_tokens_seen": 2894069760,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8301161017943005,
+      "grad_norm": 0.738002598285675,
+      "learning_rate": 9.830387989745525e-06,
+      "loss": 2.0852,
+      "num_input_tokens_seen": 2899312640,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8316172159024275,
+      "grad_norm": 0.803095817565918,
+      "learning_rate": 9.821511791105069e-06,
+      "loss": 2.1179,
+      "num_input_tokens_seen": 2904555520,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8331183300105547,
+      "grad_norm": 0.808690071105957,
+      "learning_rate": 9.812659593009853e-06,
+      "loss": 2.1157,
+      "num_input_tokens_seen": 2909798400,
+      "step": 5550
+    },
+    {
+      "epoch": 0.8346194441186818,
+      "grad_norm": 0.7011695504188538,
+      "learning_rate": 9.803831287495165e-06,
+      "loss": 2.1104,
+      "num_input_tokens_seen": 2915041280,
+      "step": 5560
+    },
+    {
+      "epoch": 0.836120558226809,
+      "grad_norm": 0.8309350609779358,
+      "learning_rate": 9.795026767275002e-06,
+      "loss": 2.0879,
+      "num_input_tokens_seen": 2920284160,
+      "step": 5570
+    },
+    {
+      "epoch": 0.8376216723349361,
+      "grad_norm": 0.7645972371101379,
+      "learning_rate": 9.78624592573661e-06,
+      "loss": 2.083,
+      "num_input_tokens_seen": 2925527040,
+      "step": 5580
+    },
+    {
+      "epoch": 0.8391227864430633,
+      "grad_norm": 0.770995557308197,
+      "learning_rate": 9.777488656935059e-06,
+      "loss": 2.1311,
+      "num_input_tokens_seen": 2930769920,
+      "step": 5590
+    },
+    {
+      "epoch": 0.8406239005511903,
+      "grad_norm": 0.7518852949142456,
+      "learning_rate": 9.768754855587863e-06,
+      "loss": 2.0951,
+      "num_input_tokens_seen": 2936012800,
+      "step": 5600
+    },
+    {
+      "epoch": 0.8406239005511903,
+      "eval_accuracy": 0.5732901912901913,
+      "eval_loss": 2.0876457691192627,
+      "eval_runtime": 94.3441,
+      "eval_samples_per_second": 3.18,
+      "eval_steps_per_second": 0.795,
+      "num_input_tokens_seen": 2936012800,
+      "step": 5600
+    },
+    {
+      "epoch": 0.8421250146593174,
+      "grad_norm": 0.8328472375869751,
+      "learning_rate": 9.760044417069675e-06,
+      "loss": 2.1032,
+      "num_input_tokens_seen": 2941255680,
+      "step": 5610
+    },
+    {
+      "epoch": 0.8436261287674446,
+      "grad_norm": 0.803054690361023,
+      "learning_rate": 9.75135723740702e-06,
+      "loss": 2.0893,
+      "num_input_tokens_seen": 2946498560,
+      "step": 5620
+    },
+    {
+      "epoch": 0.8451272428755717,
+      "grad_norm": 1.0215765237808228,
+      "learning_rate": 9.742693213273084e-06,
+      "loss": 2.091,
+      "num_input_tokens_seen": 2951741440,
+      "step": 5630
+    },
+    {
+      "epoch": 0.8466283569836989,
+      "grad_norm": 0.669141948223114,
+      "learning_rate": 9.734052241982545e-06,
+      "loss": 2.1025,
+      "num_input_tokens_seen": 2956984320,
+      "step": 5640
+    },
+    {
+      "epoch": 0.848129471091826,
+      "grad_norm": 0.7541179656982422,
+      "learning_rate": 9.725434221486473e-06,
+      "loss": 2.1037,
+      "num_input_tokens_seen": 2962227200,
+      "step": 5650
+    },
+    {
+      "epoch": 0.849630585199953,
+      "grad_norm": 0.8103038668632507,
+      "learning_rate": 9.716839050367259e-06,
+      "loss": 2.1022,
+      "num_input_tokens_seen": 2967470080,
+      "step": 5660
+    },
+    {
+      "epoch": 0.8511316993080802,
+      "grad_norm": 0.7163017988204956,
+      "learning_rate": 9.7082666278336e-06,
+      "loss": 2.1019,
+      "num_input_tokens_seen": 2972712960,
+      "step": 5670
+    },
+    {
+      "epoch": 0.8526328134162073,
+      "grad_norm": 0.7300000786781311,
+      "learning_rate": 9.699716853715554e-06,
+      "loss": 2.118,
+      "num_input_tokens_seen": 2977955840,
+      "step": 5680
+    },
+    {
+      "epoch": 0.8541339275243345,
+      "grad_norm": 0.7246140241622925,
+      "learning_rate": 9.691189628459591e-06,
+      "loss": 2.1528,
+      "num_input_tokens_seen": 2983198720,
+      "step": 5690
+    },
+    {
+      "epoch": 0.8556350416324616,
+      "grad_norm": 0.8242561221122742,
+      "learning_rate": 9.682684853123759e-06,
+      "loss": 2.1482,
+      "num_input_tokens_seen": 2988441600,
+      "step": 5700
+    },
+    {
+      "epoch": 0.8571361557405888,
+      "grad_norm": 0.7430049777030945,
+      "learning_rate": 9.674202429372843e-06,
+      "loss": 2.0871,
+      "num_input_tokens_seen": 2993684480,
+      "step": 5710
+    },
+    {
+      "epoch": 0.8586372698487158,
+      "grad_norm": 0.7430338263511658,
+      "learning_rate": 9.665742259473581e-06,
+      "loss": 2.0706,
+      "num_input_tokens_seen": 2998927360,
+      "step": 5720
+    },
+    {
+      "epoch": 0.860138383956843,
+      "grad_norm": 0.7350336313247681,
+      "learning_rate": 9.657304246289963e-06,
+      "loss": 2.1035,
+      "num_input_tokens_seen": 3004170240,
+      "step": 5730
+    },
+    {
+      "epoch": 0.8616394980649701,
+      "grad_norm": 0.5950428247451782,
+      "learning_rate": 9.64888829327852e-06,
+      "loss": 2.1055,
+      "num_input_tokens_seen": 3009413120,
+      "step": 5740
+    },
+    {
+      "epoch": 0.8631406121730972,
+      "grad_norm": 0.7408620715141296,
+      "learning_rate": 9.640494304483698e-06,
+      "loss": 2.1078,
+      "num_input_tokens_seen": 3014656000,
+      "step": 5750
+    },
+    {
+      "epoch": 0.8646417262812244,
+      "grad_norm": 0.8998221158981323,
+      "learning_rate": 9.63212218453326e-06,
+      "loss": 2.1249,
+      "num_input_tokens_seen": 3019898880,
+      "step": 5760
+    },
+    {
+      "epoch": 0.8661428403893515,
+      "grad_norm": 0.8028159141540527,
+      "learning_rate": 9.623771838633735e-06,
+      "loss": 2.0808,
+      "num_input_tokens_seen": 3025141760,
+      "step": 5770
+    },
+    {
+      "epoch": 0.8676439544974786,
+      "grad_norm": 0.7718355655670166,
+      "learning_rate": 9.615443172565908e-06,
+      "loss": 2.1335,
+      "num_input_tokens_seen": 3030384640,
+      "step": 5780
+    },
+    {
+      "epoch": 0.8691450686056057,
+      "grad_norm": 0.6967669129371643,
+      "learning_rate": 9.607136092680355e-06,
+      "loss": 2.1056,
+      "num_input_tokens_seen": 3035627520,
+      "step": 5790
+    },
+    {
+      "epoch": 0.8706461827137328,
+      "grad_norm": 0.8609460592269897,
+      "learning_rate": 9.598850505893025e-06,
+      "loss": 2.1219,
+      "num_input_tokens_seen": 3040870400,
+      "step": 5800
+    },
+    {
+      "epoch": 0.87214729682186,
+      "grad_norm": 0.7492549419403076,
+      "learning_rate": 9.590586319680857e-06,
+      "loss": 2.0746,
+      "num_input_tokens_seen": 3046113280,
+      "step": 5810
+    },
+    {
+      "epoch": 0.8736484109299871,
+      "grad_norm": 0.7335068583488464,
+      "learning_rate": 9.582343442077434e-06,
+      "loss": 2.1379,
+      "num_input_tokens_seen": 3051356160,
+      "step": 5820
+    },
+    {
+      "epoch": 0.8751495250381143,
+      "grad_norm": 0.6927693486213684,
+      "learning_rate": 9.574121781668698e-06,
+      "loss": 2.1052,
+      "num_input_tokens_seen": 3056599040,
+      "step": 5830
+    },
+    {
+      "epoch": 0.8766506391462413,
+      "grad_norm": 0.7068751454353333,
+      "learning_rate": 9.565921247588678e-06,
+      "loss": 2.0933,
+      "num_input_tokens_seen": 3061841920,
+      "step": 5840
+    },
+    {
+      "epoch": 0.8781517532543685,
+      "grad_norm": 0.7502044439315796,
+      "learning_rate": 9.557741749515278e-06,
+      "loss": 2.0825,
+      "num_input_tokens_seen": 3067084800,
+      "step": 5850
+    },
+    {
+      "epoch": 0.8796528673624956,
+      "grad_norm": 0.6573644876480103,
+      "learning_rate": 9.549583197666103e-06,
+      "loss": 2.1222,
+      "num_input_tokens_seen": 3072327680,
+      "step": 5860
+    },
+    {
+      "epoch": 0.8811539814706227,
+      "grad_norm": 0.8025186657905579,
+      "learning_rate": 9.541445502794315e-06,
+      "loss": 2.0886,
+      "num_input_tokens_seen": 3077570560,
+      "step": 5870
+    },
+    {
+      "epoch": 0.8826550955787499,
+      "grad_norm": 0.8089122176170349,
+      "learning_rate": 9.533328576184532e-06,
+      "loss": 2.1055,
+      "num_input_tokens_seen": 3082813440,
+      "step": 5880
+    },
+    {
+      "epoch": 0.884156209686877,
+      "grad_norm": 0.8042004704475403,
+      "learning_rate": 9.525232329648768e-06,
+      "loss": 2.1178,
+      "num_input_tokens_seen": 3088056320,
+      "step": 5890
+    },
+    {
+      "epoch": 0.8856573237950041,
+      "grad_norm": 0.8820722699165344,
+      "learning_rate": 9.517156675522405e-06,
+      "loss": 2.0989,
+      "num_input_tokens_seen": 3093299200,
+      "step": 5900
+    },
+    {
+      "epoch": 0.8871584379031312,
+      "grad_norm": 0.731745719909668,
+      "learning_rate": 9.509101526660216e-06,
+      "loss": 2.1112,
+      "num_input_tokens_seen": 3098542080,
+      "step": 5910
+    },
+    {
+      "epoch": 0.8886595520112583,
+      "grad_norm": 0.8687134981155396,
+      "learning_rate": 9.501066796432403e-06,
+      "loss": 2.0962,
+      "num_input_tokens_seen": 3103784960,
+      "step": 5920
+    },
+    {
+      "epoch": 0.8901606661193855,
+      "grad_norm": 0.7933871150016785,
+      "learning_rate": 9.493052398720693e-06,
+      "loss": 2.1376,
+      "num_input_tokens_seen": 3109027840,
+      "step": 5930
+    },
+    {
+      "epoch": 0.8916617802275126,
+      "grad_norm": 0.8090782165527344,
+      "learning_rate": 9.485058247914453e-06,
+      "loss": 2.1105,
+      "num_input_tokens_seen": 3114270720,
+      "step": 5940
+    },
+    {
+      "epoch": 0.8931628943356398,
+      "grad_norm": 0.7016152739524841,
+      "learning_rate": 9.477084258906861e-06,
+      "loss": 2.0717,
+      "num_input_tokens_seen": 3119513600,
+      "step": 5950
+    },
+    {
+      "epoch": 0.8946640084437668,
+      "grad_norm": 0.7488442659378052,
+      "learning_rate": 9.469130347091085e-06,
+      "loss": 2.1083,
+      "num_input_tokens_seen": 3124756480,
+      "step": 5960
+    },
+    {
+      "epoch": 0.896165122551894,
+      "grad_norm": 0.8162315487861633,
+      "learning_rate": 9.461196428356533e-06,
+      "loss": 2.0822,
+      "num_input_tokens_seen": 3129999360,
+      "step": 5970
+    },
+    {
+      "epoch": 0.8976662366600211,
+      "grad_norm": 0.6994791030883789,
+      "learning_rate": 9.453282419085091e-06,
+      "loss": 2.1665,
+      "num_input_tokens_seen": 3135242240,
+      "step": 5980
+    },
+    {
+      "epoch": 0.8991673507681482,
+      "grad_norm": 0.7622292041778564,
+      "learning_rate": 9.445388236147448e-06,
+      "loss": 2.0986,
+      "num_input_tokens_seen": 3140485120,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9006684648762754,
+      "grad_norm": 0.7731397747993469,
+      "learning_rate": 9.437513796899408e-06,
+      "loss": 2.1335,
+      "num_input_tokens_seen": 3145728000,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9006684648762754,
+      "eval_accuracy": 0.5739422059422059,
+      "eval_loss": 2.0838024616241455,
+      "eval_runtime": 93.3731,
+      "eval_samples_per_second": 3.213,
+      "eval_steps_per_second": 0.803,
+      "num_input_tokens_seen": 3145728000,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9021695789844025,
+      "grad_norm": 0.7499359250068665,
+      "learning_rate": 9.429659019178268e-06,
+      "loss": 2.1091,
+      "num_input_tokens_seen": 3150970880,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9036706930925296,
+      "grad_norm": 1.0717604160308838,
+      "learning_rate": 9.421823821299214e-06,
+      "loss": 2.0924,
+      "num_input_tokens_seen": 3156213760,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9051718072006567,
+      "grad_norm": 0.8998821973800659,
+      "learning_rate": 9.414008122051756e-06,
+      "loss": 2.108,
+      "num_input_tokens_seen": 3161456640,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9066729213087839,
+      "grad_norm": 0.9980584383010864,
+      "learning_rate": 9.406211840696178e-06,
+      "loss": 2.0913,
+      "num_input_tokens_seen": 3166699520,
+      "step": 6040
+    },
+    {
+      "epoch": 0.908174035416911,
+      "grad_norm": 0.8425881862640381,
+      "learning_rate": 9.398434896960062e-06,
+      "loss": 2.1174,
+      "num_input_tokens_seen": 3171942400,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9096751495250381,
+      "grad_norm": 0.7694622874259949,
+      "learning_rate": 9.390677211034795e-06,
+      "loss": 2.1214,
+      "num_input_tokens_seen": 3177185280,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9111762636331653,
+      "grad_norm": 0.8509136438369751,
+      "learning_rate": 9.382938703572126e-06,
+      "loss": 2.1062,
+      "num_input_tokens_seen": 3182428160,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9126773777412923,
+      "grad_norm": 0.688194990158081,
+      "learning_rate": 9.375219295680784e-06,
+      "loss": 2.0959,
+      "num_input_tokens_seen": 3187671040,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9141784918494195,
+      "grad_norm": 0.9447828531265259,
+      "learning_rate": 9.367518908923069e-06,
+      "loss": 2.1215,
+      "num_input_tokens_seen": 3192913920,
+      "step": 6090
+    },
+    {
+      "epoch": 0.9156796059575466,
+      "grad_norm": 0.7632228136062622,
+      "learning_rate": 9.35983746531152e-06,
+      "loss": 2.0893,
+      "num_input_tokens_seen": 3198156800,
+      "step": 6100
+    },
+    {
+      "epoch": 0.9171807200656737,
+      "grad_norm": 0.6607562899589539,
+      "learning_rate": 9.352174887305604e-06,
+      "loss": 2.1476,
+      "num_input_tokens_seen": 3203399680,
+      "step": 6110
+    },
+    {
+      "epoch": 0.9186818341738009,
+      "grad_norm": 0.7273072004318237,
+      "learning_rate": 9.344531097808414e-06,
+      "loss": 2.0986,
+      "num_input_tokens_seen": 3208642560,
+      "step": 6120
+    },
+    {
+      "epoch": 0.920182948281928,
+      "grad_norm": 0.7447370290756226,
+      "learning_rate": 9.336906020163414e-06,
+      "loss": 2.1146,
+      "num_input_tokens_seen": 3213885440,
+      "step": 6130
+    },
+    {
+      "epoch": 0.9216840623900551,
+      "grad_norm": 0.6748001575469971,
+      "learning_rate": 9.329299578151221e-06,
+      "loss": 2.0587,
+      "num_input_tokens_seen": 3219128320,
+      "step": 6140
+    },
+    {
+      "epoch": 0.9231851764981822,
+      "grad_norm": 0.874092698097229,
+      "learning_rate": 9.321711695986389e-06,
+      "loss": 2.1088,
+      "num_input_tokens_seen": 3224371200,
+      "step": 6150
+    },
+    {
+      "epoch": 0.9246862906063094,
+      "grad_norm": 0.7109676003456116,
+      "learning_rate": 9.314142298314256e-06,
+      "loss": 2.1308,
+      "num_input_tokens_seen": 3229614080,
+      "step": 6160
+    },
+    {
+      "epoch": 0.9261874047144365,
+      "grad_norm": 0.9448009133338928,
+      "learning_rate": 9.306591310207784e-06,
+      "loss": 2.1008,
+      "num_input_tokens_seen": 3234856960,
+      "step": 6170
+    },
+    {
+      "epoch": 0.9276885188225636,
+      "grad_norm": 1.0548337697982788,
+      "learning_rate": 9.29905865716445e-06,
+      "loss": 2.1113,
+      "num_input_tokens_seen": 3240099840,
+      "step": 6180
+    },
+    {
+      "epoch": 0.9291896329306908,
+      "grad_norm": 0.8548759818077087,
+      "learning_rate": 9.291544265103168e-06,
+      "loss": 2.0949,
+      "num_input_tokens_seen": 3245342720,
+      "step": 6190
+    },
+    {
+      "epoch": 0.9306907470388178,
+      "grad_norm": 0.7196878790855408,
+      "learning_rate": 9.284048060361212e-06,
+      "loss": 2.1348,
+      "num_input_tokens_seen": 3250585600,
+      "step": 6200
+    },
+    {
+      "epoch": 0.932191861146945,
+      "grad_norm": 0.970755934715271,
+      "learning_rate": 9.276569969691194e-06,
+      "loss": 2.1198,
+      "num_input_tokens_seen": 3255828480,
+      "step": 6210
+    },
+    {
+      "epoch": 0.9336929752550721,
+      "grad_norm": 0.7153322100639343,
+      "learning_rate": 9.26910992025806e-06,
+      "loss": 2.1145,
+      "num_input_tokens_seen": 3261071360,
+      "step": 6220
+    },
+    {
+      "epoch": 0.9351940893631993,
+      "grad_norm": 0.9428135752677917,
+      "learning_rate": 9.2616678396361e-06,
+      "loss": 2.0888,
+      "num_input_tokens_seen": 3266314240,
+      "step": 6230
+    },
+    {
+      "epoch": 0.9366952034713264,
+      "grad_norm": 0.8652244806289673,
+      "learning_rate": 9.254243655806003e-06,
+      "loss": 2.0739,
+      "num_input_tokens_seen": 3271557120,
+      "step": 6240
+    },
+    {
+      "epoch": 0.9381963175794535,
+      "grad_norm": 1.0319374799728394,
+      "learning_rate": 9.24683729715193e-06,
+      "loss": 2.127,
+      "num_input_tokens_seen": 3276800000,
+      "step": 6250
+    },
+    {
+      "epoch": 0.9396974316875806,
+      "grad_norm": 1.1255468130111694,
+      "learning_rate": 9.239448692458609e-06,
+      "loss": 2.1033,
+      "num_input_tokens_seen": 3282042880,
+      "step": 6260
+    },
+    {
+      "epoch": 0.9411985457957077,
+      "grad_norm": 0.7438716888427734,
+      "learning_rate": 9.232077770908458e-06,
+      "loss": 2.1018,
+      "num_input_tokens_seen": 3287285760,
+      "step": 6270
+    },
+    {
+      "epoch": 0.9426996599038349,
+      "grad_norm": 0.7200952768325806,
+      "learning_rate": 9.22472446207874e-06,
+      "loss": 2.1389,
+      "num_input_tokens_seen": 3292528640,
+      "step": 6280
+    },
+    {
+      "epoch": 0.944200774011962,
+      "grad_norm": 0.6740290522575378,
+      "learning_rate": 9.21738869593873e-06,
+      "loss": 2.0873,
+      "num_input_tokens_seen": 3297771520,
+      "step": 6290
+    },
+    {
+      "epoch": 0.9457018881200892,
+      "grad_norm": 0.7916491031646729,
+      "learning_rate": 9.210070402846921e-06,
+      "loss": 2.0943,
+      "num_input_tokens_seen": 3303014400,
+      "step": 6300
+    },
+    {
+      "epoch": 0.9472030022282163,
+      "grad_norm": 0.9598779678344727,
+      "learning_rate": 9.202769513548237e-06,
+      "loss": 2.0996,
+      "num_input_tokens_seen": 3308257280,
+      "step": 6310
+    },
+    {
+      "epoch": 0.9487041163363433,
+      "grad_norm": 0.8254228234291077,
+      "learning_rate": 9.195485959171296e-06,
+      "loss": 2.0863,
+      "num_input_tokens_seen": 3313500160,
+      "step": 6320
+    },
+    {
+      "epoch": 0.9502052304444705,
+      "grad_norm": 0.6629865169525146,
+      "learning_rate": 9.188219671225665e-06,
+      "loss": 2.0822,
+      "num_input_tokens_seen": 3318743040,
+      "step": 6330
+    },
+    {
+      "epoch": 0.9517063445525976,
+      "grad_norm": 0.7633962035179138,
+      "learning_rate": 9.180970581599163e-06,
+      "loss": 2.1215,
+      "num_input_tokens_seen": 3323985920,
+      "step": 6340
+    },
+    {
+      "epoch": 0.9532074586607248,
+      "grad_norm": 0.7902926206588745,
+      "learning_rate": 9.17373862255518e-06,
+      "loss": 2.1073,
+      "num_input_tokens_seen": 3329228800,
+      "step": 6350
+    },
+    {
+      "epoch": 0.9547085727688519,
+      "grad_norm": 0.7041093111038208,
+      "learning_rate": 9.16652372673002e-06,
+      "loss": 2.1416,
+      "num_input_tokens_seen": 3334471680,
+      "step": 6360
+    },
+    {
+      "epoch": 0.956209686876979,
+      "grad_norm": 0.7147159576416016,
+      "learning_rate": 9.159325827130255e-06,
+      "loss": 2.0744,
+      "num_input_tokens_seen": 3339714560,
+      "step": 6370
+    },
+    {
+      "epoch": 0.9577108009851061,
+      "grad_norm": 0.7527651786804199,
+      "learning_rate": 9.152144857130128e-06,
+      "loss": 2.1002,
+      "num_input_tokens_seen": 3344957440,
+      "step": 6380
+    },
+    {
+      "epoch": 0.9592119150932332,
+      "grad_norm": 0.7740973830223083,
+      "learning_rate": 9.144980750468947e-06,
+      "loss": 2.0972,
+      "num_input_tokens_seen": 3350200320,
+      "step": 6390
+    },
+    {
+      "epoch": 0.9607130292013604,
+      "grad_norm": 0.8114108443260193,
+      "learning_rate": 9.137833441248534e-06,
+      "loss": 2.0731,
+      "num_input_tokens_seen": 3355443200,
+      "step": 6400
+    },
+    {
+      "epoch": 0.9607130292013604,
+      "eval_accuracy": 0.5744249084249085,
+      "eval_loss": 2.080249071121216,
+      "eval_runtime": 93.7101,
+      "eval_samples_per_second": 3.201,
+      "eval_steps_per_second": 0.8,
+      "num_input_tokens_seen": 3355443200,
+      "step": 6400
+    },
+    {
+      "epoch": 0.9622141433094875,
+      "grad_norm": 0.7784215211868286,
+      "learning_rate": 9.130702863930661e-06,
+      "loss": 2.0991,
+      "num_input_tokens_seen": 3360686080,
+      "step": 6410
+    },
+    {
+      "epoch": 0.9637152574176147,
+      "grad_norm": 0.883834958076477,
+      "learning_rate": 9.123588953334543e-06,
+      "loss": 2.127,
+      "num_input_tokens_seen": 3365928960,
+      "step": 6420
+    },
+    {
+      "epoch": 0.9652163715257418,
+      "grad_norm": 0.6606909036636353,
+      "learning_rate": 9.116491644634323e-06,
+      "loss": 2.0898,
+      "num_input_tokens_seen": 3371171840,
+      "step": 6430
+    },
+    {
+      "epoch": 0.9667174856338688,
+      "grad_norm": 0.8438331484794617,
+      "learning_rate": 9.109410873356595e-06,
+      "loss": 2.097,
+      "num_input_tokens_seen": 3376414720,
+      "step": 6440
+    },
+    {
+      "epoch": 0.968218599741996,
+      "grad_norm": 0.7696127891540527,
+      "learning_rate": 9.102346575377945e-06,
+      "loss": 2.0882,
+      "num_input_tokens_seen": 3381657600,
+      "step": 6450
+    },
+    {
+      "epoch": 0.9697197138501231,
+      "grad_norm": 0.7825368642807007,
+      "learning_rate": 9.095298686922513e-06,
+      "loss": 2.099,
+      "num_input_tokens_seen": 3386900480,
+      "step": 6460
+    },
+    {
+      "epoch": 0.9712208279582503,
+      "grad_norm": 1.1647248268127441,
+      "learning_rate": 9.088267144559575e-06,
+      "loss": 2.0652,
+      "num_input_tokens_seen": 3392143360,
+      "step": 6470
+    },
+    {
+      "epoch": 0.9727219420663774,
+      "grad_norm": 0.8124563694000244,
+      "learning_rate": 9.081251885201133e-06,
+      "loss": 2.0914,
+      "num_input_tokens_seen": 3397386240,
+      "step": 6480
+    },
+    {
+      "epoch": 0.9742230561745046,
+      "grad_norm": 0.8535528779029846,
+      "learning_rate": 9.074252846099567e-06,
+      "loss": 2.0706,
+      "num_input_tokens_seen": 3402629120,
+      "step": 6490
+    },
+    {
+      "epoch": 0.9757241702826316,
+      "grad_norm": 0.8512585759162903,
+      "learning_rate": 9.067269964845241e-06,
+      "loss": 2.1627,
+      "num_input_tokens_seen": 3407872000,
+      "step": 6500
+    },
+    {
+      "epoch": 0.9772252843907587,
+      "grad_norm": 0.7466162443161011,
+      "learning_rate": 9.060303179364195e-06,
+      "loss": 2.0776,
+      "num_input_tokens_seen": 3413114880,
+      "step": 6510
+    },
+    {
+      "epoch": 0.9787263984988859,
+      "grad_norm": 0.8283933997154236,
+      "learning_rate": 9.053352427915811e-06,
+      "loss": 2.0799,
+      "num_input_tokens_seen": 3418357760,
+      "step": 6520
+    },
+    {
+      "epoch": 0.980227512607013,
+      "grad_norm": 0.6737175583839417,
+      "learning_rate": 9.04641764909052e-06,
+      "loss": 2.0664,
+      "num_input_tokens_seen": 3423600640,
+      "step": 6530
+    },
+    {
+      "epoch": 0.9817286267151402,
+      "grad_norm": 0.6306325197219849,
+      "learning_rate": 9.039498781807522e-06,
+      "loss": 2.0973,
+      "num_input_tokens_seen": 3428843520,
+      "step": 6540
+    },
+    {
+      "epoch": 0.9832297408232673,
+      "grad_norm": 0.7176161408424377,
+      "learning_rate": 9.032595765312539e-06,
+      "loss": 2.0812,
+      "num_input_tokens_seen": 3434086400,
+      "step": 6550
+    },
+    {
+      "epoch": 0.9847308549313943,
+      "grad_norm": 0.7962211966514587,
+      "learning_rate": 9.025708539175545e-06,
+      "loss": 2.084,
+      "num_input_tokens_seen": 3439329280,
+      "step": 6560
+    },
+    {
+      "epoch": 0.9862319690395215,
+      "grad_norm": 0.9422264099121094,
+      "learning_rate": 9.018837043288575e-06,
+      "loss": 2.1091,
+      "num_input_tokens_seen": 3444572160,
+      "step": 6570
+    },
+    {
+      "epoch": 0.9877330831476486,
+      "grad_norm": 0.7303385734558105,
+      "learning_rate": 9.011981217863507e-06,
+      "loss": 2.113,
+      "num_input_tokens_seen": 3449815040,
+      "step": 6580
+    },
+    {
+      "epoch": 0.9892341972557758,
+      "grad_norm": 0.8498770594596863,
+      "learning_rate": 9.005141003429877e-06,
+      "loss": 2.1057,
+      "num_input_tokens_seen": 3455057920,
+      "step": 6590
+    },
+    {
+      "epoch": 0.9907353113639029,
+      "grad_norm": 0.945923924446106,
+      "learning_rate": 8.99831634083271e-06,
+      "loss": 2.0949,
+      "num_input_tokens_seen": 3460300800,
+      "step": 6600
+    },
+    {
+      "epoch": 0.9922364254720301,
+      "grad_norm": 0.8327839374542236,
+      "learning_rate": 8.991507171230386e-06,
+      "loss": 2.1105,
+      "num_input_tokens_seen": 3465543680,
+      "step": 6610
+    },
+    {
+      "epoch": 0.9937375395801572,
+      "grad_norm": 0.8914210200309753,
+      "learning_rate": 8.98471343609249e-06,
+      "loss": 2.1146,
+      "num_input_tokens_seen": 3470786560,
+      "step": 6620
+    },
+    {
+      "epoch": 0.9952386536882842,
+      "grad_norm": 0.8601353168487549,
+      "learning_rate": 8.977935077197712e-06,
+      "loss": 2.1059,
+      "num_input_tokens_seen": 3476029440,
+      "step": 6630
+    },
+    {
+      "epoch": 0.9967397677964114,
+      "grad_norm": 0.8136289715766907,
+      "learning_rate": 8.971172036631744e-06,
+      "loss": 2.0737,
+      "num_input_tokens_seen": 3481272320,
+      "step": 6640
+    },
+    {
+      "epoch": 0.9982408819045385,
+      "grad_norm": 0.980107843875885,
+      "learning_rate": 8.964424256785211e-06,
+      "loss": 2.0816,
+      "num_input_tokens_seen": 3486515200,
+      "step": 6650
+    },
+    {
+      "epoch": 0.9997419960126657,
+      "grad_norm": 0.812223494052887,
+      "learning_rate": 8.9576916803516e-06,
+      "loss": 2.0981,
+      "num_input_tokens_seen": 3491758080,
+      "step": 6660
+    },
+    {
+      "epoch": 0.9998921074234783,
+      "num_input_tokens_seen": 3492282368,
+      "step": 6661,
+      "total_flos": 4.5723602603621745e+18,
+      "train_loss": 2.154333936901462,
+      "train_runtime": 99522.0612,
+      "train_samples_per_second": 8.568,
+      "train_steps_per_second": 0.067
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 6661,
+  "num_input_tokens_seen": 3492282368,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "total_flos": 4.5723602603621745e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}