{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1395,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.021528525296017224,
      "grad_norm": 3.766150712966919,
      "learning_rate": 2.9996919063517692e-05,
      "loss": 0.6799,
      "num_input_tokens_seen": 37152,
      "step": 10,
      "train_runtime": 20.683,
      "train_tokens_per_second": 1796.26
    },
    {
      "epoch": 0.04305705059203445,
      "grad_norm": 3.9944393634796143,
      "learning_rate": 2.9986270537572827e-05,
      "loss": 0.1361,
      "num_input_tokens_seen": 73856,
      "step": 20,
      "train_runtime": 32.7469,
      "train_tokens_per_second": 2255.36
    },
    {
      "epoch": 0.06458557588805167,
      "grad_norm": 10.329977989196777,
      "learning_rate": 2.9968021784983436e-05,
      "loss": 0.1558,
      "num_input_tokens_seen": 111232,
      "step": 30,
      "train_runtime": 45.4717,
      "train_tokens_per_second": 2446.179
    },
    {
      "epoch": 0.0861141011840689,
      "grad_norm": 4.059694766998291,
      "learning_rate": 2.994218206053077e-05,
      "loss": 0.1344,
      "num_input_tokens_seen": 148992,
      "step": 40,
      "train_runtime": 57.6575,
      "train_tokens_per_second": 2584.086
    },
    {
      "epoch": 0.10764262648008611,
      "grad_norm": 2.61093807220459,
      "learning_rate": 2.9908764468726846e-05,
      "loss": 0.1262,
      "num_input_tokens_seen": 186176,
      "step": 50,
      "train_runtime": 69.6848,
      "train_tokens_per_second": 2671.688
    },
    {
      "epoch": 0.12917115177610333,
      "grad_norm": 4.661616325378418,
      "learning_rate": 2.986778595716853e-05,
      "loss": 0.1182,
      "num_input_tokens_seen": 223840,
      "step": 60,
      "train_runtime": 81.8303,
      "train_tokens_per_second": 2735.416
    },
    {
      "epoch": 0.15069967707212056,
      "grad_norm": 2.2738418579101562,
      "learning_rate": 2.981926730794264e-05,
      "loss": 0.1333,
      "num_input_tokens_seen": 260960,
      "step": 70,
      "train_runtime": 93.9745,
      "train_tokens_per_second": 2776.923
    },
    {
      "epoch": 0.1722282023681378,
      "grad_norm": 2.6716361045837402,
      "learning_rate": 2.9763233127086394e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 298048,
      "step": 80,
      "train_runtime": 106.1694,
      "train_tokens_per_second": 2807.287
    },
    {
      "epoch": 0.193756727664155,
      "grad_norm": 2.849599599838257,
      "learning_rate": 2.9699711832108555e-05,
      "loss": 0.0886,
      "num_input_tokens_seen": 334496,
      "step": 90,
      "train_runtime": 118.0166,
      "train_tokens_per_second": 2834.313
    },
    {
      "epoch": 0.21528525296017223,
      "grad_norm": 1.1690081357955933,
      "learning_rate": 2.9628735637577612e-05,
      "loss": 0.1127,
      "num_input_tokens_seen": 371712,
      "step": 100,
      "train_runtime": 130.0666,
      "train_tokens_per_second": 2857.86
    },
    {
      "epoch": 0.23681377825618946,
      "grad_norm": 1.8079500198364258,
      "learning_rate": 2.955034053878428e-05,
      "loss": 0.1018,
      "num_input_tokens_seen": 408512,
      "step": 110,
      "train_runtime": 142.0073,
      "train_tokens_per_second": 2876.697
    },
    {
      "epoch": 0.25834230355220666,
      "grad_norm": 4.173007965087891,
      "learning_rate": 2.9464566293486653e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 446048,
      "step": 120,
      "train_runtime": 154.1596,
      "train_tokens_per_second": 2893.418
    },
    {
      "epoch": 0.2798708288482239,
      "grad_norm": 3.0781972408294678,
      "learning_rate": 2.9371456401747193e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 483616,
      "step": 130,
      "train_runtime": 166.2926,
      "train_tokens_per_second": 2908.222
    },
    {
      "epoch": 0.3013993541442411,
      "grad_norm": 3.6592869758605957,
      "learning_rate": 2.9271058083871864e-05,
      "loss": 0.1076,
      "num_input_tokens_seen": 520768,
      "step": 140,
      "train_runtime": 178.3589,
      "train_tokens_per_second": 2919.776
    },
    {
      "epoch": 0.32292787944025836,
      "grad_norm": 1.6870366334915161,
      "learning_rate": 2.916342225646254e-05,
      "loss": 0.0886,
      "num_input_tokens_seen": 558240,
      "step": 150,
      "train_runtime": 190.6338,
      "train_tokens_per_second": 2928.338
    },
    {
      "epoch": 0.3444564047362756,
      "grad_norm": 3.0077579021453857,
      "learning_rate": 2.904860350659487e-05,
      "loss": 0.0861,
      "num_input_tokens_seen": 595648,
      "step": 160,
      "train_runtime": 202.803,
      "train_tokens_per_second": 2937.076
    },
    {
      "epoch": 0.36598493003229277,
      "grad_norm": 2.7459771633148193,
      "learning_rate": 2.8926660064134637e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 632352,
      "step": 170,
      "train_runtime": 214.8858,
      "train_tokens_per_second": 2942.735
    },
    {
      "epoch": 0.38751345532831,
      "grad_norm": 1.1326333284378052,
      "learning_rate": 2.8797653772206726e-05,
      "loss": 0.0931,
      "num_input_tokens_seen": 669440,
      "step": 180,
      "train_runtime": 226.9184,
      "train_tokens_per_second": 2950.136
    },
    {
      "epoch": 0.40904198062432723,
      "grad_norm": 2.5396974086761475,
      "learning_rate": 2.8661650055831645e-05,
      "loss": 0.0964,
      "num_input_tokens_seen": 706784,
      "step": 190,
      "train_runtime": 239.0179,
      "train_tokens_per_second": 2957.034
    },
    {
      "epoch": 0.43057050592034446,
      "grad_norm": 2.9514100551605225,
      "learning_rate": 2.8518717888745486e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 744480,
      "step": 200,
      "train_runtime": 251.1851,
      "train_tokens_per_second": 2963.87
    },
    {
      "epoch": 0.4520990312163617,
      "grad_norm": 2.363633871078491,
      "learning_rate": 2.836892975842015e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 781632,
      "step": 210,
      "train_runtime": 263.3225,
      "train_tokens_per_second": 2968.345
    },
    {
      "epoch": 0.4736275565123789,
      "grad_norm": 2.0013911724090576,
      "learning_rate": 2.8212361629301655e-05,
      "loss": 0.0897,
      "num_input_tokens_seen": 819264,
      "step": 220,
      "train_runtime": 275.4833,
      "train_tokens_per_second": 2973.916
    },
    {
      "epoch": 0.4951560818083961,
      "grad_norm": 2.076155185699463,
      "learning_rate": 2.8049092904285033e-05,
      "loss": 0.0783,
      "num_input_tokens_seen": 855936,
      "step": 230,
      "train_runtime": 287.5153,
      "train_tokens_per_second": 2977.01
    },
    {
      "epoch": 0.5166846071044133,
      "grad_norm": 3.730830192565918,
      "learning_rate": 2.7879206384445526e-05,
      "loss": 0.092,
      "num_input_tokens_seen": 893536,
      "step": 240,
      "train_runtime": 299.6624,
      "train_tokens_per_second": 2981.809
    },
    {
      "epoch": 0.5382131324004306,
      "grad_norm": 1.213758111000061,
      "learning_rate": 2.7702788227046286e-05,
      "loss": 0.1014,
      "num_input_tokens_seen": 930272,
      "step": 250,
      "train_runtime": 311.6586,
      "train_tokens_per_second": 2984.907
    },
    {
      "epoch": 0.5597416576964478,
      "grad_norm": 1.2503005266189575,
      "learning_rate": 2.751992790184413e-05,
      "loss": 0.1032,
      "num_input_tokens_seen": 968768,
      "step": 260,
      "train_runtime": 324.0216,
      "train_tokens_per_second": 2989.825
    },
    {
      "epoch": 0.581270182992465,
      "grad_norm": 2.627034902572632,
      "learning_rate": 2.733071814571527e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 1005888,
      "step": 270,
      "train_runtime": 336.1097,
      "train_tokens_per_second": 2992.737
    },
    {
      "epoch": 0.6027987082884823,
      "grad_norm": 1.8291891813278198,
      "learning_rate": 2.7135254915624213e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 1043744,
      "step": 280,
      "train_runtime": 348.2992,
      "train_tokens_per_second": 2996.688
    },
    {
      "epoch": 0.6243272335844995,
      "grad_norm": 2.0011653900146484,
      "learning_rate": 2.693363733995954e-05,
      "loss": 0.1012,
      "num_input_tokens_seen": 1081920,
      "step": 290,
      "train_runtime": 360.6262,
      "train_tokens_per_second": 3000.115
    },
    {
      "epoch": 0.6458557588805167,
      "grad_norm": 1.9434913396835327,
      "learning_rate": 2.672596766826138e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 1118784,
      "step": 300,
      "train_runtime": 372.7516,
      "train_tokens_per_second": 3001.419
    },
    {
      "epoch": 0.667384284176534,
      "grad_norm": 1.8366467952728271,
      "learning_rate": 2.6512351219365927e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 1155392,
      "step": 310,
      "train_runtime": 384.8491,
      "train_tokens_per_second": 3002.195
    },
    {
      "epoch": 0.6889128094725512,
      "grad_norm": 1.1592612266540527,
      "learning_rate": 2.6292896327993467e-05,
      "loss": 0.0907,
      "num_input_tokens_seen": 1192768,
      "step": 320,
      "train_runtime": 397.0198,
      "train_tokens_per_second": 3004.303
    },
    {
      "epoch": 0.7104413347685683,
      "grad_norm": 1.4585574865341187,
      "learning_rate": 2.606771428980681e-05,
      "loss": 0.0786,
      "num_input_tokens_seen": 1230464,
      "step": 330,
      "train_runtime": 409.2176,
      "train_tokens_per_second": 3006.87
    },
    {
      "epoch": 0.7319698600645855,
      "grad_norm": 1.2248650789260864,
      "learning_rate": 2.5836919304968184e-05,
      "loss": 0.0863,
      "num_input_tokens_seen": 1267520,
      "step": 340,
      "train_runtime": 421.3156,
      "train_tokens_per_second": 3008.481
    },
    {
      "epoch": 0.7534983853606028,
      "grad_norm": 2.204815149307251,
      "learning_rate": 2.5600628420223057e-05,
      "loss": 0.0976,
      "num_input_tokens_seen": 1306592,
      "step": 350,
      "train_runtime": 433.7939,
      "train_tokens_per_second": 3012.011
    },
    {
      "epoch": 0.77502691065662,
      "grad_norm": 1.4170422554016113,
      "learning_rate": 2.535896146954037e-05,
      "loss": 0.0914,
      "num_input_tokens_seen": 1343840,
      "step": 360,
      "train_runtime": 445.8983,
      "train_tokens_per_second": 3013.781
    },
    {
      "epoch": 0.7965554359526372,
      "grad_norm": 2.460742712020874,
      "learning_rate": 2.5112041013339212e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 1382816,
      "step": 370,
      "train_runtime": 458.3847,
      "train_tokens_per_second": 3016.715
    },
    {
      "epoch": 0.8180839612486545,
      "grad_norm": 0.7948861718177795,
      "learning_rate": 2.48599922763328e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 1420352,
      "step": 380,
      "train_runtime": 470.464,
      "train_tokens_per_second": 3019.045
    },
    {
      "epoch": 0.8396124865446717,
      "grad_norm": 0.855792760848999,
      "learning_rate": 2.4602943084021286e-05,
      "loss": 0.1062,
      "num_input_tokens_seen": 1457280,
      "step": 390,
      "train_runtime": 482.5408,
      "train_tokens_per_second": 3020.014
    },
    {
      "epoch": 0.8611410118406889,
      "grad_norm": 1.3703863620758057,
      "learning_rate": 2.434102379786553e-05,
      "loss": 0.0579,
      "num_input_tokens_seen": 1493984,
      "step": 400,
      "train_runtime": 494.602,
      "train_tokens_per_second": 3020.578
    },
    {
      "epoch": 0.8826695371367062,
      "grad_norm": 1.5039470195770264,
      "learning_rate": 2.4074367249174864e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 1531392,
      "step": 410,
      "train_runtime": 506.7325,
      "train_tokens_per_second": 3022.091
    },
    {
      "epoch": 0.9041980624327234,
      "grad_norm": 2.4852614402770996,
      "learning_rate": 2.380310867174219e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 1568992,
      "step": 420,
      "train_runtime": 518.936,
      "train_tokens_per_second": 3023.479
    },
    {
      "epoch": 0.9257265877287406,
      "grad_norm": 0.6510267853736877,
      "learning_rate": 2.3527385633260724e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 1606144,
      "step": 430,
      "train_runtime": 530.9278,
      "train_tokens_per_second": 3025.165
    },
    {
      "epoch": 0.9472551130247578,
      "grad_norm": 1.678964376449585,
      "learning_rate": 2.3247337965557097e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 1641696,
      "step": 440,
      "train_runtime": 542.7133,
      "train_tokens_per_second": 3024.979
    },
    {
      "epoch": 0.9687836383207751,
      "grad_norm": 0.5749807953834534,
      "learning_rate": 2.2963107693676192e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 1678816,
      "step": 450,
      "train_runtime": 554.8525,
      "train_tokens_per_second": 3025.698
    },
    {
      "epoch": 0.9903121636167922,
      "grad_norm": 1.0846511125564575,
      "learning_rate": 2.2674838963853738e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 1715808,
      "step": 460,
      "train_runtime": 566.8619,
      "train_tokens_per_second": 3026.854
    },
    {
      "epoch": 1.0107642626480087,
      "grad_norm": 4.8021769523620605,
      "learning_rate": 2.2382677970413097e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 1751552,
      "step": 470,
      "train_runtime": 578.3306,
      "train_tokens_per_second": 3028.634
    },
    {
      "epoch": 1.0322927879440258,
      "grad_norm": 1.0518341064453125,
      "learning_rate": 2.2086772881623455e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 1788544,
      "step": 480,
      "train_runtime": 590.3999,
      "train_tokens_per_second": 3029.377
    },
    {
      "epoch": 1.0538213132400431,
      "grad_norm": 2.102010488510132,
      "learning_rate": 2.1787273764556827e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 1826496,
      "step": 490,
      "train_runtime": 602.6517,
      "train_tokens_per_second": 3030.765
    },
    {
      "epoch": 1.0753498385360603,
      "grad_norm": 0.5412517786026001,
      "learning_rate": 2.1484332508982167e-05,
      "loss": 0.0534,
      "num_input_tokens_seen": 1864544,
      "step": 500,
      "train_runtime": 614.9379,
      "train_tokens_per_second": 3032.085
    },
    {
      "epoch": 1.0968783638320776,
      "grad_norm": 0.6883963942527771,
      "learning_rate": 2.11781027503351e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 1901856,
      "step": 510,
      "train_runtime": 627.0846,
      "train_tokens_per_second": 3032.854
    },
    {
      "epoch": 1.1184068891280947,
      "grad_norm": 2.446444034576416,
      "learning_rate": 2.0868739791802312e-05,
      "loss": 0.0747,
      "num_input_tokens_seen": 1939264,
      "step": 520,
      "train_runtime": 639.2551,
      "train_tokens_per_second": 3033.631
    },
    {
      "epoch": 1.1399354144241118,
      "grad_norm": 0.9430311322212219,
      "learning_rate": 2.0556400525560168e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 1975872,
      "step": 530,
      "train_runtime": 651.2018,
      "train_tokens_per_second": 3034.193
    },
    {
      "epoch": 1.1614639397201292,
      "grad_norm": 2.4045143127441406,
      "learning_rate": 2.0241243353207484e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 2012928,
      "step": 540,
      "train_runtime": 663.3226,
      "train_tokens_per_second": 3034.614
    },
    {
      "epoch": 1.1829924650161463,
      "grad_norm": 1.0207197666168213,
      "learning_rate": 1.9923428105432796e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 2051168,
      "step": 550,
      "train_runtime": 675.6641,
      "train_tokens_per_second": 3035.78
    },
    {
      "epoch": 1.2045209903121636,
      "grad_norm": 7.416859149932861,
      "learning_rate": 1.9603115960956852e-05,
      "loss": 0.0696,
      "num_input_tokens_seen": 2088576,
      "step": 560,
      "train_runtime": 687.7762,
      "train_tokens_per_second": 3036.709
    },
    {
      "epoch": 1.2260495156081808,
      "grad_norm": 1.7133443355560303,
      "learning_rate": 1.9280469364791505e-05,
      "loss": 0.0553,
      "num_input_tokens_seen": 2125216,
      "step": 570,
      "train_runtime": 699.7307,
      "train_tokens_per_second": 3037.191
    },
    {
      "epoch": 1.247578040904198,
      "grad_norm": 2.1229183673858643,
      "learning_rate": 1.8955651945856354e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 2162816,
      "step": 580,
      "train_runtime": 711.9331,
      "train_tokens_per_second": 3037.948
    },
    {
      "epoch": 1.2691065662002152,
      "grad_norm": 0.5658122897148132,
      "learning_rate": 1.8628828433995013e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 2199712,
      "step": 590,
      "train_runtime": 724.0149,
      "train_tokens_per_second": 3038.214
    },
    {
      "epoch": 1.2906350914962326,
      "grad_norm": 2.8044610023498535,
      "learning_rate": 1.8300164576433035e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 2236416,
      "step": 600,
      "train_runtime": 736.0531,
      "train_tokens_per_second": 3038.389
    },
    {
      "epoch": 1.3121636167922497,
      "grad_norm": 1.5163172483444214,
      "learning_rate": 1.7969827053719862e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 2274272,
      "step": 610,
      "train_runtime": 748.3131,
      "train_tokens_per_second": 3039.198
    },
    {
      "epoch": 1.333692142088267,
      "grad_norm": 2.5300729274749756,
      "learning_rate": 1.7637983395197465e-05,
      "loss": 0.0683,
      "num_input_tokens_seen": 2310688,
      "step": 620,
      "train_runtime": 760.313,
      "train_tokens_per_second": 3039.127
    },
    {
      "epoch": 1.3552206673842842,
      "grad_norm": 2.793215036392212,
      "learning_rate": 1.7304801894038522e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 2348608,
      "step": 630,
      "train_runtime": 772.6051,
      "train_tokens_per_second": 3039.856
    },
    {
      "epoch": 1.3767491926803013,
      "grad_norm": 1.3523025512695312,
      "learning_rate": 1.6970451521897198e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 2386144,
      "step": 640,
      "train_runtime": 784.8469,
      "train_tokens_per_second": 3040.267
    },
    {
      "epoch": 1.3982777179763186,
      "grad_norm": 2.378476858139038,
      "learning_rate": 1.663510184321587e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 2422912,
      "step": 650,
      "train_runtime": 796.9416,
      "train_tokens_per_second": 3040.263
    },
    {
      "epoch": 1.419806243272336,
      "grad_norm": 1.0586169958114624,
      "learning_rate": 1.6298922929231194e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 2459744,
      "step": 660,
      "train_runtime": 809.0893,
      "train_tokens_per_second": 3040.139
    },
    {
      "epoch": 1.441334768568353,
      "grad_norm": 1.8658058643341064,
      "learning_rate": 1.5962085271723158e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 2497184,
      "step": 670,
      "train_runtime": 821.239,
      "train_tokens_per_second": 3040.752
    },
    {
      "epoch": 1.4628632938643702,
      "grad_norm": 2.6104471683502197,
      "learning_rate": 1.5624759696550874e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 2535456,
      "step": 680,
      "train_runtime": 833.6348,
      "train_tokens_per_second": 3041.447
    },
    {
      "epoch": 1.4843918191603875,
      "grad_norm": 1.810530662536621,
      "learning_rate": 1.528711727701893e-05,
      "loss": 0.0338,
      "num_input_tokens_seen": 2572704,
      "step": 690,
      "train_runtime": 845.7584,
      "train_tokens_per_second": 3041.89
    },
    {
      "epoch": 1.5059203444564049,
      "grad_norm": 2.5149903297424316,
      "learning_rate": 1.4949329247118258e-05,
      "loss": 0.0621,
      "num_input_tokens_seen": 2609696,
      "step": 700,
      "train_runtime": 857.7592,
      "train_tokens_per_second": 3042.458
    },
    {
      "epoch": 1.527448869752422,
      "grad_norm": 1.2988933324813843,
      "learning_rate": 1.4611566914685489e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 2646976,
      "step": 710,
      "train_runtime": 869.9041,
      "train_tokens_per_second": 3042.837
    },
    {
      "epoch": 1.5489773950484391,
      "grad_norm": 1.6637475490570068,
      "learning_rate": 1.4274001574524894e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 2684576,
      "step": 720,
      "train_runtime": 882.1134,
      "train_tokens_per_second": 3043.346
    },
    {
      "epoch": 1.5705059203444565,
      "grad_norm": 2.7324392795562744,
      "learning_rate": 1.3936804421536906e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 2720736,
      "step": 730,
      "train_runtime": 894.5163,
      "train_tokens_per_second": 3041.572
    },
    {
      "epoch": 1.5920344456404736,
      "grad_norm": 0.791470468044281,
      "learning_rate": 1.3600146463897327e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 2757728,
      "step": 740,
      "train_runtime": 906.5686,
      "train_tokens_per_second": 3041.941
    },
    {
      "epoch": 1.6135629709364907,
      "grad_norm": 1.0436009168624878,
      "learning_rate": 1.3264198436331193e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 2795168,
      "step": 750,
      "train_runtime": 918.6492,
      "train_tokens_per_second": 3042.694
    },
    {
      "epoch": 1.635091496232508,
      "grad_norm": 1.297885775566101,
      "learning_rate": 1.2929130713525396e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 2832000,
      "step": 760,
      "train_runtime": 930.6759,
      "train_tokens_per_second": 3042.95
    },
    {
      "epoch": 1.6566200215285254,
      "grad_norm": 1.0669901371002197,
      "learning_rate": 1.2595113223723788e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 2869696,
      "step": 770,
      "train_runtime": 942.933,
      "train_tokens_per_second": 3043.372
    },
    {
      "epoch": 1.6781485468245425,
      "grad_norm": 2.101264715194702,
      "learning_rate": 1.2262315362548825e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 2907840,
      "step": 780,
      "train_runtime": 955.2475,
      "train_tokens_per_second": 3044.07
    },
    {
      "epoch": 1.6996770721205596,
      "grad_norm": 3.501140594482422,
      "learning_rate": 1.1930905907093196e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 2943232,
      "step": 790,
      "train_runtime": 967.0703,
      "train_tokens_per_second": 3043.452
    },
    {
      "epoch": 1.721205597416577,
      "grad_norm": 5.079594612121582,
      "learning_rate": 1.1601052930325236e-05,
      "loss": 0.0509,
      "num_input_tokens_seen": 2980608,
      "step": 800,
      "train_runtime": 979.2158,
      "train_tokens_per_second": 3043.872
    },
    {
      "epoch": 1.7427341227125943,
      "grad_norm": 1.612389326095581,
      "learning_rate": 1.127292371585141e-05,
      "loss": 0.0674,
      "num_input_tokens_seen": 3018080,
      "step": 810,
      "train_runtime": 991.3276,
      "train_tokens_per_second": 3044.483
    },
    {
      "epoch": 1.7642626480086114,
      "grad_norm": 4.1869120597839355,
      "learning_rate": 1.0946684673079074e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 3055520,
      "step": 820,
      "train_runtime": 1003.4787,
      "train_tokens_per_second": 3044.928
    },
    {
      "epoch": 1.7857911733046286,
      "grad_norm": 0.6242759823799133,
      "learning_rate": 1.0622501252822628e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 3092800,
      "step": 830,
      "train_runtime": 1015.5301,
      "train_tokens_per_second": 3045.503
    },
    {
      "epoch": 1.807319698600646,
      "grad_norm": 1.8685392141342163,
      "learning_rate": 1.0300537863395823e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 3130656,
      "step": 840,
      "train_runtime": 1027.7868,
      "train_tokens_per_second": 3046.017
    },
    {
      "epoch": 1.8288482238966632,
      "grad_norm": 1.3861064910888672,
      "learning_rate": 9.98095778723271e-06,
      "loss": 0.0516,
      "num_input_tokens_seen": 3167744,
      "step": 850,
      "train_runtime": 1039.7962,
      "train_tokens_per_second": 3046.505
    },
    {
      "epoch": 1.8503767491926801,
      "grad_norm": 2.453305244445801,
      "learning_rate": 9.663923098079656e-06,
      "loss": 0.0628,
      "num_input_tokens_seen": 3205376,
      "step": 860,
      "train_runtime": 1051.9233,
      "train_tokens_per_second": 3047.157
    },
    {
      "epoch": 1.8719052744886975,
      "grad_norm": 0.682584822177887,
      "learning_rate": 9.349594578800259e-06,
      "loss": 0.0733,
      "num_input_tokens_seen": 3242976,
      "step": 870,
      "train_runtime": 1064.1457,
      "train_tokens_per_second": 3047.492
    },
    {
      "epoch": 1.8934337997847148,
      "grad_norm": 1.5370641946792603,
      "learning_rate": 9.038131639834977e-06,
      "loss": 0.0727,
      "num_input_tokens_seen": 3280096,
      "step": 880,
      "train_runtime": 1076.18,
      "train_tokens_per_second": 3047.906
    },
    {
      "epoch": 1.914962325080732,
      "grad_norm": 0.9756731390953064,
      "learning_rate": 8.7296922383568e-06,
      "loss": 0.058,
      "num_input_tokens_seen": 3316960,
      "step": 890,
      "train_runtime": 1088.1737,
      "train_tokens_per_second": 3048.19
    },
    {
      "epoch": 1.936490850376749,
      "grad_norm": 1.8065251111984253,
      "learning_rate": 8.424432798163838e-06,
      "loss": 0.0588,
      "num_input_tokens_seen": 3353600,
      "step": 900,
      "train_runtime": 1100.1193,
      "train_tokens_per_second": 3048.397
    },
    {
      "epoch": 1.9580193756727664,
      "grad_norm": 1.448608636856079,
      "learning_rate": 8.122508130349711e-06,
      "loss": 0.0375,
      "num_input_tokens_seen": 3390592,
      "step": 910,
      "train_runtime": 1112.0701,
      "train_tokens_per_second": 3048.901
    },
    {
      "epoch": 1.9795479009687837,
      "grad_norm": 1.7946593761444092,
      "learning_rate": 7.82407135479164e-06,
      "loss": 0.0644,
      "num_input_tokens_seen": 3427840,
      "step": 920,
      "train_runtime": 1124.2394,
      "train_tokens_per_second": 3049.03
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.08534291386604309,
      "learning_rate": 7.529273822496379e-06,
      "loss": 0.0347,
      "num_input_tokens_seen": 3464032,
      "step": 930,
      "train_runtime": 1135.8636,
      "train_tokens_per_second": 3049.69
    },
    {
      "epoch": 2.0215285252960173,
      "grad_norm": 3.5694122314453125,
      "learning_rate": 7.238265038843143e-06,
      "loss": 0.0491,
      "num_input_tokens_seen": 3501056,
      "step": 940,
      "train_runtime": 1147.9804,
      "train_tokens_per_second": 3049.752
    },
    {
      "epoch": 2.0430570505920342,
      "grad_norm": 1.0579440593719482,
      "learning_rate": 6.951192587762555e-06,
      "loss": 0.0263,
      "num_input_tokens_seen": 3538464,
      "step": 950,
      "train_runtime": 1160.1923,
      "train_tokens_per_second": 3049.894
    },
    {
      "epoch": 2.0645855758880516,
      "grad_norm": 1.1105297803878784,
      "learning_rate": 6.668202056890052e-06,
      "loss": 0.029,
      "num_input_tokens_seen": 3575552,
      "step": 960,
      "train_runtime": 1172.3064,
      "train_tokens_per_second": 3050.015
    },
    {
      "epoch": 2.086114101184069,
      "grad_norm": 2.4604287147521973,
      "learning_rate": 6.389436963731707e-06,
      "loss": 0.0338,
      "num_input_tokens_seen": 3612736,
      "step": 970,
      "train_runtime": 1184.4757,
      "train_tokens_per_second": 3050.072
    },
    {
      "epoch": 2.1076426264800863,
      "grad_norm": 2.9901983737945557,
      "learning_rate": 6.115038682879904e-06,
      "loss": 0.0427,
      "num_input_tokens_seen": 3650464,
      "step": 980,
      "train_runtime": 1196.7382,
      "train_tokens_per_second": 3050.345
    },
    {
      "epoch": 2.129171151776103,
      "grad_norm": 1.1089911460876465,
      "learning_rate": 5.845146374315781e-06,
      "loss": 0.0323,
      "num_input_tokens_seen": 3687392,
      "step": 990,
      "train_runtime": 1208.888,
      "train_tokens_per_second": 3050.235
    },
    {
      "epoch": 2.1506996770721205,
      "grad_norm": 4.3608317375183105,
      "learning_rate": 5.579896912834807e-06,
      "loss": 0.0504,
      "num_input_tokens_seen": 3725152,
      "step": 1000,
      "train_runtime": 1221.1038,
      "train_tokens_per_second": 3050.643
    },
    {
      "epoch": 2.172228202368138,
      "grad_norm": 2.5201330184936523,
      "learning_rate": 5.319424818631293e-06,
      "loss": 0.043,
      "num_input_tokens_seen": 3762752,
      "step": 1010,
      "train_runtime": 1234.1135,
      "train_tokens_per_second": 3048.951
    },
    {
      "epoch": 2.193756727664155,
      "grad_norm": 2.4340291023254395,
      "learning_rate": 5.063862189076988e-06,
      "loss": 0.0377,
      "num_input_tokens_seen": 3800736,
      "step": 1020,
      "train_runtime": 1246.4039,
      "train_tokens_per_second": 3049.361
    },
    {
      "epoch": 2.215285252960172,
      "grad_norm": 1.2626358270645142,
      "learning_rate": 4.813338631728471e-06,
      "loss": 0.0408,
      "num_input_tokens_seen": 3838432,
      "step": 1030,
      "train_runtime": 1258.5862,
      "train_tokens_per_second": 3049.797
    },
    {
      "epoch": 2.2368137782561894,
      "grad_norm": 2.4242489337921143,
      "learning_rate": 4.567981198597168e-06,
      "loss": 0.0328,
      "num_input_tokens_seen": 3876224,
      "step": 1040,
      "train_runtime": 1270.8593,
      "train_tokens_per_second": 3050.081
    },
    {
      "epoch": 2.2583423035522068,
      "grad_norm": 2.391545534133911,
      "learning_rate": 4.327914321715459e-06,
      "loss": 0.0259,
      "num_input_tokens_seen": 3913216,
      "step": 1050,
      "train_runtime": 1282.8604,
      "train_tokens_per_second": 3050.383
    },
    {
      "epoch": 2.2798708288482237,
      "grad_norm": 2.008685827255249,
      "learning_rate": 4.09325975003146e-06,
      "loss": 0.027,
      "num_input_tokens_seen": 3950208,
      "step": 1060,
      "train_runtime": 1294.9339,
      "train_tokens_per_second": 3050.509
    },
    {
      "epoch": 2.301399354144241,
      "grad_norm": 5.242519855499268,
      "learning_rate": 3.864136487664538e-06,
      "loss": 0.0347,
      "num_input_tokens_seen": 3986752,
      "step": 1070,
      "train_runtime": 1306.9829,
      "train_tokens_per_second": 3050.348
    },
    {
      "epoch": 2.3229278794402584,
      "grad_norm": 0.17025278508663177,
      "learning_rate": 3.640660733552863e-06,
      "loss": 0.0189,
      "num_input_tokens_seen": 4023360,
      "step": 1080,
      "train_runtime": 1318.9572,
      "train_tokens_per_second": 3050.41
    },
    {
      "epoch": 2.3444564047362757,
      "grad_norm": 3.2421324253082275,
      "learning_rate": 3.4229458225235696e-06,
      "loss": 0.0186,
      "num_input_tokens_seen": 4061856,
      "step": 1090,
      "train_runtime": 1331.3712,
      "train_tokens_per_second": 3050.882
    },
    {
      "epoch": 2.3659849300322926,
      "grad_norm": 1.6071536540985107,
      "learning_rate": 3.211102167815469e-06,
      "loss": 0.0366,
      "num_input_tokens_seen": 4098464,
      "step": 1100,
      "train_runtime": 1343.347,
      "train_tokens_per_second": 3050.935
    },
    {
      "epoch": 2.38751345532831,
      "grad_norm": 1.8836613893508911,
      "learning_rate": 3.0052372050834025e-06,
      "loss": 0.0333,
      "num_input_tokens_seen": 4136576,
      "step": 1110,
      "train_runtime": 1355.7033,
      "train_tokens_per_second": 3051.24
    },
    {
      "epoch": 2.4090419806243273,
      "grad_norm": 2.0569138526916504,
      "learning_rate": 2.805455337912722e-06,
      "loss": 0.0259,
      "num_input_tokens_seen": 4174528,
      "step": 1120,
      "train_runtime": 1367.887,
      "train_tokens_per_second": 3051.808
    },
    {
      "epoch": 2.4305705059203446,
      "grad_norm": 0.382569283246994,
      "learning_rate": 2.611857884871379e-06,
      "loss": 0.0377,
      "num_input_tokens_seen": 4211712,
      "step": 1130,
      "train_runtime": 1380.0113,
      "train_tokens_per_second": 3051.94
    },
    {
      "epoch": 2.4520990312163615,
      "grad_norm": 0.1618432253599167,
      "learning_rate": 2.424543028126661e-06,
      "loss": 0.0457,
      "num_input_tokens_seen": 4248800,
      "step": 1140,
      "train_runtime": 1392.1418,
      "train_tokens_per_second": 3051.988
    },
    {
      "epoch": 2.473627556512379,
      "grad_norm": 1.5764034986495972,
      "learning_rate": 2.243605763652489e-06,
      "loss": 0.0341,
      "num_input_tokens_seen": 4286496,
      "step": 1150,
      "train_runtime": 1404.3544,
      "train_tokens_per_second": 3052.289
    },
    {
      "epoch": 2.495156081808396,
      "grad_norm": 5.99455451965332,
      "learning_rate": 2.0691378530525933e-06,
      "loss": 0.0311,
      "num_input_tokens_seen": 4323328,
      "step": 1160,
      "train_runtime": 1416.3868,
      "train_tokens_per_second": 3052.364
    },
    {
      "epoch": 2.5166846071044136,
      "grad_norm": 3.6439156532287598,
      "learning_rate": 1.9012277770239829e-06,
      "loss": 0.0373,
      "num_input_tokens_seen": 4360384,
      "step": 1170,
      "train_runtime": 1428.4267,
      "train_tokens_per_second": 3052.578
    },
    {
      "epoch": 2.5382131324004304,
      "grad_norm": 3.2437760829925537,
      "learning_rate": 1.7399606904843224e-06,
      "loss": 0.0176,
      "num_input_tokens_seen": 4397440,
      "step": 1180,
      "train_runtime": 1440.5831,
      "train_tokens_per_second": 3052.542
    },
    {
      "epoch": 2.559741657696448,
      "grad_norm": 2.4674410820007324,
      "learning_rate": 1.5854183793859577e-06,
      "loss": 0.0264,
      "num_input_tokens_seen": 4435584,
      "step": 1190,
      "train_runtime": 1452.9748,
      "train_tokens_per_second": 3052.76
    },
    {
      "epoch": 2.581270182992465,
      "grad_norm": 1.2190111875534058,
      "learning_rate": 1.437679219238498e-06,
      "loss": 0.0211,
      "num_input_tokens_seen": 4472800,
      "step": 1200,
      "train_runtime": 1465.0586,
      "train_tokens_per_second": 3052.984
    },
    {
      "epoch": 2.602798708288482,
      "grad_norm": 2.0457441806793213,
      "learning_rate": 1.2968181353609854e-06,
      "loss": 0.0264,
      "num_input_tokens_seen": 4510464,
      "step": 1210,
      "train_runtime": 1477.2352,
      "train_tokens_per_second": 3053.315
    },
    {
      "epoch": 2.6243272335844994,
      "grad_norm": 1.473620057106018,
      "learning_rate": 1.1629065648838495e-06,
      "loss": 0.0369,
      "num_input_tokens_seen": 4546592,
      "step": 1220,
      "train_runtime": 1489.2585,
      "train_tokens_per_second": 3052.923
    },
    {
      "epoch": 2.6458557588805167,
      "grad_norm": 0.5360433459281921,
      "learning_rate": 1.0360124205198102e-06,
      "loss": 0.0285,
      "num_input_tokens_seen": 4583680,
      "step": 1230,
      "train_runtime": 1501.3452,
      "train_tokens_per_second": 3053.049
    },
    {
      "epoch": 2.667384284176534,
      "grad_norm": 3.2004950046539307,
      "learning_rate": 9.162000561222672e-07,
      "loss": 0.0452,
      "num_input_tokens_seen": 4622080,
      "step": 1240,
      "train_runtime": 1513.7232,
      "train_tokens_per_second": 3053.451
    },
    {
      "epoch": 2.6889128094725514,
      "grad_norm": 0.3612094819545746,
      "learning_rate": 8.035302340484646e-07,
      "loss": 0.051,
      "num_input_tokens_seen": 4658848,
      "step": 1250,
      "train_runtime": 1525.7739,
      "train_tokens_per_second": 3053.433
    },
    {
      "epoch": 2.7104413347685683,
      "grad_norm": 1.8036645650863647,
      "learning_rate": 6.980600943441273e-07,
      "loss": 0.0217,
      "num_input_tokens_seen": 4695968,
      "step": 1260,
      "train_runtime": 1537.831,
      "train_tokens_per_second": 3053.631
    },
    {
      "epoch": 2.7319698600645856,
      "grad_norm": 2.2066173553466797,
      "learning_rate": 5.998431257651132e-07,
      "loss": 0.0414,
      "num_input_tokens_seen": 4733344,
      "step": 1270,
      "train_runtime": 1550.0453,
      "train_tokens_per_second": 3053.681
    },
    {
      "epoch": 2.7534983853606025,
      "grad_norm": 2.491161823272705,
      "learning_rate": 5.089291386508055e-07,
      "loss": 0.0315,
      "num_input_tokens_seen": 4770496,
      "step": 1280,
      "train_runtime": 1562.1107,
      "train_tokens_per_second": 3053.878
    },
    {
      "epoch": 2.77502691065662,
      "grad_norm": 1.4545611143112183,
      "learning_rate": 4.2536423966299256e-07,
      "loss": 0.0336,
      "num_input_tokens_seen": 4807456,
      "step": 1290,
      "train_runtime": 1574.2633,
      "train_tokens_per_second": 3053.781
    },
    {
      "epoch": 2.7965554359526372,
      "grad_norm": 1.0926755666732788,
      "learning_rate": 3.491908084030776e-07,
      "loss": 0.0426,
      "num_input_tokens_seen": 4845440,
      "step": 1300,
      "train_runtime": 1586.6432,
      "train_tokens_per_second": 3053.894
    },
    {
      "epoch": 2.8180839612486546,
      "grad_norm": 2.7122678756713867,
      "learning_rate": 2.804474759194176e-07,
      "loss": 0.0434,
      "num_input_tokens_seen": 4883392,
      "step": 1310,
      "train_runtime": 1598.92,
      "train_tokens_per_second": 3054.182
    },
    {
      "epoch": 2.839612486544672,
      "grad_norm": 1.7297043800354004,
      "learning_rate": 2.1916910511575627e-07,
      "loss": 0.0297,
      "num_input_tokens_seen": 4920384,
      "step": 1320,
      "train_runtime": 1610.9432,
      "train_tokens_per_second": 3054.35
    },
    {
      "epoch": 2.861141011840689,
      "grad_norm": 2.814833641052246,
      "learning_rate": 1.6538677307062933e-07,
      "loss": 0.036,
      "num_input_tokens_seen": 4957888,
      "step": 1330,
      "train_runtime": 1623.1577,
      "train_tokens_per_second": 3054.471
    },
    {
      "epoch": 2.882669537136706,
      "grad_norm": 3.3779215812683105,
      "learning_rate": 1.1912775527676278e-07,
      "loss": 0.0209,
      "num_input_tokens_seen": 4994944,
      "step": 1340,
      "train_runtime": 1635.1442,
      "train_tokens_per_second": 3054.742
    },
    {
      "epoch": 2.9041980624327235,
      "grad_norm": 1.425120234489441,
      "learning_rate": 8.04155118084099e-08,
      "loss": 0.0467,
      "num_input_tokens_seen": 5031584,
      "step": 1350,
      "train_runtime": 1647.097,
      "train_tokens_per_second": 3054.82
    },
    {
      "epoch": 2.9257265877287404,
      "grad_norm": 0.8811171650886536,
      "learning_rate": 4.9269675423671556e-08,
      "loss": 0.0414,
      "num_input_tokens_seen": 5068928,
      "step": 1360,
      "train_runtime": 1659.2581,
      "train_tokens_per_second": 3054.936
    },
    {
      "epoch": 2.9472551130247577,
      "grad_norm": 1.2112094163894653,
      "learning_rate": 2.570604160781753e-08,
      "loss": 0.0182,
      "num_input_tokens_seen": 5105984,
      "step": 1370,
      "train_runtime": 1671.367,
      "train_tokens_per_second": 3054.975
    },
    {
      "epoch": 2.968783638320775,
      "grad_norm": 2.05094313621521,
      "learning_rate": 9.736560562675912e-09,
      "loss": 0.0236,
      "num_input_tokens_seen": 5143296,
      "step": 1380,
      "train_runtime": 1683.5361,
      "train_tokens_per_second": 3055.055
    },
    {
      "epoch": 2.9903121636167924,
      "grad_norm": 0.11877042800188065,
      "learning_rate": 1.3693311461326908e-09,
      "loss": 0.0323,
      "num_input_tokens_seen": 5181056,
      "step": 1390,
      "train_runtime": 1695.7436,
      "train_tokens_per_second": 3055.33
    },
    {
      "epoch": 3.0,
      "num_input_tokens_seen": 5197824,
      "step": 1395,
      "total_flos": 2.2251190147979674e+17,
      "train_loss": 0.06651546376153132,
      "train_runtime": 1702.0416,
      "train_samples_per_second": 6.55,
      "train_steps_per_second": 0.82
    }
  ],
  "logging_steps": 10,
  "max_steps": 1395,
  "num_input_tokens_seen": 5197824,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2251190147979674e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}