{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 468,
  "global_step": 4680,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004273960893257827,
      "grad_norm": 9.625906944274902,
      "learning_rate": 0.0,
      "loss": 0.2226095199584961,
      "num_input_tokens_seen": 1154,
      "step": 1,
      "train_runtime": 7.7061,
      "train_tokens_per_second": 149.752
    },
    {
      "epoch": 0.0008547921786515654,
      "grad_norm": 15.64644718170166,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.37749505043029785,
      "num_input_tokens_seen": 1926,
      "step": 2,
      "train_runtime": 9.704,
      "train_tokens_per_second": 198.474
    },
    {
      "epoch": 0.001282188267977348,
      "grad_norm": 6.848448753356934,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.24094641208648682,
      "num_input_tokens_seen": 2966,
      "step": 3,
      "train_runtime": 11.7533,
      "train_tokens_per_second": 252.355
    },
    {
      "epoch": 0.0017095843573031308,
      "grad_norm": 4.784849643707275,
      "learning_rate": 1.2e-05,
      "loss": 0.19649389386177063,
      "num_input_tokens_seen": 3778,
      "step": 4,
      "train_runtime": 13.7348,
      "train_tokens_per_second": 275.069
    },
    {
      "epoch": 0.0021369804466289132,
      "grad_norm": 6.467296600341797,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.20225900411605835,
      "num_input_tokens_seen": 4340,
      "step": 5,
      "train_runtime": 15.6871,
      "train_tokens_per_second": 276.66
    },
    {
      "epoch": 0.002564376535954696,
      "grad_norm": 4.957343578338623,
      "learning_rate": 2e-05,
      "loss": 0.26445555686950684,
      "num_input_tokens_seen": 5598,
      "step": 6,
      "train_runtime": 17.7764,
      "train_tokens_per_second": 314.912
    },
    {
      "epoch": 0.0029917726252804786,
      "grad_norm": 5.587215900421143,
      "learning_rate": 1.999572192513369e-05,
      "loss": 0.11980922520160675,
      "num_input_tokens_seen": 6630,
      "step": 7,
      "train_runtime": 19.843,
      "train_tokens_per_second": 334.124
    },
    {
      "epoch": 0.0034191687146062615,
      "grad_norm": 3.824303388595581,
      "learning_rate": 1.999144385026738e-05,
      "loss": 0.1420525759458542,
      "num_input_tokens_seen": 7824,
      "step": 8,
      "train_runtime": 21.9807,
      "train_tokens_per_second": 355.948
    },
    {
      "epoch": 0.003846564803932044,
      "grad_norm": 3.2600786685943604,
      "learning_rate": 1.9987165775401073e-05,
      "loss": 0.09598482400178909,
      "num_input_tokens_seen": 8920,
      "step": 9,
      "train_runtime": 24.064,
      "train_tokens_per_second": 370.678
    },
    {
      "epoch": 0.0042739608932578265,
      "grad_norm": 3.5554864406585693,
      "learning_rate": 1.998288770053476e-05,
      "loss": 0.18107062578201294,
      "num_input_tokens_seen": 10266,
      "step": 10,
      "train_runtime": 26.2377,
      "train_tokens_per_second": 391.269
    },
    {
      "epoch": 0.004701356982583609,
      "grad_norm": 3.7602057456970215,
      "learning_rate": 1.997860962566845e-05,
      "loss": 0.1214386597275734,
      "num_input_tokens_seen": 11024,
      "step": 11,
      "train_runtime": 28.1906,
      "train_tokens_per_second": 391.052
    },
    {
      "epoch": 0.005128753071909392,
      "grad_norm": 4.019107818603516,
      "learning_rate": 1.997433155080214e-05,
      "loss": 0.23680613934993744,
      "num_input_tokens_seen": 12374,
      "step": 12,
      "train_runtime": 30.3292,
      "train_tokens_per_second": 407.99
    },
    {
      "epoch": 0.005556149161235174,
      "grad_norm": 4.988264083862305,
      "learning_rate": 1.997005347593583e-05,
      "loss": 0.1975032091140747,
      "num_input_tokens_seen": 13532,
      "step": 13,
      "train_runtime": 32.4343,
      "train_tokens_per_second": 417.213
    },
    {
      "epoch": 0.005983545250560957,
      "grad_norm": 5.2693681716918945,
      "learning_rate": 1.996577540106952e-05,
      "loss": 0.14342525601387024,
      "num_input_tokens_seen": 14356,
      "step": 14,
      "train_runtime": 34.4337,
      "train_tokens_per_second": 416.917
    },
    {
      "epoch": 0.00641094133988674,
      "grad_norm": 4.687433242797852,
      "learning_rate": 1.996149732620321e-05,
      "loss": 0.20123495161533356,
      "num_input_tokens_seen": 15532,
      "step": 15,
      "train_runtime": 36.5139,
      "train_tokens_per_second": 425.373
    },
    {
      "epoch": 0.006838337429212523,
      "grad_norm": 3.9584639072418213,
      "learning_rate": 1.99572192513369e-05,
      "loss": 0.16135592758655548,
      "num_input_tokens_seen": 16276,
      "step": 16,
      "train_runtime": 38.4918,
      "train_tokens_per_second": 422.843
    },
    {
      "epoch": 0.007265733518538305,
      "grad_norm": 4.302605628967285,
      "learning_rate": 1.995294117647059e-05,
      "loss": 0.0981002077460289,
      "num_input_tokens_seen": 17076,
      "step": 17,
      "train_runtime": 40.501,
      "train_tokens_per_second": 421.619
    },
    {
      "epoch": 0.007693129607864088,
      "grad_norm": 3.010422706604004,
      "learning_rate": 1.994866310160428e-05,
      "loss": 0.09488782286643982,
      "num_input_tokens_seen": 18224,
      "step": 18,
      "train_runtime": 42.5906,
      "train_tokens_per_second": 427.888
    },
    {
      "epoch": 0.00812052569718987,
      "grad_norm": 3.195756435394287,
      "learning_rate": 1.9944385026737968e-05,
      "loss": 0.11447030305862427,
      "num_input_tokens_seen": 19792,
      "step": 19,
      "train_runtime": 44.7866,
      "train_tokens_per_second": 441.918
    },
    {
      "epoch": 0.008547921786515653,
      "grad_norm": 3.139535903930664,
      "learning_rate": 1.9940106951871657e-05,
      "loss": 0.13146063685417175,
      "num_input_tokens_seen": 20890,
      "step": 20,
      "train_runtime": 46.8964,
      "train_tokens_per_second": 445.45
    },
    {
      "epoch": 0.008975317875841436,
      "grad_norm": 4.932249069213867,
      "learning_rate": 1.993582887700535e-05,
      "loss": 0.1503104567527771,
      "num_input_tokens_seen": 21788,
      "step": 21,
      "train_runtime": 48.9228,
      "train_tokens_per_second": 445.354
    },
    {
      "epoch": 0.009402713965167219,
      "grad_norm": 4.747905254364014,
      "learning_rate": 1.993155080213904e-05,
      "loss": 0.1295323520898819,
      "num_input_tokens_seen": 22586,
      "step": 22,
      "train_runtime": 50.898,
      "train_tokens_per_second": 443.751
    },
    {
      "epoch": 0.009830110054493002,
      "grad_norm": 3.491011619567871,
      "learning_rate": 1.992727272727273e-05,
      "loss": 0.1252719759941101,
      "num_input_tokens_seen": 23474,
      "step": 23,
      "train_runtime": 52.9039,
      "train_tokens_per_second": 443.71
    },
    {
      "epoch": 0.010257506143818785,
      "grad_norm": 5.479099273681641,
      "learning_rate": 1.992299465240642e-05,
      "loss": 0.19048258662223816,
      "num_input_tokens_seen": 24318,
      "step": 24,
      "train_runtime": 54.9243,
      "train_tokens_per_second": 442.755
    },
    {
      "epoch": 0.010684902233144567,
      "grad_norm": 3.316023826599121,
      "learning_rate": 1.9918716577540108e-05,
      "loss": 0.19907110929489136,
      "num_input_tokens_seen": 25596,
      "step": 25,
      "train_runtime": 57.0225,
      "train_tokens_per_second": 448.875
    },
    {
      "epoch": 0.011112298322470349,
      "grad_norm": 3.3836982250213623,
      "learning_rate": 1.9914438502673797e-05,
      "loss": 0.10873676836490631,
      "num_input_tokens_seen": 27330,
      "step": 26,
      "train_runtime": 59.3028,
      "train_tokens_per_second": 460.855
    },
    {
      "epoch": 0.011539694411796132,
      "grad_norm": 4.329500198364258,
      "learning_rate": 1.9910160427807487e-05,
      "loss": 0.1561865508556366,
      "num_input_tokens_seen": 28246,
      "step": 27,
      "train_runtime": 61.28,
      "train_tokens_per_second": 460.934
    },
    {
      "epoch": 0.011967090501121914,
      "grad_norm": 4.218974590301514,
      "learning_rate": 1.9905882352941176e-05,
      "loss": 0.09617075324058533,
      "num_input_tokens_seen": 29652,
      "step": 28,
      "train_runtime": 63.4359,
      "train_tokens_per_second": 467.433
    },
    {
      "epoch": 0.012394486590447697,
      "grad_norm": 5.29585599899292,
      "learning_rate": 1.990160427807487e-05,
      "loss": 0.16989362239837646,
      "num_input_tokens_seen": 30714,
      "step": 29,
      "train_runtime": 65.5293,
      "train_tokens_per_second": 468.706
    },
    {
      "epoch": 0.01282188267977348,
      "grad_norm": 2.316746234893799,
      "learning_rate": 1.989732620320856e-05,
      "loss": 0.11536379158496857,
      "num_input_tokens_seen": 32788,
      "step": 30,
      "train_runtime": 67.8454,
      "train_tokens_per_second": 483.275
    },
    {
      "epoch": 0.013249278769099263,
      "grad_norm": 3.1590397357940674,
      "learning_rate": 1.9893048128342248e-05,
      "loss": 0.0931893065571785,
      "num_input_tokens_seen": 34168,
      "step": 31,
      "train_runtime": 77.2674,
      "train_tokens_per_second": 442.205
    },
    {
      "epoch": 0.013676674858425046,
      "grad_norm": 4.110548496246338,
      "learning_rate": 1.9888770053475937e-05,
      "loss": 0.138652965426445,
      "num_input_tokens_seen": 35112,
      "step": 32,
      "train_runtime": 79.3693,
      "train_tokens_per_second": 442.388
    },
    {
      "epoch": 0.014104070947750827,
      "grad_norm": 3.5470504760742188,
      "learning_rate": 1.9884491978609627e-05,
      "loss": 0.14397886395454407,
      "num_input_tokens_seen": 36280,
      "step": 33,
      "train_runtime": 81.4417,
      "train_tokens_per_second": 445.472
    },
    {
      "epoch": 0.01453146703707661,
      "grad_norm": 2.6868910789489746,
      "learning_rate": 1.9880213903743316e-05,
      "loss": 0.11197999119758606,
      "num_input_tokens_seen": 37392,
      "step": 34,
      "train_runtime": 83.5772,
      "train_tokens_per_second": 447.395
    },
    {
      "epoch": 0.014958863126402393,
      "grad_norm": 2.9456000328063965,
      "learning_rate": 1.9875935828877006e-05,
      "loss": 0.13160617649555206,
      "num_input_tokens_seen": 38426,
      "step": 35,
      "train_runtime": 85.6386,
      "train_tokens_per_second": 448.699
    },
    {
      "epoch": 0.015386259215728176,
      "grad_norm": 5.697889804840088,
      "learning_rate": 1.9871657754010695e-05,
      "loss": 0.19218449294567108,
      "num_input_tokens_seen": 39142,
      "step": 36,
      "train_runtime": 87.593,
      "train_tokens_per_second": 446.862
    },
    {
      "epoch": 0.015813655305053957,
      "grad_norm": 3.2537574768066406,
      "learning_rate": 1.9867379679144388e-05,
      "loss": 0.20149178802967072,
      "num_input_tokens_seen": 40582,
      "step": 37,
      "train_runtime": 89.7798,
      "train_tokens_per_second": 452.017
    },
    {
      "epoch": 0.01624105139437974,
      "grad_norm": 3.8951613903045654,
      "learning_rate": 1.9863101604278077e-05,
      "loss": 0.16156522929668427,
      "num_input_tokens_seen": 41558,
      "step": 38,
      "train_runtime": 91.8047,
      "train_tokens_per_second": 452.678
    },
    {
      "epoch": 0.016668447483705523,
      "grad_norm": 4.315296649932861,
      "learning_rate": 1.9858823529411767e-05,
      "loss": 0.14697720110416412,
      "num_input_tokens_seen": 43008,
      "step": 39,
      "train_runtime": 93.967,
      "train_tokens_per_second": 457.692
    },
    {
      "epoch": 0.017095843573031306,
      "grad_norm": 4.937653541564941,
      "learning_rate": 1.9854545454545456e-05,
      "loss": 0.1605147272348404,
      "num_input_tokens_seen": 43638,
      "step": 40,
      "train_runtime": 95.91,
      "train_tokens_per_second": 454.989
    },
    {
      "epoch": 0.01752323966235709,
      "grad_norm": 3.789418935775757,
      "learning_rate": 1.9850267379679146e-05,
      "loss": 0.1526976227760315,
      "num_input_tokens_seen": 44638,
      "step": 41,
      "train_runtime": 98.04,
      "train_tokens_per_second": 455.304
    },
    {
      "epoch": 0.01795063575168287,
      "grad_norm": 2.424595832824707,
      "learning_rate": 1.9845989304812835e-05,
      "loss": 0.064969502389431,
      "num_input_tokens_seen": 45956,
      "step": 42,
      "train_runtime": 100.1482,
      "train_tokens_per_second": 458.88
    },
    {
      "epoch": 0.018378031841008655,
      "grad_norm": 3.9266443252563477,
      "learning_rate": 1.9841711229946525e-05,
      "loss": 0.14873147010803223,
      "num_input_tokens_seen": 46938,
      "step": 43,
      "train_runtime": 102.2242,
      "train_tokens_per_second": 459.167
    },
    {
      "epoch": 0.018805427930334438,
      "grad_norm": 3.1564815044403076,
      "learning_rate": 1.9837433155080214e-05,
      "loss": 0.1474442481994629,
      "num_input_tokens_seen": 48790,
      "step": 44,
      "train_runtime": 104.4988,
      "train_tokens_per_second": 466.895
    },
    {
      "epoch": 0.01923282401966022,
      "grad_norm": 5.3332719802856445,
      "learning_rate": 1.9833155080213907e-05,
      "loss": 0.15616430342197418,
      "num_input_tokens_seen": 49788,
      "step": 45,
      "train_runtime": 106.5338,
      "train_tokens_per_second": 467.344
    },
    {
      "epoch": 0.019660220108986003,
      "grad_norm": 4.83596134185791,
      "learning_rate": 1.9828877005347596e-05,
      "loss": 0.14318448305130005,
      "num_input_tokens_seen": 50982,
      "step": 46,
      "train_runtime": 108.6462,
      "train_tokens_per_second": 469.248
    },
    {
      "epoch": 0.020087616198311786,
      "grad_norm": 4.82612943649292,
      "learning_rate": 1.9824598930481286e-05,
      "loss": 0.16587546467781067,
      "num_input_tokens_seen": 51774,
      "step": 47,
      "train_runtime": 110.6641,
      "train_tokens_per_second": 467.848
    },
    {
      "epoch": 0.02051501228763757,
      "grad_norm": 5.32522439956665,
      "learning_rate": 1.9820320855614975e-05,
      "loss": 0.1672104001045227,
      "num_input_tokens_seen": 52456,
      "step": 48,
      "train_runtime": 112.6502,
      "train_tokens_per_second": 465.654
    },
    {
      "epoch": 0.020942408376963352,
      "grad_norm": 4.326712608337402,
      "learning_rate": 1.9816042780748665e-05,
      "loss": 0.13658718764781952,
      "num_input_tokens_seen": 53518,
      "step": 49,
      "train_runtime": 114.762,
      "train_tokens_per_second": 466.339
    },
    {
      "epoch": 0.021369804466289135,
      "grad_norm": 3.391953945159912,
      "learning_rate": 1.9811764705882354e-05,
      "loss": 0.08182261884212494,
      "num_input_tokens_seen": 54192,
      "step": 50,
      "train_runtime": 116.7819,
      "train_tokens_per_second": 464.044
    },
    {
      "epoch": 0.021797200555614914,
      "grad_norm": 5.363497734069824,
      "learning_rate": 1.9807486631016044e-05,
      "loss": 0.18326592445373535,
      "num_input_tokens_seen": 55238,
      "step": 51,
      "train_runtime": 118.8673,
      "train_tokens_per_second": 464.703
    },
    {
      "epoch": 0.022224596644940697,
      "grad_norm": 4.2625732421875,
      "learning_rate": 1.9803208556149733e-05,
      "loss": 0.11891967803239822,
      "num_input_tokens_seen": 55944,
      "step": 52,
      "train_runtime": 120.8285,
      "train_tokens_per_second": 463.003
    },
    {
      "epoch": 0.02265199273426648,
      "grad_norm": 2.9568066596984863,
      "learning_rate": 1.9798930481283426e-05,
      "loss": 0.12739425897598267,
      "num_input_tokens_seen": 57432,
      "step": 53,
      "train_runtime": 123.0333,
      "train_tokens_per_second": 466.801
    },
    {
      "epoch": 0.023079388823592263,
      "grad_norm": 4.947142601013184,
      "learning_rate": 1.9794652406417115e-05,
      "loss": 0.11887576431035995,
      "num_input_tokens_seen": 58132,
      "step": 54,
      "train_runtime": 125.0444,
      "train_tokens_per_second": 464.891
    },
    {
      "epoch": 0.023506784912918046,
      "grad_norm": 4.3053202629089355,
      "learning_rate": 1.9790374331550805e-05,
      "loss": 0.2513520121574402,
      "num_input_tokens_seen": 59050,
      "step": 55,
      "train_runtime": 127.0689,
      "train_tokens_per_second": 464.708
    },
    {
      "epoch": 0.02393418100224383,
      "grad_norm": 3.2881264686584473,
      "learning_rate": 1.9786096256684494e-05,
      "loss": 0.14226624369621277,
      "num_input_tokens_seen": 60142,
      "step": 56,
      "train_runtime": 129.1144,
      "train_tokens_per_second": 465.804
    },
    {
      "epoch": 0.024361577091569612,
      "grad_norm": 4.641565799713135,
      "learning_rate": 1.9781818181818184e-05,
      "loss": 0.1254645586013794,
      "num_input_tokens_seen": 60816,
      "step": 57,
      "train_runtime": 131.0719,
      "train_tokens_per_second": 463.99
    },
    {
      "epoch": 0.024788973180895395,
      "grad_norm": 4.213507175445557,
      "learning_rate": 1.9777540106951873e-05,
      "loss": 0.20696590840816498,
      "num_input_tokens_seen": 61912,
      "step": 58,
      "train_runtime": 133.1641,
      "train_tokens_per_second": 464.93
    },
    {
      "epoch": 0.025216369270221178,
      "grad_norm": 3.822394609451294,
      "learning_rate": 1.9773262032085563e-05,
      "loss": 0.12147180736064911,
      "num_input_tokens_seen": 63102,
      "step": 59,
      "train_runtime": 135.3075,
      "train_tokens_per_second": 466.36
    },
    {
      "epoch": 0.02564376535954696,
      "grad_norm": 3.7647974491119385,
      "learning_rate": 1.9768983957219252e-05,
      "loss": 0.10511690378189087,
      "num_input_tokens_seen": 64178,
      "step": 60,
      "train_runtime": 137.4084,
      "train_tokens_per_second": 467.06
    },
    {
      "epoch": 0.026071161448872743,
      "grad_norm": 3.5489540100097656,
      "learning_rate": 1.9764705882352945e-05,
      "loss": 0.16315722465515137,
      "num_input_tokens_seen": 65130,
      "step": 61,
      "train_runtime": 145.9682,
      "train_tokens_per_second": 446.193
    },
    {
      "epoch": 0.026498557538198526,
      "grad_norm": 3.647955894470215,
      "learning_rate": 1.9760427807486634e-05,
      "loss": 0.15680347383022308,
      "num_input_tokens_seen": 66482,
      "step": 62,
      "train_runtime": 148.0724,
      "train_tokens_per_second": 448.983
    },
    {
      "epoch": 0.02692595362752431,
      "grad_norm": 5.077506065368652,
      "learning_rate": 1.9756149732620324e-05,
      "loss": 0.2151620090007782,
      "num_input_tokens_seen": 67294,
      "step": 63,
      "train_runtime": 150.1535,
      "train_tokens_per_second": 448.168
    },
    {
      "epoch": 0.027353349716850092,
      "grad_norm": 4.495523929595947,
      "learning_rate": 1.975187165775401e-05,
      "loss": 0.17818322777748108,
      "num_input_tokens_seen": 68030,
      "step": 64,
      "train_runtime": 152.131,
      "train_tokens_per_second": 447.18
    },
    {
      "epoch": 0.027780745806175875,
      "grad_norm": 2.6798179149627686,
      "learning_rate": 1.9747593582887703e-05,
      "loss": 0.11909501999616623,
      "num_input_tokens_seen": 69248,
      "step": 65,
      "train_runtime": 154.238,
      "train_tokens_per_second": 448.968
    },
    {
      "epoch": 0.028208141895501655,
      "grad_norm": 3.572542667388916,
      "learning_rate": 1.9743315508021392e-05,
      "loss": 0.13439372181892395,
      "num_input_tokens_seen": 70370,
      "step": 66,
      "train_runtime": 156.3623,
      "train_tokens_per_second": 450.044
    },
    {
      "epoch": 0.028635537984827437,
      "grad_norm": 4.2027082443237305,
      "learning_rate": 1.973903743315508e-05,
      "loss": 0.1301841139793396,
      "num_input_tokens_seen": 71626,
      "step": 67,
      "train_runtime": 158.4737,
      "train_tokens_per_second": 451.974
    },
    {
      "epoch": 0.02906293407415322,
      "grad_norm": 2.78859281539917,
      "learning_rate": 1.973475935828877e-05,
      "loss": 0.11929918080568314,
      "num_input_tokens_seen": 73088,
      "step": 68,
      "train_runtime": 160.6123,
      "train_tokens_per_second": 455.058
    },
    {
      "epoch": 0.029490330163479003,
      "grad_norm": 4.005275726318359,
      "learning_rate": 1.9730481283422464e-05,
      "loss": 0.11357735097408295,
      "num_input_tokens_seen": 73794,
      "step": 69,
      "train_runtime": 162.6037,
      "train_tokens_per_second": 453.827
    },
    {
      "epoch": 0.029917726252804786,
      "grad_norm": 5.254064559936523,
      "learning_rate": 1.9726203208556153e-05,
      "loss": 0.14673492312431335,
      "num_input_tokens_seen": 74550,
      "step": 70,
      "train_runtime": 164.6007,
      "train_tokens_per_second": 452.914
    },
    {
      "epoch": 0.03034512234213057,
      "grad_norm": 3.9961674213409424,
      "learning_rate": 1.972192513368984e-05,
      "loss": 0.217291459441185,
      "num_input_tokens_seen": 75664,
      "step": 71,
      "train_runtime": 166.7055,
      "train_tokens_per_second": 453.878
    },
    {
      "epoch": 0.030772518431456352,
      "grad_norm": 5.1501922607421875,
      "learning_rate": 1.971764705882353e-05,
      "loss": 0.1771675944328308,
      "num_input_tokens_seen": 76658,
      "step": 72,
      "train_runtime": 168.7704,
      "train_tokens_per_second": 454.215
    },
    {
      "epoch": 0.031199914520782135,
      "grad_norm": 4.168330192565918,
      "learning_rate": 1.971336898395722e-05,
      "loss": 0.1108478456735611,
      "num_input_tokens_seen": 77340,
      "step": 73,
      "train_runtime": 170.8413,
      "train_tokens_per_second": 452.701
    },
    {
      "epoch": 0.031627310610107914,
      "grad_norm": 3.990004777908325,
      "learning_rate": 1.970909090909091e-05,
      "loss": 0.22652441263198853,
      "num_input_tokens_seen": 78280,
      "step": 74,
      "train_runtime": 172.8849,
      "train_tokens_per_second": 452.787
    },
    {
      "epoch": 0.0320547066994337,
      "grad_norm": 6.247471332550049,
      "learning_rate": 1.97048128342246e-05,
      "loss": 0.2036120593547821,
      "num_input_tokens_seen": 79984,
      "step": 75,
      "train_runtime": 175.2028,
      "train_tokens_per_second": 456.523
    },
    {
      "epoch": 0.03248210278875948,
      "grad_norm": 3.5871591567993164,
      "learning_rate": 1.970053475935829e-05,
      "loss": 0.1707800179719925,
      "num_input_tokens_seen": 81050,
      "step": 76,
      "train_runtime": 177.2701,
      "train_tokens_per_second": 457.212
    },
    {
      "epoch": 0.032909498878085267,
      "grad_norm": 3.6746106147766113,
      "learning_rate": 1.9696256684491983e-05,
      "loss": 0.1824970841407776,
      "num_input_tokens_seen": 82148,
      "step": 77,
      "train_runtime": 179.3628,
      "train_tokens_per_second": 457.999
    },
    {
      "epoch": 0.033336894967411046,
      "grad_norm": 3.5396571159362793,
      "learning_rate": 1.969197860962567e-05,
      "loss": 0.17174741625785828,
      "num_input_tokens_seen": 83480,
      "step": 78,
      "train_runtime": 181.5274,
      "train_tokens_per_second": 459.875
    },
    {
      "epoch": 0.03376429105673683,
      "grad_norm": 4.849555015563965,
      "learning_rate": 1.9687700534759358e-05,
      "loss": 0.19618165493011475,
      "num_input_tokens_seen": 84510,
      "step": 79,
      "train_runtime": 183.6277,
      "train_tokens_per_second": 460.225
    },
    {
      "epoch": 0.03419168714606261,
      "grad_norm": 3.5431630611419678,
      "learning_rate": 1.9683422459893048e-05,
      "loss": 0.1181945949792862,
      "num_input_tokens_seen": 85856,
      "step": 80,
      "train_runtime": 185.743,
      "train_tokens_per_second": 462.23
    },
    {
      "epoch": 0.0346190832353884,
      "grad_norm": 3.853771686553955,
      "learning_rate": 1.967914438502674e-05,
      "loss": 0.11620461940765381,
      "num_input_tokens_seen": 86658,
      "step": 81,
      "train_runtime": 187.7391,
      "train_tokens_per_second": 461.587
    },
    {
      "epoch": 0.03504647932471418,
      "grad_norm": 3.4037678241729736,
      "learning_rate": 1.967486631016043e-05,
      "loss": 0.13744604587554932,
      "num_input_tokens_seen": 88316,
      "step": 82,
      "train_runtime": 189.9384,
      "train_tokens_per_second": 464.972
    },
    {
      "epoch": 0.035473875414039964,
      "grad_norm": 3.401839256286621,
      "learning_rate": 1.967058823529412e-05,
      "loss": 0.10648271441459656,
      "num_input_tokens_seen": 89176,
      "step": 83,
      "train_runtime": 191.9721,
      "train_tokens_per_second": 464.526
    },
    {
      "epoch": 0.03590127150336574,
      "grad_norm": 4.106384754180908,
      "learning_rate": 1.966631016042781e-05,
      "loss": 0.16465850174427032,
      "num_input_tokens_seen": 90190,
      "step": 84,
      "train_runtime": 194.0187,
      "train_tokens_per_second": 464.852
    },
    {
      "epoch": 0.03632866759269153,
      "grad_norm": 3.0353920459747314,
      "learning_rate": 1.96620320855615e-05,
      "loss": 0.12087035179138184,
      "num_input_tokens_seen": 91646,
      "step": 85,
      "train_runtime": 196.1609,
      "train_tokens_per_second": 467.198
    },
    {
      "epoch": 0.03675606368201731,
      "grad_norm": 4.285816192626953,
      "learning_rate": 1.9657754010695188e-05,
      "loss": 0.12293040752410889,
      "num_input_tokens_seen": 92540,
      "step": 86,
      "train_runtime": 198.1857,
      "train_tokens_per_second": 466.936
    },
    {
      "epoch": 0.037183459771343096,
      "grad_norm": 4.207911968231201,
      "learning_rate": 1.9653475935828877e-05,
      "loss": 0.18469344079494476,
      "num_input_tokens_seen": 93406,
      "step": 87,
      "train_runtime": 200.1647,
      "train_tokens_per_second": 466.646
    },
    {
      "epoch": 0.037610855860668875,
      "grad_norm": 4.45616340637207,
      "learning_rate": 1.9649197860962567e-05,
      "loss": 0.11300913989543915,
      "num_input_tokens_seen": 94042,
      "step": 88,
      "train_runtime": 202.1109,
      "train_tokens_per_second": 465.299
    },
    {
      "epoch": 0.038038251949994654,
      "grad_norm": 3.776406764984131,
      "learning_rate": 1.964491978609626e-05,
      "loss": 0.16000477969646454,
      "num_input_tokens_seen": 94768,
      "step": 89,
      "train_runtime": 204.0899,
      "train_tokens_per_second": 464.344
    },
    {
      "epoch": 0.03846564803932044,
      "grad_norm": 4.5214457511901855,
      "learning_rate": 1.964064171122995e-05,
      "loss": 0.14950791001319885,
      "num_input_tokens_seen": 95510,
      "step": 90,
      "train_runtime": 206.0628,
      "train_tokens_per_second": 463.5
    },
    {
      "epoch": 0.03889304412864622,
      "grad_norm": 3.776623010635376,
      "learning_rate": 1.963636363636364e-05,
      "loss": 0.1654057800769806,
      "num_input_tokens_seen": 96404,
      "step": 91,
      "train_runtime": 214.542,
      "train_tokens_per_second": 449.348
    },
    {
      "epoch": 0.03932044021797201,
      "grad_norm": 4.654515743255615,
      "learning_rate": 1.9632085561497328e-05,
      "loss": 0.30709636211395264,
      "num_input_tokens_seen": 97326,
      "step": 92,
      "train_runtime": 216.5679,
      "train_tokens_per_second": 449.402
    },
    {
      "epoch": 0.039747836307297786,
      "grad_norm": 3.743364095687866,
      "learning_rate": 1.9627807486631017e-05,
      "loss": 0.1282462775707245,
      "num_input_tokens_seen": 98670,
      "step": 93,
      "train_runtime": 218.7124,
      "train_tokens_per_second": 451.14
    },
    {
      "epoch": 0.04017523239662357,
      "grad_norm": 2.7732701301574707,
      "learning_rate": 1.9623529411764707e-05,
      "loss": 0.14717315137386322,
      "num_input_tokens_seen": 99802,
      "step": 94,
      "train_runtime": 220.7705,
      "train_tokens_per_second": 452.062
    },
    {
      "epoch": 0.04060262848594935,
      "grad_norm": 4.163231372833252,
      "learning_rate": 1.9619251336898396e-05,
      "loss": 0.13073277473449707,
      "num_input_tokens_seen": 100602,
      "step": 95,
      "train_runtime": 222.7312,
      "train_tokens_per_second": 451.675
    },
    {
      "epoch": 0.04103002457527514,
      "grad_norm": 4.084043979644775,
      "learning_rate": 1.9614973262032086e-05,
      "loss": 0.211775004863739,
      "num_input_tokens_seen": 101492,
      "step": 96,
      "train_runtime": 224.7527,
      "train_tokens_per_second": 451.572
    },
    {
      "epoch": 0.04145742066460092,
      "grad_norm": 2.473571538925171,
      "learning_rate": 1.961069518716578e-05,
      "loss": 0.11081207543611526,
      "num_input_tokens_seen": 102878,
      "step": 97,
      "train_runtime": 226.8868,
      "train_tokens_per_second": 453.433
    },
    {
      "epoch": 0.041884816753926704,
      "grad_norm": 3.7261149883270264,
      "learning_rate": 1.9606417112299468e-05,
      "loss": 0.13922911882400513,
      "num_input_tokens_seen": 103492,
      "step": 98,
      "train_runtime": 228.8324,
      "train_tokens_per_second": 452.261
    },
    {
      "epoch": 0.042312212843252484,
      "grad_norm": 3.1666626930236816,
      "learning_rate": 1.9602139037433157e-05,
      "loss": 0.1227794736623764,
      "num_input_tokens_seen": 104662,
      "step": 99,
      "train_runtime": 230.915,
      "train_tokens_per_second": 453.249
    },
    {
      "epoch": 0.04273960893257827,
      "grad_norm": 5.331000804901123,
      "learning_rate": 1.9597860962566847e-05,
      "loss": 0.2264232635498047,
      "num_input_tokens_seen": 105546,
      "step": 100,
      "train_runtime": 232.9534,
      "train_tokens_per_second": 453.078
    },
    {
      "epoch": 0.04316700502190405,
      "grad_norm": 3.614386558532715,
      "learning_rate": 1.9593582887700536e-05,
      "loss": 0.12469115853309631,
      "num_input_tokens_seen": 106606,
      "step": 101,
      "train_runtime": 235.0222,
      "train_tokens_per_second": 453.6
    },
    {
      "epoch": 0.04359440111122983,
      "grad_norm": 5.98403787612915,
      "learning_rate": 1.9589304812834226e-05,
      "loss": 0.13769271969795227,
      "num_input_tokens_seen": 107334,
      "step": 102,
      "train_runtime": 237.0603,
      "train_tokens_per_second": 452.771
    },
    {
      "epoch": 0.044021797200555615,
      "grad_norm": 3.6699941158294678,
      "learning_rate": 1.9585026737967915e-05,
      "loss": 0.13916221261024475,
      "num_input_tokens_seen": 108792,
      "step": 103,
      "train_runtime": 239.2198,
      "train_tokens_per_second": 454.779
    },
    {
      "epoch": 0.044449193289881395,
      "grad_norm": 4.686874866485596,
      "learning_rate": 1.9580748663101605e-05,
      "loss": 0.1531163901090622,
      "num_input_tokens_seen": 109786,
      "step": 104,
      "train_runtime": 241.2466,
      "train_tokens_per_second": 455.078
    },
    {
      "epoch": 0.04487658937920718,
      "grad_norm": 3.2465693950653076,
      "learning_rate": 1.9576470588235297e-05,
      "loss": 0.1527305245399475,
      "num_input_tokens_seen": 111036,
      "step": 105,
      "train_runtime": 243.3561,
      "train_tokens_per_second": 456.27
    },
    {
      "epoch": 0.04530398546853296,
      "grad_norm": 2.720834493637085,
      "learning_rate": 1.9572192513368987e-05,
      "loss": 0.09989616274833679,
      "num_input_tokens_seen": 112130,
      "step": 106,
      "train_runtime": 245.4039,
      "train_tokens_per_second": 456.92
    },
    {
      "epoch": 0.04573138155785875,
      "grad_norm": 4.310506820678711,
      "learning_rate": 1.9567914438502676e-05,
      "loss": 0.21421846747398376,
      "num_input_tokens_seen": 112996,
      "step": 107,
      "train_runtime": 247.4173,
      "train_tokens_per_second": 456.702
    },
    {
      "epoch": 0.046158777647184526,
      "grad_norm": 2.9606587886810303,
      "learning_rate": 1.9563636363636366e-05,
      "loss": 0.10993418842554092,
      "num_input_tokens_seen": 113962,
      "step": 108,
      "train_runtime": 249.4791,
      "train_tokens_per_second": 456.8
    },
    {
      "epoch": 0.04658617373651031,
      "grad_norm": 3.8038618564605713,
      "learning_rate": 1.9559358288770055e-05,
      "loss": 0.1932617723941803,
      "num_input_tokens_seen": 115666,
      "step": 109,
      "train_runtime": 251.7418,
      "train_tokens_per_second": 459.463
    },
    {
      "epoch": 0.04701356982583609,
      "grad_norm": 4.124566555023193,
      "learning_rate": 1.9555080213903745e-05,
      "loss": 0.1846320629119873,
      "num_input_tokens_seen": 116624,
      "step": 110,
      "train_runtime": 253.7975,
      "train_tokens_per_second": 459.516
    },
    {
      "epoch": 0.04744096591516188,
      "grad_norm": 3.0976104736328125,
      "learning_rate": 1.9550802139037434e-05,
      "loss": 0.09712183475494385,
      "num_input_tokens_seen": 117428,
      "step": 111,
      "train_runtime": 255.8459,
      "train_tokens_per_second": 458.979
    },
    {
      "epoch": 0.04786836200448766,
      "grad_norm": 2.8251500129699707,
      "learning_rate": 1.9546524064171124e-05,
      "loss": 0.11785268783569336,
      "num_input_tokens_seen": 118680,
      "step": 112,
      "train_runtime": 257.9633,
      "train_tokens_per_second": 460.065
    },
    {
      "epoch": 0.048295758093813444,
      "grad_norm": 2.817950487136841,
      "learning_rate": 1.9542245989304816e-05,
      "loss": 0.15405887365341187,
      "num_input_tokens_seen": 119692,
      "step": 113,
      "train_runtime": 260.0271,
      "train_tokens_per_second": 460.306
    },
    {
      "epoch": 0.048723154183139224,
      "grad_norm": 3.703707456588745,
      "learning_rate": 1.9537967914438506e-05,
      "loss": 0.13587014377117157,
      "num_input_tokens_seen": 120922,
      "step": 114,
      "train_runtime": 262.1133,
      "train_tokens_per_second": 461.335
    },
    {
      "epoch": 0.04915055027246501,
      "grad_norm": 4.744223594665527,
      "learning_rate": 1.9533689839572195e-05,
      "loss": 0.16133588552474976,
      "num_input_tokens_seen": 121918,
      "step": 115,
      "train_runtime": 264.1712,
      "train_tokens_per_second": 461.511
    },
    {
      "epoch": 0.04957794636179079,
      "grad_norm": 3.794447898864746,
      "learning_rate": 1.9529411764705885e-05,
      "loss": 0.2528957724571228,
      "num_input_tokens_seen": 122906,
      "step": 116,
      "train_runtime": 266.1873,
      "train_tokens_per_second": 461.727
    },
    {
      "epoch": 0.05000534245111657,
      "grad_norm": 3.831779718399048,
      "learning_rate": 1.9525133689839574e-05,
      "loss": 0.12381330132484436,
      "num_input_tokens_seen": 123910,
      "step": 117,
      "train_runtime": 268.2229,
      "train_tokens_per_second": 461.967
    },
    {
      "epoch": 0.050432738540442355,
      "grad_norm": 2.9667787551879883,
      "learning_rate": 1.9520855614973264e-05,
      "loss": 0.13823501765727997,
      "num_input_tokens_seen": 125102,
      "step": 118,
      "train_runtime": 270.3734,
      "train_tokens_per_second": 462.701
    },
    {
      "epoch": 0.050860134629768135,
      "grad_norm": 5.961679935455322,
      "learning_rate": 1.9516577540106953e-05,
      "loss": 0.25376904010772705,
      "num_input_tokens_seen": 126132,
      "step": 119,
      "train_runtime": 272.4895,
      "train_tokens_per_second": 462.888
    },
    {
      "epoch": 0.05128753071909392,
      "grad_norm": 5.949172019958496,
      "learning_rate": 1.9512299465240642e-05,
      "loss": 0.18413153290748596,
      "num_input_tokens_seen": 127110,
      "step": 120,
      "train_runtime": 274.5537,
      "train_tokens_per_second": 462.97
    },
    {
      "epoch": 0.0517149268084197,
      "grad_norm": 3.9526119232177734,
      "learning_rate": 1.9508021390374332e-05,
      "loss": 0.14880630373954773,
      "num_input_tokens_seen": 127798,
      "step": 121,
      "train_runtime": 283.068,
      "train_tokens_per_second": 451.475
    },
    {
      "epoch": 0.05214232289774549,
      "grad_norm": 4.685976505279541,
      "learning_rate": 1.9503743315508025e-05,
      "loss": 0.2541481852531433,
      "num_input_tokens_seen": 128626,
      "step": 122,
      "train_runtime": 285.0617,
      "train_tokens_per_second": 451.222
    },
    {
      "epoch": 0.052569718987071266,
      "grad_norm": 3.6252996921539307,
      "learning_rate": 1.9499465240641714e-05,
      "loss": 0.17328515648841858,
      "num_input_tokens_seen": 129792,
      "step": 123,
      "train_runtime": 287.1561,
      "train_tokens_per_second": 451.991
    },
    {
      "epoch": 0.05299711507639705,
      "grad_norm": 4.383351802825928,
      "learning_rate": 1.9495187165775404e-05,
      "loss": 0.17913982272148132,
      "num_input_tokens_seen": 131134,
      "step": 124,
      "train_runtime": 289.2781,
      "train_tokens_per_second": 453.315
    },
    {
      "epoch": 0.05342451116572283,
      "grad_norm": 4.613114833831787,
      "learning_rate": 1.9490909090909093e-05,
      "loss": 0.1824687123298645,
      "num_input_tokens_seen": 132050,
      "step": 125,
      "train_runtime": 291.3056,
      "train_tokens_per_second": 453.304
    },
    {
      "epoch": 0.05385190725504862,
      "grad_norm": 3.256249189376831,
      "learning_rate": 1.9486631016042783e-05,
      "loss": 0.12739244103431702,
      "num_input_tokens_seen": 133038,
      "step": 126,
      "train_runtime": 293.3833,
      "train_tokens_per_second": 453.461
    },
    {
      "epoch": 0.0542793033443744,
      "grad_norm": 4.473914623260498,
      "learning_rate": 1.9482352941176472e-05,
      "loss": 0.274713397026062,
      "num_input_tokens_seen": 134130,
      "step": 127,
      "train_runtime": 295.4583,
      "train_tokens_per_second": 453.973
    },
    {
      "epoch": 0.054706699433700184,
      "grad_norm": 3.998966693878174,
      "learning_rate": 1.947807486631016e-05,
      "loss": 0.1771961748600006,
      "num_input_tokens_seen": 134706,
      "step": 128,
      "train_runtime": 297.3794,
      "train_tokens_per_second": 452.977
    },
    {
      "epoch": 0.055134095523025964,
      "grad_norm": 2.9364333152770996,
      "learning_rate": 1.947379679144385e-05,
      "loss": 0.1560664176940918,
      "num_input_tokens_seen": 135978,
      "step": 129,
      "train_runtime": 299.5065,
      "train_tokens_per_second": 454.007
    },
    {
      "epoch": 0.05556149161235175,
      "grad_norm": 3.1906626224517822,
      "learning_rate": 1.9469518716577544e-05,
      "loss": 0.14533768594264984,
      "num_input_tokens_seen": 137126,
      "step": 130,
      "train_runtime": 301.6252,
      "train_tokens_per_second": 454.624
    },
    {
      "epoch": 0.05598888770167753,
      "grad_norm": 2.6740944385528564,
      "learning_rate": 1.9465240641711233e-05,
      "loss": 0.1516144573688507,
      "num_input_tokens_seen": 138358,
      "step": 131,
      "train_runtime": 303.7418,
      "train_tokens_per_second": 455.512
    },
    {
      "epoch": 0.05641628379100331,
      "grad_norm": 3.373631477355957,
      "learning_rate": 1.946096256684492e-05,
      "loss": 0.16386520862579346,
      "num_input_tokens_seen": 139642,
      "step": 132,
      "train_runtime": 305.8914,
      "train_tokens_per_second": 456.508
    },
    {
      "epoch": 0.056843679880329095,
      "grad_norm": 4.113182544708252,
      "learning_rate": 1.945668449197861e-05,
      "loss": 0.1717812865972519,
      "num_input_tokens_seen": 140602,
      "step": 133,
      "train_runtime": 308.0106,
      "train_tokens_per_second": 456.484
    },
    {
      "epoch": 0.057271075969654875,
      "grad_norm": 4.087924003601074,
      "learning_rate": 1.94524064171123e-05,
      "loss": 0.1626863032579422,
      "num_input_tokens_seen": 141428,
      "step": 134,
      "train_runtime": 310.0902,
      "train_tokens_per_second": 456.087
    },
    {
      "epoch": 0.05769847205898066,
      "grad_norm": 4.410618782043457,
      "learning_rate": 1.944812834224599e-05,
      "loss": 0.22780272364616394,
      "num_input_tokens_seen": 142422,
      "step": 135,
      "train_runtime": 312.1452,
      "train_tokens_per_second": 456.268
    },
    {
      "epoch": 0.05812586814830644,
      "grad_norm": 3.759174108505249,
      "learning_rate": 1.944385026737968e-05,
      "loss": 0.19898691773414612,
      "num_input_tokens_seen": 143346,
      "step": 136,
      "train_runtime": 314.2361,
      "train_tokens_per_second": 456.173
    },
    {
      "epoch": 0.05855326423763223,
      "grad_norm": 3.668994903564453,
      "learning_rate": 1.943957219251337e-05,
      "loss": 0.13790655136108398,
      "num_input_tokens_seen": 144322,
      "step": 137,
      "train_runtime": 316.297,
      "train_tokens_per_second": 456.286
    },
    {
      "epoch": 0.05898066032695801,
      "grad_norm": 2.9102942943573,
      "learning_rate": 1.9435294117647063e-05,
      "loss": 0.15275537967681885,
      "num_input_tokens_seen": 145602,
      "step": 138,
      "train_runtime": 318.4458,
      "train_tokens_per_second": 457.227
    },
    {
      "epoch": 0.05940805641628379,
      "grad_norm": 3.5204591751098633,
      "learning_rate": 1.943101604278075e-05,
      "loss": 0.15551750361919403,
      "num_input_tokens_seen": 146480,
      "step": 139,
      "train_runtime": 320.5139,
      "train_tokens_per_second": 457.016
    },
    {
      "epoch": 0.05983545250560957,
      "grad_norm": 3.573094129562378,
      "learning_rate": 1.9426737967914438e-05,
      "loss": 0.1385328769683838,
      "num_input_tokens_seen": 147280,
      "step": 140,
      "train_runtime": 322.5122,
      "train_tokens_per_second": 456.665
    },
    {
      "epoch": 0.06026284859493536,
      "grad_norm": 2.3515357971191406,
      "learning_rate": 1.9422459893048128e-05,
      "loss": 0.08091916143894196,
      "num_input_tokens_seen": 148260,
      "step": 141,
      "train_runtime": 324.5442,
      "train_tokens_per_second": 456.825
    },
    {
      "epoch": 0.06069024468426114,
      "grad_norm": 3.5823183059692383,
      "learning_rate": 1.941818181818182e-05,
      "loss": 0.2145853489637375,
      "num_input_tokens_seen": 149334,
      "step": 142,
      "train_runtime": 326.6077,
      "train_tokens_per_second": 457.227
    },
    {
      "epoch": 0.061117640773586925,
      "grad_norm": 3.408435344696045,
      "learning_rate": 1.941390374331551e-05,
      "loss": 0.18264466524124146,
      "num_input_tokens_seen": 150540,
      "step": 143,
      "train_runtime": 328.7099,
      "train_tokens_per_second": 457.972
    },
    {
      "epoch": 0.061545036862912704,
      "grad_norm": 4.721390724182129,
      "learning_rate": 1.94096256684492e-05,
      "loss": 0.1382511407136917,
      "num_input_tokens_seen": 151166,
      "step": 144,
      "train_runtime": 330.6693,
      "train_tokens_per_second": 457.152
    },
    {
      "epoch": 0.06197243295223849,
      "grad_norm": 3.0328853130340576,
      "learning_rate": 1.940534759358289e-05,
      "loss": 0.14752715826034546,
      "num_input_tokens_seen": 152380,
      "step": 145,
      "train_runtime": 332.7832,
      "train_tokens_per_second": 457.896
    },
    {
      "epoch": 0.06239982904156427,
      "grad_norm": 4.689001560211182,
      "learning_rate": 1.9401069518716578e-05,
      "loss": 0.25872471928596497,
      "num_input_tokens_seen": 153262,
      "step": 146,
      "train_runtime": 334.8068,
      "train_tokens_per_second": 457.763
    },
    {
      "epoch": 0.06282722513089005,
      "grad_norm": 4.3762078285217285,
      "learning_rate": 1.9396791443850268e-05,
      "loss": 0.25528281927108765,
      "num_input_tokens_seen": 154144,
      "step": 147,
      "train_runtime": 336.8297,
      "train_tokens_per_second": 457.632
    },
    {
      "epoch": 0.06325462122021583,
      "grad_norm": 4.54288911819458,
      "learning_rate": 1.9392513368983957e-05,
      "loss": 0.24058307707309723,
      "num_input_tokens_seen": 155234,
      "step": 148,
      "train_runtime": 338.9007,
      "train_tokens_per_second": 458.052
    },
    {
      "epoch": 0.06368201730954162,
      "grad_norm": 3.4448704719543457,
      "learning_rate": 1.9388235294117647e-05,
      "loss": 0.13716450333595276,
      "num_input_tokens_seen": 156324,
      "step": 149,
      "train_runtime": 341.0108,
      "train_tokens_per_second": 458.414
    },
    {
      "epoch": 0.0641094133988674,
      "grad_norm": 3.4153506755828857,
      "learning_rate": 1.938395721925134e-05,
      "loss": 0.12832224369049072,
      "num_input_tokens_seen": 157768,
      "step": 150,
      "train_runtime": 343.2138,
      "train_tokens_per_second": 459.678
    },
    {
      "epoch": 0.06453680948819318,
      "grad_norm": 5.297286033630371,
      "learning_rate": 1.937967914438503e-05,
      "loss": 0.1760442703962326,
      "num_input_tokens_seen": 158492,
      "step": 151,
      "train_runtime": 351.9666,
      "train_tokens_per_second": 450.304
    },
    {
      "epoch": 0.06496420557751896,
      "grad_norm": 3.6865053176879883,
      "learning_rate": 1.9375401069518718e-05,
      "loss": 0.15094995498657227,
      "num_input_tokens_seen": 159674,
      "step": 152,
      "train_runtime": 354.1275,
      "train_tokens_per_second": 450.894
    },
    {
      "epoch": 0.06539160166684475,
      "grad_norm": 6.42773962020874,
      "learning_rate": 1.9371122994652408e-05,
      "loss": 0.19943645596504211,
      "num_input_tokens_seen": 160804,
      "step": 153,
      "train_runtime": 356.214,
      "train_tokens_per_second": 451.425
    },
    {
      "epoch": 0.06581899775617053,
      "grad_norm": 3.904852867126465,
      "learning_rate": 1.9366844919786097e-05,
      "loss": 0.12334674596786499,
      "num_input_tokens_seen": 161480,
      "step": 154,
      "train_runtime": 358.2,
      "train_tokens_per_second": 450.81
    },
    {
      "epoch": 0.06624639384549631,
      "grad_norm": 2.9093892574310303,
      "learning_rate": 1.9362566844919787e-05,
      "loss": 0.14115239679813385,
      "num_input_tokens_seen": 162474,
      "step": 155,
      "train_runtime": 360.282,
      "train_tokens_per_second": 450.963
    },
    {
      "epoch": 0.06667378993482209,
      "grad_norm": 2.907697916030884,
      "learning_rate": 1.9358288770053476e-05,
      "loss": 0.13603979349136353,
      "num_input_tokens_seen": 163560,
      "step": 156,
      "train_runtime": 362.3778,
      "train_tokens_per_second": 451.352
    },
    {
      "epoch": 0.06710118602414789,
      "grad_norm": 3.3026914596557617,
      "learning_rate": 1.9354010695187166e-05,
      "loss": 0.14058558642864227,
      "num_input_tokens_seen": 164948,
      "step": 157,
      "train_runtime": 364.5367,
      "train_tokens_per_second": 452.487
    },
    {
      "epoch": 0.06752858211347366,
      "grad_norm": 4.028748989105225,
      "learning_rate": 1.934973262032086e-05,
      "loss": 0.18366575241088867,
      "num_input_tokens_seen": 166248,
      "step": 158,
      "train_runtime": 366.6538,
      "train_tokens_per_second": 453.42
    },
    {
      "epoch": 0.06795597820279944,
      "grad_norm": 3.1829679012298584,
      "learning_rate": 1.9345454545454548e-05,
      "loss": 0.17139247059822083,
      "num_input_tokens_seen": 167720,
      "step": 159,
      "train_runtime": 368.8367,
      "train_tokens_per_second": 454.727
    },
    {
      "epoch": 0.06838337429212522,
      "grad_norm": 3.4020562171936035,
      "learning_rate": 1.9341176470588237e-05,
      "loss": 0.14117524027824402,
      "num_input_tokens_seen": 168542,
      "step": 160,
      "train_runtime": 370.8654,
      "train_tokens_per_second": 454.456
    },
    {
      "epoch": 0.068810770381451,
      "grad_norm": 3.8880434036254883,
      "learning_rate": 1.9336898395721927e-05,
      "loss": 0.18208377063274384,
      "num_input_tokens_seen": 169602,
      "step": 161,
      "train_runtime": 372.959,
      "train_tokens_per_second": 454.747
    },
    {
      "epoch": 0.0692381664707768,
      "grad_norm": 3.70001220703125,
      "learning_rate": 1.9332620320855616e-05,
      "loss": 0.21020248532295227,
      "num_input_tokens_seen": 170802,
      "step": 162,
      "train_runtime": 375.2126,
      "train_tokens_per_second": 455.214
    },
    {
      "epoch": 0.06966556256010258,
      "grad_norm": 4.150683879852295,
      "learning_rate": 1.9328342245989306e-05,
      "loss": 0.19267253577709198,
      "num_input_tokens_seen": 171650,
      "step": 163,
      "train_runtime": 377.2294,
      "train_tokens_per_second": 455.028
    },
    {
      "epoch": 0.07009295864942836,
      "grad_norm": 4.252824783325195,
      "learning_rate": 1.9324064171122995e-05,
      "loss": 0.15652082860469818,
      "num_input_tokens_seen": 172970,
      "step": 164,
      "train_runtime": 379.3629,
      "train_tokens_per_second": 455.949
    },
    {
      "epoch": 0.07052035473875413,
      "grad_norm": 3.9441115856170654,
      "learning_rate": 1.9319786096256684e-05,
      "loss": 0.12914994359016418,
      "num_input_tokens_seen": 173750,
      "step": 165,
      "train_runtime": 381.3812,
      "train_tokens_per_second": 455.581
    },
    {
      "epoch": 0.07094775082807993,
      "grad_norm": 4.179220676422119,
      "learning_rate": 1.9315508021390377e-05,
      "loss": 0.12721984088420868,
      "num_input_tokens_seen": 174504,
      "step": 166,
      "train_runtime": 383.3764,
      "train_tokens_per_second": 455.177
    },
    {
      "epoch": 0.07137514691740571,
      "grad_norm": 2.911614418029785,
      "learning_rate": 1.9311229946524067e-05,
      "loss": 0.1691466122865677,
      "num_input_tokens_seen": 176096,
      "step": 167,
      "train_runtime": 385.5704,
      "train_tokens_per_second": 456.716
    },
    {
      "epoch": 0.07180254300673149,
      "grad_norm": 3.517160415649414,
      "learning_rate": 1.9306951871657756e-05,
      "loss": 0.08075811713933945,
      "num_input_tokens_seen": 176826,
      "step": 168,
      "train_runtime": 387.5449,
      "train_tokens_per_second": 456.272
    },
    {
      "epoch": 0.07222993909605727,
      "grad_norm": 5.577939510345459,
      "learning_rate": 1.9302673796791446e-05,
      "loss": 0.18033367395401,
      "num_input_tokens_seen": 177404,
      "step": 169,
      "train_runtime": 389.5037,
      "train_tokens_per_second": 455.462
    },
    {
      "epoch": 0.07265733518538306,
      "grad_norm": 2.962841749191284,
      "learning_rate": 1.9298395721925135e-05,
      "loss": 0.19206425547599792,
      "num_input_tokens_seen": 179500,
      "step": 170,
      "train_runtime": 391.8038,
      "train_tokens_per_second": 458.137
    },
    {
      "epoch": 0.07308473127470884,
      "grad_norm": 3.3081376552581787,
      "learning_rate": 1.9294117647058825e-05,
      "loss": 0.14741308987140656,
      "num_input_tokens_seen": 180552,
      "step": 171,
      "train_runtime": 393.8775,
      "train_tokens_per_second": 458.396
    },
    {
      "epoch": 0.07351212736403462,
      "grad_norm": 3.485266923904419,
      "learning_rate": 1.9289839572192514e-05,
      "loss": 0.1569867581129074,
      "num_input_tokens_seen": 181674,
      "step": 172,
      "train_runtime": 395.9655,
      "train_tokens_per_second": 458.813
    },
    {
      "epoch": 0.0739395234533604,
      "grad_norm": 2.7815213203430176,
      "learning_rate": 1.9285561497326203e-05,
      "loss": 0.19787417352199554,
      "num_input_tokens_seen": 183070,
      "step": 173,
      "train_runtime": 398.0846,
      "train_tokens_per_second": 459.877
    },
    {
      "epoch": 0.07436691954268619,
      "grad_norm": 2.8555409908294678,
      "learning_rate": 1.9281283422459896e-05,
      "loss": 0.08727098256349564,
      "num_input_tokens_seen": 184102,
      "step": 174,
      "train_runtime": 400.1283,
      "train_tokens_per_second": 460.107
    },
    {
      "epoch": 0.07479431563201197,
      "grad_norm": 3.37363338470459,
      "learning_rate": 1.9277005347593586e-05,
      "loss": 0.09344178438186646,
      "num_input_tokens_seen": 184738,
      "step": 175,
      "train_runtime": 402.0703,
      "train_tokens_per_second": 459.467
    },
    {
      "epoch": 0.07522171172133775,
      "grad_norm": 3.6110727787017822,
      "learning_rate": 1.9272727272727275e-05,
      "loss": 0.15122315287590027,
      "num_input_tokens_seen": 185658,
      "step": 176,
      "train_runtime": 404.0902,
      "train_tokens_per_second": 459.447
    },
    {
      "epoch": 0.07564910781066353,
      "grad_norm": 4.353687286376953,
      "learning_rate": 1.9268449197860965e-05,
      "loss": 0.20483148097991943,
      "num_input_tokens_seen": 186548,
      "step": 177,
      "train_runtime": 406.1061,
      "train_tokens_per_second": 459.358
    },
    {
      "epoch": 0.07607650389998931,
      "grad_norm": 3.902709484100342,
      "learning_rate": 1.9264171122994654e-05,
      "loss": 0.1338711380958557,
      "num_input_tokens_seen": 187216,
      "step": 178,
      "train_runtime": 408.055,
      "train_tokens_per_second": 458.801
    },
    {
      "epoch": 0.0765038999893151,
      "grad_norm": 4.904794692993164,
      "learning_rate": 1.9259893048128344e-05,
      "loss": 0.14415208995342255,
      "num_input_tokens_seen": 187852,
      "step": 179,
      "train_runtime": 409.9733,
      "train_tokens_per_second": 458.205
    },
    {
      "epoch": 0.07693129607864088,
      "grad_norm": 2.637103796005249,
      "learning_rate": 1.9255614973262033e-05,
      "loss": 0.09054088592529297,
      "num_input_tokens_seen": 188776,
      "step": 180,
      "train_runtime": 412.1239,
      "train_tokens_per_second": 458.056
    },
    {
      "epoch": 0.07735869216796666,
      "grad_norm": 3.8350894451141357,
      "learning_rate": 1.9251336898395722e-05,
      "loss": 0.1061239242553711,
      "num_input_tokens_seen": 189660,
      "step": 181,
      "train_runtime": 420.8459,
      "train_tokens_per_second": 450.664
    },
    {
      "epoch": 0.07778608825729244,
      "grad_norm": 3.1410531997680664,
      "learning_rate": 1.9247058823529415e-05,
      "loss": 0.1559763252735138,
      "num_input_tokens_seen": 190660,
      "step": 182,
      "train_runtime": 422.8953,
      "train_tokens_per_second": 450.844
    },
    {
      "epoch": 0.07821348434661823,
      "grad_norm": 4.812954425811768,
      "learning_rate": 1.9242780748663105e-05,
      "loss": 0.19170966744422913,
      "num_input_tokens_seen": 191388,
      "step": 183,
      "train_runtime": 424.8678,
      "train_tokens_per_second": 450.465
    },
    {
      "epoch": 0.07864088043594401,
      "grad_norm": 3.0507349967956543,
      "learning_rate": 1.9238502673796794e-05,
      "loss": 0.1721566766500473,
      "num_input_tokens_seen": 192522,
      "step": 184,
      "train_runtime": 426.9841,
      "train_tokens_per_second": 450.888
    },
    {
      "epoch": 0.07906827652526979,
      "grad_norm": 2.3724539279937744,
      "learning_rate": 1.9234224598930484e-05,
      "loss": 0.0992976725101471,
      "num_input_tokens_seen": 193656,
      "step": 185,
      "train_runtime": 429.0901,
      "train_tokens_per_second": 451.318
    },
    {
      "epoch": 0.07949567261459557,
      "grad_norm": 4.356204032897949,
      "learning_rate": 1.9229946524064173e-05,
      "loss": 0.17176751792430878,
      "num_input_tokens_seen": 194402,
      "step": 186,
      "train_runtime": 431.0928,
      "train_tokens_per_second": 450.952
    },
    {
      "epoch": 0.07992306870392137,
      "grad_norm": 3.711698055267334,
      "learning_rate": 1.9225668449197862e-05,
      "loss": 0.19728952646255493,
      "num_input_tokens_seen": 195820,
      "step": 187,
      "train_runtime": 433.2757,
      "train_tokens_per_second": 451.952
    },
    {
      "epoch": 0.08035046479324714,
      "grad_norm": 3.471689224243164,
      "learning_rate": 1.9221390374331552e-05,
      "loss": 0.12796440720558167,
      "num_input_tokens_seen": 196646,
      "step": 188,
      "train_runtime": 435.2937,
      "train_tokens_per_second": 451.755
    },
    {
      "epoch": 0.08077786088257292,
      "grad_norm": 5.114136219024658,
      "learning_rate": 1.921711229946524e-05,
      "loss": 0.16669723391532898,
      "num_input_tokens_seen": 197934,
      "step": 189,
      "train_runtime": 437.4352,
      "train_tokens_per_second": 452.488
    },
    {
      "epoch": 0.0812052569718987,
      "grad_norm": 4.230232238769531,
      "learning_rate": 1.9212834224598934e-05,
      "loss": 0.16396945714950562,
      "num_input_tokens_seen": 198668,
      "step": 190,
      "train_runtime": 439.4562,
      "train_tokens_per_second": 452.077
    },
    {
      "epoch": 0.08163265306122448,
      "grad_norm": 3.165710210800171,
      "learning_rate": 1.9208556149732624e-05,
      "loss": 0.12626244127750397,
      "num_input_tokens_seen": 199602,
      "step": 191,
      "train_runtime": 441.5322,
      "train_tokens_per_second": 452.067
    },
    {
      "epoch": 0.08206004915055028,
      "grad_norm": 3.7299795150756836,
      "learning_rate": 1.9204278074866313e-05,
      "loss": 0.20722240209579468,
      "num_input_tokens_seen": 200542,
      "step": 192,
      "train_runtime": 443.5588,
      "train_tokens_per_second": 452.12
    },
    {
      "epoch": 0.08248744523987606,
      "grad_norm": 4.410600662231445,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.2075149416923523,
      "num_input_tokens_seen": 201758,
      "step": 193,
      "train_runtime": 445.6815,
      "train_tokens_per_second": 452.696
    },
    {
      "epoch": 0.08291484132920184,
      "grad_norm": 3.4818227291107178,
      "learning_rate": 1.9195721925133692e-05,
      "loss": 0.19731946289539337,
      "num_input_tokens_seen": 203172,
      "step": 194,
      "train_runtime": 447.8314,
      "train_tokens_per_second": 453.68
    },
    {
      "epoch": 0.08334223741852761,
      "grad_norm": 3.2609193325042725,
      "learning_rate": 1.919144385026738e-05,
      "loss": 0.1395885944366455,
      "num_input_tokens_seen": 204078,
      "step": 195,
      "train_runtime": 449.9033,
      "train_tokens_per_second": 453.604
    },
    {
      "epoch": 0.08376963350785341,
      "grad_norm": 3.4876768589019775,
      "learning_rate": 1.918716577540107e-05,
      "loss": 0.16306047141551971,
      "num_input_tokens_seen": 204922,
      "step": 196,
      "train_runtime": 451.9484,
      "train_tokens_per_second": 453.419
    },
    {
      "epoch": 0.08419702959717919,
      "grad_norm": 2.7984631061553955,
      "learning_rate": 1.918288770053476e-05,
      "loss": 0.11668877303600311,
      "num_input_tokens_seen": 206042,
      "step": 197,
      "train_runtime": 454.031,
      "train_tokens_per_second": 453.806
    },
    {
      "epoch": 0.08462442568650497,
      "grad_norm": 2.316157579421997,
      "learning_rate": 1.9178609625668453e-05,
      "loss": 0.07432881742715836,
      "num_input_tokens_seen": 207324,
      "step": 198,
      "train_runtime": 456.1505,
      "train_tokens_per_second": 454.508
    },
    {
      "epoch": 0.08505182177583075,
      "grad_norm": 5.142146110534668,
      "learning_rate": 1.9174331550802143e-05,
      "loss": 0.32583293318748474,
      "num_input_tokens_seen": 207952,
      "step": 199,
      "train_runtime": 458.116,
      "train_tokens_per_second": 453.929
    },
    {
      "epoch": 0.08547921786515654,
      "grad_norm": 4.039165496826172,
      "learning_rate": 1.917005347593583e-05,
      "loss": 0.12693336606025696,
      "num_input_tokens_seen": 208652,
      "step": 200,
      "train_runtime": 460.1111,
      "train_tokens_per_second": 453.482
    },
    {
      "epoch": 0.08590661395448232,
      "grad_norm": 2.3335464000701904,
      "learning_rate": 1.9165775401069518e-05,
      "loss": 0.0983176976442337,
      "num_input_tokens_seen": 210218,
      "step": 201,
      "train_runtime": 462.3251,
      "train_tokens_per_second": 454.697
    },
    {
      "epoch": 0.0863340100438081,
      "grad_norm": 4.560773849487305,
      "learning_rate": 1.916149732620321e-05,
      "loss": 0.18655210733413696,
      "num_input_tokens_seen": 211226,
      "step": 202,
      "train_runtime": 464.3615,
      "train_tokens_per_second": 454.874
    },
    {
      "epoch": 0.08676140613313388,
      "grad_norm": 4.267395973205566,
      "learning_rate": 1.91572192513369e-05,
      "loss": 0.2624877095222473,
      "num_input_tokens_seen": 212022,
      "step": 203,
      "train_runtime": 466.3547,
      "train_tokens_per_second": 454.637
    },
    {
      "epoch": 0.08718880222245966,
      "grad_norm": 4.703396320343018,
      "learning_rate": 1.915294117647059e-05,
      "loss": 0.21235334873199463,
      "num_input_tokens_seen": 212970,
      "step": 204,
      "train_runtime": 468.4126,
      "train_tokens_per_second": 454.663
    },
    {
      "epoch": 0.08761619831178545,
      "grad_norm": 2.8434624671936035,
      "learning_rate": 1.914866310160428e-05,
      "loss": 0.15252263844013214,
      "num_input_tokens_seen": 214772,
      "step": 205,
      "train_runtime": 470.6775,
      "train_tokens_per_second": 456.304
    },
    {
      "epoch": 0.08804359440111123,
      "grad_norm": 3.717108964920044,
      "learning_rate": 1.9144385026737972e-05,
      "loss": 0.17825952172279358,
      "num_input_tokens_seen": 215602,
      "step": 206,
      "train_runtime": 472.6784,
      "train_tokens_per_second": 456.128
    },
    {
      "epoch": 0.08847099049043701,
      "grad_norm": 2.9121856689453125,
      "learning_rate": 1.9140106951871658e-05,
      "loss": 0.15489216148853302,
      "num_input_tokens_seen": 216490,
      "step": 207,
      "train_runtime": 474.6974,
      "train_tokens_per_second": 456.059
    },
    {
      "epoch": 0.08889838657976279,
      "grad_norm": 3.9117062091827393,
      "learning_rate": 1.9135828877005348e-05,
      "loss": 0.1688605397939682,
      "num_input_tokens_seen": 217204,
      "step": 208,
      "train_runtime": 476.6486,
      "train_tokens_per_second": 455.69
    },
    {
      "epoch": 0.08932578266908858,
      "grad_norm": 3.487771987915039,
      "learning_rate": 1.9131550802139037e-05,
      "loss": 0.1434318572282791,
      "num_input_tokens_seen": 218624,
      "step": 209,
      "train_runtime": 478.8243,
      "train_tokens_per_second": 456.585
    },
    {
      "epoch": 0.08975317875841436,
      "grad_norm": 2.691993474960327,
      "learning_rate": 1.912727272727273e-05,
      "loss": 0.13809660077095032,
      "num_input_tokens_seen": 220324,
      "step": 210,
      "train_runtime": 481.061,
      "train_tokens_per_second": 457.996
    },
    {
      "epoch": 0.09018057484774014,
      "grad_norm": 3.8660988807678223,
      "learning_rate": 1.912299465240642e-05,
      "loss": 0.19372284412384033,
      "num_input_tokens_seen": 221488,
      "step": 211,
      "train_runtime": 489.7622,
      "train_tokens_per_second": 452.236
    },
    {
      "epoch": 0.09060797093706592,
      "grad_norm": 2.50512433052063,
      "learning_rate": 1.911871657754011e-05,
      "loss": 0.09720499813556671,
      "num_input_tokens_seen": 222612,
      "step": 212,
      "train_runtime": 491.8538,
      "train_tokens_per_second": 452.598
    },
    {
      "epoch": 0.09103536702639171,
      "grad_norm": 4.057933807373047,
      "learning_rate": 1.9114438502673798e-05,
      "loss": 0.16461125016212463,
      "num_input_tokens_seen": 223292,
      "step": 213,
      "train_runtime": 493.8672,
      "train_tokens_per_second": 452.13
    },
    {
      "epoch": 0.0914627631157175,
      "grad_norm": 2.801384449005127,
      "learning_rate": 1.9110160427807488e-05,
      "loss": 0.14904817938804626,
      "num_input_tokens_seen": 224310,
      "step": 214,
      "train_runtime": 495.8975,
      "train_tokens_per_second": 452.331
    },
    {
      "epoch": 0.09189015920504327,
      "grad_norm": 3.8825550079345703,
      "learning_rate": 1.9105882352941177e-05,
      "loss": 0.2628094255924225,
      "num_input_tokens_seen": 225312,
      "step": 215,
      "train_runtime": 497.9456,
      "train_tokens_per_second": 452.483
    },
    {
      "epoch": 0.09231755529436905,
      "grad_norm": 3.3652145862579346,
      "learning_rate": 1.9101604278074867e-05,
      "loss": 0.22525590658187866,
      "num_input_tokens_seen": 226316,
      "step": 216,
      "train_runtime": 500.0115,
      "train_tokens_per_second": 452.622
    },
    {
      "epoch": 0.09274495138369485,
      "grad_norm": 2.7347452640533447,
      "learning_rate": 1.9097326203208556e-05,
      "loss": 0.14280107617378235,
      "num_input_tokens_seen": 227250,
      "step": 217,
      "train_runtime": 502.106,
      "train_tokens_per_second": 452.594
    },
    {
      "epoch": 0.09317234747302063,
      "grad_norm": 5.178623676300049,
      "learning_rate": 1.909304812834225e-05,
      "loss": 0.17395099997520447,
      "num_input_tokens_seen": 228122,
      "step": 218,
      "train_runtime": 504.1438,
      "train_tokens_per_second": 452.494
    },
    {
      "epoch": 0.0935997435623464,
      "grad_norm": 3.997356414794922,
      "learning_rate": 1.9088770053475938e-05,
      "loss": 0.1781000792980194,
      "num_input_tokens_seen": 229052,
      "step": 219,
      "train_runtime": 506.2813,
      "train_tokens_per_second": 452.42
    },
    {
      "epoch": 0.09402713965167218,
      "grad_norm": 3.2037484645843506,
      "learning_rate": 1.9084491978609628e-05,
      "loss": 0.1139114573597908,
      "num_input_tokens_seen": 229902,
      "step": 220,
      "train_runtime": 508.2989,
      "train_tokens_per_second": 452.297
    },
    {
      "epoch": 0.09445453574099796,
      "grad_norm": 4.276766777038574,
      "learning_rate": 1.9080213903743317e-05,
      "loss": 0.27040156722068787,
      "num_input_tokens_seen": 230608,
      "step": 221,
      "train_runtime": 510.271,
      "train_tokens_per_second": 451.932
    },
    {
      "epoch": 0.09488193183032376,
      "grad_norm": 2.974623441696167,
      "learning_rate": 1.9075935828877007e-05,
      "loss": 0.14041531085968018,
      "num_input_tokens_seen": 231556,
      "step": 222,
      "train_runtime": 512.302,
      "train_tokens_per_second": 451.991
    },
    {
      "epoch": 0.09530932791964954,
      "grad_norm": 3.309757709503174,
      "learning_rate": 1.9071657754010696e-05,
      "loss": 0.17925582826137543,
      "num_input_tokens_seen": 232788,
      "step": 223,
      "train_runtime": 514.3899,
      "train_tokens_per_second": 452.552
    },
    {
      "epoch": 0.09573672400897532,
      "grad_norm": 4.808985233306885,
      "learning_rate": 1.9067379679144386e-05,
      "loss": 0.1371680200099945,
      "num_input_tokens_seen": 233850,
      "step": 224,
      "train_runtime": 516.4849,
      "train_tokens_per_second": 452.772
    },
    {
      "epoch": 0.0961641200983011,
      "grad_norm": 2.974057197570801,
      "learning_rate": 1.9063101604278075e-05,
      "loss": 0.1604079306125641,
      "num_input_tokens_seen": 234864,
      "step": 225,
      "train_runtime": 518.6088,
      "train_tokens_per_second": 452.873
    },
    {
      "epoch": 0.09659151618762689,
      "grad_norm": 3.4561188220977783,
      "learning_rate": 1.9058823529411764e-05,
      "loss": 0.1315935105085373,
      "num_input_tokens_seen": 235708,
      "step": 226,
      "train_runtime": 520.6123,
      "train_tokens_per_second": 452.751
    },
    {
      "epoch": 0.09701891227695267,
      "grad_norm": 3.487989902496338,
      "learning_rate": 1.9054545454545457e-05,
      "loss": 0.2212015688419342,
      "num_input_tokens_seen": 236636,
      "step": 227,
      "train_runtime": 522.6007,
      "train_tokens_per_second": 452.805
    },
    {
      "epoch": 0.09744630836627845,
      "grad_norm": 3.0441582202911377,
      "learning_rate": 1.9050267379679147e-05,
      "loss": 0.09992823004722595,
      "num_input_tokens_seen": 237612,
      "step": 228,
      "train_runtime": 524.6806,
      "train_tokens_per_second": 452.87
    },
    {
      "epoch": 0.09787370445560423,
      "grad_norm": 3.008308172225952,
      "learning_rate": 1.9045989304812836e-05,
      "loss": 0.1457144320011139,
      "num_input_tokens_seen": 238602,
      "step": 229,
      "train_runtime": 526.7311,
      "train_tokens_per_second": 452.986
    },
    {
      "epoch": 0.09830110054493002,
      "grad_norm": 3.072152853012085,
      "learning_rate": 1.9041711229946526e-05,
      "loss": 0.16858015954494476,
      "num_input_tokens_seen": 239698,
      "step": 230,
      "train_runtime": 528.7856,
      "train_tokens_per_second": 453.299
    },
    {
      "epoch": 0.0987284966342558,
      "grad_norm": 3.4098429679870605,
      "learning_rate": 1.9037433155080215e-05,
      "loss": 0.17184992134571075,
      "num_input_tokens_seen": 240628,
      "step": 231,
      "train_runtime": 530.8023,
      "train_tokens_per_second": 453.329
    },
    {
      "epoch": 0.09915589272358158,
      "grad_norm": 3.863887071609497,
      "learning_rate": 1.9033155080213904e-05,
      "loss": 0.17547230422496796,
      "num_input_tokens_seen": 241414,
      "step": 232,
      "train_runtime": 532.7911,
      "train_tokens_per_second": 453.112
    },
    {
      "epoch": 0.09958328881290736,
      "grad_norm": 4.293150424957275,
      "learning_rate": 1.9028877005347594e-05,
      "loss": 0.1179661750793457,
      "num_input_tokens_seen": 242032,
      "step": 233,
      "train_runtime": 534.7148,
      "train_tokens_per_second": 452.638
    },
    {
      "epoch": 0.10001068490223314,
      "grad_norm": 2.6410746574401855,
      "learning_rate": 1.9024598930481283e-05,
      "loss": 0.126165509223938,
      "num_input_tokens_seen": 243242,
      "step": 234,
      "train_runtime": 536.7878,
      "train_tokens_per_second": 453.144
    },
    {
      "epoch": 0.10043808099155893,
      "grad_norm": 3.003371477127075,
      "learning_rate": 1.9020320855614976e-05,
      "loss": 0.15211904048919678,
      "num_input_tokens_seen": 244358,
      "step": 235,
      "train_runtime": 538.8639,
      "train_tokens_per_second": 453.469
    },
    {
      "epoch": 0.10086547708088471,
      "grad_norm": 6.7676873207092285,
      "learning_rate": 1.9016042780748666e-05,
      "loss": 0.1265118569135666,
      "num_input_tokens_seen": 245862,
      "step": 236,
      "train_runtime": 541.0146,
      "train_tokens_per_second": 454.446
    },
    {
      "epoch": 0.10129287317021049,
      "grad_norm": 4.040054798126221,
      "learning_rate": 1.9011764705882355e-05,
      "loss": 0.1766088455915451,
      "num_input_tokens_seen": 246806,
      "step": 237,
      "train_runtime": 543.0326,
      "train_tokens_per_second": 454.496
    },
    {
      "epoch": 0.10172026925953627,
      "grad_norm": 3.4513156414031982,
      "learning_rate": 1.9007486631016045e-05,
      "loss": 0.15230941772460938,
      "num_input_tokens_seen": 248106,
      "step": 238,
      "train_runtime": 545.187,
      "train_tokens_per_second": 455.084
    },
    {
      "epoch": 0.10214766534886206,
      "grad_norm": 3.84849214553833,
      "learning_rate": 1.9003208556149734e-05,
      "loss": 0.1926119327545166,
      "num_input_tokens_seen": 249094,
      "step": 239,
      "train_runtime": 547.2525,
      "train_tokens_per_second": 455.172
    },
    {
      "epoch": 0.10257506143818784,
      "grad_norm": 3.4827184677124023,
      "learning_rate": 1.8998930481283423e-05,
      "loss": 0.10055464506149292,
      "num_input_tokens_seen": 249848,
      "step": 240,
      "train_runtime": 549.2401,
      "train_tokens_per_second": 454.898
    },
    {
      "epoch": 0.10300245752751362,
      "grad_norm": 3.3936355113983154,
      "learning_rate": 1.8994652406417113e-05,
      "loss": 0.1272302269935608,
      "num_input_tokens_seen": 250600,
      "step": 241,
      "train_runtime": 557.7911,
      "train_tokens_per_second": 449.272
    },
    {
      "epoch": 0.1034298536168394,
      "grad_norm": 3.431915760040283,
      "learning_rate": 1.8990374331550802e-05,
      "loss": 0.16067558526992798,
      "num_input_tokens_seen": 251420,
      "step": 242,
      "train_runtime": 559.8074,
      "train_tokens_per_second": 449.119
    },
    {
      "epoch": 0.1038572497061652,
      "grad_norm": 3.095625638961792,
      "learning_rate": 1.8986096256684495e-05,
      "loss": 0.16233760118484497,
      "num_input_tokens_seen": 252652,
      "step": 243,
      "train_runtime": 562.0576,
      "train_tokens_per_second": 449.513
    },
    {
      "epoch": 0.10428464579549097,
      "grad_norm": 4.025144100189209,
      "learning_rate": 1.8981818181818185e-05,
      "loss": 0.17672306299209595,
      "num_input_tokens_seen": 253610,
      "step": 244,
      "train_runtime": 564.1197,
      "train_tokens_per_second": 449.568
    },
    {
      "epoch": 0.10471204188481675,
      "grad_norm": 3.420201539993286,
      "learning_rate": 1.8977540106951874e-05,
      "loss": 0.1417708694934845,
      "num_input_tokens_seen": 254638,
      "step": 245,
      "train_runtime": 566.1637,
      "train_tokens_per_second": 449.76
    },
    {
      "epoch": 0.10513943797414253,
      "grad_norm": 4.7547383308410645,
      "learning_rate": 1.8973262032085563e-05,
      "loss": 0.1744721382856369,
      "num_input_tokens_seen": 255274,
      "step": 246,
      "train_runtime": 568.1083,
      "train_tokens_per_second": 449.34
    },
    {
      "epoch": 0.10556683406346833,
      "grad_norm": 3.470569610595703,
      "learning_rate": 1.8968983957219253e-05,
      "loss": 0.19191935658454895,
      "num_input_tokens_seen": 256368,
      "step": 247,
      "train_runtime": 570.161,
      "train_tokens_per_second": 449.641
    },
    {
      "epoch": 0.1059942301527941,
      "grad_norm": 4.300562381744385,
      "learning_rate": 1.8964705882352942e-05,
      "loss": 0.1759026050567627,
      "num_input_tokens_seen": 257026,
      "step": 248,
      "train_runtime": 572.1358,
      "train_tokens_per_second": 449.24
    },
    {
      "epoch": 0.10642162624211988,
      "grad_norm": 4.417792320251465,
      "learning_rate": 1.8960427807486632e-05,
      "loss": 0.10642141103744507,
      "num_input_tokens_seen": 257790,
      "step": 249,
      "train_runtime": 574.1049,
      "train_tokens_per_second": 449.029
    },
    {
      "epoch": 0.10684902233144566,
      "grad_norm": 3.709242820739746,
      "learning_rate": 1.895614973262032e-05,
      "loss": 0.16776417195796967,
      "num_input_tokens_seen": 259262,
      "step": 250,
      "train_runtime": 576.2591,
      "train_tokens_per_second": 449.905
    },
    {
      "epoch": 0.10727641842077144,
      "grad_norm": 3.8251121044158936,
      "learning_rate": 1.8951871657754014e-05,
      "loss": 0.15895061194896698,
      "num_input_tokens_seen": 260110,
      "step": 251,
      "train_runtime": 578.3145,
      "train_tokens_per_second": 449.773
    },
    {
      "epoch": 0.10770381451009724,
      "grad_norm": 3.2912485599517822,
      "learning_rate": 1.8947593582887704e-05,
      "loss": 0.14582106471061707,
      "num_input_tokens_seen": 261606,
      "step": 252,
      "train_runtime": 580.4902,
      "train_tokens_per_second": 450.664
    },
    {
      "epoch": 0.10813121059942302,
      "grad_norm": 2.9822049140930176,
      "learning_rate": 1.8943315508021393e-05,
      "loss": 0.1275487244129181,
      "num_input_tokens_seen": 262526,
      "step": 253,
      "train_runtime": 582.5099,
      "train_tokens_per_second": 450.681
    },
    {
      "epoch": 0.1085586066887488,
      "grad_norm": 4.629343032836914,
      "learning_rate": 1.8939037433155082e-05,
      "loss": 0.16082903742790222,
      "num_input_tokens_seen": 263480,
      "step": 254,
      "train_runtime": 584.55,
      "train_tokens_per_second": 450.74
    },
    {
      "epoch": 0.10898600277807458,
      "grad_norm": 4.594621658325195,
      "learning_rate": 1.8934759358288772e-05,
      "loss": 0.15713101625442505,
      "num_input_tokens_seen": 264340,
      "step": 255,
      "train_runtime": 586.5634,
      "train_tokens_per_second": 450.659
    },
    {
      "epoch": 0.10941339886740037,
      "grad_norm": 2.8088879585266113,
      "learning_rate": 1.893048128342246e-05,
      "loss": 0.17190785706043243,
      "num_input_tokens_seen": 265376,
      "step": 256,
      "train_runtime": 588.6112,
      "train_tokens_per_second": 450.851
    },
    {
      "epoch": 0.10984079495672615,
      "grad_norm": 4.621315956115723,
      "learning_rate": 1.892620320855615e-05,
      "loss": 0.21718566119670868,
      "num_input_tokens_seen": 266354,
      "step": 257,
      "train_runtime": 590.71,
      "train_tokens_per_second": 450.905
    },
    {
      "epoch": 0.11026819104605193,
      "grad_norm": 2.8214497566223145,
      "learning_rate": 1.892192513368984e-05,
      "loss": 0.15230679512023926,
      "num_input_tokens_seen": 267484,
      "step": 258,
      "train_runtime": 592.7989,
      "train_tokens_per_second": 451.222
    },
    {
      "epoch": 0.11069558713537771,
      "grad_norm": 3.5636396408081055,
      "learning_rate": 1.8917647058823533e-05,
      "loss": 0.18290041387081146,
      "num_input_tokens_seen": 268488,
      "step": 259,
      "train_runtime": 594.8901,
      "train_tokens_per_second": 451.324
    },
    {
      "epoch": 0.1111229832247035,
      "grad_norm": 3.2929975986480713,
      "learning_rate": 1.8913368983957222e-05,
      "loss": 0.16350272297859192,
      "num_input_tokens_seen": 269740,
      "step": 260,
      "train_runtime": 597.0846,
      "train_tokens_per_second": 451.762
    },
    {
      "epoch": 0.11155037931402928,
      "grad_norm": 2.861598014831543,
      "learning_rate": 1.8909090909090912e-05,
      "loss": 0.14524902403354645,
      "num_input_tokens_seen": 270722,
      "step": 261,
      "train_runtime": 599.15,
      "train_tokens_per_second": 451.843
    },
    {
      "epoch": 0.11197777540335506,
      "grad_norm": 3.105642795562744,
      "learning_rate": 1.8904812834224598e-05,
      "loss": 0.16657987236976624,
      "num_input_tokens_seen": 271516,
      "step": 262,
      "train_runtime": 601.1344,
      "train_tokens_per_second": 451.673
    },
    {
      "epoch": 0.11240517149268084,
      "grad_norm": 4.586379528045654,
      "learning_rate": 1.890053475935829e-05,
      "loss": 0.14546626806259155,
      "num_input_tokens_seen": 272224,
      "step": 263,
      "train_runtime": 603.1388,
      "train_tokens_per_second": 451.346
    },
    {
      "epoch": 0.11283256758200662,
      "grad_norm": 3.1321394443511963,
      "learning_rate": 1.889625668449198e-05,
      "loss": 0.1691616326570511,
      "num_input_tokens_seen": 273658,
      "step": 264,
      "train_runtime": 605.2636,
      "train_tokens_per_second": 452.13
    },
    {
      "epoch": 0.11325996367133241,
      "grad_norm": 4.213000774383545,
      "learning_rate": 1.889197860962567e-05,
      "loss": 0.17923316359519958,
      "num_input_tokens_seen": 274538,
      "step": 265,
      "train_runtime": 607.291,
      "train_tokens_per_second": 452.07
    },
    {
      "epoch": 0.11368735976065819,
      "grad_norm": 4.686862945556641,
      "learning_rate": 1.888770053475936e-05,
      "loss": 0.2047138214111328,
      "num_input_tokens_seen": 275238,
      "step": 266,
      "train_runtime": 609.2469,
      "train_tokens_per_second": 451.768
    },
    {
      "epoch": 0.11411475584998397,
      "grad_norm": 3.2828869819641113,
      "learning_rate": 1.8883422459893052e-05,
      "loss": 0.2003101110458374,
      "num_input_tokens_seen": 276596,
      "step": 267,
      "train_runtime": 611.3822,
      "train_tokens_per_second": 452.411
    },
    {
      "epoch": 0.11454215193930975,
      "grad_norm": 3.989023208618164,
      "learning_rate": 1.8879144385026738e-05,
      "loss": 0.14641880989074707,
      "num_input_tokens_seen": 277522,
      "step": 268,
      "train_runtime": 613.4081,
      "train_tokens_per_second": 452.426
    },
    {
      "epoch": 0.11496954802863554,
      "grad_norm": 3.4730944633483887,
      "learning_rate": 1.8874866310160427e-05,
      "loss": 0.20197799801826477,
      "num_input_tokens_seen": 279086,
      "step": 269,
      "train_runtime": 615.5842,
      "train_tokens_per_second": 453.368
    },
    {
      "epoch": 0.11539694411796132,
      "grad_norm": 3.5127711296081543,
      "learning_rate": 1.8870588235294117e-05,
      "loss": 0.1377829611301422,
      "num_input_tokens_seen": 280000,
      "step": 270,
      "train_runtime": 617.6133,
      "train_tokens_per_second": 453.358
    },
    {
      "epoch": 0.1158243402072871,
      "grad_norm": 3.486196756362915,
      "learning_rate": 1.886631016042781e-05,
      "loss": 0.12544484436511993,
      "num_input_tokens_seen": 280854,
      "step": 271,
      "train_runtime": 626.7152,
      "train_tokens_per_second": 448.137
    },
    {
      "epoch": 0.11625173629661288,
      "grad_norm": 4.045497417449951,
      "learning_rate": 1.88620320855615e-05,
      "loss": 0.1471298187971115,
      "num_input_tokens_seen": 281708,
      "step": 272,
      "train_runtime": 628.752,
      "train_tokens_per_second": 448.043
    },
    {
      "epoch": 0.11667913238593867,
      "grad_norm": 2.6336021423339844,
      "learning_rate": 1.885775401069519e-05,
      "loss": 0.1305892914533615,
      "num_input_tokens_seen": 283286,
      "step": 273,
      "train_runtime": 630.9488,
      "train_tokens_per_second": 448.984
    },
    {
      "epoch": 0.11710652847526445,
      "grad_norm": 3.9806833267211914,
      "learning_rate": 1.8853475935828878e-05,
      "loss": 0.1073172464966774,
      "num_input_tokens_seen": 283902,
      "step": 274,
      "train_runtime": 632.931,
      "train_tokens_per_second": 448.551
    },
    {
      "epoch": 0.11753392456459023,
      "grad_norm": 4.577077865600586,
      "learning_rate": 1.8849197860962568e-05,
      "loss": 0.13305075466632843,
      "num_input_tokens_seen": 284622,
      "step": 275,
      "train_runtime": 634.8798,
      "train_tokens_per_second": 448.309
    },
    {
      "epoch": 0.11796132065391601,
      "grad_norm": 2.967496395111084,
      "learning_rate": 1.8844919786096257e-05,
      "loss": 0.21203455328941345,
      "num_input_tokens_seen": 286416,
      "step": 276,
      "train_runtime": 637.0768,
      "train_tokens_per_second": 449.578
    },
    {
      "epoch": 0.11838871674324179,
      "grad_norm": 3.238326072692871,
      "learning_rate": 1.8840641711229946e-05,
      "loss": 0.11164429038763046,
      "num_input_tokens_seen": 287294,
      "step": 277,
      "train_runtime": 639.0896,
      "train_tokens_per_second": 449.536
    },
    {
      "epoch": 0.11881611283256759,
      "grad_norm": 3.6869401931762695,
      "learning_rate": 1.8836363636363636e-05,
      "loss": 0.10534940659999847,
      "num_input_tokens_seen": 288488,
      "step": 278,
      "train_runtime": 641.1629,
      "train_tokens_per_second": 449.945
    },
    {
      "epoch": 0.11924350892189337,
      "grad_norm": 3.2805209159851074,
      "learning_rate": 1.883208556149733e-05,
      "loss": 0.12838447093963623,
      "num_input_tokens_seen": 289728,
      "step": 279,
      "train_runtime": 643.2919,
      "train_tokens_per_second": 450.383
    },
    {
      "epoch": 0.11967090501121914,
      "grad_norm": 4.235654830932617,
      "learning_rate": 1.8827807486631018e-05,
      "loss": 0.13418586552143097,
      "num_input_tokens_seen": 290828,
      "step": 280,
      "train_runtime": 645.357,
      "train_tokens_per_second": 450.647
    },
    {
      "epoch": 0.12009830110054492,
      "grad_norm": 2.6712050437927246,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 0.08912317454814911,
      "num_input_tokens_seen": 291796,
      "step": 281,
      "train_runtime": 647.3817,
      "train_tokens_per_second": 450.733
    },
    {
      "epoch": 0.12052569718987072,
      "grad_norm": 3.322579860687256,
      "learning_rate": 1.8819251336898397e-05,
      "loss": 0.14409010112285614,
      "num_input_tokens_seen": 292782,
      "step": 282,
      "train_runtime": 649.4428,
      "train_tokens_per_second": 450.82
    },
    {
      "epoch": 0.1209530932791965,
      "grad_norm": 6.024010181427002,
      "learning_rate": 1.8814973262032087e-05,
      "loss": 0.12049269676208496,
      "num_input_tokens_seen": 294044,
      "step": 283,
      "train_runtime": 651.5748,
      "train_tokens_per_second": 451.282
    },
    {
      "epoch": 0.12138048936852228,
      "grad_norm": 3.4060072898864746,
      "learning_rate": 1.8810695187165776e-05,
      "loss": 0.15825903415679932,
      "num_input_tokens_seen": 295040,
      "step": 284,
      "train_runtime": 653.6757,
      "train_tokens_per_second": 451.355
    },
    {
      "epoch": 0.12180788545784806,
      "grad_norm": 2.8432233333587646,
      "learning_rate": 1.8806417112299465e-05,
      "loss": 0.12002027779817581,
      "num_input_tokens_seen": 296080,
      "step": 285,
      "train_runtime": 655.7539,
      "train_tokens_per_second": 451.511
    },
    {
      "epoch": 0.12223528154717385,
      "grad_norm": 3.7648465633392334,
      "learning_rate": 1.8802139037433155e-05,
      "loss": 0.3100983202457428,
      "num_input_tokens_seen": 296966,
      "step": 286,
      "train_runtime": 657.7684,
      "train_tokens_per_second": 451.475
    },
    {
      "epoch": 0.12266267763649963,
      "grad_norm": 4.5041399002075195,
      "learning_rate": 1.8797860962566848e-05,
      "loss": 0.13915427029132843,
      "num_input_tokens_seen": 297852,
      "step": 287,
      "train_runtime": 659.8361,
      "train_tokens_per_second": 451.403
    },
    {
      "epoch": 0.12309007372582541,
      "grad_norm": 4.070438385009766,
      "learning_rate": 1.8793582887700537e-05,
      "loss": 0.221987783908844,
      "num_input_tokens_seen": 298816,
      "step": 288,
      "train_runtime": 661.8856,
      "train_tokens_per_second": 451.462
    },
    {
      "epoch": 0.12351746981515119,
      "grad_norm": 4.3437089920043945,
      "learning_rate": 1.8789304812834227e-05,
      "loss": 0.1698419451713562,
      "num_input_tokens_seen": 299528,
      "step": 289,
      "train_runtime": 663.8516,
      "train_tokens_per_second": 451.197
    },
    {
      "epoch": 0.12394486590447698,
      "grad_norm": 3.5586817264556885,
      "learning_rate": 1.8785026737967916e-05,
      "loss": 0.20045310258865356,
      "num_input_tokens_seen": 300536,
      "step": 290,
      "train_runtime": 665.9378,
      "train_tokens_per_second": 451.297
    },
    {
      "epoch": 0.12437226199380276,
      "grad_norm": 3.427109718322754,
      "learning_rate": 1.8780748663101605e-05,
      "loss": 0.1372043788433075,
      "num_input_tokens_seen": 301440,
      "step": 291,
      "train_runtime": 667.979,
      "train_tokens_per_second": 451.272
    },
    {
      "epoch": 0.12479965808312854,
      "grad_norm": 3.365877866744995,
      "learning_rate": 1.8776470588235295e-05,
      "loss": 0.15128262341022491,
      "num_input_tokens_seen": 302300,
      "step": 292,
      "train_runtime": 670.0594,
      "train_tokens_per_second": 451.154
    },
    {
      "epoch": 0.12522705417245433,
      "grad_norm": 3.641059398651123,
      "learning_rate": 1.8772192513368984e-05,
      "loss": 0.1684384047985077,
      "num_input_tokens_seen": 303418,
      "step": 293,
      "train_runtime": 672.15,
      "train_tokens_per_second": 451.414
    },
    {
      "epoch": 0.1256544502617801,
      "grad_norm": 3.037754535675049,
      "learning_rate": 1.8767914438502674e-05,
      "loss": 0.15381601452827454,
      "num_input_tokens_seen": 304350,
      "step": 294,
      "train_runtime": 674.2031,
      "train_tokens_per_second": 451.422
    },
    {
      "epoch": 0.1260818463511059,
      "grad_norm": 2.938688278198242,
      "learning_rate": 1.8763636363636367e-05,
      "loss": 0.15069277584552765,
      "num_input_tokens_seen": 305476,
      "step": 295,
      "train_runtime": 676.2531,
      "train_tokens_per_second": 451.718
    },
    {
      "epoch": 0.12650924244043166,
      "grad_norm": 3.0169084072113037,
      "learning_rate": 1.8759358288770056e-05,
      "loss": 0.14942556619644165,
      "num_input_tokens_seen": 306308,
      "step": 296,
      "train_runtime": 678.2757,
      "train_tokens_per_second": 451.598
    },
    {
      "epoch": 0.12693663852975745,
      "grad_norm": 2.7263107299804688,
      "learning_rate": 1.8755080213903746e-05,
      "loss": 0.09780525416135788,
      "num_input_tokens_seen": 307260,
      "step": 297,
      "train_runtime": 680.3243,
      "train_tokens_per_second": 451.638
    },
    {
      "epoch": 0.12736403461908324,
      "grad_norm": 4.067149639129639,
      "learning_rate": 1.8750802139037435e-05,
      "loss": 0.17667019367218018,
      "num_input_tokens_seen": 307968,
      "step": 298,
      "train_runtime": 682.2915,
      "train_tokens_per_second": 451.373
    },
    {
      "epoch": 0.127791430708409,
      "grad_norm": 2.900526523590088,
      "learning_rate": 1.8746524064171124e-05,
      "loss": 0.10331056267023087,
      "num_input_tokens_seen": 309040,
      "step": 299,
      "train_runtime": 684.3739,
      "train_tokens_per_second": 451.566
    },
    {
      "epoch": 0.1282188267977348,
      "grad_norm": 3.4931366443634033,
      "learning_rate": 1.8742245989304814e-05,
      "loss": 0.1742061972618103,
      "num_input_tokens_seen": 310064,
      "step": 300,
      "train_runtime": 686.432,
      "train_tokens_per_second": 451.704
    },
    {
      "epoch": 0.1286462228870606,
      "grad_norm": 3.780637741088867,
      "learning_rate": 1.8737967914438503e-05,
      "loss": 0.17963796854019165,
      "num_input_tokens_seen": 311010,
      "step": 301,
      "train_runtime": 695.3613,
      "train_tokens_per_second": 447.264
    },
    {
      "epoch": 0.12907361897638636,
      "grad_norm": 2.2857017517089844,
      "learning_rate": 1.8733689839572193e-05,
      "loss": 0.10262749344110489,
      "num_input_tokens_seen": 312498,
      "step": 302,
      "train_runtime": 697.5074,
      "train_tokens_per_second": 448.021
    },
    {
      "epoch": 0.12950101506571216,
      "grad_norm": 2.307969570159912,
      "learning_rate": 1.8729411764705886e-05,
      "loss": 0.12309729307889938,
      "num_input_tokens_seen": 313832,
      "step": 303,
      "train_runtime": 699.6924,
      "train_tokens_per_second": 448.529
    },
    {
      "epoch": 0.12992841115503792,
      "grad_norm": 4.083592414855957,
      "learning_rate": 1.8725133689839575e-05,
      "loss": 0.1447969675064087,
      "num_input_tokens_seen": 314562,
      "step": 304,
      "train_runtime": 701.6911,
      "train_tokens_per_second": 448.291
    },
    {
      "epoch": 0.13035580724436371,
      "grad_norm": 3.1060731410980225,
      "learning_rate": 1.8720855614973264e-05,
      "loss": 0.13127095997333527,
      "num_input_tokens_seen": 315366,
      "step": 305,
      "train_runtime": 703.7314,
      "train_tokens_per_second": 448.134
    },
    {
      "epoch": 0.1307832033336895,
      "grad_norm": 2.7246339321136475,
      "learning_rate": 1.8716577540106954e-05,
      "loss": 0.08170565962791443,
      "num_input_tokens_seen": 316344,
      "step": 306,
      "train_runtime": 705.8149,
      "train_tokens_per_second": 448.197
    },
    {
      "epoch": 0.13121059942301527,
      "grad_norm": 2.998915433883667,
      "learning_rate": 1.8712299465240643e-05,
      "loss": 0.12575949728488922,
      "num_input_tokens_seen": 317502,
      "step": 307,
      "train_runtime": 707.9242,
      "train_tokens_per_second": 448.497
    },
    {
      "epoch": 0.13163799551234107,
      "grad_norm": 3.3328864574432373,
      "learning_rate": 1.8708021390374333e-05,
      "loss": 0.1568697690963745,
      "num_input_tokens_seen": 318744,
      "step": 308,
      "train_runtime": 710.1045,
      "train_tokens_per_second": 448.869
    },
    {
      "epoch": 0.13206539160166683,
      "grad_norm": 3.177522897720337,
      "learning_rate": 1.8703743315508022e-05,
      "loss": 0.12127576023340225,
      "num_input_tokens_seen": 319662,
      "step": 309,
      "train_runtime": 712.1831,
      "train_tokens_per_second": 448.848
    },
    {
      "epoch": 0.13249278769099262,
      "grad_norm": 3.746555805206299,
      "learning_rate": 1.8699465240641712e-05,
      "loss": 0.20369161665439606,
      "num_input_tokens_seen": 320918,
      "step": 310,
      "train_runtime": 714.2756,
      "train_tokens_per_second": 449.292
    },
    {
      "epoch": 0.13292018378031842,
      "grad_norm": 3.511276960372925,
      "learning_rate": 1.8695187165775405e-05,
      "loss": 0.12065492570400238,
      "num_input_tokens_seen": 321682,
      "step": 311,
      "train_runtime": 716.2958,
      "train_tokens_per_second": 449.091
    },
    {
      "epoch": 0.13334757986964418,
      "grad_norm": 3.698040723800659,
      "learning_rate": 1.8690909090909094e-05,
      "loss": 0.1961260735988617,
      "num_input_tokens_seen": 322496,
      "step": 312,
      "train_runtime": 718.2899,
      "train_tokens_per_second": 448.978
    },
    {
      "epoch": 0.13377497595896998,
      "grad_norm": 3.9469563961029053,
      "learning_rate": 1.8686631016042783e-05,
      "loss": 0.201097771525383,
      "num_input_tokens_seen": 323340,
      "step": 313,
      "train_runtime": 720.3017,
      "train_tokens_per_second": 448.895
    },
    {
      "epoch": 0.13420237204829577,
      "grad_norm": 2.2532620429992676,
      "learning_rate": 1.8682352941176473e-05,
      "loss": 0.09061051160097122,
      "num_input_tokens_seen": 324558,
      "step": 314,
      "train_runtime": 722.4288,
      "train_tokens_per_second": 449.26
    },
    {
      "epoch": 0.13462976813762154,
      "grad_norm": 4.383615016937256,
      "learning_rate": 1.8678074866310162e-05,
      "loss": 0.18581202626228333,
      "num_input_tokens_seen": 325174,
      "step": 315,
      "train_runtime": 724.3555,
      "train_tokens_per_second": 448.915
    },
    {
      "epoch": 0.13505716422694733,
      "grad_norm": 7.402688026428223,
      "learning_rate": 1.8673796791443852e-05,
      "loss": 0.11511622369289398,
      "num_input_tokens_seen": 326062,
      "step": 316,
      "train_runtime": 726.3913,
      "train_tokens_per_second": 448.879
    },
    {
      "epoch": 0.1354845603162731,
      "grad_norm": 4.005338668823242,
      "learning_rate": 1.866951871657754e-05,
      "loss": 0.16818778216838837,
      "num_input_tokens_seen": 326946,
      "step": 317,
      "train_runtime": 728.4069,
      "train_tokens_per_second": 448.851
    },
    {
      "epoch": 0.1359119564055989,
      "grad_norm": 2.95188045501709,
      "learning_rate": 1.866524064171123e-05,
      "loss": 0.08211202919483185,
      "num_input_tokens_seen": 327712,
      "step": 318,
      "train_runtime": 730.3682,
      "train_tokens_per_second": 448.694
    },
    {
      "epoch": 0.13633935249492468,
      "grad_norm": 2.805638551712036,
      "learning_rate": 1.8660962566844923e-05,
      "loss": 0.2649306356906891,
      "num_input_tokens_seen": 329000,
      "step": 319,
      "train_runtime": 732.4667,
      "train_tokens_per_second": 449.167
    },
    {
      "epoch": 0.13676674858425045,
      "grad_norm": 3.8310837745666504,
      "learning_rate": 1.8656684491978613e-05,
      "loss": 0.13509626686573029,
      "num_input_tokens_seen": 329852,
      "step": 320,
      "train_runtime": 734.48,
      "train_tokens_per_second": 449.096
    },
    {
      "epoch": 0.13719414467357624,
      "grad_norm": 3.723883628845215,
      "learning_rate": 1.8652406417112302e-05,
      "loss": 0.08581741154193878,
      "num_input_tokens_seen": 330872,
      "step": 321,
      "train_runtime": 736.5671,
      "train_tokens_per_second": 449.208
    },
    {
      "epoch": 0.137621540762902,
      "grad_norm": 3.8647916316986084,
      "learning_rate": 1.8648128342245992e-05,
      "loss": 0.09364113956689835,
      "num_input_tokens_seen": 331688,
      "step": 322,
      "train_runtime": 738.5579,
      "train_tokens_per_second": 449.102
    },
    {
      "epoch": 0.1380489368522278,
      "grad_norm": 4.102155685424805,
      "learning_rate": 1.864385026737968e-05,
      "loss": 0.2538776993751526,
      "num_input_tokens_seen": 333338,
      "step": 323,
      "train_runtime": 740.7915,
      "train_tokens_per_second": 449.975
    },
    {
      "epoch": 0.1384763329415536,
      "grad_norm": 4.6270856857299805,
      "learning_rate": 1.863957219251337e-05,
      "loss": 0.2844879627227783,
      "num_input_tokens_seen": 334224,
      "step": 324,
      "train_runtime": 742.7862,
      "train_tokens_per_second": 449.96
    },
    {
      "epoch": 0.13890372903087936,
      "grad_norm": 2.7475757598876953,
      "learning_rate": 1.863529411764706e-05,
      "loss": 0.10463880002498627,
      "num_input_tokens_seen": 335028,
      "step": 325,
      "train_runtime": 744.7779,
      "train_tokens_per_second": 449.836
    },
    {
      "epoch": 0.13933112512020515,
      "grad_norm": 4.592890739440918,
      "learning_rate": 1.863101604278075e-05,
      "loss": 0.16257499158382416,
      "num_input_tokens_seen": 336004,
      "step": 326,
      "train_runtime": 746.7945,
      "train_tokens_per_second": 449.928
    },
    {
      "epoch": 0.13975852120953094,
      "grad_norm": 5.129707336425781,
      "learning_rate": 1.8626737967914442e-05,
      "loss": 0.13594549894332886,
      "num_input_tokens_seen": 336610,
      "step": 327,
      "train_runtime": 748.7692,
      "train_tokens_per_second": 449.551
    },
    {
      "epoch": 0.1401859172988567,
      "grad_norm": 3.1647987365722656,
      "learning_rate": 1.8622459893048132e-05,
      "loss": 0.08152987062931061,
      "num_input_tokens_seen": 337282,
      "step": 328,
      "train_runtime": 750.8125,
      "train_tokens_per_second": 449.223
    },
    {
      "epoch": 0.1406133133881825,
      "grad_norm": 2.9010794162750244,
      "learning_rate": 1.861818181818182e-05,
      "loss": 0.1412004679441452,
      "num_input_tokens_seen": 338422,
      "step": 329,
      "train_runtime": 752.8648,
      "train_tokens_per_second": 449.512
    },
    {
      "epoch": 0.14104070947750827,
      "grad_norm": 4.177113056182861,
      "learning_rate": 1.8613903743315507e-05,
      "loss": 0.19049745798110962,
      "num_input_tokens_seen": 339422,
      "step": 330,
      "train_runtime": 754.9166,
      "train_tokens_per_second": 449.615
    },
    {
      "epoch": 0.14146810556683406,
      "grad_norm": 2.712876081466675,
      "learning_rate": 1.86096256684492e-05,
      "loss": 0.1313464641571045,
      "num_input_tokens_seen": 340936,
      "step": 331,
      "train_runtime": 763.8042,
      "train_tokens_per_second": 446.366
    },
    {
      "epoch": 0.14189550165615986,
      "grad_norm": 2.798196315765381,
      "learning_rate": 1.860534759358289e-05,
      "loss": 0.11945220082998276,
      "num_input_tokens_seen": 341784,
      "step": 332,
      "train_runtime": 765.8969,
      "train_tokens_per_second": 446.253
    },
    {
      "epoch": 0.14232289774548562,
      "grad_norm": 3.371490955352783,
      "learning_rate": 1.860106951871658e-05,
      "loss": 0.11920512467622757,
      "num_input_tokens_seen": 343004,
      "step": 333,
      "train_runtime": 768.2008,
      "train_tokens_per_second": 446.503
    },
    {
      "epoch": 0.14275029383481141,
      "grad_norm": 2.9165289402008057,
      "learning_rate": 1.859679144385027e-05,
      "loss": 0.3008955419063568,
      "num_input_tokens_seen": 345688,
      "step": 334,
      "train_runtime": 770.7273,
      "train_tokens_per_second": 448.522
    },
    {
      "epoch": 0.14317768992413718,
      "grad_norm": 3.924025297164917,
      "learning_rate": 1.8592513368983958e-05,
      "loss": 0.07638221979141235,
      "num_input_tokens_seen": 347148,
      "step": 335,
      "train_runtime": 772.897,
      "train_tokens_per_second": 449.152
    },
    {
      "epoch": 0.14360508601346297,
      "grad_norm": 3.8307268619537354,
      "learning_rate": 1.8588235294117647e-05,
      "loss": 0.17071478068828583,
      "num_input_tokens_seen": 348164,
      "step": 336,
      "train_runtime": 774.927,
      "train_tokens_per_second": 449.286
    },
    {
      "epoch": 0.14403248210278877,
      "grad_norm": 2.630080223083496,
      "learning_rate": 1.8583957219251337e-05,
      "loss": 0.14733010530471802,
      "num_input_tokens_seen": 349356,
      "step": 337,
      "train_runtime": 777.0438,
      "train_tokens_per_second": 449.596
    },
    {
      "epoch": 0.14445987819211453,
      "grad_norm": 3.4713733196258545,
      "learning_rate": 1.8579679144385026e-05,
      "loss": 0.1895836889743805,
      "num_input_tokens_seen": 350252,
      "step": 338,
      "train_runtime": 779.1156,
      "train_tokens_per_second": 449.551
    },
    {
      "epoch": 0.14488727428144033,
      "grad_norm": 5.767294883728027,
      "learning_rate": 1.8575401069518716e-05,
      "loss": 0.21986524760723114,
      "num_input_tokens_seen": 350854,
      "step": 339,
      "train_runtime": 781.1106,
      "train_tokens_per_second": 449.173
    },
    {
      "epoch": 0.14531467037076612,
      "grad_norm": 2.541485071182251,
      "learning_rate": 1.857112299465241e-05,
      "loss": 0.09996365755796432,
      "num_input_tokens_seen": 351776,
      "step": 340,
      "train_runtime": 783.1609,
      "train_tokens_per_second": 449.175
    },
    {
      "epoch": 0.14574206646009188,
      "grad_norm": 3.624178886413574,
      "learning_rate": 1.8566844919786098e-05,
      "loss": 0.23689131438732147,
      "num_input_tokens_seen": 352826,
      "step": 341,
      "train_runtime": 785.25,
      "train_tokens_per_second": 449.317
    },
    {
      "epoch": 0.14616946254941768,
      "grad_norm": 3.0588555335998535,
      "learning_rate": 1.8562566844919788e-05,
      "loss": 0.17323711514472961,
      "num_input_tokens_seen": 353882,
      "step": 342,
      "train_runtime": 787.3151,
      "train_tokens_per_second": 449.48
    },
    {
      "epoch": 0.14659685863874344,
      "grad_norm": 3.714693546295166,
      "learning_rate": 1.8558288770053477e-05,
      "loss": 0.1840294450521469,
      "num_input_tokens_seen": 354970,
      "step": 343,
      "train_runtime": 789.4612,
      "train_tokens_per_second": 449.636
    },
    {
      "epoch": 0.14702425472806924,
      "grad_norm": 3.6947803497314453,
      "learning_rate": 1.8554010695187166e-05,
      "loss": 0.1884085237979889,
      "num_input_tokens_seen": 356082,
      "step": 344,
      "train_runtime": 791.5332,
      "train_tokens_per_second": 449.864
    },
    {
      "epoch": 0.14745165081739503,
      "grad_norm": 4.252649784088135,
      "learning_rate": 1.8549732620320856e-05,
      "loss": 0.13772591948509216,
      "num_input_tokens_seen": 357028,
      "step": 345,
      "train_runtime": 793.5993,
      "train_tokens_per_second": 449.884
    },
    {
      "epoch": 0.1478790469067208,
      "grad_norm": 3.3578357696533203,
      "learning_rate": 1.8545454545454545e-05,
      "loss": 0.2070823609828949,
      "num_input_tokens_seen": 358174,
      "step": 346,
      "train_runtime": 795.6966,
      "train_tokens_per_second": 450.139
    },
    {
      "epoch": 0.1483064429960466,
      "grad_norm": 4.760191917419434,
      "learning_rate": 1.8541176470588235e-05,
      "loss": 0.14720162749290466,
      "num_input_tokens_seen": 359022,
      "step": 347,
      "train_runtime": 797.7208,
      "train_tokens_per_second": 450.06
    },
    {
      "epoch": 0.14873383908537238,
      "grad_norm": 4.793249130249023,
      "learning_rate": 1.8536898395721928e-05,
      "loss": 0.17956335842609406,
      "num_input_tokens_seen": 359836,
      "step": 348,
      "train_runtime": 799.7329,
      "train_tokens_per_second": 449.945
    },
    {
      "epoch": 0.14916123517469815,
      "grad_norm": 3.8941121101379395,
      "learning_rate": 1.8532620320855617e-05,
      "loss": 0.16894230246543884,
      "num_input_tokens_seen": 360908,
      "step": 349,
      "train_runtime": 801.8407,
      "train_tokens_per_second": 450.099
    },
    {
      "epoch": 0.14958863126402394,
      "grad_norm": 2.622894525527954,
      "learning_rate": 1.8528342245989306e-05,
      "loss": 0.1034291535615921,
      "num_input_tokens_seen": 361784,
      "step": 350,
      "train_runtime": 803.9208,
      "train_tokens_per_second": 450.024
    },
    {
      "epoch": 0.1500160273533497,
      "grad_norm": 3.6771161556243896,
      "learning_rate": 1.8524064171122996e-05,
      "loss": 0.1609831303358078,
      "num_input_tokens_seen": 362754,
      "step": 351,
      "train_runtime": 805.9475,
      "train_tokens_per_second": 450.096
    },
    {
      "epoch": 0.1504434234426755,
      "grad_norm": 3.5313310623168945,
      "learning_rate": 1.8519786096256685e-05,
      "loss": 0.18430784344673157,
      "num_input_tokens_seen": 363890,
      "step": 352,
      "train_runtime": 808.0758,
      "train_tokens_per_second": 450.317
    },
    {
      "epoch": 0.1508708195320013,
      "grad_norm": 2.998183488845825,
      "learning_rate": 1.8515508021390375e-05,
      "loss": 0.13345034420490265,
      "num_input_tokens_seen": 365082,
      "step": 353,
      "train_runtime": 810.2033,
      "train_tokens_per_second": 450.605
    },
    {
      "epoch": 0.15129821562132706,
      "grad_norm": 3.4583919048309326,
      "learning_rate": 1.8511229946524064e-05,
      "loss": 0.1690501719713211,
      "num_input_tokens_seen": 365994,
      "step": 354,
      "train_runtime": 812.2927,
      "train_tokens_per_second": 450.569
    },
    {
      "epoch": 0.15172561171065285,
      "grad_norm": 3.4179177284240723,
      "learning_rate": 1.8506951871657754e-05,
      "loss": 0.15503522753715515,
      "num_input_tokens_seen": 366826,
      "step": 355,
      "train_runtime": 814.3792,
      "train_tokens_per_second": 450.436
    },
    {
      "epoch": 0.15215300779997862,
      "grad_norm": 4.092080116271973,
      "learning_rate": 1.8502673796791447e-05,
      "loss": 0.16115663945674896,
      "num_input_tokens_seen": 367598,
      "step": 356,
      "train_runtime": 816.3637,
      "train_tokens_per_second": 450.287
    },
    {
      "epoch": 0.1525804038893044,
      "grad_norm": 2.182504892349243,
      "learning_rate": 1.8498395721925136e-05,
      "loss": 0.08430016785860062,
      "num_input_tokens_seen": 369124,
      "step": 357,
      "train_runtime": 818.5705,
      "train_tokens_per_second": 450.937
    },
    {
      "epoch": 0.1530077999786302,
      "grad_norm": 3.81746244430542,
      "learning_rate": 1.8494117647058825e-05,
      "loss": 0.15793293714523315,
      "num_input_tokens_seen": 369900,
      "step": 358,
      "train_runtime": 820.5765,
      "train_tokens_per_second": 450.781
    },
    {
      "epoch": 0.15343519606795597,
      "grad_norm": 2.8908159732818604,
      "learning_rate": 1.8489839572192515e-05,
      "loss": 0.11062908172607422,
      "num_input_tokens_seen": 371024,
      "step": 359,
      "train_runtime": 822.6396,
      "train_tokens_per_second": 451.016
    },
    {
      "epoch": 0.15386259215728176,
      "grad_norm": 3.6751654148101807,
      "learning_rate": 1.8485561497326204e-05,
      "loss": 0.09459831565618515,
      "num_input_tokens_seen": 372068,
      "step": 360,
      "train_runtime": 824.72,
      "train_tokens_per_second": 451.145
    },
    {
      "epoch": 0.15428998824660756,
      "grad_norm": 2.9600813388824463,
      "learning_rate": 1.8481283422459894e-05,
      "loss": 0.11612650752067566,
      "num_input_tokens_seen": 372806,
      "step": 361,
      "train_runtime": 833.6043,
      "train_tokens_per_second": 447.222
    },
    {
      "epoch": 0.15471738433593332,
      "grad_norm": 3.6384379863739014,
      "learning_rate": 1.8477005347593583e-05,
      "loss": 0.16494277119636536,
      "num_input_tokens_seen": 373880,
      "step": 362,
      "train_runtime": 835.743,
      "train_tokens_per_second": 447.362
    },
    {
      "epoch": 0.15514478042525912,
      "grad_norm": 2.7391035556793213,
      "learning_rate": 1.8472727272727273e-05,
      "loss": 0.1711333692073822,
      "num_input_tokens_seen": 375686,
      "step": 363,
      "train_runtime": 838.0615,
      "train_tokens_per_second": 448.28
    },
    {
      "epoch": 0.15557217651458488,
      "grad_norm": 2.278661012649536,
      "learning_rate": 1.8468449197860965e-05,
      "loss": 0.12076908349990845,
      "num_input_tokens_seen": 377046,
      "step": 364,
      "train_runtime": 840.2695,
      "train_tokens_per_second": 448.72
    },
    {
      "epoch": 0.15599957260391067,
      "grad_norm": 4.426092624664307,
      "learning_rate": 1.8464171122994655e-05,
      "loss": 0.2757834792137146,
      "num_input_tokens_seen": 377830,
      "step": 365,
      "train_runtime": 842.2549,
      "train_tokens_per_second": 448.593
    },
    {
      "epoch": 0.15642696869323647,
      "grad_norm": 3.152621269226074,
      "learning_rate": 1.8459893048128344e-05,
      "loss": 0.16674590110778809,
      "num_input_tokens_seen": 379008,
      "step": 366,
      "train_runtime": 844.3522,
      "train_tokens_per_second": 448.874
    },
    {
      "epoch": 0.15685436478256223,
      "grad_norm": 3.77046275138855,
      "learning_rate": 1.8455614973262034e-05,
      "loss": 0.2336619347333908,
      "num_input_tokens_seen": 379960,
      "step": 367,
      "train_runtime": 846.3598,
      "train_tokens_per_second": 448.934
    },
    {
      "epoch": 0.15728176087188803,
      "grad_norm": 3.0744481086730957,
      "learning_rate": 1.8451336898395723e-05,
      "loss": 0.16401953995227814,
      "num_input_tokens_seen": 381224,
      "step": 368,
      "train_runtime": 848.4986,
      "train_tokens_per_second": 449.292
    },
    {
      "epoch": 0.1577091569612138,
      "grad_norm": 2.2774553298950195,
      "learning_rate": 1.8447058823529413e-05,
      "loss": 0.080047607421875,
      "num_input_tokens_seen": 381954,
      "step": 369,
      "train_runtime": 850.4565,
      "train_tokens_per_second": 449.116
    },
    {
      "epoch": 0.15813655305053959,
      "grad_norm": 3.4942691326141357,
      "learning_rate": 1.8442780748663102e-05,
      "loss": 0.16062352061271667,
      "num_input_tokens_seen": 382930,
      "step": 370,
      "train_runtime": 852.5438,
      "train_tokens_per_second": 449.162
    },
    {
      "epoch": 0.15856394913986538,
      "grad_norm": 3.2642838954925537,
      "learning_rate": 1.843850267379679e-05,
      "loss": 0.17898786067962646,
      "num_input_tokens_seen": 384038,
      "step": 371,
      "train_runtime": 854.6314,
      "train_tokens_per_second": 449.361
    },
    {
      "epoch": 0.15899134522919114,
      "grad_norm": 3.8704471588134766,
      "learning_rate": 1.8434224598930484e-05,
      "loss": 0.24493300914764404,
      "num_input_tokens_seen": 385318,
      "step": 372,
      "train_runtime": 856.7494,
      "train_tokens_per_second": 449.744
    },
    {
      "epoch": 0.15941874131851694,
      "grad_norm": 3.457505941390991,
      "learning_rate": 1.8429946524064174e-05,
      "loss": 0.1411478966474533,
      "num_input_tokens_seen": 386132,
      "step": 373,
      "train_runtime": 858.8073,
      "train_tokens_per_second": 449.614
    },
    {
      "epoch": 0.15984613740784273,
      "grad_norm": 3.7425549030303955,
      "learning_rate": 1.8425668449197863e-05,
      "loss": 0.125191792845726,
      "num_input_tokens_seen": 386856,
      "step": 374,
      "train_runtime": 860.8107,
      "train_tokens_per_second": 449.409
    },
    {
      "epoch": 0.1602735334971685,
      "grad_norm": 3.255617618560791,
      "learning_rate": 1.8421390374331553e-05,
      "loss": 0.17602336406707764,
      "num_input_tokens_seen": 388868,
      "step": 375,
      "train_runtime": 863.2429,
      "train_tokens_per_second": 450.473
    },
    {
      "epoch": 0.1607009295864943,
      "grad_norm": 3.0290393829345703,
      "learning_rate": 1.8417112299465242e-05,
      "loss": 0.12505877017974854,
      "num_input_tokens_seen": 389734,
      "step": 376,
      "train_runtime": 865.3651,
      "train_tokens_per_second": 450.369
    },
    {
      "epoch": 0.16112832567582006,
      "grad_norm": 2.4295220375061035,
      "learning_rate": 1.841283422459893e-05,
      "loss": 0.14126133918762207,
      "num_input_tokens_seen": 391344,
      "step": 377,
      "train_runtime": 867.5522,
      "train_tokens_per_second": 451.09
    },
    {
      "epoch": 0.16155572176514585,
      "grad_norm": 4.606272220611572,
      "learning_rate": 1.840855614973262e-05,
      "loss": 0.13228780031204224,
      "num_input_tokens_seen": 392618,
      "step": 378,
      "train_runtime": 869.6981,
      "train_tokens_per_second": 451.442
    },
    {
      "epoch": 0.16198311785447164,
      "grad_norm": 3.0182430744171143,
      "learning_rate": 1.840427807486631e-05,
      "loss": 0.12397598475217819,
      "num_input_tokens_seen": 393650,
      "step": 379,
      "train_runtime": 871.7806,
      "train_tokens_per_second": 451.547
    },
    {
      "epoch": 0.1624105139437974,
      "grad_norm": 3.784773349761963,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.1671362966299057,
      "num_input_tokens_seen": 394766,
      "step": 380,
      "train_runtime": 873.8605,
      "train_tokens_per_second": 451.749
    },
    {
      "epoch": 0.1628379100331232,
      "grad_norm": 3.4677631855010986,
      "learning_rate": 1.8395721925133693e-05,
      "loss": 0.1889929324388504,
      "num_input_tokens_seen": 396408,
      "step": 381,
      "train_runtime": 876.0321,
      "train_tokens_per_second": 452.504
    },
    {
      "epoch": 0.16326530612244897,
      "grad_norm": 4.223570346832275,
      "learning_rate": 1.8391443850267382e-05,
      "loss": 0.279923677444458,
      "num_input_tokens_seen": 397408,
      "step": 382,
      "train_runtime": 878.089,
      "train_tokens_per_second": 452.583
    },
    {
      "epoch": 0.16369270221177476,
      "grad_norm": 2.817697763442993,
      "learning_rate": 1.8387165775401072e-05,
      "loss": 0.1628781259059906,
      "num_input_tokens_seen": 398320,
      "step": 383,
      "train_runtime": 880.123,
      "train_tokens_per_second": 452.573
    },
    {
      "epoch": 0.16412009830110055,
      "grad_norm": 3.717114210128784,
      "learning_rate": 1.838288770053476e-05,
      "loss": 0.22462254762649536,
      "num_input_tokens_seen": 399114,
      "step": 384,
      "train_runtime": 882.0883,
      "train_tokens_per_second": 452.465
    },
    {
      "epoch": 0.16454749439042632,
      "grad_norm": 3.7585573196411133,
      "learning_rate": 1.837860962566845e-05,
      "loss": 0.20147722959518433,
      "num_input_tokens_seen": 400164,
      "step": 385,
      "train_runtime": 884.1495,
      "train_tokens_per_second": 452.598
    },
    {
      "epoch": 0.1649748904797521,
      "grad_norm": 3.3615450859069824,
      "learning_rate": 1.837433155080214e-05,
      "loss": 0.1268632709980011,
      "num_input_tokens_seen": 401158,
      "step": 386,
      "train_runtime": 886.1683,
      "train_tokens_per_second": 452.688
    },
    {
      "epoch": 0.1654022865690779,
      "grad_norm": 2.9951412677764893,
      "learning_rate": 1.837005347593583e-05,
      "loss": 0.16406765580177307,
      "num_input_tokens_seen": 402288,
      "step": 387,
      "train_runtime": 888.2214,
      "train_tokens_per_second": 452.914
    },
    {
      "epoch": 0.16582968265840367,
      "grad_norm": 3.7813777923583984,
      "learning_rate": 1.8365775401069522e-05,
      "loss": 0.21073871850967407,
      "num_input_tokens_seen": 403888,
      "step": 388,
      "train_runtime": 890.4253,
      "train_tokens_per_second": 453.59
    },
    {
      "epoch": 0.16625707874772946,
      "grad_norm": 2.657719612121582,
      "learning_rate": 1.8361497326203212e-05,
      "loss": 0.10052626579999924,
      "num_input_tokens_seen": 404850,
      "step": 389,
      "train_runtime": 892.4732,
      "train_tokens_per_second": 453.627
    },
    {
      "epoch": 0.16668447483705523,
      "grad_norm": 3.0095977783203125,
      "learning_rate": 1.83572192513369e-05,
      "loss": 0.13588152825832367,
      "num_input_tokens_seen": 405874,
      "step": 390,
      "train_runtime": 894.5257,
      "train_tokens_per_second": 453.731
    },
    {
      "epoch": 0.16711187092638102,
      "grad_norm": 2.6088316440582275,
      "learning_rate": 1.8352941176470587e-05,
      "loss": 0.13496223092079163,
      "num_input_tokens_seen": 407232,
      "step": 391,
      "train_runtime": 903.2552,
      "train_tokens_per_second": 450.849
    },
    {
      "epoch": 0.16753926701570682,
      "grad_norm": 3.707267999649048,
      "learning_rate": 1.834866310160428e-05,
      "loss": 0.22445905208587646,
      "num_input_tokens_seen": 408164,
      "step": 392,
      "train_runtime": 905.2755,
      "train_tokens_per_second": 450.873
    },
    {
      "epoch": 0.16796666310503258,
      "grad_norm": 3.083017349243164,
      "learning_rate": 1.834438502673797e-05,
      "loss": 0.14702178537845612,
      "num_input_tokens_seen": 409462,
      "step": 393,
      "train_runtime": 907.4003,
      "train_tokens_per_second": 451.247
    },
    {
      "epoch": 0.16839405919435838,
      "grad_norm": 3.6747193336486816,
      "learning_rate": 1.834010695187166e-05,
      "loss": 0.13536429405212402,
      "num_input_tokens_seen": 410536,
      "step": 394,
      "train_runtime": 909.4546,
      "train_tokens_per_second": 451.409
    },
    {
      "epoch": 0.16882145528368414,
      "grad_norm": 3.5335593223571777,
      "learning_rate": 1.833582887700535e-05,
      "loss": 0.08958180993795395,
      "num_input_tokens_seen": 411574,
      "step": 395,
      "train_runtime": 911.5201,
      "train_tokens_per_second": 451.525
    },
    {
      "epoch": 0.16924885137300993,
      "grad_norm": 4.21598482131958,
      "learning_rate": 1.833155080213904e-05,
      "loss": 0.12426377832889557,
      "num_input_tokens_seen": 412522,
      "step": 396,
      "train_runtime": 913.5858,
      "train_tokens_per_second": 451.542
    },
    {
      "epoch": 0.16967624746233573,
      "grad_norm": 3.440768241882324,
      "learning_rate": 1.832727272727273e-05,
      "loss": 0.10499773174524307,
      "num_input_tokens_seen": 413270,
      "step": 397,
      "train_runtime": 915.595,
      "train_tokens_per_second": 451.368
    },
    {
      "epoch": 0.1701036435516615,
      "grad_norm": 3.2018141746520996,
      "learning_rate": 1.8322994652406417e-05,
      "loss": 0.15937253832817078,
      "num_input_tokens_seen": 414482,
      "step": 398,
      "train_runtime": 917.826,
      "train_tokens_per_second": 451.591
    },
    {
      "epoch": 0.1705310396409873,
      "grad_norm": 3.3081984519958496,
      "learning_rate": 1.8318716577540106e-05,
      "loss": 0.16216382384300232,
      "num_input_tokens_seen": 415424,
      "step": 399,
      "train_runtime": 919.8549,
      "train_tokens_per_second": 451.619
    },
    {
      "epoch": 0.17095843573031308,
      "grad_norm": 3.623704671859741,
      "learning_rate": 1.83144385026738e-05,
      "loss": 0.11215674877166748,
      "num_input_tokens_seen": 416154,
      "step": 400,
      "train_runtime": 921.8503,
      "train_tokens_per_second": 451.433
    },
    {
      "epoch": 0.17138583181963885,
      "grad_norm": 3.782424211502075,
      "learning_rate": 1.831016042780749e-05,
      "loss": 0.1894763559103012,
      "num_input_tokens_seen": 417020,
      "step": 401,
      "train_runtime": 923.871,
      "train_tokens_per_second": 451.383
    },
    {
      "epoch": 0.17181322790896464,
      "grad_norm": 3.312699794769287,
      "learning_rate": 1.8305882352941178e-05,
      "loss": 0.14521507918834686,
      "num_input_tokens_seen": 418036,
      "step": 402,
      "train_runtime": 925.8911,
      "train_tokens_per_second": 451.496
    },
    {
      "epoch": 0.1722406239982904,
      "grad_norm": 2.5472071170806885,
      "learning_rate": 1.8301604278074867e-05,
      "loss": 0.12104392051696777,
      "num_input_tokens_seen": 419384,
      "step": 403,
      "train_runtime": 927.9956,
      "train_tokens_per_second": 451.925
    },
    {
      "epoch": 0.1726680200876162,
      "grad_norm": 3.5174648761749268,
      "learning_rate": 1.8297326203208557e-05,
      "loss": 0.18970666825771332,
      "num_input_tokens_seen": 420338,
      "step": 404,
      "train_runtime": 930.0683,
      "train_tokens_per_second": 451.943
    },
    {
      "epoch": 0.173095416176942,
      "grad_norm": 3.4436521530151367,
      "learning_rate": 1.8293048128342246e-05,
      "loss": 0.17532992362976074,
      "num_input_tokens_seen": 421290,
      "step": 405,
      "train_runtime": 932.0969,
      "train_tokens_per_second": 451.981
    },
    {
      "epoch": 0.17352281226626776,
      "grad_norm": 3.6591150760650635,
      "learning_rate": 1.8288770053475936e-05,
      "loss": 0.20689848065376282,
      "num_input_tokens_seen": 422422,
      "step": 406,
      "train_runtime": 934.1993,
      "train_tokens_per_second": 452.175
    },
    {
      "epoch": 0.17395020835559355,
      "grad_norm": 2.3392040729522705,
      "learning_rate": 1.8284491978609625e-05,
      "loss": 0.13393135368824005,
      "num_input_tokens_seen": 423902,
      "step": 407,
      "train_runtime": 936.3749,
      "train_tokens_per_second": 452.705
    },
    {
      "epoch": 0.17437760444491932,
      "grad_norm": 3.121617555618286,
      "learning_rate": 1.8280213903743318e-05,
      "loss": 0.16588927805423737,
      "num_input_tokens_seen": 425214,
      "step": 408,
      "train_runtime": 938.4945,
      "train_tokens_per_second": 453.081
    },
    {
      "epoch": 0.1748050005342451,
      "grad_norm": 2.4870142936706543,
      "learning_rate": 1.8275935828877007e-05,
      "loss": 0.07691413909196854,
      "num_input_tokens_seen": 426314,
      "step": 409,
      "train_runtime": 940.5254,
      "train_tokens_per_second": 453.272
    },
    {
      "epoch": 0.1752323966235709,
      "grad_norm": 2.869314193725586,
      "learning_rate": 1.8271657754010697e-05,
      "loss": 0.1472451090812683,
      "num_input_tokens_seen": 427306,
      "step": 410,
      "train_runtime": 942.5554,
      "train_tokens_per_second": 453.348
    },
    {
      "epoch": 0.17565979271289667,
      "grad_norm": 2.8579304218292236,
      "learning_rate": 1.8267379679144386e-05,
      "loss": 0.12922635674476624,
      "num_input_tokens_seen": 428258,
      "step": 411,
      "train_runtime": 944.6391,
      "train_tokens_per_second": 453.356
    },
    {
      "epoch": 0.17608718880222246,
      "grad_norm": 2.791553020477295,
      "learning_rate": 1.8263101604278076e-05,
      "loss": 0.14650341868400574,
      "num_input_tokens_seen": 429488,
      "step": 412,
      "train_runtime": 946.7275,
      "train_tokens_per_second": 453.655
    },
    {
      "epoch": 0.17651458489154825,
      "grad_norm": 3.428823709487915,
      "learning_rate": 1.8258823529411765e-05,
      "loss": 0.20975293219089508,
      "num_input_tokens_seen": 430388,
      "step": 413,
      "train_runtime": 948.8014,
      "train_tokens_per_second": 453.612
    },
    {
      "epoch": 0.17694198098087402,
      "grad_norm": 4.322518825531006,
      "learning_rate": 1.8254545454545455e-05,
      "loss": 0.286907434463501,
      "num_input_tokens_seen": 431344,
      "step": 414,
      "train_runtime": 950.8826,
      "train_tokens_per_second": 453.625
    },
    {
      "epoch": 0.1773693770701998,
      "grad_norm": 2.7757461071014404,
      "learning_rate": 1.8250267379679144e-05,
      "loss": 0.19005164504051208,
      "num_input_tokens_seen": 432614,
      "step": 415,
      "train_runtime": 953.0084,
      "train_tokens_per_second": 453.946
    },
    {
      "epoch": 0.17779677315952558,
      "grad_norm": 3.2586443424224854,
      "learning_rate": 1.8245989304812837e-05,
      "loss": 0.11516334861516953,
      "num_input_tokens_seen": 433394,
      "step": 416,
      "train_runtime": 955.0325,
      "train_tokens_per_second": 453.8
    },
    {
      "epoch": 0.17822416924885137,
      "grad_norm": 3.8034894466400146,
      "learning_rate": 1.8241711229946526e-05,
      "loss": 0.23013776540756226,
      "num_input_tokens_seen": 434194,
      "step": 417,
      "train_runtime": 957.0669,
      "train_tokens_per_second": 453.672
    },
    {
      "epoch": 0.17865156533817717,
      "grad_norm": 3.1263651847839355,
      "learning_rate": 1.8237433155080216e-05,
      "loss": 0.19704614579677582,
      "num_input_tokens_seen": 435574,
      "step": 418,
      "train_runtime": 959.2001,
      "train_tokens_per_second": 454.101
    },
    {
      "epoch": 0.17907896142750293,
      "grad_norm": 2.3848934173583984,
      "learning_rate": 1.8233155080213905e-05,
      "loss": 0.10682269930839539,
      "num_input_tokens_seen": 436530,
      "step": 419,
      "train_runtime": 961.2195,
      "train_tokens_per_second": 454.142
    },
    {
      "epoch": 0.17950635751682872,
      "grad_norm": 2.6706292629241943,
      "learning_rate": 1.8228877005347595e-05,
      "loss": 0.13585788011550903,
      "num_input_tokens_seen": 437646,
      "step": 420,
      "train_runtime": 963.3213,
      "train_tokens_per_second": 454.309
    },
    {
      "epoch": 0.17993375360615452,
      "grad_norm": 2.38997745513916,
      "learning_rate": 1.8224598930481284e-05,
      "loss": 0.11742988228797913,
      "num_input_tokens_seen": 439542,
      "step": 421,
      "train_runtime": 972.385,
      "train_tokens_per_second": 452.025
    },
    {
      "epoch": 0.18036114969548028,
      "grad_norm": 3.0404889583587646,
      "learning_rate": 1.8220320855614974e-05,
      "loss": 0.12024495005607605,
      "num_input_tokens_seen": 440376,
      "step": 422,
      "train_runtime": 974.517,
      "train_tokens_per_second": 451.892
    },
    {
      "epoch": 0.18078854578480608,
      "grad_norm": 3.102108955383301,
      "learning_rate": 1.8216042780748663e-05,
      "loss": 0.16088958084583282,
      "num_input_tokens_seen": 441150,
      "step": 423,
      "train_runtime": 976.5189,
      "train_tokens_per_second": 451.758
    },
    {
      "epoch": 0.18121594187413184,
      "grad_norm": 2.958465337753296,
      "learning_rate": 1.8211764705882356e-05,
      "loss": 0.11718589812517166,
      "num_input_tokens_seen": 441932,
      "step": 424,
      "train_runtime": 978.5149,
      "train_tokens_per_second": 451.635
    },
    {
      "epoch": 0.18164333796345764,
      "grad_norm": 2.6293013095855713,
      "learning_rate": 1.8207486631016045e-05,
      "loss": 0.09669187664985657,
      "num_input_tokens_seen": 442800,
      "step": 425,
      "train_runtime": 980.5422,
      "train_tokens_per_second": 451.587
    },
    {
      "epoch": 0.18207073405278343,
      "grad_norm": 3.878410816192627,
      "learning_rate": 1.8203208556149735e-05,
      "loss": 0.21885468065738678,
      "num_input_tokens_seen": 443690,
      "step": 426,
      "train_runtime": 982.5882,
      "train_tokens_per_second": 451.552
    },
    {
      "epoch": 0.1824981301421092,
      "grad_norm": 3.7683587074279785,
      "learning_rate": 1.8198930481283424e-05,
      "loss": 0.19215959310531616,
      "num_input_tokens_seen": 445006,
      "step": 427,
      "train_runtime": 984.7498,
      "train_tokens_per_second": 451.898
    },
    {
      "epoch": 0.182925526231435,
      "grad_norm": 4.942471027374268,
      "learning_rate": 1.8194652406417114e-05,
      "loss": 0.250957190990448,
      "num_input_tokens_seen": 445768,
      "step": 428,
      "train_runtime": 986.7483,
      "train_tokens_per_second": 451.754
    },
    {
      "epoch": 0.18335292232076075,
      "grad_norm": 3.3123409748077393,
      "learning_rate": 1.8190374331550803e-05,
      "loss": 0.23503629863262177,
      "num_input_tokens_seen": 446872,
      "step": 429,
      "train_runtime": 988.8509,
      "train_tokens_per_second": 451.91
    },
    {
      "epoch": 0.18378031841008655,
      "grad_norm": 3.3676066398620605,
      "learning_rate": 1.8186096256684493e-05,
      "loss": 0.20039096474647522,
      "num_input_tokens_seen": 448346,
      "step": 430,
      "train_runtime": 991.0611,
      "train_tokens_per_second": 452.39
    },
    {
      "epoch": 0.18420771449941234,
      "grad_norm": 4.725825309753418,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.22150349617004395,
      "num_input_tokens_seen": 449010,
      "step": 431,
      "train_runtime": 993.0153,
      "train_tokens_per_second": 452.168
    },
    {
      "epoch": 0.1846351105887381,
      "grad_norm": 3.901179790496826,
      "learning_rate": 1.8177540106951875e-05,
      "loss": 0.14798009395599365,
      "num_input_tokens_seen": 449648,
      "step": 432,
      "train_runtime": 994.9649,
      "train_tokens_per_second": 451.923
    },
    {
      "epoch": 0.1850625066780639,
      "grad_norm": 2.6945748329162598,
      "learning_rate": 1.8173262032085564e-05,
      "loss": 0.15738606452941895,
      "num_input_tokens_seen": 450730,
      "step": 433,
      "train_runtime": 997.1186,
      "train_tokens_per_second": 452.032
    },
    {
      "epoch": 0.1854899027673897,
      "grad_norm": 3.1521708965301514,
      "learning_rate": 1.8168983957219254e-05,
      "loss": 0.16670338809490204,
      "num_input_tokens_seen": 451750,
      "step": 434,
      "train_runtime": 999.2258,
      "train_tokens_per_second": 452.1
    },
    {
      "epoch": 0.18591729885671546,
      "grad_norm": 3.082484006881714,
      "learning_rate": 1.8164705882352943e-05,
      "loss": 0.13141794502735138,
      "num_input_tokens_seen": 452766,
      "step": 435,
      "train_runtime": 1001.2886,
      "train_tokens_per_second": 452.183
    },
    {
      "epoch": 0.18634469494604125,
      "grad_norm": 3.484811782836914,
      "learning_rate": 1.8160427807486633e-05,
      "loss": 0.11263776570558548,
      "num_input_tokens_seen": 453596,
      "step": 436,
      "train_runtime": 1003.2861,
      "train_tokens_per_second": 452.11
    },
    {
      "epoch": 0.18677209103536702,
      "grad_norm": 2.980639934539795,
      "learning_rate": 1.8156149732620322e-05,
      "loss": 0.19117361307144165,
      "num_input_tokens_seen": 454666,
      "step": 437,
      "train_runtime": 1005.3559,
      "train_tokens_per_second": 452.244
    },
    {
      "epoch": 0.1871994871246928,
      "grad_norm": 3.6414475440979004,
      "learning_rate": 1.815187165775401e-05,
      "loss": 0.16954579949378967,
      "num_input_tokens_seen": 455646,
      "step": 438,
      "train_runtime": 1007.4016,
      "train_tokens_per_second": 452.298
    },
    {
      "epoch": 0.1876268832140186,
      "grad_norm": 2.5613105297088623,
      "learning_rate": 1.81475935828877e-05,
      "loss": 0.1326548159122467,
      "num_input_tokens_seen": 457336,
      "step": 439,
      "train_runtime": 1009.6099,
      "train_tokens_per_second": 452.983
    },
    {
      "epoch": 0.18805427930334437,
      "grad_norm": 3.301633358001709,
      "learning_rate": 1.8143315508021394e-05,
      "loss": 0.173603817820549,
      "num_input_tokens_seen": 458736,
      "step": 440,
      "train_runtime": 1011.7589,
      "train_tokens_per_second": 453.404
    },
    {
      "epoch": 0.18848167539267016,
      "grad_norm": 2.6853365898132324,
      "learning_rate": 1.8139037433155083e-05,
      "loss": 0.12467221915721893,
      "num_input_tokens_seen": 459680,
      "step": 441,
      "train_runtime": 1013.8682,
      "train_tokens_per_second": 453.392
    },
    {
      "epoch": 0.18890907148199593,
      "grad_norm": 2.6115994453430176,
      "learning_rate": 1.8134759358288773e-05,
      "loss": 0.14297708868980408,
      "num_input_tokens_seen": 460690,
      "step": 442,
      "train_runtime": 1015.9842,
      "train_tokens_per_second": 453.442
    },
    {
      "epoch": 0.18933646757132172,
      "grad_norm": 1.9338431358337402,
      "learning_rate": 1.8130481283422462e-05,
      "loss": 0.05107812583446503,
      "num_input_tokens_seen": 461756,
      "step": 443,
      "train_runtime": 1018.1832,
      "train_tokens_per_second": 453.51
    },
    {
      "epoch": 0.1897638636606475,
      "grad_norm": 2.7567906379699707,
      "learning_rate": 1.812620320855615e-05,
      "loss": 0.105326347053051,
      "num_input_tokens_seen": 462770,
      "step": 444,
      "train_runtime": 1020.3097,
      "train_tokens_per_second": 453.558
    },
    {
      "epoch": 0.19019125974997328,
      "grad_norm": 2.0738582611083984,
      "learning_rate": 1.812192513368984e-05,
      "loss": 0.08425863832235336,
      "num_input_tokens_seen": 463976,
      "step": 445,
      "train_runtime": 1022.4078,
      "train_tokens_per_second": 453.807
    },
    {
      "epoch": 0.19061865583929907,
      "grad_norm": 3.048781156539917,
      "learning_rate": 1.811764705882353e-05,
      "loss": 0.15324826538562775,
      "num_input_tokens_seen": 464878,
      "step": 446,
      "train_runtime": 1024.4659,
      "train_tokens_per_second": 453.776
    },
    {
      "epoch": 0.19104605192862487,
      "grad_norm": 3.7193727493286133,
      "learning_rate": 1.811336898395722e-05,
      "loss": 0.16986773908138275,
      "num_input_tokens_seen": 465974,
      "step": 447,
      "train_runtime": 1026.5798,
      "train_tokens_per_second": 453.909
    },
    {
      "epoch": 0.19147344801795063,
      "grad_norm": 2.749384641647339,
      "learning_rate": 1.810909090909091e-05,
      "loss": 0.18088173866271973,
      "num_input_tokens_seen": 467186,
      "step": 448,
      "train_runtime": 1028.6981,
      "train_tokens_per_second": 454.153
    },
    {
      "epoch": 0.19190084410727642,
      "grad_norm": 2.837508201599121,
      "learning_rate": 1.8104812834224602e-05,
      "loss": 0.1461426317691803,
      "num_input_tokens_seen": 468308,
      "step": 449,
      "train_runtime": 1030.7883,
      "train_tokens_per_second": 454.32
    },
    {
      "epoch": 0.1923282401966022,
      "grad_norm": 2.9812381267547607,
      "learning_rate": 1.8100534759358292e-05,
      "loss": 0.13258974254131317,
      "num_input_tokens_seen": 469070,
      "step": 450,
      "train_runtime": 1032.8413,
      "train_tokens_per_second": 454.155
    },
    {
      "epoch": 0.19275563628592798,
      "grad_norm": 5.960050106048584,
      "learning_rate": 1.809625668449198e-05,
      "loss": 0.1815432608127594,
      "num_input_tokens_seen": 469884,
      "step": 451,
      "train_runtime": 1041.6011,
      "train_tokens_per_second": 451.117
    },
    {
      "epoch": 0.19318303237525378,
      "grad_norm": 3.958709716796875,
      "learning_rate": 1.8091978609625667e-05,
      "loss": 0.1256699413061142,
      "num_input_tokens_seen": 470606,
      "step": 452,
      "train_runtime": 1043.6271,
      "train_tokens_per_second": 450.933
    },
    {
      "epoch": 0.19361042846457954,
      "grad_norm": 3.3820741176605225,
      "learning_rate": 1.808770053475936e-05,
      "loss": 0.1685105860233307,
      "num_input_tokens_seen": 471720,
      "step": 453,
      "train_runtime": 1045.7944,
      "train_tokens_per_second": 451.064
    },
    {
      "epoch": 0.19403782455390534,
      "grad_norm": 4.036883354187012,
      "learning_rate": 1.808342245989305e-05,
      "loss": 0.16119134426116943,
      "num_input_tokens_seen": 472782,
      "step": 454,
      "train_runtime": 1047.8862,
      "train_tokens_per_second": 451.177
    },
    {
      "epoch": 0.1944652206432311,
      "grad_norm": 4.750235557556152,
      "learning_rate": 1.807914438502674e-05,
      "loss": 0.19967752695083618,
      "num_input_tokens_seen": 473422,
      "step": 455,
      "train_runtime": 1049.809,
      "train_tokens_per_second": 450.96
    },
    {
      "epoch": 0.1948926167325569,
      "grad_norm": 3.6832988262176514,
      "learning_rate": 1.807486631016043e-05,
      "loss": 0.16555368900299072,
      "num_input_tokens_seen": 474068,
      "step": 456,
      "train_runtime": 1051.8073,
      "train_tokens_per_second": 450.718
    },
    {
      "epoch": 0.1953200128218827,
      "grad_norm": 3.330173969268799,
      "learning_rate": 1.807058823529412e-05,
      "loss": 0.1114584356546402,
      "num_input_tokens_seen": 475158,
      "step": 457,
      "train_runtime": 1053.9072,
      "train_tokens_per_second": 450.854
    },
    {
      "epoch": 0.19574740891120845,
      "grad_norm": 3.2561466693878174,
      "learning_rate": 1.806631016042781e-05,
      "loss": 0.1865159571170807,
      "num_input_tokens_seen": 476750,
      "step": 458,
      "train_runtime": 1056.1027,
      "train_tokens_per_second": 451.424
    },
    {
      "epoch": 0.19617480500053425,
      "grad_norm": 3.319549083709717,
      "learning_rate": 1.8062032085561497e-05,
      "loss": 0.1691288948059082,
      "num_input_tokens_seen": 477686,
      "step": 459,
      "train_runtime": 1058.131,
      "train_tokens_per_second": 451.443
    },
    {
      "epoch": 0.19660220108986004,
      "grad_norm": 3.0110888481140137,
      "learning_rate": 1.8057754010695186e-05,
      "loss": 0.13728314638137817,
      "num_input_tokens_seen": 478550,
      "step": 460,
      "train_runtime": 1060.1336,
      "train_tokens_per_second": 451.405
    },
    {
      "epoch": 0.1970295971791858,
      "grad_norm": 3.8415775299072266,
      "learning_rate": 1.805347593582888e-05,
      "loss": 0.24058830738067627,
      "num_input_tokens_seen": 479448,
      "step": 461,
      "train_runtime": 1062.1543,
      "train_tokens_per_second": 451.392
    },
    {
      "epoch": 0.1974569932685116,
      "grad_norm": 3.1477935314178467,
      "learning_rate": 1.804919786096257e-05,
      "loss": 0.12435933202505112,
      "num_input_tokens_seen": 480304,
      "step": 462,
      "train_runtime": 1064.1871,
      "train_tokens_per_second": 451.334
    },
    {
      "epoch": 0.19788438935783736,
      "grad_norm": 2.164750337600708,
      "learning_rate": 1.8044919786096258e-05,
      "loss": 0.10824844241142273,
      "num_input_tokens_seen": 481772,
      "step": 463,
      "train_runtime": 1066.4032,
      "train_tokens_per_second": 451.773
    },
    {
      "epoch": 0.19831178544716316,
      "grad_norm": 4.262755393981934,
      "learning_rate": 1.8040641711229947e-05,
      "loss": 0.1483963429927826,
      "num_input_tokens_seen": 482828,
      "step": 464,
      "train_runtime": 1068.4999,
      "train_tokens_per_second": 451.875
    },
    {
      "epoch": 0.19873918153648895,
      "grad_norm": 3.3177640438079834,
      "learning_rate": 1.803636363636364e-05,
      "loss": 0.12290341407060623,
      "num_input_tokens_seen": 483550,
      "step": 465,
      "train_runtime": 1070.4703,
      "train_tokens_per_second": 451.717
    },
    {
      "epoch": 0.19916657762581472,
      "grad_norm": 4.042696952819824,
      "learning_rate": 1.8032085561497326e-05,
      "loss": 0.18432125449180603,
      "num_input_tokens_seen": 484462,
      "step": 466,
      "train_runtime": 1072.54,
      "train_tokens_per_second": 451.696
    },
    {
      "epoch": 0.1995939737151405,
      "grad_norm": 4.409545421600342,
      "learning_rate": 1.8027807486631016e-05,
      "loss": 0.18324634432792664,
      "num_input_tokens_seen": 485672,
      "step": 467,
      "train_runtime": 1074.6616,
      "train_tokens_per_second": 451.93
    },
    {
      "epoch": 0.20002136980446628,
      "grad_norm": 2.6798605918884277,
      "learning_rate": 1.8023529411764705e-05,
      "loss": 0.10559245944023132,
      "num_input_tokens_seen": 486932,
      "step": 468,
      "train_runtime": 1076.832,
      "train_tokens_per_second": 452.189
    },
    {
      "epoch": 0.20002136980446628,
      "eval_loss": 0.5899712443351746,
      "eval_runtime": 58.14,
      "eval_samples_per_second": 17.165,
      "eval_steps_per_second": 8.583,
      "num_input_tokens_seen": 486932,
      "step": 468
    },
    {
      "epoch": 0.20044876589379207,
      "grad_norm": 3.4593505859375,
      "learning_rate": 1.8019251336898398e-05,
      "loss": 0.2516295611858368,
      "num_input_tokens_seen": 487912,
      "step": 469,
      "train_runtime": 1137.0689,
      "train_tokens_per_second": 429.096
    },
    {
      "epoch": 0.20087616198311786,
      "grad_norm": 3.63491153717041,
      "learning_rate": 1.8014973262032087e-05,
      "loss": 0.13701091706752777,
      "num_input_tokens_seen": 488624,
      "step": 470,
      "train_runtime": 1139.046,
      "train_tokens_per_second": 428.977
    },
    {
      "epoch": 0.20130355807244363,
      "grad_norm": 3.00518798828125,
      "learning_rate": 1.8010695187165777e-05,
      "loss": 0.14881111681461334,
      "num_input_tokens_seen": 489490,
      "step": 471,
      "train_runtime": 1141.0921,
      "train_tokens_per_second": 428.966
    },
    {
      "epoch": 0.20173095416176942,
      "grad_norm": 3.599620819091797,
      "learning_rate": 1.8006417112299466e-05,
      "loss": 0.2012540102005005,
      "num_input_tokens_seen": 490284,
      "step": 472,
      "train_runtime": 1143.0646,
      "train_tokens_per_second": 428.921
    },
    {
      "epoch": 0.20215835025109521,
      "grad_norm": 3.7925641536712646,
      "learning_rate": 1.8002139037433156e-05,
      "loss": 0.13354450464248657,
      "num_input_tokens_seen": 491452,
      "step": 473,
      "train_runtime": 1145.1771,
      "train_tokens_per_second": 429.149
    },
    {
      "epoch": 0.20258574634042098,
      "grad_norm": 3.233987331390381,
      "learning_rate": 1.7997860962566845e-05,
      "loss": 0.13340525329113007,
      "num_input_tokens_seen": 492384,
      "step": 474,
      "train_runtime": 1147.2256,
      "train_tokens_per_second": 429.195
    },
    {
      "epoch": 0.20301314242974677,
      "grad_norm": 18.15591049194336,
      "learning_rate": 1.7993582887700535e-05,
      "loss": 0.16208212077617645,
      "num_input_tokens_seen": 493588,
      "step": 475,
      "train_runtime": 1149.3054,
      "train_tokens_per_second": 429.466
    },
    {
      "epoch": 0.20344053851907254,
      "grad_norm": 2.619872808456421,
      "learning_rate": 1.7989304812834224e-05,
      "loss": 0.14249897003173828,
      "num_input_tokens_seen": 495088,
      "step": 476,
      "train_runtime": 1151.4864,
      "train_tokens_per_second": 429.956
    },
    {
      "epoch": 0.20386793460839833,
      "grad_norm": 2.6133358478546143,
      "learning_rate": 1.7985026737967917e-05,
      "loss": 0.18429988622665405,
      "num_input_tokens_seen": 496728,
      "step": 477,
      "train_runtime": 1153.6992,
      "train_tokens_per_second": 430.552
    },
    {
      "epoch": 0.20429533069772413,
      "grad_norm": 2.314570426940918,
      "learning_rate": 1.7980748663101606e-05,
      "loss": 0.10642392188310623,
      "num_input_tokens_seen": 497664,
      "step": 478,
      "train_runtime": 1155.7282,
      "train_tokens_per_second": 430.606
    },
    {
      "epoch": 0.2047227267870499,
      "grad_norm": 2.7060182094573975,
      "learning_rate": 1.7976470588235296e-05,
      "loss": 0.13580118119716644,
      "num_input_tokens_seen": 498506,
      "step": 479,
      "train_runtime": 1157.7879,
      "train_tokens_per_second": 430.568
    },
    {
      "epoch": 0.20515012287637568,
      "grad_norm": 2.408872127532959,
      "learning_rate": 1.7972192513368985e-05,
      "loss": 0.09444678574800491,
      "num_input_tokens_seen": 499478,
      "step": 480,
      "train_runtime": 1159.8664,
      "train_tokens_per_second": 430.634
    },
    {
      "epoch": 0.20557751896570145,
      "grad_norm": 3.5546629428863525,
      "learning_rate": 1.7967914438502675e-05,
      "loss": 0.12948663532733917,
      "num_input_tokens_seen": 500350,
      "step": 481,
      "train_runtime": 1168.4237,
      "train_tokens_per_second": 428.227
    },
    {
      "epoch": 0.20600491505502724,
      "grad_norm": 3.448176145553589,
      "learning_rate": 1.7963636363636364e-05,
      "loss": 0.1861601024866104,
      "num_input_tokens_seen": 501142,
      "step": 482,
      "train_runtime": 1170.5777,
      "train_tokens_per_second": 428.115
    },
    {
      "epoch": 0.20643231114435304,
      "grad_norm": 3.6159188747406006,
      "learning_rate": 1.7959358288770054e-05,
      "loss": 0.1830386519432068,
      "num_input_tokens_seen": 502198,
      "step": 483,
      "train_runtime": 1172.7358,
      "train_tokens_per_second": 428.228
    },
    {
      "epoch": 0.2068597072336788,
      "grad_norm": 3.0487568378448486,
      "learning_rate": 1.7955080213903743e-05,
      "loss": 0.1396036595106125,
      "num_input_tokens_seen": 503676,
      "step": 484,
      "train_runtime": 1174.9427,
      "train_tokens_per_second": 428.681
    },
    {
      "epoch": 0.2072871033230046,
      "grad_norm": 4.191316604614258,
      "learning_rate": 1.7950802139037436e-05,
      "loss": 0.23336780071258545,
      "num_input_tokens_seen": 504508,
      "step": 485,
      "train_runtime": 1176.8983,
      "train_tokens_per_second": 428.676
    },
    {
      "epoch": 0.2077144994123304,
      "grad_norm": 2.661525011062622,
      "learning_rate": 1.7946524064171125e-05,
      "loss": 0.1197209283709526,
      "num_input_tokens_seen": 505448,
      "step": 486,
      "train_runtime": 1178.9343,
      "train_tokens_per_second": 428.733
    },
    {
      "epoch": 0.20814189550165615,
      "grad_norm": 5.443660259246826,
      "learning_rate": 1.7942245989304815e-05,
      "loss": 0.16842028498649597,
      "num_input_tokens_seen": 506284,
      "step": 487,
      "train_runtime": 1180.9598,
      "train_tokens_per_second": 428.706
    },
    {
      "epoch": 0.20856929159098195,
      "grad_norm": 3.885406732559204,
      "learning_rate": 1.7937967914438504e-05,
      "loss": 0.23998695611953735,
      "num_input_tokens_seen": 507260,
      "step": 488,
      "train_runtime": 1183.0058,
      "train_tokens_per_second": 428.789
    },
    {
      "epoch": 0.2089966876803077,
      "grad_norm": 3.6050217151641846,
      "learning_rate": 1.7933689839572194e-05,
      "loss": 0.10298469662666321,
      "num_input_tokens_seen": 508176,
      "step": 489,
      "train_runtime": 1185.0292,
      "train_tokens_per_second": 428.83
    },
    {
      "epoch": 0.2094240837696335,
      "grad_norm": 2.4668757915496826,
      "learning_rate": 1.7929411764705883e-05,
      "loss": 0.1151868999004364,
      "num_input_tokens_seen": 509374,
      "step": 490,
      "train_runtime": 1187.1425,
      "train_tokens_per_second": 429.076
    },
    {
      "epoch": 0.2098514798589593,
      "grad_norm": 3.3567590713500977,
      "learning_rate": 1.7925133689839573e-05,
      "loss": 0.18225990235805511,
      "num_input_tokens_seen": 510504,
      "step": 491,
      "train_runtime": 1189.2331,
      "train_tokens_per_second": 429.272
    },
    {
      "epoch": 0.21027887594828507,
      "grad_norm": 3.4371345043182373,
      "learning_rate": 1.7920855614973262e-05,
      "loss": 0.18509897589683533,
      "num_input_tokens_seen": 512370,
      "step": 492,
      "train_runtime": 1191.4364,
      "train_tokens_per_second": 430.044
    },
    {
      "epoch": 0.21070627203761086,
      "grad_norm": 3.046013832092285,
      "learning_rate": 1.7916577540106955e-05,
      "loss": 0.15275262296199799,
      "num_input_tokens_seen": 513418,
      "step": 493,
      "train_runtime": 1193.5018,
      "train_tokens_per_second": 430.178
    },
    {
      "epoch": 0.21113366812693665,
      "grad_norm": 3.2314887046813965,
      "learning_rate": 1.7912299465240644e-05,
      "loss": 0.09924692660570145,
      "num_input_tokens_seen": 514308,
      "step": 494,
      "train_runtime": 1195.5559,
      "train_tokens_per_second": 430.183
    },
    {
      "epoch": 0.21156106421626242,
      "grad_norm": 3.24064564704895,
      "learning_rate": 1.7908021390374334e-05,
      "loss": 0.08571077883243561,
      "num_input_tokens_seen": 515252,
      "step": 495,
      "train_runtime": 1197.6177,
      "train_tokens_per_second": 430.231
    },
    {
      "epoch": 0.2119884603055882,
      "grad_norm": 3.3843514919281006,
      "learning_rate": 1.7903743315508023e-05,
      "loss": 0.15649192035198212,
      "num_input_tokens_seen": 516116,
      "step": 496,
      "train_runtime": 1199.6125,
      "train_tokens_per_second": 430.236
    },
    {
      "epoch": 0.21241585639491398,
      "grad_norm": 3.2591617107391357,
      "learning_rate": 1.7899465240641713e-05,
      "loss": 0.24604164063930511,
      "num_input_tokens_seen": 517416,
      "step": 497,
      "train_runtime": 1201.7331,
      "train_tokens_per_second": 430.558
    },
    {
      "epoch": 0.21284325248423977,
      "grad_norm": 2.597635269165039,
      "learning_rate": 1.7895187165775402e-05,
      "loss": 0.11903823912143707,
      "num_input_tokens_seen": 518206,
      "step": 498,
      "train_runtime": 1203.7242,
      "train_tokens_per_second": 430.502
    },
    {
      "epoch": 0.21327064857356556,
      "grad_norm": 4.272196292877197,
      "learning_rate": 1.789090909090909e-05,
      "loss": 0.19234110414981842,
      "num_input_tokens_seen": 519112,
      "step": 499,
      "train_runtime": 1205.7309,
      "train_tokens_per_second": 430.537
    },
    {
      "epoch": 0.21369804466289133,
      "grad_norm": 4.057042121887207,
      "learning_rate": 1.788663101604278e-05,
      "loss": 0.16333362460136414,
      "num_input_tokens_seen": 520030,
      "step": 500,
      "train_runtime": 1207.7767,
      "train_tokens_per_second": 430.568
    },
    {
      "epoch": 0.21412544075221712,
      "grad_norm": 2.749988079071045,
      "learning_rate": 1.7882352941176474e-05,
      "loss": 0.15354587137699127,
      "num_input_tokens_seen": 521166,
      "step": 501,
      "train_runtime": 1209.9138,
      "train_tokens_per_second": 430.746
    },
    {
      "epoch": 0.2145528368415429,
      "grad_norm": 1.7050426006317139,
      "learning_rate": 1.7878074866310163e-05,
      "loss": 0.08256734907627106,
      "num_input_tokens_seen": 522844,
      "step": 502,
      "train_runtime": 1212.1243,
      "train_tokens_per_second": 431.345
    },
    {
      "epoch": 0.21498023293086868,
      "grad_norm": 3.860900402069092,
      "learning_rate": 1.7873796791443853e-05,
      "loss": 0.1697292923927307,
      "num_input_tokens_seen": 523784,
      "step": 503,
      "train_runtime": 1214.1465,
      "train_tokens_per_second": 431.401
    },
    {
      "epoch": 0.21540762902019447,
      "grad_norm": 3.3950910568237305,
      "learning_rate": 1.7869518716577542e-05,
      "loss": 0.19019439816474915,
      "num_input_tokens_seen": 524676,
      "step": 504,
      "train_runtime": 1216.2315,
      "train_tokens_per_second": 431.395
    },
    {
      "epoch": 0.21583502510952024,
      "grad_norm": 2.1655850410461426,
      "learning_rate": 1.786524064171123e-05,
      "loss": 0.10404959321022034,
      "num_input_tokens_seen": 525942,
      "step": 505,
      "train_runtime": 1218.3528,
      "train_tokens_per_second": 431.683
    },
    {
      "epoch": 0.21626242119884603,
      "grad_norm": 2.452092170715332,
      "learning_rate": 1.786096256684492e-05,
      "loss": 0.10681039094924927,
      "num_input_tokens_seen": 526914,
      "step": 506,
      "train_runtime": 1220.4178,
      "train_tokens_per_second": 431.749
    },
    {
      "epoch": 0.21668981728817183,
      "grad_norm": 3.558967113494873,
      "learning_rate": 1.785668449197861e-05,
      "loss": 0.1856197565793991,
      "num_input_tokens_seen": 527780,
      "step": 507,
      "train_runtime": 1222.4641,
      "train_tokens_per_second": 431.735
    },
    {
      "epoch": 0.2171172133774976,
      "grad_norm": 2.9209251403808594,
      "learning_rate": 1.78524064171123e-05,
      "loss": 0.11624643951654434,
      "num_input_tokens_seen": 529420,
      "step": 508,
      "train_runtime": 1224.6732,
      "train_tokens_per_second": 432.295
    },
    {
      "epoch": 0.21754460946682339,
      "grad_norm": 4.511392116546631,
      "learning_rate": 1.7848128342245993e-05,
      "loss": 0.19399908185005188,
      "num_input_tokens_seen": 530426,
      "step": 509,
      "train_runtime": 1226.7199,
      "train_tokens_per_second": 432.394
    },
    {
      "epoch": 0.21797200555614915,
      "grad_norm": 3.1867551803588867,
      "learning_rate": 1.7843850267379682e-05,
      "loss": 0.19393324851989746,
      "num_input_tokens_seen": 531432,
      "step": 510,
      "train_runtime": 1228.8197,
      "train_tokens_per_second": 432.474
    },
    {
      "epoch": 0.21839940164547494,
      "grad_norm": 3.1381099224090576,
      "learning_rate": 1.783957219251337e-05,
      "loss": 0.1334000676870346,
      "num_input_tokens_seen": 532300,
      "step": 511,
      "train_runtime": 1237.4063,
      "train_tokens_per_second": 430.174
    },
    {
      "epoch": 0.21882679773480074,
      "grad_norm": 2.9225268363952637,
      "learning_rate": 1.783529411764706e-05,
      "loss": 0.160324364900589,
      "num_input_tokens_seen": 533464,
      "step": 512,
      "train_runtime": 1239.5207,
      "train_tokens_per_second": 430.379
    },
    {
      "epoch": 0.2192541938241265,
      "grad_norm": 2.1712334156036377,
      "learning_rate": 1.783101604278075e-05,
      "loss": 0.14365139603614807,
      "num_input_tokens_seen": 534688,
      "step": 513,
      "train_runtime": 1241.8109,
      "train_tokens_per_second": 430.571
    },
    {
      "epoch": 0.2196815899134523,
      "grad_norm": 3.938077926635742,
      "learning_rate": 1.782673796791444e-05,
      "loss": 0.18016456067562103,
      "num_input_tokens_seen": 535718,
      "step": 514,
      "train_runtime": 1243.8776,
      "train_tokens_per_second": 430.684
    },
    {
      "epoch": 0.22010898600277806,
      "grad_norm": 3.6298506259918213,
      "learning_rate": 1.782245989304813e-05,
      "loss": 0.208083376288414,
      "num_input_tokens_seen": 536774,
      "step": 515,
      "train_runtime": 1245.9594,
      "train_tokens_per_second": 430.812
    },
    {
      "epoch": 0.22053638209210386,
      "grad_norm": 3.076723575592041,
      "learning_rate": 1.781818181818182e-05,
      "loss": 0.09675504267215729,
      "num_input_tokens_seen": 537520,
      "step": 516,
      "train_runtime": 1247.945,
      "train_tokens_per_second": 430.724
    },
    {
      "epoch": 0.22096377818142965,
      "grad_norm": 2.7116281986236572,
      "learning_rate": 1.781390374331551e-05,
      "loss": 0.14069856703281403,
      "num_input_tokens_seen": 538532,
      "step": 517,
      "train_runtime": 1250.0419,
      "train_tokens_per_second": 430.811
    },
    {
      "epoch": 0.22139117427075541,
      "grad_norm": 3.195251941680908,
      "learning_rate": 1.78096256684492e-05,
      "loss": 0.1874803602695465,
      "num_input_tokens_seen": 539472,
      "step": 518,
      "train_runtime": 1252.0806,
      "train_tokens_per_second": 430.86
    },
    {
      "epoch": 0.2218185703600812,
      "grad_norm": 2.814730644226074,
      "learning_rate": 1.780534759358289e-05,
      "loss": 0.15296101570129395,
      "num_input_tokens_seen": 540582,
      "step": 519,
      "train_runtime": 1254.1608,
      "train_tokens_per_second": 431.031
    },
    {
      "epoch": 0.222245966449407,
      "grad_norm": 3.212073802947998,
      "learning_rate": 1.7801069518716577e-05,
      "loss": 0.15851101279258728,
      "num_input_tokens_seen": 541854,
      "step": 520,
      "train_runtime": 1256.28,
      "train_tokens_per_second": 431.316
    },
    {
      "epoch": 0.22267336253873277,
      "grad_norm": 3.0334653854370117,
      "learning_rate": 1.779679144385027e-05,
      "loss": 0.15528877079486847,
      "num_input_tokens_seen": 543082,
      "step": 521,
      "train_runtime": 1258.3884,
      "train_tokens_per_second": 431.569
    },
    {
      "epoch": 0.22310075862805856,
      "grad_norm": 3.363664388656616,
      "learning_rate": 1.779251336898396e-05,
      "loss": 0.17860281467437744,
      "num_input_tokens_seen": 543988,
      "step": 522,
      "train_runtime": 1260.41,
      "train_tokens_per_second": 431.596
    },
    {
      "epoch": 0.22352815471738433,
      "grad_norm": 3.103485107421875,
      "learning_rate": 1.778823529411765e-05,
      "loss": 0.20006173849105835,
      "num_input_tokens_seen": 545036,
      "step": 523,
      "train_runtime": 1262.4745,
      "train_tokens_per_second": 431.72
    },
    {
      "epoch": 0.22395555080671012,
      "grad_norm": 3.38698673248291,
      "learning_rate": 1.7783957219251338e-05,
      "loss": 0.15314073860645294,
      "num_input_tokens_seen": 545972,
      "step": 524,
      "train_runtime": 1264.5436,
      "train_tokens_per_second": 431.754
    },
    {
      "epoch": 0.2243829468960359,
      "grad_norm": 2.5583505630493164,
      "learning_rate": 1.777967914438503e-05,
      "loss": 0.14785772562026978,
      "num_input_tokens_seen": 547502,
      "step": 525,
      "train_runtime": 1266.8216,
      "train_tokens_per_second": 432.186
    },
    {
      "epoch": 0.22481034298536168,
      "grad_norm": 4.153393745422363,
      "learning_rate": 1.777540106951872e-05,
      "loss": 0.15759462118148804,
      "num_input_tokens_seen": 548188,
      "step": 526,
      "train_runtime": 1268.8247,
      "train_tokens_per_second": 432.044
    },
    {
      "epoch": 0.22523773907468747,
      "grad_norm": 3.2607686519622803,
      "learning_rate": 1.7771122994652406e-05,
      "loss": 0.1772948056459427,
      "num_input_tokens_seen": 549974,
      "step": 527,
      "train_runtime": 1271.0825,
      "train_tokens_per_second": 432.682
    },
    {
      "epoch": 0.22566513516401324,
      "grad_norm": 5.678312301635742,
      "learning_rate": 1.7766844919786096e-05,
      "loss": 0.29141029715538025,
      "num_input_tokens_seen": 550696,
      "step": 528,
      "train_runtime": 1273.0966,
      "train_tokens_per_second": 432.564
    },
    {
      "epoch": 0.22609253125333903,
      "grad_norm": 3.297149181365967,
      "learning_rate": 1.776256684491979e-05,
      "loss": 0.17609791457653046,
      "num_input_tokens_seen": 551666,
      "step": 529,
      "train_runtime": 1275.1678,
      "train_tokens_per_second": 432.622
    },
    {
      "epoch": 0.22651992734266482,
      "grad_norm": 2.048459053039551,
      "learning_rate": 1.7758288770053478e-05,
      "loss": 0.07686586678028107,
      "num_input_tokens_seen": 552572,
      "step": 530,
      "train_runtime": 1277.1922,
      "train_tokens_per_second": 432.646
    },
    {
      "epoch": 0.2269473234319906,
      "grad_norm": 2.443927049636841,
      "learning_rate": 1.7754010695187167e-05,
      "loss": 0.17237553000450134,
      "num_input_tokens_seen": 554080,
      "step": 531,
      "train_runtime": 1279.3657,
      "train_tokens_per_second": 433.09
    },
    {
      "epoch": 0.22737471952131638,
      "grad_norm": 2.345522403717041,
      "learning_rate": 1.7749732620320857e-05,
      "loss": 0.15595164895057678,
      "num_input_tokens_seen": 555458,
      "step": 532,
      "train_runtime": 1281.6239,
      "train_tokens_per_second": 433.402
    },
    {
      "epoch": 0.22780211561064218,
      "grad_norm": 3.5701868534088135,
      "learning_rate": 1.774545454545455e-05,
      "loss": 0.1228233128786087,
      "num_input_tokens_seen": 556286,
      "step": 533,
      "train_runtime": 1283.6419,
      "train_tokens_per_second": 433.365
    },
    {
      "epoch": 0.22822951169996794,
      "grad_norm": 3.4031097888946533,
      "learning_rate": 1.7741176470588236e-05,
      "loss": 0.1412114053964615,
      "num_input_tokens_seen": 557110,
      "step": 534,
      "train_runtime": 1285.6336,
      "train_tokens_per_second": 433.335
    },
    {
      "epoch": 0.22865690778929373,
      "grad_norm": 2.7743172645568848,
      "learning_rate": 1.7736898395721925e-05,
      "loss": 0.17318317294120789,
      "num_input_tokens_seen": 558288,
      "step": 535,
      "train_runtime": 1287.7462,
      "train_tokens_per_second": 433.539
    },
    {
      "epoch": 0.2290843038786195,
      "grad_norm": 1.9242955446243286,
      "learning_rate": 1.7732620320855615e-05,
      "loss": 0.1027928814291954,
      "num_input_tokens_seen": 559406,
      "step": 536,
      "train_runtime": 1289.9349,
      "train_tokens_per_second": 433.67
    },
    {
      "epoch": 0.2295116999679453,
      "grad_norm": 2.708173990249634,
      "learning_rate": 1.7728342245989307e-05,
      "loss": 0.11816737055778503,
      "num_input_tokens_seen": 560272,
      "step": 537,
      "train_runtime": 1291.9337,
      "train_tokens_per_second": 433.669
    },
    {
      "epoch": 0.2299390960572711,
      "grad_norm": 3.260429620742798,
      "learning_rate": 1.7724064171122997e-05,
      "loss": 0.14136207103729248,
      "num_input_tokens_seen": 561538,
      "step": 538,
      "train_runtime": 1294.0379,
      "train_tokens_per_second": 433.942
    },
    {
      "epoch": 0.23036649214659685,
      "grad_norm": 4.084691524505615,
      "learning_rate": 1.7719786096256686e-05,
      "loss": 0.21478018164634705,
      "num_input_tokens_seen": 562386,
      "step": 539,
      "train_runtime": 1296.0728,
      "train_tokens_per_second": 433.915
    },
    {
      "epoch": 0.23079388823592265,
      "grad_norm": 4.207942008972168,
      "learning_rate": 1.7715508021390376e-05,
      "loss": 0.18661078810691833,
      "num_input_tokens_seen": 563408,
      "step": 540,
      "train_runtime": 1298.1743,
      "train_tokens_per_second": 434.0
    },
    {
      "epoch": 0.2312212843252484,
      "grad_norm": 3.6092209815979004,
      "learning_rate": 1.7711229946524065e-05,
      "loss": 0.17875447869300842,
      "num_input_tokens_seen": 564254,
      "step": 541,
      "train_runtime": 1307.0205,
      "train_tokens_per_second": 431.71
    },
    {
      "epoch": 0.2316486804145742,
      "grad_norm": 3.1494710445404053,
      "learning_rate": 1.7706951871657755e-05,
      "loss": 0.20335906744003296,
      "num_input_tokens_seen": 565498,
      "step": 542,
      "train_runtime": 1309.2062,
      "train_tokens_per_second": 431.94
    },
    {
      "epoch": 0.2320760765039,
      "grad_norm": 2.553554058074951,
      "learning_rate": 1.7702673796791444e-05,
      "loss": 0.11096488684415817,
      "num_input_tokens_seen": 566718,
      "step": 543,
      "train_runtime": 1311.3465,
      "train_tokens_per_second": 432.165
    },
    {
      "epoch": 0.23250347259322576,
      "grad_norm": 3.310425281524658,
      "learning_rate": 1.7698395721925133e-05,
      "loss": 0.17454029619693756,
      "num_input_tokens_seen": 567764,
      "step": 544,
      "train_runtime": 1313.4259,
      "train_tokens_per_second": 432.277
    },
    {
      "epoch": 0.23293086868255156,
      "grad_norm": 2.9846858978271484,
      "learning_rate": 1.7694117647058826e-05,
      "loss": 0.11867425590753555,
      "num_input_tokens_seen": 568674,
      "step": 545,
      "train_runtime": 1315.4473,
      "train_tokens_per_second": 432.305
    },
    {
      "epoch": 0.23335826477187735,
      "grad_norm": 3.943613290786743,
      "learning_rate": 1.7689839572192516e-05,
      "loss": 0.15576013922691345,
      "num_input_tokens_seen": 569620,
      "step": 546,
      "train_runtime": 1317.554,
      "train_tokens_per_second": 432.331
    },
    {
      "epoch": 0.23378566086120312,
      "grad_norm": 3.6025476455688477,
      "learning_rate": 1.7685561497326205e-05,
      "loss": 0.18531107902526855,
      "num_input_tokens_seen": 570572,
      "step": 547,
      "train_runtime": 1319.6007,
      "train_tokens_per_second": 432.382
    },
    {
      "epoch": 0.2342130569505289,
      "grad_norm": 4.696913719177246,
      "learning_rate": 1.7681283422459895e-05,
      "loss": 0.19636806845664978,
      "num_input_tokens_seen": 571354,
      "step": 548,
      "train_runtime": 1321.5975,
      "train_tokens_per_second": 432.321
    },
    {
      "epoch": 0.23464045303985467,
      "grad_norm": 3.6664090156555176,
      "learning_rate": 1.7677005347593584e-05,
      "loss": 0.15613016486167908,
      "num_input_tokens_seen": 572478,
      "step": 549,
      "train_runtime": 1323.6787,
      "train_tokens_per_second": 432.49
    },
    {
      "epoch": 0.23506784912918047,
      "grad_norm": 2.894364833831787,
      "learning_rate": 1.7672727272727274e-05,
      "loss": 0.09920540452003479,
      "num_input_tokens_seen": 573442,
      "step": 550,
      "train_runtime": 1325.7228,
      "train_tokens_per_second": 432.55
    },
    {
      "epoch": 0.23549524521850626,
      "grad_norm": 3.000265598297119,
      "learning_rate": 1.7668449197860963e-05,
      "loss": 0.14117231965065002,
      "num_input_tokens_seen": 574304,
      "step": 551,
      "train_runtime": 1327.7262,
      "train_tokens_per_second": 432.547
    },
    {
      "epoch": 0.23592264130783203,
      "grad_norm": 3.5130443572998047,
      "learning_rate": 1.7664171122994652e-05,
      "loss": 0.16135336458683014,
      "num_input_tokens_seen": 575224,
      "step": 552,
      "train_runtime": 1329.7215,
      "train_tokens_per_second": 432.59
    },
    {
      "epoch": 0.23635003739715782,
      "grad_norm": 3.135456085205078,
      "learning_rate": 1.7659893048128342e-05,
      "loss": 0.14004194736480713,
      "num_input_tokens_seen": 576216,
      "step": 553,
      "train_runtime": 1331.8927,
      "train_tokens_per_second": 432.629
    },
    {
      "epoch": 0.23677743348648359,
      "grad_norm": 2.7119812965393066,
      "learning_rate": 1.7655614973262035e-05,
      "loss": 0.1117730513215065,
      "num_input_tokens_seen": 577130,
      "step": 554,
      "train_runtime": 1334.0155,
      "train_tokens_per_second": 432.626
    },
    {
      "epoch": 0.23720482957580938,
      "grad_norm": 3.049938440322876,
      "learning_rate": 1.7651336898395724e-05,
      "loss": 0.13727469742298126,
      "num_input_tokens_seen": 578188,
      "step": 555,
      "train_runtime": 1336.0635,
      "train_tokens_per_second": 432.755
    },
    {
      "epoch": 0.23763222566513517,
      "grad_norm": 4.262728214263916,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.12640132009983063,
      "num_input_tokens_seen": 579398,
      "step": 556,
      "train_runtime": 1338.1345,
      "train_tokens_per_second": 432.989
    },
    {
      "epoch": 0.23805962175446094,
      "grad_norm": 4.481271743774414,
      "learning_rate": 1.7642780748663103e-05,
      "loss": 0.14547674357891083,
      "num_input_tokens_seen": 581002,
      "step": 557,
      "train_runtime": 1340.3065,
      "train_tokens_per_second": 433.484
    },
    {
      "epoch": 0.23848701784378673,
      "grad_norm": 3.1068921089172363,
      "learning_rate": 1.7638502673796792e-05,
      "loss": 0.18444842100143433,
      "num_input_tokens_seen": 582120,
      "step": 558,
      "train_runtime": 1342.3854,
      "train_tokens_per_second": 433.646
    },
    {
      "epoch": 0.23891441393311252,
      "grad_norm": 4.638261318206787,
      "learning_rate": 1.7634224598930482e-05,
      "loss": 0.18371248245239258,
      "num_input_tokens_seen": 582940,
      "step": 559,
      "train_runtime": 1344.4098,
      "train_tokens_per_second": 433.603
    },
    {
      "epoch": 0.2393418100224383,
      "grad_norm": 2.936429738998413,
      "learning_rate": 1.762994652406417e-05,
      "loss": 0.15735919773578644,
      "num_input_tokens_seen": 583904,
      "step": 560,
      "train_runtime": 1346.4826,
      "train_tokens_per_second": 433.651
    },
    {
      "epoch": 0.23976920611176408,
      "grad_norm": 2.95801043510437,
      "learning_rate": 1.762566844919786e-05,
      "loss": 0.1881670504808426,
      "num_input_tokens_seen": 585096,
      "step": 561,
      "train_runtime": 1348.558,
      "train_tokens_per_second": 433.868
    },
    {
      "epoch": 0.24019660220108985,
      "grad_norm": 4.456726551055908,
      "learning_rate": 1.7621390374331554e-05,
      "loss": 0.15737006068229675,
      "num_input_tokens_seen": 586010,
      "step": 562,
      "train_runtime": 1350.5337,
      "train_tokens_per_second": 433.91
    },
    {
      "epoch": 0.24062399829041564,
      "grad_norm": 2.182157278060913,
      "learning_rate": 1.7617112299465243e-05,
      "loss": 0.08489621430635452,
      "num_input_tokens_seen": 587124,
      "step": 563,
      "train_runtime": 1352.6023,
      "train_tokens_per_second": 434.07
    },
    {
      "epoch": 0.24105139437974143,
      "grad_norm": 3.1126081943511963,
      "learning_rate": 1.7612834224598933e-05,
      "loss": 0.19863459467887878,
      "num_input_tokens_seen": 588444,
      "step": 564,
      "train_runtime": 1354.6708,
      "train_tokens_per_second": 434.382
    },
    {
      "epoch": 0.2414787904690672,
      "grad_norm": 2.538266181945801,
      "learning_rate": 1.7608556149732622e-05,
      "loss": 0.15238778293132782,
      "num_input_tokens_seen": 589462,
      "step": 565,
      "train_runtime": 1356.7107,
      "train_tokens_per_second": 434.479
    },
    {
      "epoch": 0.241906186558393,
      "grad_norm": 3.10831356048584,
      "learning_rate": 1.760427807486631e-05,
      "loss": 0.10415173321962357,
      "num_input_tokens_seen": 590564,
      "step": 566,
      "train_runtime": 1358.7672,
      "train_tokens_per_second": 434.632
    },
    {
      "epoch": 0.2423335826477188,
      "grad_norm": 2.7880208492279053,
      "learning_rate": 1.76e-05,
      "loss": 0.17961712181568146,
      "num_input_tokens_seen": 591790,
      "step": 567,
      "train_runtime": 1360.8769,
      "train_tokens_per_second": 434.859
    },
    {
      "epoch": 0.24276097873704455,
      "grad_norm": 2.3337337970733643,
      "learning_rate": 1.759572192513369e-05,
      "loss": 0.13482636213302612,
      "num_input_tokens_seen": 592800,
      "step": 568,
      "train_runtime": 1362.9172,
      "train_tokens_per_second": 434.949
    },
    {
      "epoch": 0.24318837482637035,
      "grad_norm": 2.6206247806549072,
      "learning_rate": 1.759144385026738e-05,
      "loss": 0.16304020583629608,
      "num_input_tokens_seen": 593980,
      "step": 569,
      "train_runtime": 1365.0253,
      "train_tokens_per_second": 435.142
    },
    {
      "epoch": 0.2436157709156961,
      "grad_norm": 2.3684146404266357,
      "learning_rate": 1.7587165775401073e-05,
      "loss": 0.13877791166305542,
      "num_input_tokens_seen": 595218,
      "step": 570,
      "train_runtime": 1367.1138,
      "train_tokens_per_second": 435.383
    },
    {
      "epoch": 0.2440431670050219,
      "grad_norm": 4.515527725219727,
      "learning_rate": 1.7582887700534762e-05,
      "loss": 0.16203367710113525,
      "num_input_tokens_seen": 595882,
      "step": 571,
      "train_runtime": 1375.6555,
      "train_tokens_per_second": 433.162
    },
    {
      "epoch": 0.2444705630943477,
      "grad_norm": 2.327942371368408,
      "learning_rate": 1.757860962566845e-05,
      "loss": 0.07233120501041412,
      "num_input_tokens_seen": 597056,
      "step": 572,
      "train_runtime": 1377.8128,
      "train_tokens_per_second": 433.336
    },
    {
      "epoch": 0.24489795918367346,
      "grad_norm": 3.114610195159912,
      "learning_rate": 1.757433155080214e-05,
      "loss": 0.13980111479759216,
      "num_input_tokens_seen": 598454,
      "step": 573,
      "train_runtime": 1379.9889,
      "train_tokens_per_second": 433.666
    },
    {
      "epoch": 0.24532535527299926,
      "grad_norm": 3.724837303161621,
      "learning_rate": 1.757005347593583e-05,
      "loss": 0.1479569375514984,
      "num_input_tokens_seen": 599250,
      "step": 574,
      "train_runtime": 1382.085,
      "train_tokens_per_second": 433.584
    },
    {
      "epoch": 0.24575275136232502,
      "grad_norm": 2.8897786140441895,
      "learning_rate": 1.756577540106952e-05,
      "loss": 0.1169058084487915,
      "num_input_tokens_seen": 600164,
      "step": 575,
      "train_runtime": 1384.1381,
      "train_tokens_per_second": 433.601
    },
    {
      "epoch": 0.24618014745165082,
      "grad_norm": 3.322599172592163,
      "learning_rate": 1.756149732620321e-05,
      "loss": 0.11802787333726883,
      "num_input_tokens_seen": 600880,
      "step": 576,
      "train_runtime": 1386.1272,
      "train_tokens_per_second": 433.496
    },
    {
      "epoch": 0.2466075435409766,
      "grad_norm": 2.489203929901123,
      "learning_rate": 1.75572192513369e-05,
      "loss": 0.1123126819729805,
      "num_input_tokens_seen": 602352,
      "step": 577,
      "train_runtime": 1388.2822,
      "train_tokens_per_second": 433.883
    },
    {
      "epoch": 0.24703493963030237,
      "grad_norm": 3.2752487659454346,
      "learning_rate": 1.755294117647059e-05,
      "loss": 0.1498231589794159,
      "num_input_tokens_seen": 603496,
      "step": 578,
      "train_runtime": 1390.3652,
      "train_tokens_per_second": 434.056
    },
    {
      "epoch": 0.24746233571962817,
      "grad_norm": 2.8954977989196777,
      "learning_rate": 1.754866310160428e-05,
      "loss": 0.18885569274425507,
      "num_input_tokens_seen": 604872,
      "step": 579,
      "train_runtime": 1392.485,
      "train_tokens_per_second": 434.383
    },
    {
      "epoch": 0.24788973180895396,
      "grad_norm": 3.355302333831787,
      "learning_rate": 1.754438502673797e-05,
      "loss": 0.1503232717514038,
      "num_input_tokens_seen": 605866,
      "step": 580,
      "train_runtime": 1394.5079,
      "train_tokens_per_second": 434.466
    },
    {
      "epoch": 0.24831712789827973,
      "grad_norm": 2.6362667083740234,
      "learning_rate": 1.754010695187166e-05,
      "loss": 0.08661763370037079,
      "num_input_tokens_seen": 606936,
      "step": 581,
      "train_runtime": 1396.5518,
      "train_tokens_per_second": 434.596
    },
    {
      "epoch": 0.24874452398760552,
      "grad_norm": 3.7280542850494385,
      "learning_rate": 1.753582887700535e-05,
      "loss": 0.2307671159505844,
      "num_input_tokens_seen": 607974,
      "step": 582,
      "train_runtime": 1398.6036,
      "train_tokens_per_second": 434.701
    },
    {
      "epoch": 0.24917192007693129,
      "grad_norm": 5.050597667694092,
      "learning_rate": 1.753155080213904e-05,
      "loss": 0.28117018938064575,
      "num_input_tokens_seen": 609022,
      "step": 583,
      "train_runtime": 1400.6804,
      "train_tokens_per_second": 434.804
    },
    {
      "epoch": 0.24959931616625708,
      "grad_norm": 2.0937132835388184,
      "learning_rate": 1.7527272727272728e-05,
      "loss": 0.09031204134225845,
      "num_input_tokens_seen": 610356,
      "step": 584,
      "train_runtime": 1402.7949,
      "train_tokens_per_second": 435.1
    },
    {
      "epoch": 0.2500267122555829,
      "grad_norm": 4.441471576690674,
      "learning_rate": 1.7522994652406418e-05,
      "loss": 0.15239763259887695,
      "num_input_tokens_seen": 611014,
      "step": 585,
      "train_runtime": 1404.7819,
      "train_tokens_per_second": 434.953
    },
    {
      "epoch": 0.25045410834490867,
      "grad_norm": 3.232152223587036,
      "learning_rate": 1.751871657754011e-05,
      "loss": 0.19042807817459106,
      "num_input_tokens_seen": 611958,
      "step": 586,
      "train_runtime": 1406.7982,
      "train_tokens_per_second": 435.001
    },
    {
      "epoch": 0.2508815044342344,
      "grad_norm": 3.8009955883026123,
      "learning_rate": 1.75144385026738e-05,
      "loss": 0.16365259885787964,
      "num_input_tokens_seen": 613254,
      "step": 587,
      "train_runtime": 1408.9415,
      "train_tokens_per_second": 435.259
    },
    {
      "epoch": 0.2513089005235602,
      "grad_norm": 2.9389092922210693,
      "learning_rate": 1.7510160427807486e-05,
      "loss": 0.204757422208786,
      "num_input_tokens_seen": 614336,
      "step": 588,
      "train_runtime": 1411.0397,
      "train_tokens_per_second": 435.378
    },
    {
      "epoch": 0.251736296612886,
      "grad_norm": 3.7831485271453857,
      "learning_rate": 1.7505882352941175e-05,
      "loss": 0.15953440964221954,
      "num_input_tokens_seen": 615766,
      "step": 589,
      "train_runtime": 1413.2612,
      "train_tokens_per_second": 435.706
    },
    {
      "epoch": 0.2521636927022118,
      "grad_norm": 2.512909173965454,
      "learning_rate": 1.750160427807487e-05,
      "loss": 0.1422572284936905,
      "num_input_tokens_seen": 617264,
      "step": 590,
      "train_runtime": 1415.4185,
      "train_tokens_per_second": 436.1
    },
    {
      "epoch": 0.2525910887915376,
      "grad_norm": 3.1766417026519775,
      "learning_rate": 1.7497326203208558e-05,
      "loss": 0.1679745465517044,
      "num_input_tokens_seen": 618526,
      "step": 591,
      "train_runtime": 1417.5736,
      "train_tokens_per_second": 436.327
    },
    {
      "epoch": 0.2530184848808633,
      "grad_norm": 2.4790332317352295,
      "learning_rate": 1.7493048128342247e-05,
      "loss": 0.16574200987815857,
      "num_input_tokens_seen": 620828,
      "step": 592,
      "train_runtime": 1419.9065,
      "train_tokens_per_second": 437.232
    },
    {
      "epoch": 0.2534458809701891,
      "grad_norm": 2.5413501262664795,
      "learning_rate": 1.7488770053475937e-05,
      "loss": 0.17381271719932556,
      "num_input_tokens_seen": 621908,
      "step": 593,
      "train_runtime": 1421.9728,
      "train_tokens_per_second": 437.356
    },
    {
      "epoch": 0.2538732770595149,
      "grad_norm": 3.3934741020202637,
      "learning_rate": 1.748449197860963e-05,
      "loss": 0.14800743758678436,
      "num_input_tokens_seen": 623304,
      "step": 594,
      "train_runtime": 1424.131,
      "train_tokens_per_second": 437.673
    },
    {
      "epoch": 0.2543006731488407,
      "grad_norm": 2.9373390674591064,
      "learning_rate": 1.7480213903743316e-05,
      "loss": 0.1242058053612709,
      "num_input_tokens_seen": 624342,
      "step": 595,
      "train_runtime": 1426.2062,
      "train_tokens_per_second": 437.764
    },
    {
      "epoch": 0.2547280692381665,
      "grad_norm": 2.9254722595214844,
      "learning_rate": 1.7475935828877005e-05,
      "loss": 0.22464588284492493,
      "num_input_tokens_seen": 625290,
      "step": 596,
      "train_runtime": 1428.226,
      "train_tokens_per_second": 437.809
    },
    {
      "epoch": 0.2551554653274922,
      "grad_norm": 4.2407307624816895,
      "learning_rate": 1.7471657754010694e-05,
      "loss": 0.15919914841651917,
      "num_input_tokens_seen": 626950,
      "step": 597,
      "train_runtime": 1430.4349,
      "train_tokens_per_second": 438.293
    },
    {
      "epoch": 0.255582861416818,
      "grad_norm": 2.609807252883911,
      "learning_rate": 1.7467379679144387e-05,
      "loss": 0.19275696575641632,
      "num_input_tokens_seen": 628778,
      "step": 598,
      "train_runtime": 1432.7002,
      "train_tokens_per_second": 438.876
    },
    {
      "epoch": 0.2560102575061438,
      "grad_norm": 3.7087695598602295,
      "learning_rate": 1.7463101604278077e-05,
      "loss": 0.23309898376464844,
      "num_input_tokens_seen": 629784,
      "step": 599,
      "train_runtime": 1434.7557,
      "train_tokens_per_second": 438.949
    },
    {
      "epoch": 0.2564376535954696,
      "grad_norm": 2.742621660232544,
      "learning_rate": 1.7458823529411766e-05,
      "loss": 0.13614144921302795,
      "num_input_tokens_seen": 630786,
      "step": 600,
      "train_runtime": 1436.8607,
      "train_tokens_per_second": 439.003
    },
    {
      "epoch": 0.2568650496847954,
      "grad_norm": 3.2122294902801514,
      "learning_rate": 1.7454545454545456e-05,
      "loss": 0.13549478352069855,
      "num_input_tokens_seen": 631604,
      "step": 601,
      "train_runtime": 1445.2476,
      "train_tokens_per_second": 437.021
    },
    {
      "epoch": 0.2572924457741212,
      "grad_norm": 3.1866111755371094,
      "learning_rate": 1.7450267379679145e-05,
      "loss": 0.1817619800567627,
      "num_input_tokens_seen": 632476,
      "step": 602,
      "train_runtime": 1447.3105,
      "train_tokens_per_second": 437.001
    },
    {
      "epoch": 0.25771984186344693,
      "grad_norm": 4.4216203689575195,
      "learning_rate": 1.7445989304812834e-05,
      "loss": 0.12546993792057037,
      "num_input_tokens_seen": 633386,
      "step": 603,
      "train_runtime": 1449.3432,
      "train_tokens_per_second": 437.016
    },
    {
      "epoch": 0.2581472379527727,
      "grad_norm": 2.9314122200012207,
      "learning_rate": 1.7441711229946524e-05,
      "loss": 0.1516924649477005,
      "num_input_tokens_seen": 634266,
      "step": 604,
      "train_runtime": 1451.3951,
      "train_tokens_per_second": 437.004
    },
    {
      "epoch": 0.2585746340420985,
      "grad_norm": 3.23160719871521,
      "learning_rate": 1.7437433155080213e-05,
      "loss": 0.1365823894739151,
      "num_input_tokens_seen": 635158,
      "step": 605,
      "train_runtime": 1453.4169,
      "train_tokens_per_second": 437.01
    },
    {
      "epoch": 0.2590020301314243,
      "grad_norm": 2.3937151432037354,
      "learning_rate": 1.7433155080213906e-05,
      "loss": 0.1260988414287567,
      "num_input_tokens_seen": 636060,
      "step": 606,
      "train_runtime": 1455.5372,
      "train_tokens_per_second": 436.993
    },
    {
      "epoch": 0.2594294262207501,
      "grad_norm": 2.9081015586853027,
      "learning_rate": 1.7428877005347596e-05,
      "loss": 0.1591911017894745,
      "num_input_tokens_seen": 637656,
      "step": 607,
      "train_runtime": 1457.7674,
      "train_tokens_per_second": 437.42
    },
    {
      "epoch": 0.25985682231007584,
      "grad_norm": 2.369614601135254,
      "learning_rate": 1.7424598930481285e-05,
      "loss": 0.08819835633039474,
      "num_input_tokens_seen": 638622,
      "step": 608,
      "train_runtime": 1459.7932,
      "train_tokens_per_second": 437.474
    },
    {
      "epoch": 0.26028421839940163,
      "grad_norm": 3.0980114936828613,
      "learning_rate": 1.7420320855614975e-05,
      "loss": 0.16393521428108215,
      "num_input_tokens_seen": 639516,
      "step": 609,
      "train_runtime": 1461.7836,
      "train_tokens_per_second": 437.49
    },
    {
      "epoch": 0.26071161448872743,
      "grad_norm": 2.485929012298584,
      "learning_rate": 1.7416042780748664e-05,
      "loss": 0.1554461121559143,
      "num_input_tokens_seen": 641198,
      "step": 610,
      "train_runtime": 1464.0289,
      "train_tokens_per_second": 437.968
    },
    {
      "epoch": 0.2611390105780532,
      "grad_norm": 4.0961012840271,
      "learning_rate": 1.7411764705882353e-05,
      "loss": 0.1821095049381256,
      "num_input_tokens_seen": 642066,
      "step": 611,
      "train_runtime": 1466.0505,
      "train_tokens_per_second": 437.956
    },
    {
      "epoch": 0.261566406667379,
      "grad_norm": 2.3384289741516113,
      "learning_rate": 1.7407486631016043e-05,
      "loss": 0.13961754739284515,
      "num_input_tokens_seen": 643672,
      "step": 612,
      "train_runtime": 1468.3779,
      "train_tokens_per_second": 438.356
    },
    {
      "epoch": 0.26199380275670475,
      "grad_norm": 2.7055482864379883,
      "learning_rate": 1.7403208556149732e-05,
      "loss": 0.1763858199119568,
      "num_input_tokens_seen": 644672,
      "step": 613,
      "train_runtime": 1470.4545,
      "train_tokens_per_second": 438.417
    },
    {
      "epoch": 0.26242119884603055,
      "grad_norm": 3.583223819732666,
      "learning_rate": 1.7398930481283425e-05,
      "loss": 0.21828489005565643,
      "num_input_tokens_seen": 645518,
      "step": 614,
      "train_runtime": 1472.4824,
      "train_tokens_per_second": 438.388
    },
    {
      "epoch": 0.26284859493535634,
      "grad_norm": 2.7754290103912354,
      "learning_rate": 1.7394652406417115e-05,
      "loss": 0.114726223051548,
      "num_input_tokens_seen": 646582,
      "step": 615,
      "train_runtime": 1474.5747,
      "train_tokens_per_second": 438.487
    },
    {
      "epoch": 0.26327599102468213,
      "grad_norm": 4.047714710235596,
      "learning_rate": 1.7390374331550804e-05,
      "loss": 0.18886765837669373,
      "num_input_tokens_seen": 647430,
      "step": 616,
      "train_runtime": 1476.6078,
      "train_tokens_per_second": 438.458
    },
    {
      "epoch": 0.2637033871140079,
      "grad_norm": 5.156888961791992,
      "learning_rate": 1.7386096256684494e-05,
      "loss": 0.1598343849182129,
      "num_input_tokens_seen": 648152,
      "step": 617,
      "train_runtime": 1478.6533,
      "train_tokens_per_second": 438.339
    },
    {
      "epoch": 0.26413078320333366,
      "grad_norm": 4.466670036315918,
      "learning_rate": 1.7381818181818183e-05,
      "loss": 0.19553837180137634,
      "num_input_tokens_seen": 648804,
      "step": 618,
      "train_runtime": 1480.6539,
      "train_tokens_per_second": 438.187
    },
    {
      "epoch": 0.26455817929265946,
      "grad_norm": 3.7706260681152344,
      "learning_rate": 1.7377540106951872e-05,
      "loss": 0.20222875475883484,
      "num_input_tokens_seen": 650052,
      "step": 619,
      "train_runtime": 1482.8124,
      "train_tokens_per_second": 438.391
    },
    {
      "epoch": 0.26498557538198525,
      "grad_norm": 2.714515209197998,
      "learning_rate": 1.7373262032085562e-05,
      "loss": 0.12274238467216492,
      "num_input_tokens_seen": 651060,
      "step": 620,
      "train_runtime": 1484.8986,
      "train_tokens_per_second": 438.454
    },
    {
      "epoch": 0.26541297147131104,
      "grad_norm": 3.983781099319458,
      "learning_rate": 1.736898395721925e-05,
      "loss": 0.17818422615528107,
      "num_input_tokens_seen": 651978,
      "step": 621,
      "train_runtime": 1486.946,
      "train_tokens_per_second": 438.468
    },
    {
      "epoch": 0.26584036756063684,
      "grad_norm": 8.49150562286377,
      "learning_rate": 1.7364705882352944e-05,
      "loss": 0.2724238336086273,
      "num_input_tokens_seen": 652776,
      "step": 622,
      "train_runtime": 1488.9792,
      "train_tokens_per_second": 438.405
    },
    {
      "epoch": 0.26626776364996263,
      "grad_norm": 2.6343836784362793,
      "learning_rate": 1.7360427807486634e-05,
      "loss": 0.13595706224441528,
      "num_input_tokens_seen": 654328,
      "step": 623,
      "train_runtime": 1491.1695,
      "train_tokens_per_second": 438.802
    },
    {
      "epoch": 0.26669515973928837,
      "grad_norm": 5.991724491119385,
      "learning_rate": 1.7356149732620323e-05,
      "loss": 0.3250628411769867,
      "num_input_tokens_seen": 655520,
      "step": 624,
      "train_runtime": 1493.2362,
      "train_tokens_per_second": 438.993
    },
    {
      "epoch": 0.26712255582861416,
      "grad_norm": 3.857046127319336,
      "learning_rate": 1.7351871657754012e-05,
      "loss": 0.1667994111776352,
      "num_input_tokens_seen": 656314,
      "step": 625,
      "train_runtime": 1495.2358,
      "train_tokens_per_second": 438.937
    },
    {
      "epoch": 0.26754995191793995,
      "grad_norm": 3.1027472019195557,
      "learning_rate": 1.7347593582887702e-05,
      "loss": 0.17843560874462128,
      "num_input_tokens_seen": 657426,
      "step": 626,
      "train_runtime": 1497.3008,
      "train_tokens_per_second": 439.074
    },
    {
      "epoch": 0.26797734800726575,
      "grad_norm": 3.7872979640960693,
      "learning_rate": 1.734331550802139e-05,
      "loss": 0.16721181571483612,
      "num_input_tokens_seen": 658524,
      "step": 627,
      "train_runtime": 1499.414,
      "train_tokens_per_second": 439.188
    },
    {
      "epoch": 0.26840474409659154,
      "grad_norm": 3.251497745513916,
      "learning_rate": 1.733903743315508e-05,
      "loss": 0.11725930869579315,
      "num_input_tokens_seen": 659636,
      "step": 628,
      "train_runtime": 1501.4782,
      "train_tokens_per_second": 439.324
    },
    {
      "epoch": 0.2688321401859173,
      "grad_norm": 2.6759772300720215,
      "learning_rate": 1.733475935828877e-05,
      "loss": 0.1434108167886734,
      "num_input_tokens_seen": 660798,
      "step": 629,
      "train_runtime": 1503.5833,
      "train_tokens_per_second": 439.482
    },
    {
      "epoch": 0.26925953627524307,
      "grad_norm": 2.004434585571289,
      "learning_rate": 1.7330481283422463e-05,
      "loss": 0.11522868275642395,
      "num_input_tokens_seen": 662582,
      "step": 630,
      "train_runtime": 1505.8284,
      "train_tokens_per_second": 440.012
    },
    {
      "epoch": 0.26968693236456887,
      "grad_norm": 2.673767566680908,
      "learning_rate": 1.7326203208556153e-05,
      "loss": 0.15051482617855072,
      "num_input_tokens_seen": 663804,
      "step": 631,
      "train_runtime": 1514.1449,
      "train_tokens_per_second": 438.402
    },
    {
      "epoch": 0.27011432845389466,
      "grad_norm": 2.9031167030334473,
      "learning_rate": 1.7321925133689842e-05,
      "loss": 0.1311330795288086,
      "num_input_tokens_seen": 664674,
      "step": 632,
      "train_runtime": 1516.172,
      "train_tokens_per_second": 438.39
    },
    {
      "epoch": 0.27054172454322045,
      "grad_norm": 3.364966869354248,
      "learning_rate": 1.731764705882353e-05,
      "loss": 0.15498673915863037,
      "num_input_tokens_seen": 665636,
      "step": 633,
      "train_runtime": 1518.283,
      "train_tokens_per_second": 438.414
    },
    {
      "epoch": 0.2709691206325462,
      "grad_norm": 2.531022787094116,
      "learning_rate": 1.731336898395722e-05,
      "loss": 0.11715377122163773,
      "num_input_tokens_seen": 666736,
      "step": 634,
      "train_runtime": 1520.35,
      "train_tokens_per_second": 438.541
    },
    {
      "epoch": 0.271396516721872,
      "grad_norm": 2.71030592918396,
      "learning_rate": 1.730909090909091e-05,
      "loss": 0.12561234831809998,
      "num_input_tokens_seen": 668148,
      "step": 635,
      "train_runtime": 1522.4985,
      "train_tokens_per_second": 438.85
    },
    {
      "epoch": 0.2718239128111978,
      "grad_norm": 2.7606797218322754,
      "learning_rate": 1.73048128342246e-05,
      "loss": 0.1251692771911621,
      "num_input_tokens_seen": 669176,
      "step": 636,
      "train_runtime": 1524.5212,
      "train_tokens_per_second": 438.942
    },
    {
      "epoch": 0.27225130890052357,
      "grad_norm": 2.743720293045044,
      "learning_rate": 1.730053475935829e-05,
      "loss": 0.12467562407255173,
      "num_input_tokens_seen": 670002,
      "step": 637,
      "train_runtime": 1526.5528,
      "train_tokens_per_second": 438.899
    },
    {
      "epoch": 0.27267870498984936,
      "grad_norm": 2.1312413215637207,
      "learning_rate": 1.7296256684491982e-05,
      "loss": 0.13293921947479248,
      "num_input_tokens_seen": 671710,
      "step": 638,
      "train_runtime": 1528.7995,
      "train_tokens_per_second": 439.371
    },
    {
      "epoch": 0.2731061010791751,
      "grad_norm": 3.113715887069702,
      "learning_rate": 1.729197860962567e-05,
      "loss": 0.1551579385995865,
      "num_input_tokens_seen": 672736,
      "step": 639,
      "train_runtime": 1530.8652,
      "train_tokens_per_second": 439.448
    },
    {
      "epoch": 0.2735334971685009,
      "grad_norm": 2.57463002204895,
      "learning_rate": 1.728770053475936e-05,
      "loss": 0.15595224499702454,
      "num_input_tokens_seen": 673894,
      "step": 640,
      "train_runtime": 1532.981,
      "train_tokens_per_second": 439.597
    },
    {
      "epoch": 0.2739608932578267,
      "grad_norm": 3.000229597091675,
      "learning_rate": 1.728342245989305e-05,
      "loss": 0.2024955004453659,
      "num_input_tokens_seen": 675060,
      "step": 641,
      "train_runtime": 1535.1013,
      "train_tokens_per_second": 439.749
    },
    {
      "epoch": 0.2743882893471525,
      "grad_norm": 3.2893600463867188,
      "learning_rate": 1.727914438502674e-05,
      "loss": 0.18899883329868317,
      "num_input_tokens_seen": 676204,
      "step": 642,
      "train_runtime": 1537.1613,
      "train_tokens_per_second": 439.904
    },
    {
      "epoch": 0.2748156854364783,
      "grad_norm": 2.9903957843780518,
      "learning_rate": 1.727486631016043e-05,
      "loss": 0.11259229481220245,
      "num_input_tokens_seen": 677100,
      "step": 643,
      "train_runtime": 1539.1614,
      "train_tokens_per_second": 439.915
    },
    {
      "epoch": 0.275243081525804,
      "grad_norm": 3.6501784324645996,
      "learning_rate": 1.727058823529412e-05,
      "loss": 0.14545176923274994,
      "num_input_tokens_seen": 678234,
      "step": 644,
      "train_runtime": 1541.2244,
      "train_tokens_per_second": 440.062
    },
    {
      "epoch": 0.2756704776151298,
      "grad_norm": 3.6744797229766846,
      "learning_rate": 1.7266310160427808e-05,
      "loss": 0.1335563063621521,
      "num_input_tokens_seen": 678926,
      "step": 645,
      "train_runtime": 1543.2733,
      "train_tokens_per_second": 439.926
    },
    {
      "epoch": 0.2760978737044556,
      "grad_norm": 2.7332375049591064,
      "learning_rate": 1.72620320855615e-05,
      "loss": 0.1322796642780304,
      "num_input_tokens_seen": 679898,
      "step": 646,
      "train_runtime": 1545.2938,
      "train_tokens_per_second": 439.98
    },
    {
      "epoch": 0.2765252697937814,
      "grad_norm": 2.5545904636383057,
      "learning_rate": 1.725775401069519e-05,
      "loss": 0.15949004888534546,
      "num_input_tokens_seen": 680990,
      "step": 647,
      "train_runtime": 1547.3526,
      "train_tokens_per_second": 440.1
    },
    {
      "epoch": 0.2769526658831072,
      "grad_norm": 2.9079134464263916,
      "learning_rate": 1.725347593582888e-05,
      "loss": 0.12913838028907776,
      "num_input_tokens_seen": 682104,
      "step": 648,
      "train_runtime": 1549.4315,
      "train_tokens_per_second": 440.229
    },
    {
      "epoch": 0.277380061972433,
      "grad_norm": 2.6565921306610107,
      "learning_rate": 1.724919786096257e-05,
      "loss": 0.17941144108772278,
      "num_input_tokens_seen": 683204,
      "step": 649,
      "train_runtime": 1551.504,
      "train_tokens_per_second": 440.349
    },
    {
      "epoch": 0.2778074580617587,
      "grad_norm": 3.9844510555267334,
      "learning_rate": 1.724491978609626e-05,
      "loss": 0.22910510003566742,
      "num_input_tokens_seen": 683966,
      "step": 650,
      "train_runtime": 1553.4987,
      "train_tokens_per_second": 440.275
    },
    {
      "epoch": 0.2782348541510845,
      "grad_norm": 3.6703052520751953,
      "learning_rate": 1.7240641711229948e-05,
      "loss": 0.11243750900030136,
      "num_input_tokens_seen": 684992,
      "step": 651,
      "train_runtime": 1555.5168,
      "train_tokens_per_second": 440.363
    },
    {
      "epoch": 0.2786622502404103,
      "grad_norm": 2.7850770950317383,
      "learning_rate": 1.7236363636363638e-05,
      "loss": 0.18875238299369812,
      "num_input_tokens_seen": 686164,
      "step": 652,
      "train_runtime": 1557.6351,
      "train_tokens_per_second": 440.517
    },
    {
      "epoch": 0.2790896463297361,
      "grad_norm": 2.373117208480835,
      "learning_rate": 1.7232085561497327e-05,
      "loss": 0.14671288430690765,
      "num_input_tokens_seen": 687650,
      "step": 653,
      "train_runtime": 1559.8517,
      "train_tokens_per_second": 440.843
    },
    {
      "epoch": 0.2795170424190619,
      "grad_norm": 3.4782025814056396,
      "learning_rate": 1.722780748663102e-05,
      "loss": 0.23798540234565735,
      "num_input_tokens_seen": 688598,
      "step": 654,
      "train_runtime": 1561.9064,
      "train_tokens_per_second": 440.87
    },
    {
      "epoch": 0.2799444385083876,
      "grad_norm": 4.702680587768555,
      "learning_rate": 1.722352941176471e-05,
      "loss": 0.1546967476606369,
      "num_input_tokens_seen": 689498,
      "step": 655,
      "train_runtime": 1563.9556,
      "train_tokens_per_second": 440.868
    },
    {
      "epoch": 0.2803718345977134,
      "grad_norm": 3.1064441204071045,
      "learning_rate": 1.7219251336898395e-05,
      "loss": 0.12857335805892944,
      "num_input_tokens_seen": 690404,
      "step": 656,
      "train_runtime": 1565.993,
      "train_tokens_per_second": 440.873
    },
    {
      "epoch": 0.2807992306870392,
      "grad_norm": 2.365166187286377,
      "learning_rate": 1.7214973262032085e-05,
      "loss": 0.12610051035881042,
      "num_input_tokens_seen": 692104,
      "step": 657,
      "train_runtime": 1568.275,
      "train_tokens_per_second": 441.315
    },
    {
      "epoch": 0.281226626776365,
      "grad_norm": 3.49941086769104,
      "learning_rate": 1.7210695187165778e-05,
      "loss": 0.14522790908813477,
      "num_input_tokens_seen": 693036,
      "step": 658,
      "train_runtime": 1570.2917,
      "train_tokens_per_second": 441.342
    },
    {
      "epoch": 0.2816540228656908,
      "grad_norm": 3.1161298751831055,
      "learning_rate": 1.7206417112299467e-05,
      "loss": 0.1213645339012146,
      "num_input_tokens_seen": 694052,
      "step": 659,
      "train_runtime": 1572.3719,
      "train_tokens_per_second": 441.404
    },
    {
      "epoch": 0.28208141895501654,
      "grad_norm": 3.147714138031006,
      "learning_rate": 1.7202139037433157e-05,
      "loss": 0.13667607307434082,
      "num_input_tokens_seen": 695124,
      "step": 660,
      "train_runtime": 1574.4334,
      "train_tokens_per_second": 441.507
    },
    {
      "epoch": 0.28250881504434233,
      "grad_norm": 4.764846324920654,
      "learning_rate": 1.7197860962566846e-05,
      "loss": 0.1356460452079773,
      "num_input_tokens_seen": 696154,
      "step": 661,
      "train_runtime": 1583.007,
      "train_tokens_per_second": 439.767
    },
    {
      "epoch": 0.2829362111336681,
      "grad_norm": 2.6429355144500732,
      "learning_rate": 1.7193582887700536e-05,
      "loss": 0.13607126474380493,
      "num_input_tokens_seen": 697178,
      "step": 662,
      "train_runtime": 1585.1046,
      "train_tokens_per_second": 439.831
    },
    {
      "epoch": 0.2833636072229939,
      "grad_norm": 4.223821640014648,
      "learning_rate": 1.7189304812834225e-05,
      "loss": 0.19682806730270386,
      "num_input_tokens_seen": 697816,
      "step": 663,
      "train_runtime": 1587.0357,
      "train_tokens_per_second": 439.698
    },
    {
      "epoch": 0.2837910033123197,
      "grad_norm": 3.172499179840088,
      "learning_rate": 1.7185026737967914e-05,
      "loss": 0.13376055657863617,
      "num_input_tokens_seen": 698672,
      "step": 664,
      "train_runtime": 1589.1194,
      "train_tokens_per_second": 439.66
    },
    {
      "epoch": 0.28421839940164545,
      "grad_norm": 3.1004250049591064,
      "learning_rate": 1.7180748663101604e-05,
      "loss": 0.19494307041168213,
      "num_input_tokens_seen": 699868,
      "step": 665,
      "train_runtime": 1591.2378,
      "train_tokens_per_second": 439.826
    },
    {
      "epoch": 0.28464579549097124,
      "grad_norm": 2.8081493377685547,
      "learning_rate": 1.7176470588235293e-05,
      "loss": 0.12997466325759888,
      "num_input_tokens_seen": 700964,
      "step": 666,
      "train_runtime": 1593.3424,
      "train_tokens_per_second": 439.933
    },
    {
      "epoch": 0.28507319158029704,
      "grad_norm": 2.8225746154785156,
      "learning_rate": 1.7172192513368986e-05,
      "loss": 0.18104994297027588,
      "num_input_tokens_seen": 702766,
      "step": 667,
      "train_runtime": 1595.5428,
      "train_tokens_per_second": 440.456
    },
    {
      "epoch": 0.28550058766962283,
      "grad_norm": 3.4809491634368896,
      "learning_rate": 1.7167914438502676e-05,
      "loss": 0.22088421881198883,
      "num_input_tokens_seen": 703934,
      "step": 668,
      "train_runtime": 1597.6296,
      "train_tokens_per_second": 440.612
    },
    {
      "epoch": 0.2859279837589486,
      "grad_norm": 3.1476643085479736,
      "learning_rate": 1.7163636363636365e-05,
      "loss": 0.10465161502361298,
      "num_input_tokens_seen": 704658,
      "step": 669,
      "train_runtime": 1599.6027,
      "train_tokens_per_second": 440.521
    },
    {
      "epoch": 0.28635537984827436,
      "grad_norm": 3.01611328125,
      "learning_rate": 1.7159358288770054e-05,
      "loss": 0.14153167605400085,
      "num_input_tokens_seen": 706188,
      "step": 670,
      "train_runtime": 1601.782,
      "train_tokens_per_second": 440.876
    },
    {
      "epoch": 0.28678277593760015,
      "grad_norm": 1.938707947731018,
      "learning_rate": 1.7155080213903744e-05,
      "loss": 0.10746683180332184,
      "num_input_tokens_seen": 707622,
      "step": 671,
      "train_runtime": 1603.9052,
      "train_tokens_per_second": 441.187
    },
    {
      "epoch": 0.28721017202692595,
      "grad_norm": 5.22975492477417,
      "learning_rate": 1.7150802139037433e-05,
      "loss": 0.20209376513957977,
      "num_input_tokens_seen": 708358,
      "step": 672,
      "train_runtime": 1605.9684,
      "train_tokens_per_second": 441.078
    },
    {
      "epoch": 0.28763756811625174,
      "grad_norm": 2.241725206375122,
      "learning_rate": 1.7146524064171123e-05,
      "loss": 0.10788868367671967,
      "num_input_tokens_seen": 709372,
      "step": 673,
      "train_runtime": 1608.0325,
      "train_tokens_per_second": 441.143
    },
    {
      "epoch": 0.28806496420557753,
      "grad_norm": 4.1165642738342285,
      "learning_rate": 1.7142245989304812e-05,
      "loss": 0.10829375684261322,
      "num_input_tokens_seen": 709858,
      "step": 674,
      "train_runtime": 1609.9182,
      "train_tokens_per_second": 440.928
    },
    {
      "epoch": 0.2884923602949033,
      "grad_norm": 2.7129435539245605,
      "learning_rate": 1.7137967914438505e-05,
      "loss": 0.11378169059753418,
      "num_input_tokens_seen": 710642,
      "step": 675,
      "train_runtime": 1611.9211,
      "train_tokens_per_second": 440.866
    },
    {
      "epoch": 0.28891975638422907,
      "grad_norm": 2.3255112171173096,
      "learning_rate": 1.7133689839572195e-05,
      "loss": 0.1456914097070694,
      "num_input_tokens_seen": 711836,
      "step": 676,
      "train_runtime": 1614.1808,
      "train_tokens_per_second": 440.989
    },
    {
      "epoch": 0.28934715247355486,
      "grad_norm": 3.522317409515381,
      "learning_rate": 1.7129411764705884e-05,
      "loss": 0.151935875415802,
      "num_input_tokens_seen": 712588,
      "step": 677,
      "train_runtime": 1616.1754,
      "train_tokens_per_second": 440.91
    },
    {
      "epoch": 0.28977454856288065,
      "grad_norm": 3.6937501430511475,
      "learning_rate": 1.7125133689839573e-05,
      "loss": 0.16675081849098206,
      "num_input_tokens_seen": 713650,
      "step": 678,
      "train_runtime": 1618.24,
      "train_tokens_per_second": 441.004
    },
    {
      "epoch": 0.29020194465220644,
      "grad_norm": 4.128882884979248,
      "learning_rate": 1.7120855614973263e-05,
      "loss": 0.20945796370506287,
      "num_input_tokens_seen": 714694,
      "step": 679,
      "train_runtime": 1620.3033,
      "train_tokens_per_second": 441.087
    },
    {
      "epoch": 0.29062934074153224,
      "grad_norm": 1.9825953245162964,
      "learning_rate": 1.7116577540106952e-05,
      "loss": 0.084213986992836,
      "num_input_tokens_seen": 715872,
      "step": 680,
      "train_runtime": 1622.3853,
      "train_tokens_per_second": 441.247
    },
    {
      "epoch": 0.291056736830858,
      "grad_norm": 3.867910146713257,
      "learning_rate": 1.7112299465240642e-05,
      "loss": 0.20786266028881073,
      "num_input_tokens_seen": 716616,
      "step": 681,
      "train_runtime": 1624.4679,
      "train_tokens_per_second": 441.139
    },
    {
      "epoch": 0.29148413292018377,
      "grad_norm": 3.203005790710449,
      "learning_rate": 1.710802139037433e-05,
      "loss": 0.15088096261024475,
      "num_input_tokens_seen": 717396,
      "step": 682,
      "train_runtime": 1626.5048,
      "train_tokens_per_second": 441.066
    },
    {
      "epoch": 0.29191152900950956,
      "grad_norm": 3.3211700916290283,
      "learning_rate": 1.7103743315508024e-05,
      "loss": 0.15841370820999146,
      "num_input_tokens_seen": 718324,
      "step": 683,
      "train_runtime": 1628.5301,
      "train_tokens_per_second": 441.087
    },
    {
      "epoch": 0.29233892509883536,
      "grad_norm": 4.750456809997559,
      "learning_rate": 1.7099465240641713e-05,
      "loss": 0.1749861091375351,
      "num_input_tokens_seen": 719262,
      "step": 684,
      "train_runtime": 1630.5793,
      "train_tokens_per_second": 441.108
    },
    {
      "epoch": 0.29276632118816115,
      "grad_norm": 3.024815320968628,
      "learning_rate": 1.7095187165775403e-05,
      "loss": 0.12039594352245331,
      "num_input_tokens_seen": 720384,
      "step": 685,
      "train_runtime": 1632.6689,
      "train_tokens_per_second": 441.231
    },
    {
      "epoch": 0.2931937172774869,
      "grad_norm": 3.736140012741089,
      "learning_rate": 1.7090909090909092e-05,
      "loss": 0.15887826681137085,
      "num_input_tokens_seen": 721052,
      "step": 686,
      "train_runtime": 1634.6096,
      "train_tokens_per_second": 441.116
    },
    {
      "epoch": 0.2936211133668127,
      "grad_norm": 2.7646737098693848,
      "learning_rate": 1.7086631016042782e-05,
      "loss": 0.07137889415025711,
      "num_input_tokens_seen": 721936,
      "step": 687,
      "train_runtime": 1636.6613,
      "train_tokens_per_second": 441.103
    },
    {
      "epoch": 0.2940485094561385,
      "grad_norm": 3.6937055587768555,
      "learning_rate": 1.708235294117647e-05,
      "loss": 0.22509972751140594,
      "num_input_tokens_seen": 722674,
      "step": 688,
      "train_runtime": 1638.6715,
      "train_tokens_per_second": 441.012
    },
    {
      "epoch": 0.29447590554546427,
      "grad_norm": 2.640775680541992,
      "learning_rate": 1.707807486631016e-05,
      "loss": 0.2119310200214386,
      "num_input_tokens_seen": 723766,
      "step": 689,
      "train_runtime": 1640.7739,
      "train_tokens_per_second": 441.113
    },
    {
      "epoch": 0.29490330163479006,
      "grad_norm": 1.934248447418213,
      "learning_rate": 1.707379679144385e-05,
      "loss": 0.1286829710006714,
      "num_input_tokens_seen": 725872,
      "step": 690,
      "train_runtime": 1643.0905,
      "train_tokens_per_second": 441.772
    },
    {
      "epoch": 0.2953306977241158,
      "grad_norm": 2.6751089096069336,
      "learning_rate": 1.7069518716577543e-05,
      "loss": 0.12910306453704834,
      "num_input_tokens_seen": 726914,
      "step": 691,
      "train_runtime": 1651.9994,
      "train_tokens_per_second": 440.021
    },
    {
      "epoch": 0.2957580938134416,
      "grad_norm": 2.5396509170532227,
      "learning_rate": 1.7065240641711232e-05,
      "loss": 0.21183884143829346,
      "num_input_tokens_seen": 728272,
      "step": 692,
      "train_runtime": 1654.156,
      "train_tokens_per_second": 440.268
    },
    {
      "epoch": 0.2961854899027674,
      "grad_norm": 3.9704675674438477,
      "learning_rate": 1.7060962566844922e-05,
      "loss": 0.13218924403190613,
      "num_input_tokens_seen": 729172,
      "step": 693,
      "train_runtime": 1656.184,
      "train_tokens_per_second": 440.272
    },
    {
      "epoch": 0.2966128859920932,
      "grad_norm": 3.2691850662231445,
      "learning_rate": 1.705668449197861e-05,
      "loss": 0.17206302285194397,
      "num_input_tokens_seen": 730106,
      "step": 694,
      "train_runtime": 1658.2171,
      "train_tokens_per_second": 440.296
    },
    {
      "epoch": 0.29704028208141897,
      "grad_norm": 2.426727771759033,
      "learning_rate": 1.70524064171123e-05,
      "loss": 0.09853766113519669,
      "num_input_tokens_seen": 731538,
      "step": 695,
      "train_runtime": 1660.3802,
      "train_tokens_per_second": 440.585
    },
    {
      "epoch": 0.29746767817074476,
      "grad_norm": 2.596492290496826,
      "learning_rate": 1.704812834224599e-05,
      "loss": 0.10594348609447479,
      "num_input_tokens_seen": 732798,
      "step": 696,
      "train_runtime": 1662.533,
      "train_tokens_per_second": 440.772
    },
    {
      "epoch": 0.2978950742600705,
      "grad_norm": 2.4652998447418213,
      "learning_rate": 1.704385026737968e-05,
      "loss": 0.13944201171398163,
      "num_input_tokens_seen": 734214,
      "step": 697,
      "train_runtime": 1664.7198,
      "train_tokens_per_second": 441.044
    },
    {
      "epoch": 0.2983224703493963,
      "grad_norm": 4.074872970581055,
      "learning_rate": 1.703957219251337e-05,
      "loss": 0.2254430651664734,
      "num_input_tokens_seen": 735190,
      "step": 698,
      "train_runtime": 1666.7638,
      "train_tokens_per_second": 441.088
    },
    {
      "epoch": 0.2987498664387221,
      "grad_norm": 4.0392351150512695,
      "learning_rate": 1.7035294117647062e-05,
      "loss": 0.1986958384513855,
      "num_input_tokens_seen": 736072,
      "step": 699,
      "train_runtime": 1668.7559,
      "train_tokens_per_second": 441.09
    },
    {
      "epoch": 0.2991772625280479,
      "grad_norm": 2.5483055114746094,
      "learning_rate": 1.703101604278075e-05,
      "loss": 0.12155450880527496,
      "num_input_tokens_seen": 737138,
      "step": 700,
      "train_runtime": 1670.8187,
      "train_tokens_per_second": 441.184
    },
    {
      "epoch": 0.2996046586173737,
      "grad_norm": 3.423879861831665,
      "learning_rate": 1.702673796791444e-05,
      "loss": 0.12087435275316238,
      "num_input_tokens_seen": 737838,
      "step": 701,
      "train_runtime": 1672.8099,
      "train_tokens_per_second": 441.077
    },
    {
      "epoch": 0.3000320547066994,
      "grad_norm": 3.225769281387329,
      "learning_rate": 1.702245989304813e-05,
      "loss": 0.14212779700756073,
      "num_input_tokens_seen": 738804,
      "step": 702,
      "train_runtime": 1674.8949,
      "train_tokens_per_second": 441.105
    },
    {
      "epoch": 0.3004594507960252,
      "grad_norm": 2.860381603240967,
      "learning_rate": 1.701818181818182e-05,
      "loss": 0.17546290159225464,
      "num_input_tokens_seen": 740052,
      "step": 703,
      "train_runtime": 1677.0548,
      "train_tokens_per_second": 441.281
    },
    {
      "epoch": 0.300886846885351,
      "grad_norm": 3.222641944885254,
      "learning_rate": 1.701390374331551e-05,
      "loss": 0.1763782501220703,
      "num_input_tokens_seen": 741284,
      "step": 704,
      "train_runtime": 1679.1425,
      "train_tokens_per_second": 441.466
    },
    {
      "epoch": 0.3013142429746768,
      "grad_norm": 2.3681511878967285,
      "learning_rate": 1.70096256684492e-05,
      "loss": 0.13529324531555176,
      "num_input_tokens_seen": 742224,
      "step": 705,
      "train_runtime": 1681.2782,
      "train_tokens_per_second": 441.464
    },
    {
      "epoch": 0.3017416390640026,
      "grad_norm": 3.322913646697998,
      "learning_rate": 1.7005347593582888e-05,
      "loss": 0.1992805451154709,
      "num_input_tokens_seen": 743002,
      "step": 706,
      "train_runtime": 1683.3098,
      "train_tokens_per_second": 441.393
    },
    {
      "epoch": 0.3021690351533283,
      "grad_norm": 2.673914909362793,
      "learning_rate": 1.700106951871658e-05,
      "loss": 0.13492314517498016,
      "num_input_tokens_seen": 744234,
      "step": 707,
      "train_runtime": 1685.3949,
      "train_tokens_per_second": 441.578
    },
    {
      "epoch": 0.3025964312426541,
      "grad_norm": 2.9854207038879395,
      "learning_rate": 1.699679144385027e-05,
      "loss": 0.08190875500440598,
      "num_input_tokens_seen": 744844,
      "step": 708,
      "train_runtime": 1687.3481,
      "train_tokens_per_second": 441.429
    },
    {
      "epoch": 0.3030238273319799,
      "grad_norm": 3.064990758895874,
      "learning_rate": 1.699251336898396e-05,
      "loss": 0.1609775424003601,
      "num_input_tokens_seen": 745712,
      "step": 709,
      "train_runtime": 1689.4292,
      "train_tokens_per_second": 441.399
    },
    {
      "epoch": 0.3034512234213057,
      "grad_norm": 4.355386257171631,
      "learning_rate": 1.698823529411765e-05,
      "loss": 0.20793989300727844,
      "num_input_tokens_seen": 746854,
      "step": 710,
      "train_runtime": 1691.5474,
      "train_tokens_per_second": 441.521
    },
    {
      "epoch": 0.3038786195106315,
      "grad_norm": 3.515242338180542,
      "learning_rate": 1.698395721925134e-05,
      "loss": 0.17526139318943024,
      "num_input_tokens_seen": 747604,
      "step": 711,
      "train_runtime": 1693.5558,
      "train_tokens_per_second": 441.44
    },
    {
      "epoch": 0.30430601559995724,
      "grad_norm": 2.7136337757110596,
      "learning_rate": 1.6979679144385028e-05,
      "loss": 0.10106546431779861,
      "num_input_tokens_seen": 748358,
      "step": 712,
      "train_runtime": 1695.5992,
      "train_tokens_per_second": 441.353
    },
    {
      "epoch": 0.30473341168928303,
      "grad_norm": 3.5016796588897705,
      "learning_rate": 1.6975401069518718e-05,
      "loss": 0.1796518713235855,
      "num_input_tokens_seen": 749440,
      "step": 713,
      "train_runtime": 1697.6806,
      "train_tokens_per_second": 441.449
    },
    {
      "epoch": 0.3051608077786088,
      "grad_norm": 3.995384693145752,
      "learning_rate": 1.6971122994652407e-05,
      "loss": 0.17982743680477142,
      "num_input_tokens_seen": 750418,
      "step": 714,
      "train_runtime": 1699.751,
      "train_tokens_per_second": 441.487
    },
    {
      "epoch": 0.3055882038679346,
      "grad_norm": 2.906602382659912,
      "learning_rate": 1.69668449197861e-05,
      "loss": 0.15344633162021637,
      "num_input_tokens_seen": 751646,
      "step": 715,
      "train_runtime": 1701.8386,
      "train_tokens_per_second": 441.667
    },
    {
      "epoch": 0.3060155999572604,
      "grad_norm": 4.614682197570801,
      "learning_rate": 1.696256684491979e-05,
      "loss": 0.18905337154865265,
      "num_input_tokens_seen": 752600,
      "step": 716,
      "train_runtime": 1703.9577,
      "train_tokens_per_second": 441.678
    },
    {
      "epoch": 0.30644299604658615,
      "grad_norm": 4.919049263000488,
      "learning_rate": 1.695828877005348e-05,
      "loss": 0.2674981355667114,
      "num_input_tokens_seen": 753414,
      "step": 717,
      "train_runtime": 1705.999,
      "train_tokens_per_second": 441.626
    },
    {
      "epoch": 0.30687039213591194,
      "grad_norm": 3.67826771736145,
      "learning_rate": 1.6954010695187165e-05,
      "loss": 0.16822344064712524,
      "num_input_tokens_seen": 754518,
      "step": 718,
      "train_runtime": 1708.0614,
      "train_tokens_per_second": 441.739
    },
    {
      "epoch": 0.30729778822523773,
      "grad_norm": 2.349168300628662,
      "learning_rate": 1.6949732620320858e-05,
      "loss": 0.1278461217880249,
      "num_input_tokens_seen": 755588,
      "step": 719,
      "train_runtime": 1710.1309,
      "train_tokens_per_second": 441.831
    },
    {
      "epoch": 0.3077251843145635,
      "grad_norm": 4.480087757110596,
      "learning_rate": 1.6945454545454547e-05,
      "loss": 0.11029095202684402,
      "num_input_tokens_seen": 756476,
      "step": 720,
      "train_runtime": 1712.166,
      "train_tokens_per_second": 441.824
    },
    {
      "epoch": 0.3081525804038893,
      "grad_norm": 2.819782257080078,
      "learning_rate": 1.6941176470588237e-05,
      "loss": 0.09692086279392242,
      "num_input_tokens_seen": 757168,
      "step": 721,
      "train_runtime": 1721.001,
      "train_tokens_per_second": 439.958
    },
    {
      "epoch": 0.3085799764932151,
      "grad_norm": 2.976001262664795,
      "learning_rate": 1.6936898395721926e-05,
      "loss": 0.16137419641017914,
      "num_input_tokens_seen": 758160,
      "step": 722,
      "train_runtime": 1723.127,
      "train_tokens_per_second": 439.991
    },
    {
      "epoch": 0.30900737258254085,
      "grad_norm": 3.4080145359039307,
      "learning_rate": 1.693262032085562e-05,
      "loss": 0.1251823902130127,
      "num_input_tokens_seen": 758940,
      "step": 723,
      "train_runtime": 1725.1264,
      "train_tokens_per_second": 439.933
    },
    {
      "epoch": 0.30943476867186664,
      "grad_norm": 2.275359630584717,
      "learning_rate": 1.6928342245989305e-05,
      "loss": 0.1225162148475647,
      "num_input_tokens_seen": 760078,
      "step": 724,
      "train_runtime": 1727.2558,
      "train_tokens_per_second": 440.049
    },
    {
      "epoch": 0.30986216476119244,
      "grad_norm": 3.0035042762756348,
      "learning_rate": 1.6924064171122994e-05,
      "loss": 0.12130562961101532,
      "num_input_tokens_seen": 761064,
      "step": 725,
      "train_runtime": 1729.3499,
      "train_tokens_per_second": 440.087
    },
    {
      "epoch": 0.31028956085051823,
      "grad_norm": 2.424938917160034,
      "learning_rate": 1.6919786096256684e-05,
      "loss": 0.13783925771713257,
      "num_input_tokens_seen": 762414,
      "step": 726,
      "train_runtime": 1731.4569,
      "train_tokens_per_second": 440.331
    },
    {
      "epoch": 0.310716956939844,
      "grad_norm": 2.9705116748809814,
      "learning_rate": 1.6915508021390377e-05,
      "loss": 0.1928931176662445,
      "num_input_tokens_seen": 763692,
      "step": 727,
      "train_runtime": 1733.5347,
      "train_tokens_per_second": 440.54
    },
    {
      "epoch": 0.31114435302916976,
      "grad_norm": 3.2809789180755615,
      "learning_rate": 1.6911229946524066e-05,
      "loss": 0.18057897686958313,
      "num_input_tokens_seen": 764738,
      "step": 728,
      "train_runtime": 1735.5868,
      "train_tokens_per_second": 440.622
    },
    {
      "epoch": 0.31157174911849556,
      "grad_norm": 3.0725319385528564,
      "learning_rate": 1.6906951871657755e-05,
      "loss": 0.13870379328727722,
      "num_input_tokens_seen": 766040,
      "step": 729,
      "train_runtime": 1737.7003,
      "train_tokens_per_second": 440.836
    },
    {
      "epoch": 0.31199914520782135,
      "grad_norm": 3.658844232559204,
      "learning_rate": 1.6902673796791445e-05,
      "loss": 0.18355195224285126,
      "num_input_tokens_seen": 766918,
      "step": 730,
      "train_runtime": 1739.7138,
      "train_tokens_per_second": 440.83
    },
    {
      "epoch": 0.31242654129714714,
      "grad_norm": 4.367164611816406,
      "learning_rate": 1.6898395721925134e-05,
      "loss": 0.11822845041751862,
      "num_input_tokens_seen": 767498,
      "step": 731,
      "train_runtime": 1741.6718,
      "train_tokens_per_second": 440.667
    },
    {
      "epoch": 0.31285393738647294,
      "grad_norm": 2.4613184928894043,
      "learning_rate": 1.6894117647058824e-05,
      "loss": 0.12785813212394714,
      "num_input_tokens_seen": 768748,
      "step": 732,
      "train_runtime": 1743.887,
      "train_tokens_per_second": 440.824
    },
    {
      "epoch": 0.3132813334757987,
      "grad_norm": 2.3444745540618896,
      "learning_rate": 1.6889839572192513e-05,
      "loss": 0.12883754074573517,
      "num_input_tokens_seen": 769808,
      "step": 733,
      "train_runtime": 1745.9426,
      "train_tokens_per_second": 440.913
    },
    {
      "epoch": 0.31370872956512447,
      "grad_norm": 3.0968801975250244,
      "learning_rate": 1.6885561497326203e-05,
      "loss": 0.17797592282295227,
      "num_input_tokens_seen": 770774,
      "step": 734,
      "train_runtime": 1747.9777,
      "train_tokens_per_second": 440.952
    },
    {
      "epoch": 0.31413612565445026,
      "grad_norm": 3.5281708240509033,
      "learning_rate": 1.6881283422459896e-05,
      "loss": 0.13884249329566956,
      "num_input_tokens_seen": 771764,
      "step": 735,
      "train_runtime": 1750.0652,
      "train_tokens_per_second": 440.992
    },
    {
      "epoch": 0.31456352174377605,
      "grad_norm": 7.061663627624512,
      "learning_rate": 1.6877005347593585e-05,
      "loss": 0.16773094236850739,
      "num_input_tokens_seen": 773010,
      "step": 736,
      "train_runtime": 1752.191,
      "train_tokens_per_second": 441.168
    },
    {
      "epoch": 0.31499091783310185,
      "grad_norm": 2.613240957260132,
      "learning_rate": 1.6872727272727274e-05,
      "loss": 0.13964132964611053,
      "num_input_tokens_seen": 773894,
      "step": 737,
      "train_runtime": 1754.2167,
      "train_tokens_per_second": 441.162
    },
    {
      "epoch": 0.3154183139224276,
      "grad_norm": 2.094176769256592,
      "learning_rate": 1.6868449197860964e-05,
      "loss": 0.1435990333557129,
      "num_input_tokens_seen": 775084,
      "step": 738,
      "train_runtime": 1756.3029,
      "train_tokens_per_second": 441.316
    },
    {
      "epoch": 0.3158457100117534,
      "grad_norm": 3.6297993659973145,
      "learning_rate": 1.6864171122994653e-05,
      "loss": 0.1722583919763565,
      "num_input_tokens_seen": 776132,
      "step": 739,
      "train_runtime": 1758.3479,
      "train_tokens_per_second": 441.398
    },
    {
      "epoch": 0.31627310610107917,
      "grad_norm": 4.502858638763428,
      "learning_rate": 1.6859893048128343e-05,
      "loss": 0.1423422396183014,
      "num_input_tokens_seen": 776914,
      "step": 740,
      "train_runtime": 1760.3029,
      "train_tokens_per_second": 441.352
    },
    {
      "epoch": 0.31670050219040496,
      "grad_norm": 2.884168863296509,
      "learning_rate": 1.6855614973262032e-05,
      "loss": 0.09878357499837875,
      "num_input_tokens_seen": 777732,
      "step": 741,
      "train_runtime": 1762.2569,
      "train_tokens_per_second": 441.327
    },
    {
      "epoch": 0.31712789827973076,
      "grad_norm": 2.4770572185516357,
      "learning_rate": 1.685133689839572e-05,
      "loss": 0.1018742173910141,
      "num_input_tokens_seen": 778950,
      "step": 742,
      "train_runtime": 1764.3422,
      "train_tokens_per_second": 441.496
    },
    {
      "epoch": 0.3175552943690565,
      "grad_norm": 3.128291606903076,
      "learning_rate": 1.6847058823529414e-05,
      "loss": 0.20683753490447998,
      "num_input_tokens_seen": 780304,
      "step": 743,
      "train_runtime": 1766.4888,
      "train_tokens_per_second": 441.726
    },
    {
      "epoch": 0.3179826904583823,
      "grad_norm": 3.732107162475586,
      "learning_rate": 1.6842780748663104e-05,
      "loss": 0.23228523135185242,
      "num_input_tokens_seen": 781224,
      "step": 744,
      "train_runtime": 1768.5073,
      "train_tokens_per_second": 441.742
    },
    {
      "epoch": 0.3184100865477081,
      "grad_norm": 3.751624345779419,
      "learning_rate": 1.6838502673796793e-05,
      "loss": 0.14719495177268982,
      "num_input_tokens_seen": 782016,
      "step": 745,
      "train_runtime": 1770.531,
      "train_tokens_per_second": 441.684
    },
    {
      "epoch": 0.3188374826370339,
      "grad_norm": 2.419607639312744,
      "learning_rate": 1.6834224598930483e-05,
      "loss": 0.11479254812002182,
      "num_input_tokens_seen": 783148,
      "step": 746,
      "train_runtime": 1772.6168,
      "train_tokens_per_second": 441.803
    },
    {
      "epoch": 0.31926487872635967,
      "grad_norm": 3.030592918395996,
      "learning_rate": 1.6829946524064172e-05,
      "loss": 0.10850606858730316,
      "num_input_tokens_seen": 784318,
      "step": 747,
      "train_runtime": 1774.7043,
      "train_tokens_per_second": 441.943
    },
    {
      "epoch": 0.31969227481568546,
      "grad_norm": 4.8611955642700195,
      "learning_rate": 1.6825668449197862e-05,
      "loss": 0.15850195288658142,
      "num_input_tokens_seen": 785114,
      "step": 748,
      "train_runtime": 1776.6957,
      "train_tokens_per_second": 441.896
    },
    {
      "epoch": 0.3201196709050112,
      "grad_norm": 3.4313971996307373,
      "learning_rate": 1.682139037433155e-05,
      "loss": 0.14053592085838318,
      "num_input_tokens_seen": 786298,
      "step": 749,
      "train_runtime": 1778.7778,
      "train_tokens_per_second": 442.044
    },
    {
      "epoch": 0.320547066994337,
      "grad_norm": 3.6184823513031006,
      "learning_rate": 1.681711229946524e-05,
      "loss": 0.1352541148662567,
      "num_input_tokens_seen": 787154,
      "step": 750,
      "train_runtime": 1780.7809,
      "train_tokens_per_second": 442.027
    },
    {
      "epoch": 0.3209744630836628,
      "grad_norm": 3.11987566947937,
      "learning_rate": 1.6812834224598933e-05,
      "loss": 0.18279433250427246,
      "num_input_tokens_seen": 788344,
      "step": 751,
      "train_runtime": 1789.2084,
      "train_tokens_per_second": 440.61
    },
    {
      "epoch": 0.3214018591729886,
      "grad_norm": 4.212590217590332,
      "learning_rate": 1.6808556149732623e-05,
      "loss": 0.17546242475509644,
      "num_input_tokens_seen": 789102,
      "step": 752,
      "train_runtime": 1791.1829,
      "train_tokens_per_second": 440.548
    },
    {
      "epoch": 0.3218292552623144,
      "grad_norm": 2.880328416824341,
      "learning_rate": 1.6804278074866312e-05,
      "loss": 0.15373101830482483,
      "num_input_tokens_seen": 789858,
      "step": 753,
      "train_runtime": 1793.159,
      "train_tokens_per_second": 440.484
    },
    {
      "epoch": 0.3222566513516401,
      "grad_norm": 3.7703986167907715,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.17377933859825134,
      "num_input_tokens_seen": 790640,
      "step": 754,
      "train_runtime": 1795.177,
      "train_tokens_per_second": 440.425
    },
    {
      "epoch": 0.3226840474409659,
      "grad_norm": 2.64025616645813,
      "learning_rate": 1.679572192513369e-05,
      "loss": 0.1529747098684311,
      "num_input_tokens_seen": 792164,
      "step": 755,
      "train_runtime": 1797.3855,
      "train_tokens_per_second": 440.731
    },
    {
      "epoch": 0.3231114435302917,
      "grad_norm": 3.908766269683838,
      "learning_rate": 1.679144385026738e-05,
      "loss": 0.19651946425437927,
      "num_input_tokens_seen": 793302,
      "step": 756,
      "train_runtime": 1799.5136,
      "train_tokens_per_second": 440.842
    },
    {
      "epoch": 0.3235388396196175,
      "grad_norm": 2.4419639110565186,
      "learning_rate": 1.678716577540107e-05,
      "loss": 0.08055537939071655,
      "num_input_tokens_seen": 794362,
      "step": 757,
      "train_runtime": 1801.6083,
      "train_tokens_per_second": 440.918
    },
    {
      "epoch": 0.3239662357089433,
      "grad_norm": 4.166549205780029,
      "learning_rate": 1.678288770053476e-05,
      "loss": 0.18417219817638397,
      "num_input_tokens_seen": 795272,
      "step": 758,
      "train_runtime": 1803.6547,
      "train_tokens_per_second": 440.923
    },
    {
      "epoch": 0.324393631798269,
      "grad_norm": 3.282649517059326,
      "learning_rate": 1.6778609625668452e-05,
      "loss": 0.12055408209562302,
      "num_input_tokens_seen": 796026,
      "step": 759,
      "train_runtime": 1805.6579,
      "train_tokens_per_second": 440.851
    },
    {
      "epoch": 0.3248210278875948,
      "grad_norm": 2.189688205718994,
      "learning_rate": 1.6774331550802142e-05,
      "loss": 0.14737199246883392,
      "num_input_tokens_seen": 797736,
      "step": 760,
      "train_runtime": 1807.8678,
      "train_tokens_per_second": 441.258
    },
    {
      "epoch": 0.3252484239769206,
      "grad_norm": 3.0677638053894043,
      "learning_rate": 1.677005347593583e-05,
      "loss": 0.11574704945087433,
      "num_input_tokens_seen": 798690,
      "step": 761,
      "train_runtime": 1809.908,
      "train_tokens_per_second": 441.288
    },
    {
      "epoch": 0.3256758200662464,
      "grad_norm": 3.0081984996795654,
      "learning_rate": 1.676577540106952e-05,
      "loss": 0.21783174574375153,
      "num_input_tokens_seen": 799628,
      "step": 762,
      "train_runtime": 1811.9379,
      "train_tokens_per_second": 441.311
    },
    {
      "epoch": 0.3261032161555722,
      "grad_norm": 2.4546148777008057,
      "learning_rate": 1.676149732620321e-05,
      "loss": 0.14510956406593323,
      "num_input_tokens_seen": 800854,
      "step": 763,
      "train_runtime": 1814.0349,
      "train_tokens_per_second": 441.477
    },
    {
      "epoch": 0.32653061224489793,
      "grad_norm": 3.125731945037842,
      "learning_rate": 1.67572192513369e-05,
      "loss": 0.11249782145023346,
      "num_input_tokens_seen": 801748,
      "step": 764,
      "train_runtime": 1816.1485,
      "train_tokens_per_second": 441.455
    },
    {
      "epoch": 0.3269580083342237,
      "grad_norm": 2.754300355911255,
      "learning_rate": 1.675294117647059e-05,
      "loss": 0.13277259469032288,
      "num_input_tokens_seen": 802744,
      "step": 765,
      "train_runtime": 1818.2273,
      "train_tokens_per_second": 441.498
    },
    {
      "epoch": 0.3273854044235495,
      "grad_norm": 2.4079010486602783,
      "learning_rate": 1.674866310160428e-05,
      "loss": 0.07246275991201401,
      "num_input_tokens_seen": 803590,
      "step": 766,
      "train_runtime": 1820.2412,
      "train_tokens_per_second": 441.474
    },
    {
      "epoch": 0.3278128005128753,
      "grad_norm": 3.0190322399139404,
      "learning_rate": 1.674438502673797e-05,
      "loss": 0.1391064077615738,
      "num_input_tokens_seen": 804376,
      "step": 767,
      "train_runtime": 1822.3268,
      "train_tokens_per_second": 441.401
    },
    {
      "epoch": 0.3282401966022011,
      "grad_norm": 2.4388716220855713,
      "learning_rate": 1.674010695187166e-05,
      "loss": 0.13550874590873718,
      "num_input_tokens_seen": 805452,
      "step": 768,
      "train_runtime": 1824.448,
      "train_tokens_per_second": 441.477
    },
    {
      "epoch": 0.3286675926915269,
      "grad_norm": 3.1067779064178467,
      "learning_rate": 1.673582887700535e-05,
      "loss": 0.20665422081947327,
      "num_input_tokens_seen": 806856,
      "step": 769,
      "train_runtime": 1826.6087,
      "train_tokens_per_second": 441.723
    },
    {
      "epoch": 0.32909498878085264,
      "grad_norm": 3.392644166946411,
      "learning_rate": 1.673155080213904e-05,
      "loss": 0.16381341218948364,
      "num_input_tokens_seen": 807632,
      "step": 770,
      "train_runtime": 1828.6333,
      "train_tokens_per_second": 441.659
    },
    {
      "epoch": 0.32952238487017843,
      "grad_norm": 3.523287057876587,
      "learning_rate": 1.672727272727273e-05,
      "loss": 0.13788162171840668,
      "num_input_tokens_seen": 809556,
      "step": 771,
      "train_runtime": 1830.9039,
      "train_tokens_per_second": 442.162
    },
    {
      "epoch": 0.3299497809595042,
      "grad_norm": 5.4983720779418945,
      "learning_rate": 1.672299465240642e-05,
      "loss": 0.1619395762681961,
      "num_input_tokens_seen": 810648,
      "step": 772,
      "train_runtime": 1832.9782,
      "train_tokens_per_second": 442.257
    },
    {
      "epoch": 0.33037717704883,
      "grad_norm": 2.9557442665100098,
      "learning_rate": 1.6718716577540108e-05,
      "loss": 0.13064926862716675,
      "num_input_tokens_seen": 811464,
      "step": 773,
      "train_runtime": 1834.9918,
      "train_tokens_per_second": 442.217
    },
    {
      "epoch": 0.3308045731381558,
      "grad_norm": 3.253335952758789,
      "learning_rate": 1.6714438502673797e-05,
      "loss": 0.14006176590919495,
      "num_input_tokens_seen": 812348,
      "step": 774,
      "train_runtime": 1837.0877,
      "train_tokens_per_second": 442.193
    },
    {
      "epoch": 0.33123196922748155,
      "grad_norm": 2.5818698406219482,
      "learning_rate": 1.6710160427807487e-05,
      "loss": 0.12496156990528107,
      "num_input_tokens_seen": 813656,
      "step": 775,
      "train_runtime": 1839.2479,
      "train_tokens_per_second": 442.385
    },
    {
      "epoch": 0.33165936531680734,
      "grad_norm": 3.6411750316619873,
      "learning_rate": 1.670588235294118e-05,
      "loss": 0.1782451868057251,
      "num_input_tokens_seen": 814622,
      "step": 776,
      "train_runtime": 1841.3332,
      "train_tokens_per_second": 442.409
    },
    {
      "epoch": 0.33208676140613314,
      "grad_norm": 3.092738389968872,
      "learning_rate": 1.670160427807487e-05,
      "loss": 0.1875651329755783,
      "num_input_tokens_seen": 815802,
      "step": 777,
      "train_runtime": 1843.4295,
      "train_tokens_per_second": 442.546
    },
    {
      "epoch": 0.33251415749545893,
      "grad_norm": 3.4639575481414795,
      "learning_rate": 1.669732620320856e-05,
      "loss": 0.15780915319919586,
      "num_input_tokens_seen": 816674,
      "step": 778,
      "train_runtime": 1845.5042,
      "train_tokens_per_second": 442.521
    },
    {
      "epoch": 0.3329415535847847,
      "grad_norm": 2.5314581394195557,
      "learning_rate": 1.6693048128342245e-05,
      "loss": 0.15358608961105347,
      "num_input_tokens_seen": 817948,
      "step": 779,
      "train_runtime": 1847.6382,
      "train_tokens_per_second": 442.699
    },
    {
      "epoch": 0.33336894967411046,
      "grad_norm": 2.615727424621582,
      "learning_rate": 1.6688770053475938e-05,
      "loss": 0.1357804834842682,
      "num_input_tokens_seen": 819274,
      "step": 780,
      "train_runtime": 1849.7877,
      "train_tokens_per_second": 442.902
    },
    {
      "epoch": 0.33379634576343625,
      "grad_norm": 3.8525586128234863,
      "learning_rate": 1.6684491978609627e-05,
      "loss": 0.17240309715270996,
      "num_input_tokens_seen": 820162,
      "step": 781,
      "train_runtime": 1858.3713,
      "train_tokens_per_second": 441.334
    },
    {
      "epoch": 0.33422374185276205,
      "grad_norm": 4.191407203674316,
      "learning_rate": 1.6680213903743316e-05,
      "loss": 0.23687896132469177,
      "num_input_tokens_seen": 820874,
      "step": 782,
      "train_runtime": 1860.4096,
      "train_tokens_per_second": 441.233
    },
    {
      "epoch": 0.33465113794208784,
      "grad_norm": 2.6943485736846924,
      "learning_rate": 1.6675935828877006e-05,
      "loss": 0.16705572605133057,
      "num_input_tokens_seen": 822422,
      "step": 783,
      "train_runtime": 1862.6224,
      "train_tokens_per_second": 441.54
    },
    {
      "epoch": 0.33507853403141363,
      "grad_norm": 1.860338807106018,
      "learning_rate": 1.66716577540107e-05,
      "loss": 0.07211625576019287,
      "num_input_tokens_seen": 824162,
      "step": 784,
      "train_runtime": 1864.8986,
      "train_tokens_per_second": 441.934
    },
    {
      "epoch": 0.33550593012073937,
      "grad_norm": 2.996925115585327,
      "learning_rate": 1.6667379679144388e-05,
      "loss": 0.13057556748390198,
      "num_input_tokens_seen": 825056,
      "step": 785,
      "train_runtime": 1866.9345,
      "train_tokens_per_second": 441.931
    },
    {
      "epoch": 0.33593332621006516,
      "grad_norm": 3.4369115829467773,
      "learning_rate": 1.6663101604278074e-05,
      "loss": 0.19953583180904388,
      "num_input_tokens_seen": 826424,
      "step": 786,
      "train_runtime": 1869.11,
      "train_tokens_per_second": 442.148
    },
    {
      "epoch": 0.33636072229939096,
      "grad_norm": 2.488295793533325,
      "learning_rate": 1.6658823529411764e-05,
      "loss": 0.1348193883895874,
      "num_input_tokens_seen": 827710,
      "step": 787,
      "train_runtime": 1871.2625,
      "train_tokens_per_second": 442.327
    },
    {
      "epoch": 0.33678811838871675,
      "grad_norm": 2.3568084239959717,
      "learning_rate": 1.6654545454545456e-05,
      "loss": 0.11009471118450165,
      "num_input_tokens_seen": 828636,
      "step": 788,
      "train_runtime": 1873.3483,
      "train_tokens_per_second": 442.329
    },
    {
      "epoch": 0.33721551447804254,
      "grad_norm": 3.801801919937134,
      "learning_rate": 1.6650267379679146e-05,
      "loss": 0.14569516479969025,
      "num_input_tokens_seen": 829302,
      "step": 789,
      "train_runtime": 1875.3026,
      "train_tokens_per_second": 442.223
    },
    {
      "epoch": 0.3376429105673683,
      "grad_norm": 5.081080913543701,
      "learning_rate": 1.6645989304812835e-05,
      "loss": 0.12997569143772125,
      "num_input_tokens_seen": 829892,
      "step": 790,
      "train_runtime": 1877.2649,
      "train_tokens_per_second": 442.075
    },
    {
      "epoch": 0.3380703066566941,
      "grad_norm": 3.5495529174804688,
      "learning_rate": 1.6641711229946525e-05,
      "loss": 0.16700680553913116,
      "num_input_tokens_seen": 830962,
      "step": 791,
      "train_runtime": 1879.3438,
      "train_tokens_per_second": 442.155
    },
    {
      "epoch": 0.33849770274601987,
      "grad_norm": 2.476435661315918,
      "learning_rate": 1.6637433155080214e-05,
      "loss": 0.09858560562133789,
      "num_input_tokens_seen": 831794,
      "step": 792,
      "train_runtime": 1881.4071,
      "train_tokens_per_second": 442.113
    },
    {
      "epoch": 0.33892509883534566,
      "grad_norm": 3.570925712585449,
      "learning_rate": 1.6633155080213904e-05,
      "loss": 0.08923085033893585,
      "num_input_tokens_seen": 832656,
      "step": 793,
      "train_runtime": 1883.3909,
      "train_tokens_per_second": 442.105
    },
    {
      "epoch": 0.33935249492467146,
      "grad_norm": 2.364238977432251,
      "learning_rate": 1.6628877005347593e-05,
      "loss": 0.11550764739513397,
      "num_input_tokens_seen": 834142,
      "step": 794,
      "train_runtime": 1885.815,
      "train_tokens_per_second": 442.324
    },
    {
      "epoch": 0.33977989101399725,
      "grad_norm": 3.608776807785034,
      "learning_rate": 1.6624598930481283e-05,
      "loss": 0.18349257111549377,
      "num_input_tokens_seen": 834814,
      "step": 795,
      "train_runtime": 1887.7806,
      "train_tokens_per_second": 442.22
    },
    {
      "epoch": 0.340207287103323,
      "grad_norm": 4.224783420562744,
      "learning_rate": 1.6620320855614975e-05,
      "loss": 0.15702037513256073,
      "num_input_tokens_seen": 835578,
      "step": 796,
      "train_runtime": 1889.8057,
      "train_tokens_per_second": 442.15
    },
    {
      "epoch": 0.3406346831926488,
      "grad_norm": 2.4544620513916016,
      "learning_rate": 1.6616042780748665e-05,
      "loss": 0.11303136497735977,
      "num_input_tokens_seen": 836666,
      "step": 797,
      "train_runtime": 1891.9705,
      "train_tokens_per_second": 442.219
    },
    {
      "epoch": 0.3410620792819746,
      "grad_norm": 2.1436290740966797,
      "learning_rate": 1.6611764705882354e-05,
      "loss": 0.13750866055488586,
      "num_input_tokens_seen": 837820,
      "step": 798,
      "train_runtime": 1894.1037,
      "train_tokens_per_second": 442.331
    },
    {
      "epoch": 0.34148947537130037,
      "grad_norm": 3.9511640071868896,
      "learning_rate": 1.6607486631016044e-05,
      "loss": 0.16122567653656006,
      "num_input_tokens_seen": 838552,
      "step": 799,
      "train_runtime": 1896.1225,
      "train_tokens_per_second": 442.246
    },
    {
      "epoch": 0.34191687146062616,
      "grad_norm": 3.582397937774658,
      "learning_rate": 1.6603208556149733e-05,
      "loss": 0.1547728180885315,
      "num_input_tokens_seen": 839356,
      "step": 800,
      "train_runtime": 1898.1443,
      "train_tokens_per_second": 442.198
    },
    {
      "epoch": 0.3423442675499519,
      "grad_norm": 5.425139427185059,
      "learning_rate": 1.6598930481283423e-05,
      "loss": 0.1871921718120575,
      "num_input_tokens_seen": 839988,
      "step": 801,
      "train_runtime": 1900.1262,
      "train_tokens_per_second": 442.07
    },
    {
      "epoch": 0.3427716636392777,
      "grad_norm": 2.7834746837615967,
      "learning_rate": 1.6594652406417112e-05,
      "loss": 0.15691137313842773,
      "num_input_tokens_seen": 840830,
      "step": 802,
      "train_runtime": 1902.1987,
      "train_tokens_per_second": 442.031
    },
    {
      "epoch": 0.3431990597286035,
      "grad_norm": 3.048058271408081,
      "learning_rate": 1.65903743315508e-05,
      "loss": 0.14582957327365875,
      "num_input_tokens_seen": 841794,
      "step": 803,
      "train_runtime": 1904.2632,
      "train_tokens_per_second": 442.058
    },
    {
      "epoch": 0.3436264558179293,
      "grad_norm": 3.980039596557617,
      "learning_rate": 1.6586096256684494e-05,
      "loss": 0.0996789038181305,
      "num_input_tokens_seen": 842524,
      "step": 804,
      "train_runtime": 1906.26,
      "train_tokens_per_second": 441.977
    },
    {
      "epoch": 0.34405385190725507,
      "grad_norm": 2.8752737045288086,
      "learning_rate": 1.6581818181818184e-05,
      "loss": 0.24773788452148438,
      "num_input_tokens_seen": 844360,
      "step": 805,
      "train_runtime": 1908.5343,
      "train_tokens_per_second": 442.413
    },
    {
      "epoch": 0.3444812479965808,
      "grad_norm": 3.2004787921905518,
      "learning_rate": 1.6577540106951873e-05,
      "loss": 0.2751487195491791,
      "num_input_tokens_seen": 846190,
      "step": 806,
      "train_runtime": 1910.7811,
      "train_tokens_per_second": 442.85
    },
    {
      "epoch": 0.3449086440859066,
      "grad_norm": 3.3579134941101074,
      "learning_rate": 1.6573262032085563e-05,
      "loss": 0.18909302353858948,
      "num_input_tokens_seen": 847272,
      "step": 807,
      "train_runtime": 1912.8623,
      "train_tokens_per_second": 442.934
    },
    {
      "epoch": 0.3453360401752324,
      "grad_norm": 2.9090867042541504,
      "learning_rate": 1.6568983957219252e-05,
      "loss": 0.10554330050945282,
      "num_input_tokens_seen": 848110,
      "step": 808,
      "train_runtime": 1914.9074,
      "train_tokens_per_second": 442.899
    },
    {
      "epoch": 0.3457634362645582,
      "grad_norm": 2.701162576675415,
      "learning_rate": 1.656470588235294e-05,
      "loss": 0.1586577296257019,
      "num_input_tokens_seen": 849084,
      "step": 809,
      "train_runtime": 1916.9359,
      "train_tokens_per_second": 442.938
    },
    {
      "epoch": 0.346190832353884,
      "grad_norm": 1.7255891561508179,
      "learning_rate": 1.656042780748663e-05,
      "loss": 0.0649508461356163,
      "num_input_tokens_seen": 850498,
      "step": 810,
      "train_runtime": 1919.0832,
      "train_tokens_per_second": 443.179
    },
    {
      "epoch": 0.3466182284432097,
      "grad_norm": 3.123565912246704,
      "learning_rate": 1.655614973262032e-05,
      "loss": 0.1544187217950821,
      "num_input_tokens_seen": 851400,
      "step": 811,
      "train_runtime": 1927.6204,
      "train_tokens_per_second": 441.684
    },
    {
      "epoch": 0.3470456245325355,
      "grad_norm": 3.224686861038208,
      "learning_rate": 1.6551871657754013e-05,
      "loss": 0.18803249299526215,
      "num_input_tokens_seen": 853148,
      "step": 812,
      "train_runtime": 1929.8742,
      "train_tokens_per_second": 442.074
    },
    {
      "epoch": 0.3474730206218613,
      "grad_norm": 3.279426097869873,
      "learning_rate": 1.6547593582887703e-05,
      "loss": 0.14644157886505127,
      "num_input_tokens_seen": 853884,
      "step": 813,
      "train_runtime": 1931.8241,
      "train_tokens_per_second": 442.009
    },
    {
      "epoch": 0.3479004167111871,
      "grad_norm": 2.5244524478912354,
      "learning_rate": 1.6543315508021392e-05,
      "loss": 0.1774372160434723,
      "num_input_tokens_seen": 855732,
      "step": 814,
      "train_runtime": 1934.0499,
      "train_tokens_per_second": 442.456
    },
    {
      "epoch": 0.3483278128005129,
      "grad_norm": 2.610222816467285,
      "learning_rate": 1.653903743315508e-05,
      "loss": 0.13001713156700134,
      "num_input_tokens_seen": 856782,
      "step": 815,
      "train_runtime": 1936.049,
      "train_tokens_per_second": 442.541
    },
    {
      "epoch": 0.34875520888983863,
      "grad_norm": 2.8688743114471436,
      "learning_rate": 1.653475935828877e-05,
      "loss": 0.1396550089120865,
      "num_input_tokens_seen": 857708,
      "step": 816,
      "train_runtime": 1938.091,
      "train_tokens_per_second": 442.553
    },
    {
      "epoch": 0.3491826049791644,
      "grad_norm": 4.757078647613525,
      "learning_rate": 1.653048128342246e-05,
      "loss": 0.23061591386795044,
      "num_input_tokens_seen": 858566,
      "step": 817,
      "train_runtime": 1940.1444,
      "train_tokens_per_second": 442.527
    },
    {
      "epoch": 0.3496100010684902,
      "grad_norm": 2.236754894256592,
      "learning_rate": 1.652620320855615e-05,
      "loss": 0.10535576194524765,
      "num_input_tokens_seen": 859472,
      "step": 818,
      "train_runtime": 1942.148,
      "train_tokens_per_second": 442.537
    },
    {
      "epoch": 0.350037397157816,
      "grad_norm": 3.769700288772583,
      "learning_rate": 1.652192513368984e-05,
      "loss": 0.1694517880678177,
      "num_input_tokens_seen": 860164,
      "step": 819,
      "train_runtime": 1944.1443,
      "train_tokens_per_second": 442.438
    },
    {
      "epoch": 0.3504647932471418,
      "grad_norm": 3.2518036365509033,
      "learning_rate": 1.6517647058823532e-05,
      "loss": 0.17297501862049103,
      "num_input_tokens_seen": 861046,
      "step": 820,
      "train_runtime": 1946.1328,
      "train_tokens_per_second": 442.439
    },
    {
      "epoch": 0.3508921893364676,
      "grad_norm": 2.8238017559051514,
      "learning_rate": 1.6513368983957222e-05,
      "loss": 0.1672133207321167,
      "num_input_tokens_seen": 862272,
      "step": 821,
      "train_runtime": 1948.2436,
      "train_tokens_per_second": 442.589
    },
    {
      "epoch": 0.35131958542579333,
      "grad_norm": 3.0380592346191406,
      "learning_rate": 1.650909090909091e-05,
      "loss": 0.08867887407541275,
      "num_input_tokens_seen": 862822,
      "step": 822,
      "train_runtime": 1950.1842,
      "train_tokens_per_second": 442.431
    },
    {
      "epoch": 0.35174698151511913,
      "grad_norm": 2.7388217449188232,
      "learning_rate": 1.65048128342246e-05,
      "loss": 0.13384008407592773,
      "num_input_tokens_seen": 863764,
      "step": 823,
      "train_runtime": 1952.2342,
      "train_tokens_per_second": 442.449
    },
    {
      "epoch": 0.3521743776044449,
      "grad_norm": 3.4671761989593506,
      "learning_rate": 1.650053475935829e-05,
      "loss": 0.1287350356578827,
      "num_input_tokens_seen": 864614,
      "step": 824,
      "train_runtime": 1954.3744,
      "train_tokens_per_second": 442.399
    },
    {
      "epoch": 0.3526017736937707,
      "grad_norm": 2.7498767375946045,
      "learning_rate": 1.649625668449198e-05,
      "loss": 0.13570083677768707,
      "num_input_tokens_seen": 865604,
      "step": 825,
      "train_runtime": 1956.4799,
      "train_tokens_per_second": 442.429
    },
    {
      "epoch": 0.3530291697830965,
      "grad_norm": 2.897430896759033,
      "learning_rate": 1.649197860962567e-05,
      "loss": 0.19579720497131348,
      "num_input_tokens_seen": 866716,
      "step": 826,
      "train_runtime": 1958.5376,
      "train_tokens_per_second": 442.532
    },
    {
      "epoch": 0.35345656587242225,
      "grad_norm": 2.8864974975585938,
      "learning_rate": 1.648770053475936e-05,
      "loss": 0.19011685252189636,
      "num_input_tokens_seen": 868744,
      "step": 827,
      "train_runtime": 1960.8749,
      "train_tokens_per_second": 443.039
    },
    {
      "epoch": 0.35388396196174804,
      "grad_norm": 3.6182870864868164,
      "learning_rate": 1.648342245989305e-05,
      "loss": 0.12158944457769394,
      "num_input_tokens_seen": 869632,
      "step": 828,
      "train_runtime": 1962.9246,
      "train_tokens_per_second": 443.029
    },
    {
      "epoch": 0.35431135805107383,
      "grad_norm": 4.422471046447754,
      "learning_rate": 1.647914438502674e-05,
      "loss": 0.18166618049144745,
      "num_input_tokens_seen": 870562,
      "step": 829,
      "train_runtime": 1964.9757,
      "train_tokens_per_second": 443.04
    },
    {
      "epoch": 0.3547387541403996,
      "grad_norm": 3.618467092514038,
      "learning_rate": 1.647486631016043e-05,
      "loss": 0.1652064025402069,
      "num_input_tokens_seen": 871682,
      "step": 830,
      "train_runtime": 1967.0706,
      "train_tokens_per_second": 443.137
    },
    {
      "epoch": 0.3551661502297254,
      "grad_norm": 4.541975498199463,
      "learning_rate": 1.647058823529412e-05,
      "loss": 0.16861656308174133,
      "num_input_tokens_seen": 872416,
      "step": 831,
      "train_runtime": 1969.025,
      "train_tokens_per_second": 443.07
    },
    {
      "epoch": 0.35559354631905116,
      "grad_norm": 3.375227928161621,
      "learning_rate": 1.646631016042781e-05,
      "loss": 0.18978804349899292,
      "num_input_tokens_seen": 873606,
      "step": 832,
      "train_runtime": 1971.1167,
      "train_tokens_per_second": 443.204
    },
    {
      "epoch": 0.35602094240837695,
      "grad_norm": 3.7770273685455322,
      "learning_rate": 1.64620320855615e-05,
      "loss": 0.2243930995464325,
      "num_input_tokens_seen": 874724,
      "step": 833,
      "train_runtime": 1973.1972,
      "train_tokens_per_second": 443.303
    },
    {
      "epoch": 0.35644833849770274,
      "grad_norm": 4.624592304229736,
      "learning_rate": 1.6457754010695188e-05,
      "loss": 0.1322430819272995,
      "num_input_tokens_seen": 875570,
      "step": 834,
      "train_runtime": 1975.2638,
      "train_tokens_per_second": 443.267
    },
    {
      "epoch": 0.35687573458702854,
      "grad_norm": 2.136991262435913,
      "learning_rate": 1.6453475935828877e-05,
      "loss": 0.12605592608451843,
      "num_input_tokens_seen": 876702,
      "step": 835,
      "train_runtime": 1977.3183,
      "train_tokens_per_second": 443.379
    },
    {
      "epoch": 0.35730313067635433,
      "grad_norm": 3.0726447105407715,
      "learning_rate": 1.644919786096257e-05,
      "loss": 0.11882643401622772,
      "num_input_tokens_seen": 877628,
      "step": 836,
      "train_runtime": 1979.3446,
      "train_tokens_per_second": 443.393
    },
    {
      "epoch": 0.35773052676568007,
      "grad_norm": 2.1071488857269287,
      "learning_rate": 1.644491978609626e-05,
      "loss": 0.05799455940723419,
      "num_input_tokens_seen": 878390,
      "step": 837,
      "train_runtime": 1981.3652,
      "train_tokens_per_second": 443.326
    },
    {
      "epoch": 0.35815792285500586,
      "grad_norm": 1.6122456789016724,
      "learning_rate": 1.644064171122995e-05,
      "loss": 0.08234067261219025,
      "num_input_tokens_seen": 880084,
      "step": 838,
      "train_runtime": 1983.5782,
      "train_tokens_per_second": 443.685
    },
    {
      "epoch": 0.35858531894433165,
      "grad_norm": 4.1051812171936035,
      "learning_rate": 1.643636363636364e-05,
      "loss": 0.13144336640834808,
      "num_input_tokens_seen": 881128,
      "step": 839,
      "train_runtime": 1985.5973,
      "train_tokens_per_second": 443.76
    },
    {
      "epoch": 0.35901271503365745,
      "grad_norm": 3.988614082336426,
      "learning_rate": 1.6432085561497328e-05,
      "loss": 0.23451296985149384,
      "num_input_tokens_seen": 881942,
      "step": 840,
      "train_runtime": 1987.591,
      "train_tokens_per_second": 443.724
    },
    {
      "epoch": 0.35944011112298324,
      "grad_norm": 3.1022684574127197,
      "learning_rate": 1.6427807486631017e-05,
      "loss": 0.11788983643054962,
      "num_input_tokens_seen": 882700,
      "step": 841,
      "train_runtime": 1996.1137,
      "train_tokens_per_second": 442.209
    },
    {
      "epoch": 0.35986750721230903,
      "grad_norm": 2.941065549850464,
      "learning_rate": 1.6423529411764707e-05,
      "loss": 0.16564802825450897,
      "num_input_tokens_seen": 883724,
      "step": 842,
      "train_runtime": 1998.1362,
      "train_tokens_per_second": 442.274
    },
    {
      "epoch": 0.3602949033016348,
      "grad_norm": 3.216688632965088,
      "learning_rate": 1.6419251336898396e-05,
      "loss": 0.1497475951910019,
      "num_input_tokens_seen": 884510,
      "step": 843,
      "train_runtime": 2000.1572,
      "train_tokens_per_second": 442.22
    },
    {
      "epoch": 0.36072229939096057,
      "grad_norm": 2.5600903034210205,
      "learning_rate": 1.641497326203209e-05,
      "loss": 0.12463045120239258,
      "num_input_tokens_seen": 885522,
      "step": 844,
      "train_runtime": 2002.2305,
      "train_tokens_per_second": 442.268
    },
    {
      "epoch": 0.36114969548028636,
      "grad_norm": 2.9868717193603516,
      "learning_rate": 1.641069518716578e-05,
      "loss": 0.12309199571609497,
      "num_input_tokens_seen": 886520,
      "step": 845,
      "train_runtime": 2004.2629,
      "train_tokens_per_second": 442.317
    },
    {
      "epoch": 0.36157709156961215,
      "grad_norm": 2.599369525909424,
      "learning_rate": 1.6406417112299468e-05,
      "loss": 0.1329491287469864,
      "num_input_tokens_seen": 887478,
      "step": 846,
      "train_runtime": 2006.3683,
      "train_tokens_per_second": 442.331
    },
    {
      "epoch": 0.36200448765893795,
      "grad_norm": 4.006803035736084,
      "learning_rate": 1.6402139037433154e-05,
      "loss": 0.12610451877117157,
      "num_input_tokens_seen": 888142,
      "step": 847,
      "train_runtime": 2008.3612,
      "train_tokens_per_second": 442.222
    },
    {
      "epoch": 0.3624318837482637,
      "grad_norm": 3.9792346954345703,
      "learning_rate": 1.6397860962566847e-05,
      "loss": 0.17190277576446533,
      "num_input_tokens_seen": 889178,
      "step": 848,
      "train_runtime": 2010.5036,
      "train_tokens_per_second": 442.266
    },
    {
      "epoch": 0.3628592798375895,
      "grad_norm": 2.634158134460449,
      "learning_rate": 1.6393582887700536e-05,
      "loss": 0.1466003656387329,
      "num_input_tokens_seen": 890746,
      "step": 849,
      "train_runtime": 2012.6883,
      "train_tokens_per_second": 442.565
    },
    {
      "epoch": 0.36328667592691527,
      "grad_norm": 3.4565250873565674,
      "learning_rate": 1.6389304812834226e-05,
      "loss": 0.13625645637512207,
      "num_input_tokens_seen": 891694,
      "step": 850,
      "train_runtime": 2014.7349,
      "train_tokens_per_second": 442.586
    },
    {
      "epoch": 0.36371407201624106,
      "grad_norm": 2.904454231262207,
      "learning_rate": 1.6385026737967915e-05,
      "loss": 0.21185842156410217,
      "num_input_tokens_seen": 892998,
      "step": 851,
      "train_runtime": 2016.8572,
      "train_tokens_per_second": 442.767
    },
    {
      "epoch": 0.36414146810556686,
      "grad_norm": 3.2865755558013916,
      "learning_rate": 1.6380748663101608e-05,
      "loss": 0.17540650069713593,
      "num_input_tokens_seen": 893882,
      "step": 852,
      "train_runtime": 2018.8885,
      "train_tokens_per_second": 442.759
    },
    {
      "epoch": 0.3645688641948926,
      "grad_norm": 2.5559585094451904,
      "learning_rate": 1.6376470588235298e-05,
      "loss": 0.11230617761611938,
      "num_input_tokens_seen": 894888,
      "step": 853,
      "train_runtime": 2021.0596,
      "train_tokens_per_second": 442.782
    },
    {
      "epoch": 0.3649962602842184,
      "grad_norm": 3.1157190799713135,
      "learning_rate": 1.6372192513368984e-05,
      "loss": 0.2301216870546341,
      "num_input_tokens_seen": 896216,
      "step": 854,
      "train_runtime": 2023.2594,
      "train_tokens_per_second": 442.957
    },
    {
      "epoch": 0.3654236563735442,
      "grad_norm": 2.973968505859375,
      "learning_rate": 1.6367914438502673e-05,
      "loss": 0.11380991339683533,
      "num_input_tokens_seen": 897404,
      "step": 855,
      "train_runtime": 2025.3996,
      "train_tokens_per_second": 443.075
    },
    {
      "epoch": 0.36585105246287,
      "grad_norm": 3.152690887451172,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 0.1799992322921753,
      "num_input_tokens_seen": 898318,
      "step": 856,
      "train_runtime": 2027.424,
      "train_tokens_per_second": 443.083
    },
    {
      "epoch": 0.36627844855219577,
      "grad_norm": 2.545757532119751,
      "learning_rate": 1.6359358288770055e-05,
      "loss": 0.1458369791507721,
      "num_input_tokens_seen": 899926,
      "step": 857,
      "train_runtime": 2029.6251,
      "train_tokens_per_second": 443.395
    },
    {
      "epoch": 0.3667058446415215,
      "grad_norm": 3.038119077682495,
      "learning_rate": 1.6355080213903745e-05,
      "loss": 0.1422380805015564,
      "num_input_tokens_seen": 901122,
      "step": 858,
      "train_runtime": 2031.741,
      "train_tokens_per_second": 443.522
    },
    {
      "epoch": 0.3671332407308473,
      "grad_norm": 2.1880640983581543,
      "learning_rate": 1.6350802139037434e-05,
      "loss": 0.0793108344078064,
      "num_input_tokens_seen": 902054,
      "step": 859,
      "train_runtime": 2033.8212,
      "train_tokens_per_second": 443.527
    },
    {
      "epoch": 0.3675606368201731,
      "grad_norm": 3.888615608215332,
      "learning_rate": 1.6346524064171124e-05,
      "loss": 0.15858900547027588,
      "num_input_tokens_seen": 902682,
      "step": 860,
      "train_runtime": 2035.784,
      "train_tokens_per_second": 443.408
    },
    {
      "epoch": 0.3679880329094989,
      "grad_norm": 3.678223133087158,
      "learning_rate": 1.6342245989304813e-05,
      "loss": 0.2031177282333374,
      "num_input_tokens_seen": 903482,
      "step": 861,
      "train_runtime": 2037.7453,
      "train_tokens_per_second": 443.373
    },
    {
      "epoch": 0.3684154289988247,
      "grad_norm": 5.0188517570495605,
      "learning_rate": 1.6337967914438503e-05,
      "loss": 0.16376955807209015,
      "num_input_tokens_seen": 904300,
      "step": 862,
      "train_runtime": 2039.749,
      "train_tokens_per_second": 443.339
    },
    {
      "epoch": 0.3688428250881504,
      "grad_norm": 3.451781749725342,
      "learning_rate": 1.6333689839572192e-05,
      "loss": 0.17166760563850403,
      "num_input_tokens_seen": 905288,
      "step": 863,
      "train_runtime": 2041.7884,
      "train_tokens_per_second": 443.38
    },
    {
      "epoch": 0.3692702211774762,
      "grad_norm": 3.7770071029663086,
      "learning_rate": 1.6329411764705885e-05,
      "loss": 0.20119602978229523,
      "num_input_tokens_seen": 906054,
      "step": 864,
      "train_runtime": 2043.7914,
      "train_tokens_per_second": 443.32
    },
    {
      "epoch": 0.369697617266802,
      "grad_norm": 2.81948184967041,
      "learning_rate": 1.6325133689839574e-05,
      "loss": 0.1101490929722786,
      "num_input_tokens_seen": 907008,
      "step": 865,
      "train_runtime": 2045.8559,
      "train_tokens_per_second": 443.339
    },
    {
      "epoch": 0.3701250133561278,
      "grad_norm": 2.6873080730438232,
      "learning_rate": 1.6320855614973264e-05,
      "loss": 0.13334931433200836,
      "num_input_tokens_seen": 908128,
      "step": 866,
      "train_runtime": 2048.0003,
      "train_tokens_per_second": 443.422
    },
    {
      "epoch": 0.3705524094454536,
      "grad_norm": 3.717231273651123,
      "learning_rate": 1.6316577540106953e-05,
      "loss": 0.17222562432289124,
      "num_input_tokens_seen": 909008,
      "step": 867,
      "train_runtime": 2049.9971,
      "train_tokens_per_second": 443.419
    },
    {
      "epoch": 0.3709798055347794,
      "grad_norm": 2.938831329345703,
      "learning_rate": 1.6312299465240643e-05,
      "loss": 0.17824360728263855,
      "num_input_tokens_seen": 910658,
      "step": 868,
      "train_runtime": 2052.1588,
      "train_tokens_per_second": 443.756
    },
    {
      "epoch": 0.3714072016241051,
      "grad_norm": 5.017563343048096,
      "learning_rate": 1.6308021390374332e-05,
      "loss": 0.19953244924545288,
      "num_input_tokens_seen": 911350,
      "step": 869,
      "train_runtime": 2054.1031,
      "train_tokens_per_second": 443.673
    },
    {
      "epoch": 0.3718345977134309,
      "grad_norm": 3.751927137374878,
      "learning_rate": 1.630374331550802e-05,
      "loss": 0.16339567303657532,
      "num_input_tokens_seen": 912134,
      "step": 870,
      "train_runtime": 2056.0838,
      "train_tokens_per_second": 443.627
    },
    {
      "epoch": 0.3722619938027567,
      "grad_norm": 2.96889591217041,
      "learning_rate": 1.629946524064171e-05,
      "loss": 0.20123767852783203,
      "num_input_tokens_seen": 913428,
      "step": 871,
      "train_runtime": 2064.7675,
      "train_tokens_per_second": 442.388
    },
    {
      "epoch": 0.3726893898920825,
      "grad_norm": 2.8208200931549072,
      "learning_rate": 1.6295187165775404e-05,
      "loss": 0.12837810814380646,
      "num_input_tokens_seen": 914398,
      "step": 872,
      "train_runtime": 2066.7625,
      "train_tokens_per_second": 442.43
    },
    {
      "epoch": 0.3731167859814083,
      "grad_norm": 3.2568740844726562,
      "learning_rate": 1.6290909090909093e-05,
      "loss": 0.12384645640850067,
      "num_input_tokens_seen": 915234,
      "step": 873,
      "train_runtime": 2068.7583,
      "train_tokens_per_second": 442.407
    },
    {
      "epoch": 0.37354418207073403,
      "grad_norm": 2.983619451522827,
      "learning_rate": 1.6286631016042783e-05,
      "loss": 0.1463351845741272,
      "num_input_tokens_seen": 916102,
      "step": 874,
      "train_runtime": 2070.7753,
      "train_tokens_per_second": 442.396
    },
    {
      "epoch": 0.3739715781600598,
      "grad_norm": 3.2139527797698975,
      "learning_rate": 1.6282352941176472e-05,
      "loss": 0.12374912947416306,
      "num_input_tokens_seen": 916690,
      "step": 875,
      "train_runtime": 2072.7004,
      "train_tokens_per_second": 442.268
    },
    {
      "epoch": 0.3743989742493856,
      "grad_norm": 2.230070114135742,
      "learning_rate": 1.627807486631016e-05,
      "loss": 0.10868636518716812,
      "num_input_tokens_seen": 917832,
      "step": 876,
      "train_runtime": 2074.8174,
      "train_tokens_per_second": 442.368
    },
    {
      "epoch": 0.3748263703387114,
      "grad_norm": 2.75704288482666,
      "learning_rate": 1.627379679144385e-05,
      "loss": 0.13413727283477783,
      "num_input_tokens_seen": 919210,
      "step": 877,
      "train_runtime": 2076.9287,
      "train_tokens_per_second": 442.581
    },
    {
      "epoch": 0.3752537664280372,
      "grad_norm": 2.8754377365112305,
      "learning_rate": 1.626951871657754e-05,
      "loss": 0.10905253887176514,
      "num_input_tokens_seen": 920192,
      "step": 878,
      "train_runtime": 2078.9864,
      "train_tokens_per_second": 442.616
    },
    {
      "epoch": 0.37568116251736294,
      "grad_norm": 2.77583646774292,
      "learning_rate": 1.626524064171123e-05,
      "loss": 0.0867101177573204,
      "num_input_tokens_seen": 920826,
      "step": 879,
      "train_runtime": 2080.9473,
      "train_tokens_per_second": 442.503
    },
    {
      "epoch": 0.37610855860668874,
      "grad_norm": 3.4160633087158203,
      "learning_rate": 1.626096256684492e-05,
      "loss": 0.16699495911598206,
      "num_input_tokens_seen": 922070,
      "step": 880,
      "train_runtime": 2083.0732,
      "train_tokens_per_second": 442.649
    },
    {
      "epoch": 0.37653595469601453,
      "grad_norm": 3.6048743724823,
      "learning_rate": 1.6256684491978612e-05,
      "loss": 0.28260552883148193,
      "num_input_tokens_seen": 923318,
      "step": 881,
      "train_runtime": 2085.1799,
      "train_tokens_per_second": 442.8
    },
    {
      "epoch": 0.3769633507853403,
      "grad_norm": 3.290318012237549,
      "learning_rate": 1.62524064171123e-05,
      "loss": 0.13404695689678192,
      "num_input_tokens_seen": 924142,
      "step": 882,
      "train_runtime": 2087.2232,
      "train_tokens_per_second": 442.761
    },
    {
      "epoch": 0.3773907468746661,
      "grad_norm": 3.361940860748291,
      "learning_rate": 1.624812834224599e-05,
      "loss": 0.17885807156562805,
      "num_input_tokens_seen": 925586,
      "step": 883,
      "train_runtime": 2089.5069,
      "train_tokens_per_second": 442.969
    },
    {
      "epoch": 0.37781814296399185,
      "grad_norm": 2.3778014183044434,
      "learning_rate": 1.624385026737968e-05,
      "loss": 0.10743165761232376,
      "num_input_tokens_seen": 926742,
      "step": 884,
      "train_runtime": 2091.6867,
      "train_tokens_per_second": 443.06
    },
    {
      "epoch": 0.37824553905331765,
      "grad_norm": 2.6063477993011475,
      "learning_rate": 1.623957219251337e-05,
      "loss": 0.17332960665225983,
      "num_input_tokens_seen": 928106,
      "step": 885,
      "train_runtime": 2093.8662,
      "train_tokens_per_second": 443.25
    },
    {
      "epoch": 0.37867293514264344,
      "grad_norm": 4.0899505615234375,
      "learning_rate": 1.623529411764706e-05,
      "loss": 0.19288423657417297,
      "num_input_tokens_seen": 929166,
      "step": 886,
      "train_runtime": 2095.9856,
      "train_tokens_per_second": 443.307
    },
    {
      "epoch": 0.37910033123196923,
      "grad_norm": 2.5437886714935303,
      "learning_rate": 1.623101604278075e-05,
      "loss": 0.11767511814832687,
      "num_input_tokens_seen": 930410,
      "step": 887,
      "train_runtime": 2098.1375,
      "train_tokens_per_second": 443.446
    },
    {
      "epoch": 0.379527727321295,
      "grad_norm": 3.331578493118286,
      "learning_rate": 1.622673796791444e-05,
      "loss": 0.1090846061706543,
      "num_input_tokens_seen": 931220,
      "step": 888,
      "train_runtime": 2100.2103,
      "train_tokens_per_second": 443.394
    },
    {
      "epoch": 0.37995512341062077,
      "grad_norm": 3.095195770263672,
      "learning_rate": 1.622245989304813e-05,
      "loss": 0.10828014463186264,
      "num_input_tokens_seen": 932864,
      "step": 889,
      "train_runtime": 2102.4118,
      "train_tokens_per_second": 443.711
    },
    {
      "epoch": 0.38038251949994656,
      "grad_norm": 3.322017192840576,
      "learning_rate": 1.621818181818182e-05,
      "loss": 0.17956194281578064,
      "num_input_tokens_seen": 934192,
      "step": 890,
      "train_runtime": 2104.544,
      "train_tokens_per_second": 443.893
    },
    {
      "epoch": 0.38080991558927235,
      "grad_norm": 4.183267116546631,
      "learning_rate": 1.621390374331551e-05,
      "loss": 0.20426103472709656,
      "num_input_tokens_seen": 935036,
      "step": 891,
      "train_runtime": 2106.5351,
      "train_tokens_per_second": 443.874
    },
    {
      "epoch": 0.38123731167859815,
      "grad_norm": 2.649113893508911,
      "learning_rate": 1.62096256684492e-05,
      "loss": 0.17787085473537445,
      "num_input_tokens_seen": 936516,
      "step": 892,
      "train_runtime": 2108.6962,
      "train_tokens_per_second": 444.121
    },
    {
      "epoch": 0.38166470776792394,
      "grad_norm": 2.7996020317077637,
      "learning_rate": 1.620534759358289e-05,
      "loss": 0.11850351840257645,
      "num_input_tokens_seen": 937898,
      "step": 893,
      "train_runtime": 2110.816,
      "train_tokens_per_second": 444.33
    },
    {
      "epoch": 0.38209210385724973,
      "grad_norm": 3.758162498474121,
      "learning_rate": 1.620106951871658e-05,
      "loss": 0.10957302898168564,
      "num_input_tokens_seen": 938748,
      "step": 894,
      "train_runtime": 2112.8862,
      "train_tokens_per_second": 444.297
    },
    {
      "epoch": 0.38251949994657547,
      "grad_norm": 2.9185712337493896,
      "learning_rate": 1.6196791443850268e-05,
      "loss": 0.13005822896957397,
      "num_input_tokens_seen": 939548,
      "step": 895,
      "train_runtime": 2114.9377,
      "train_tokens_per_second": 444.244
    },
    {
      "epoch": 0.38294689603590126,
      "grad_norm": 2.704132318496704,
      "learning_rate": 1.6192513368983957e-05,
      "loss": 0.11967745423316956,
      "num_input_tokens_seen": 940356,
      "step": 896,
      "train_runtime": 2116.9273,
      "train_tokens_per_second": 444.208
    },
    {
      "epoch": 0.38337429212522706,
      "grad_norm": 3.0454626083374023,
      "learning_rate": 1.618823529411765e-05,
      "loss": 0.14856570959091187,
      "num_input_tokens_seen": 941316,
      "step": 897,
      "train_runtime": 2119.003,
      "train_tokens_per_second": 444.226
    },
    {
      "epoch": 0.38380168821455285,
      "grad_norm": 3.2519073486328125,
      "learning_rate": 1.618395721925134e-05,
      "loss": 0.19943198561668396,
      "num_input_tokens_seen": 942222,
      "step": 898,
      "train_runtime": 2121.0845,
      "train_tokens_per_second": 444.217
    },
    {
      "epoch": 0.38422908430387864,
      "grad_norm": 2.854214668273926,
      "learning_rate": 1.617967914438503e-05,
      "loss": 0.15924370288848877,
      "num_input_tokens_seen": 943360,
      "step": 899,
      "train_runtime": 2123.1721,
      "train_tokens_per_second": 444.316
    },
    {
      "epoch": 0.3846564803932044,
      "grad_norm": 3.443159580230713,
      "learning_rate": 1.617540106951872e-05,
      "loss": 0.1765565574169159,
      "num_input_tokens_seen": 944406,
      "step": 900,
      "train_runtime": 2125.2337,
      "train_tokens_per_second": 444.377
    },
    {
      "epoch": 0.3850838764825302,
      "grad_norm": 2.476480484008789,
      "learning_rate": 1.6171122994652408e-05,
      "loss": 0.1342540979385376,
      "num_input_tokens_seen": 945882,
      "step": 901,
      "train_runtime": 2133.894,
      "train_tokens_per_second": 443.266
    },
    {
      "epoch": 0.38551127257185597,
      "grad_norm": 3.512960195541382,
      "learning_rate": 1.6166844919786097e-05,
      "loss": 0.18861031532287598,
      "num_input_tokens_seen": 946730,
      "step": 902,
      "train_runtime": 2135.888,
      "train_tokens_per_second": 443.249
    },
    {
      "epoch": 0.38593866866118176,
      "grad_norm": 4.3065643310546875,
      "learning_rate": 1.6162566844919787e-05,
      "loss": 0.22209052741527557,
      "num_input_tokens_seen": 947530,
      "step": 903,
      "train_runtime": 2137.8852,
      "train_tokens_per_second": 443.209
    },
    {
      "epoch": 0.38636606475050755,
      "grad_norm": 2.3028688430786133,
      "learning_rate": 1.6158288770053476e-05,
      "loss": 0.08168645203113556,
      "num_input_tokens_seen": 948602,
      "step": 904,
      "train_runtime": 2139.9571,
      "train_tokens_per_second": 443.281
    },
    {
      "epoch": 0.3867934608398333,
      "grad_norm": 2.892332077026367,
      "learning_rate": 1.615401069518717e-05,
      "loss": 0.14299827814102173,
      "num_input_tokens_seen": 949498,
      "step": 905,
      "train_runtime": 2142.0689,
      "train_tokens_per_second": 443.262
    },
    {
      "epoch": 0.3872208569291591,
      "grad_norm": 3.6618220806121826,
      "learning_rate": 1.614973262032086e-05,
      "loss": 0.1568375527858734,
      "num_input_tokens_seen": 950432,
      "step": 906,
      "train_runtime": 2144.0982,
      "train_tokens_per_second": 443.278
    },
    {
      "epoch": 0.3876482530184849,
      "grad_norm": 2.7025883197784424,
      "learning_rate": 1.6145454545454548e-05,
      "loss": 0.1164485439658165,
      "num_input_tokens_seen": 951456,
      "step": 907,
      "train_runtime": 2146.138,
      "train_tokens_per_second": 443.334
    },
    {
      "epoch": 0.38807564910781067,
      "grad_norm": 3.104348659515381,
      "learning_rate": 1.6141176470588234e-05,
      "loss": 0.10492739826440811,
      "num_input_tokens_seen": 952172,
      "step": 908,
      "train_runtime": 2148.1463,
      "train_tokens_per_second": 443.253
    },
    {
      "epoch": 0.38850304519713647,
      "grad_norm": 2.747732162475586,
      "learning_rate": 1.6136898395721927e-05,
      "loss": 0.11487283557653427,
      "num_input_tokens_seen": 953040,
      "step": 909,
      "train_runtime": 2150.1372,
      "train_tokens_per_second": 443.246
    },
    {
      "epoch": 0.3889304412864622,
      "grad_norm": 2.6833279132843018,
      "learning_rate": 1.6132620320855616e-05,
      "loss": 0.12157800048589706,
      "num_input_tokens_seen": 954068,
      "step": 910,
      "train_runtime": 2152.1638,
      "train_tokens_per_second": 443.306
    },
    {
      "epoch": 0.389357837375788,
      "grad_norm": 3.4484288692474365,
      "learning_rate": 1.6128342245989306e-05,
      "loss": 0.226801797747612,
      "num_input_tokens_seen": 954938,
      "step": 911,
      "train_runtime": 2154.196,
      "train_tokens_per_second": 443.292
    },
    {
      "epoch": 0.3897852334651138,
      "grad_norm": 3.9453911781311035,
      "learning_rate": 1.6124064171122995e-05,
      "loss": 0.14873573184013367,
      "num_input_tokens_seen": 955768,
      "step": 912,
      "train_runtime": 2156.2191,
      "train_tokens_per_second": 443.261
    },
    {
      "epoch": 0.3902126295544396,
      "grad_norm": 3.6305999755859375,
      "learning_rate": 1.6119786096256688e-05,
      "loss": 0.21297237277030945,
      "num_input_tokens_seen": 956702,
      "step": 913,
      "train_runtime": 2158.2875,
      "train_tokens_per_second": 443.269
    },
    {
      "epoch": 0.3906400256437654,
      "grad_norm": 3.201023817062378,
      "learning_rate": 1.6115508021390377e-05,
      "loss": 0.138827845454216,
      "num_input_tokens_seen": 957424,
      "step": 914,
      "train_runtime": 2160.3069,
      "train_tokens_per_second": 443.189
    },
    {
      "epoch": 0.39106742173309117,
      "grad_norm": 2.8727478981018066,
      "learning_rate": 1.6111229946524064e-05,
      "loss": 0.11612001806497574,
      "num_input_tokens_seen": 958388,
      "step": 915,
      "train_runtime": 2162.3579,
      "train_tokens_per_second": 443.214
    },
    {
      "epoch": 0.3914948178224169,
      "grad_norm": 3.085972785949707,
      "learning_rate": 1.6106951871657753e-05,
      "loss": 0.2506169080734253,
      "num_input_tokens_seen": 959302,
      "step": 916,
      "train_runtime": 2164.388,
      "train_tokens_per_second": 443.221
    },
    {
      "epoch": 0.3919222139117427,
      "grad_norm": 3.15606689453125,
      "learning_rate": 1.6102673796791446e-05,
      "loss": 0.10323436558246613,
      "num_input_tokens_seen": 960216,
      "step": 917,
      "train_runtime": 2166.4106,
      "train_tokens_per_second": 443.229
    },
    {
      "epoch": 0.3923496100010685,
      "grad_norm": 3.7641422748565674,
      "learning_rate": 1.6098395721925135e-05,
      "loss": 0.15620802342891693,
      "num_input_tokens_seen": 961178,
      "step": 918,
      "train_runtime": 2168.4328,
      "train_tokens_per_second": 443.259
    },
    {
      "epoch": 0.3927770060903943,
      "grad_norm": 3.679429531097412,
      "learning_rate": 1.6094117647058825e-05,
      "loss": 0.17789867520332336,
      "num_input_tokens_seen": 961906,
      "step": 919,
      "train_runtime": 2170.4274,
      "train_tokens_per_second": 443.187
    },
    {
      "epoch": 0.3932044021797201,
      "grad_norm": 3.106301784515381,
      "learning_rate": 1.6089839572192514e-05,
      "loss": 0.21292844414710999,
      "num_input_tokens_seen": 962894,
      "step": 920,
      "train_runtime": 2172.508,
      "train_tokens_per_second": 443.218
    },
    {
      "epoch": 0.3936317982690458,
      "grad_norm": 2.34635329246521,
      "learning_rate": 1.6085561497326207e-05,
      "loss": 0.0811031237244606,
      "num_input_tokens_seen": 963752,
      "step": 921,
      "train_runtime": 2174.4984,
      "train_tokens_per_second": 443.207
    },
    {
      "epoch": 0.3940591943583716,
      "grad_norm": 2.3631465435028076,
      "learning_rate": 1.6081283422459893e-05,
      "loss": 0.10404130071401596,
      "num_input_tokens_seen": 964612,
      "step": 922,
      "train_runtime": 2176.5207,
      "train_tokens_per_second": 443.19
    },
    {
      "epoch": 0.3944865904476974,
      "grad_norm": 3.8228330612182617,
      "learning_rate": 1.6077005347593582e-05,
      "loss": 0.173420250415802,
      "num_input_tokens_seen": 965652,
      "step": 923,
      "train_runtime": 2178.5605,
      "train_tokens_per_second": 443.252
    },
    {
      "epoch": 0.3949139865370232,
      "grad_norm": 3.662015914916992,
      "learning_rate": 1.6072727272727272e-05,
      "loss": 0.15936538577079773,
      "num_input_tokens_seen": 966668,
      "step": 924,
      "train_runtime": 2180.59,
      "train_tokens_per_second": 443.306
    },
    {
      "epoch": 0.395341382626349,
      "grad_norm": 2.7576632499694824,
      "learning_rate": 1.6068449197860965e-05,
      "loss": 0.11928541958332062,
      "num_input_tokens_seen": 967584,
      "step": 925,
      "train_runtime": 2182.5933,
      "train_tokens_per_second": 443.319
    },
    {
      "epoch": 0.39576877871567473,
      "grad_norm": 2.92537784576416,
      "learning_rate": 1.6064171122994654e-05,
      "loss": 0.18931931257247925,
      "num_input_tokens_seen": 968924,
      "step": 926,
      "train_runtime": 2184.6909,
      "train_tokens_per_second": 443.506
    },
    {
      "epoch": 0.3961961748050005,
      "grad_norm": 2.3758535385131836,
      "learning_rate": 1.6059893048128344e-05,
      "loss": 0.1301591694355011,
      "num_input_tokens_seen": 970256,
      "step": 927,
      "train_runtime": 2186.8044,
      "train_tokens_per_second": 443.687
    },
    {
      "epoch": 0.3966235708943263,
      "grad_norm": 2.2258036136627197,
      "learning_rate": 1.6055614973262033e-05,
      "loss": 0.11166122555732727,
      "num_input_tokens_seen": 971508,
      "step": 928,
      "train_runtime": 2188.9467,
      "train_tokens_per_second": 443.824
    },
    {
      "epoch": 0.3970509669836521,
      "grad_norm": 3.133446216583252,
      "learning_rate": 1.6051336898395723e-05,
      "loss": 0.14364859461784363,
      "num_input_tokens_seen": 972496,
      "step": 929,
      "train_runtime": 2190.9939,
      "train_tokens_per_second": 443.861
    },
    {
      "epoch": 0.3974783630729779,
      "grad_norm": 3.3615312576293945,
      "learning_rate": 1.6047058823529412e-05,
      "loss": 0.1700805425643921,
      "num_input_tokens_seen": 973502,
      "step": 930,
      "train_runtime": 2193.0685,
      "train_tokens_per_second": 443.899
    },
    {
      "epoch": 0.39790575916230364,
      "grad_norm": 2.510171413421631,
      "learning_rate": 1.60427807486631e-05,
      "loss": 0.09587380290031433,
      "num_input_tokens_seen": 974398,
      "step": 931,
      "train_runtime": 2201.6722,
      "train_tokens_per_second": 442.572
    },
    {
      "epoch": 0.39833315525162943,
      "grad_norm": 3.8139679431915283,
      "learning_rate": 1.603850267379679e-05,
      "loss": 0.15770947933197021,
      "num_input_tokens_seen": 975244,
      "step": 932,
      "train_runtime": 2203.6645,
      "train_tokens_per_second": 442.556
    },
    {
      "epoch": 0.3987605513409552,
      "grad_norm": 3.882706880569458,
      "learning_rate": 1.6034224598930484e-05,
      "loss": 0.1570173054933548,
      "num_input_tokens_seen": 976058,
      "step": 933,
      "train_runtime": 2205.654,
      "train_tokens_per_second": 442.525
    },
    {
      "epoch": 0.399187947430281,
      "grad_norm": 3.7372584342956543,
      "learning_rate": 1.6029946524064173e-05,
      "loss": 0.20967872440814972,
      "num_input_tokens_seen": 976876,
      "step": 934,
      "train_runtime": 2207.6456,
      "train_tokens_per_second": 442.497
    },
    {
      "epoch": 0.3996153435196068,
      "grad_norm": 5.46579647064209,
      "learning_rate": 1.6025668449197863e-05,
      "loss": 0.26738402247428894,
      "num_input_tokens_seen": 977490,
      "step": 935,
      "train_runtime": 2209.5962,
      "train_tokens_per_second": 442.384
    },
    {
      "epoch": 0.40004273960893255,
      "grad_norm": 3.54427433013916,
      "learning_rate": 1.6021390374331552e-05,
      "loss": 0.11531199514865875,
      "num_input_tokens_seen": 978154,
      "step": 936,
      "train_runtime": 2211.5918,
      "train_tokens_per_second": 442.285
    },
    {
      "epoch": 0.40004273960893255,
      "eval_loss": 0.5698503851890564,
      "eval_runtime": 57.953,
      "eval_samples_per_second": 17.221,
      "eval_steps_per_second": 8.61,
      "num_input_tokens_seen": 978154,
      "step": 936
    },
    {
      "epoch": 0.40047013569825834,
      "grad_norm": 3.129371404647827,
      "learning_rate": 1.601711229946524e-05,
      "loss": 0.1033656895160675,
      "num_input_tokens_seen": 978928,
      "step": 937,
      "train_runtime": 2271.6265,
      "train_tokens_per_second": 430.937
    },
    {
      "epoch": 0.40089753178758414,
      "grad_norm": 2.525700569152832,
      "learning_rate": 1.601283422459893e-05,
      "loss": 0.13463933765888214,
      "num_input_tokens_seen": 979884,
      "step": 938,
      "train_runtime": 2273.7054,
      "train_tokens_per_second": 430.963
    },
    {
      "epoch": 0.40132492787690993,
      "grad_norm": 3.946357011795044,
      "learning_rate": 1.600855614973262e-05,
      "loss": 0.15102031826972961,
      "num_input_tokens_seen": 980664,
      "step": 939,
      "train_runtime": 2275.765,
      "train_tokens_per_second": 430.916
    },
    {
      "epoch": 0.4017523239662357,
      "grad_norm": 4.099009037017822,
      "learning_rate": 1.600427807486631e-05,
      "loss": 0.14986948668956757,
      "num_input_tokens_seen": 981276,
      "step": 940,
      "train_runtime": 2277.8068,
      "train_tokens_per_second": 430.799
    },
    {
      "epoch": 0.4021797200555615,
      "grad_norm": 2.5861546993255615,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.17913028597831726,
      "num_input_tokens_seen": 982846,
      "step": 941,
      "train_runtime": 2280.1165,
      "train_tokens_per_second": 431.051
    },
    {
      "epoch": 0.40260711614488726,
      "grad_norm": 2.740493059158325,
      "learning_rate": 1.5995721925133692e-05,
      "loss": 0.09153739362955093,
      "num_input_tokens_seen": 984040,
      "step": 942,
      "train_runtime": 2282.2425,
      "train_tokens_per_second": 431.172
    },
    {
      "epoch": 0.40303451223421305,
      "grad_norm": 3.864326000213623,
      "learning_rate": 1.599144385026738e-05,
      "loss": 0.13687407970428467,
      "num_input_tokens_seen": 984862,
      "step": 943,
      "train_runtime": 2284.3261,
      "train_tokens_per_second": 431.139
    },
    {
      "epoch": 0.40346190832353884,
      "grad_norm": 2.4913134574890137,
      "learning_rate": 1.598716577540107e-05,
      "loss": 0.0964777022600174,
      "num_input_tokens_seen": 985594,
      "step": 944,
      "train_runtime": 2286.3824,
      "train_tokens_per_second": 431.071
    },
    {
      "epoch": 0.40388930441286464,
      "grad_norm": 4.006031036376953,
      "learning_rate": 1.598288770053476e-05,
      "loss": 0.1856725662946701,
      "num_input_tokens_seen": 986356,
      "step": 945,
      "train_runtime": 2288.4538,
      "train_tokens_per_second": 431.014
    },
    {
      "epoch": 0.40431670050219043,
      "grad_norm": 3.6885268688201904,
      "learning_rate": 1.597860962566845e-05,
      "loss": 0.20293888449668884,
      "num_input_tokens_seen": 987016,
      "step": 946,
      "train_runtime": 2290.4571,
      "train_tokens_per_second": 430.925
    },
    {
      "epoch": 0.40474409659151617,
      "grad_norm": 3.0161800384521484,
      "learning_rate": 1.597433155080214e-05,
      "loss": 0.13541850447654724,
      "num_input_tokens_seen": 987838,
      "step": 947,
      "train_runtime": 2292.6011,
      "train_tokens_per_second": 430.881
    },
    {
      "epoch": 0.40517149268084196,
      "grad_norm": 3.6430602073669434,
      "learning_rate": 1.597005347593583e-05,
      "loss": 0.13437046110630035,
      "num_input_tokens_seen": 988560,
      "step": 948,
      "train_runtime": 2294.6169,
      "train_tokens_per_second": 430.817
    },
    {
      "epoch": 0.40559888877016775,
      "grad_norm": 5.142971992492676,
      "learning_rate": 1.596577540106952e-05,
      "loss": 0.23744851350784302,
      "num_input_tokens_seen": 989722,
      "step": 949,
      "train_runtime": 2296.7552,
      "train_tokens_per_second": 430.922
    },
    {
      "epoch": 0.40602628485949355,
      "grad_norm": 3.0357255935668945,
      "learning_rate": 1.596149732620321e-05,
      "loss": 0.12808263301849365,
      "num_input_tokens_seen": 990588,
      "step": 950,
      "train_runtime": 2298.8303,
      "train_tokens_per_second": 430.91
    },
    {
      "epoch": 0.40645368094881934,
      "grad_norm": 3.43465518951416,
      "learning_rate": 1.59572192513369e-05,
      "loss": 0.15328100323677063,
      "num_input_tokens_seen": 991610,
      "step": 951,
      "train_runtime": 2300.9723,
      "train_tokens_per_second": 430.953
    },
    {
      "epoch": 0.4068810770381451,
      "grad_norm": 3.0474939346313477,
      "learning_rate": 1.595294117647059e-05,
      "loss": 0.1505572497844696,
      "num_input_tokens_seen": 993094,
      "step": 952,
      "train_runtime": 2303.2522,
      "train_tokens_per_second": 431.17
    },
    {
      "epoch": 0.40730847312747087,
      "grad_norm": 2.5601046085357666,
      "learning_rate": 1.594866310160428e-05,
      "loss": 0.15914660692214966,
      "num_input_tokens_seen": 994546,
      "step": 953,
      "train_runtime": 2305.4935,
      "train_tokens_per_second": 431.381
    },
    {
      "epoch": 0.40773586921679666,
      "grad_norm": 2.198195457458496,
      "learning_rate": 1.594438502673797e-05,
      "loss": 0.11577246338129044,
      "num_input_tokens_seen": 996130,
      "step": 954,
      "train_runtime": 2307.7314,
      "train_tokens_per_second": 431.649
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 2.4809536933898926,
      "learning_rate": 1.5940106951871658e-05,
      "loss": 0.08430467545986176,
      "num_input_tokens_seen": 996898,
      "step": 955,
      "train_runtime": 2309.7864,
      "train_tokens_per_second": 431.597
    },
    {
      "epoch": 0.40859066139544825,
      "grad_norm": 2.530029773712158,
      "learning_rate": 1.5935828877005348e-05,
      "loss": 0.08219069242477417,
      "num_input_tokens_seen": 997846,
      "step": 956,
      "train_runtime": 2311.8748,
      "train_tokens_per_second": 431.618
    },
    {
      "epoch": 0.409018057484774,
      "grad_norm": 6.292982578277588,
      "learning_rate": 1.593155080213904e-05,
      "loss": 0.1308143585920334,
      "num_input_tokens_seen": 998512,
      "step": 957,
      "train_runtime": 2313.8816,
      "train_tokens_per_second": 431.531
    },
    {
      "epoch": 0.4094454535740998,
      "grad_norm": 4.810624599456787,
      "learning_rate": 1.592727272727273e-05,
      "loss": 0.2354581654071808,
      "num_input_tokens_seen": 999580,
      "step": 958,
      "train_runtime": 2315.9758,
      "train_tokens_per_second": 431.602
    },
    {
      "epoch": 0.4098728496634256,
      "grad_norm": 2.160900115966797,
      "learning_rate": 1.592299465240642e-05,
      "loss": 0.0986616387963295,
      "num_input_tokens_seen": 1000800,
      "step": 959,
      "train_runtime": 2318.1097,
      "train_tokens_per_second": 431.731
    },
    {
      "epoch": 0.41030024575275137,
      "grad_norm": 2.5913217067718506,
      "learning_rate": 1.591871657754011e-05,
      "loss": 0.13993312418460846,
      "num_input_tokens_seen": 1002042,
      "step": 960,
      "train_runtime": 2320.2796,
      "train_tokens_per_second": 431.863
    },
    {
      "epoch": 0.41072764184207716,
      "grad_norm": 2.4269092082977295,
      "learning_rate": 1.59144385026738e-05,
      "loss": 0.13695529103279114,
      "num_input_tokens_seen": 1003804,
      "step": 961,
      "train_runtime": 2329.1783,
      "train_tokens_per_second": 430.969
    },
    {
      "epoch": 0.4111550379314029,
      "grad_norm": 3.1382129192352295,
      "learning_rate": 1.5910160427807488e-05,
      "loss": 0.1408788561820984,
      "num_input_tokens_seen": 1005234,
      "step": 962,
      "train_runtime": 2331.3922,
      "train_tokens_per_second": 431.173
    },
    {
      "epoch": 0.4115824340207287,
      "grad_norm": 3.420305013656616,
      "learning_rate": 1.5905882352941177e-05,
      "loss": 0.19164885580539703,
      "num_input_tokens_seen": 1006284,
      "step": 963,
      "train_runtime": 2333.5462,
      "train_tokens_per_second": 431.225
    },
    {
      "epoch": 0.4120098301100545,
      "grad_norm": 4.329071998596191,
      "learning_rate": 1.5901604278074867e-05,
      "loss": 0.11083001643419266,
      "num_input_tokens_seen": 1006842,
      "step": 964,
      "train_runtime": 2335.5041,
      "train_tokens_per_second": 431.103
    },
    {
      "epoch": 0.4124372261993803,
      "grad_norm": 3.0661702156066895,
      "learning_rate": 1.589732620320856e-05,
      "loss": 0.1629539281129837,
      "num_input_tokens_seen": 1007780,
      "step": 965,
      "train_runtime": 2337.6084,
      "train_tokens_per_second": 431.116
    },
    {
      "epoch": 0.4128646222887061,
      "grad_norm": 3.1654844284057617,
      "learning_rate": 1.589304812834225e-05,
      "loss": 0.12572845816612244,
      "num_input_tokens_seen": 1008514,
      "step": 966,
      "train_runtime": 2339.6499,
      "train_tokens_per_second": 431.053
    },
    {
      "epoch": 0.41329201837803187,
      "grad_norm": 2.6483960151672363,
      "learning_rate": 1.588877005347594e-05,
      "loss": 0.10801530629396439,
      "num_input_tokens_seen": 1009364,
      "step": 967,
      "train_runtime": 2341.7333,
      "train_tokens_per_second": 431.033
    },
    {
      "epoch": 0.4137194144673576,
      "grad_norm": 3.36672306060791,
      "learning_rate": 1.5884491978609628e-05,
      "loss": 0.14181117713451385,
      "num_input_tokens_seen": 1010180,
      "step": 968,
      "train_runtime": 2343.7267,
      "train_tokens_per_second": 431.014
    },
    {
      "epoch": 0.4141468105566834,
      "grad_norm": 2.8620762825012207,
      "learning_rate": 1.5880213903743317e-05,
      "loss": 0.15125736594200134,
      "num_input_tokens_seen": 1010894,
      "step": 969,
      "train_runtime": 2345.7105,
      "train_tokens_per_second": 430.954
    },
    {
      "epoch": 0.4145742066460092,
      "grad_norm": 2.9136648178100586,
      "learning_rate": 1.5875935828877007e-05,
      "loss": 0.18435488641262054,
      "num_input_tokens_seen": 1011828,
      "step": 970,
      "train_runtime": 2347.7602,
      "train_tokens_per_second": 430.976
    },
    {
      "epoch": 0.415001602735335,
      "grad_norm": 3.272649049758911,
      "learning_rate": 1.5871657754010696e-05,
      "loss": 0.12939196825027466,
      "num_input_tokens_seen": 1012502,
      "step": 971,
      "train_runtime": 2349.792,
      "train_tokens_per_second": 430.89
    },
    {
      "epoch": 0.4154289988246608,
      "grad_norm": 5.475400447845459,
      "learning_rate": 1.5867379679144386e-05,
      "loss": 0.28860902786254883,
      "num_input_tokens_seen": 1012990,
      "step": 972,
      "train_runtime": 2351.7461,
      "train_tokens_per_second": 430.74
    },
    {
      "epoch": 0.4158563949139865,
      "grad_norm": 2.724729537963867,
      "learning_rate": 1.586310160427808e-05,
      "loss": 0.18167336285114288,
      "num_input_tokens_seen": 1014720,
      "step": 973,
      "train_runtime": 2354.0773,
      "train_tokens_per_second": 431.048
    },
    {
      "epoch": 0.4162837910033123,
      "grad_norm": 3.133634090423584,
      "learning_rate": 1.5858823529411768e-05,
      "loss": 0.17339462041854858,
      "num_input_tokens_seen": 1016746,
      "step": 974,
      "train_runtime": 2356.4361,
      "train_tokens_per_second": 431.476
    },
    {
      "epoch": 0.4167111870926381,
      "grad_norm": 2.6663413047790527,
      "learning_rate": 1.5854545454545457e-05,
      "loss": 0.1064106747508049,
      "num_input_tokens_seen": 1017582,
      "step": 975,
      "train_runtime": 2358.4943,
      "train_tokens_per_second": 431.454
    },
    {
      "epoch": 0.4171385831819639,
      "grad_norm": 2.532240152359009,
      "learning_rate": 1.5850267379679143e-05,
      "loss": 0.11097608506679535,
      "num_input_tokens_seen": 1018630,
      "step": 976,
      "train_runtime": 2360.6278,
      "train_tokens_per_second": 431.508
    },
    {
      "epoch": 0.4175659792712897,
      "grad_norm": 3.13242769241333,
      "learning_rate": 1.5845989304812836e-05,
      "loss": 0.10106104612350464,
      "num_input_tokens_seen": 1019318,
      "step": 977,
      "train_runtime": 2362.6476,
      "train_tokens_per_second": 431.43
    },
    {
      "epoch": 0.4179933753606154,
      "grad_norm": 3.1643269062042236,
      "learning_rate": 1.5841711229946526e-05,
      "loss": 0.15163138508796692,
      "num_input_tokens_seen": 1020494,
      "step": 978,
      "train_runtime": 2364.833,
      "train_tokens_per_second": 431.529
    },
    {
      "epoch": 0.4184207714499412,
      "grad_norm": 2.685286045074463,
      "learning_rate": 1.5837433155080215e-05,
      "loss": 0.14863711595535278,
      "num_input_tokens_seen": 1021976,
      "step": 979,
      "train_runtime": 2367.0527,
      "train_tokens_per_second": 431.75
    },
    {
      "epoch": 0.418848167539267,
      "grad_norm": 3.5907371044158936,
      "learning_rate": 1.5833155080213905e-05,
      "loss": 0.17044982314109802,
      "num_input_tokens_seen": 1022736,
      "step": 980,
      "train_runtime": 2369.1184,
      "train_tokens_per_second": 431.695
    },
    {
      "epoch": 0.4192755636285928,
      "grad_norm": 4.0167236328125,
      "learning_rate": 1.5828877005347597e-05,
      "loss": 0.13640011847019196,
      "num_input_tokens_seen": 1023696,
      "step": 981,
      "train_runtime": 2371.1763,
      "train_tokens_per_second": 431.725
    },
    {
      "epoch": 0.4197029597179186,
      "grad_norm": 2.88775897026062,
      "learning_rate": 1.5824598930481287e-05,
      "loss": 0.17530319094657898,
      "num_input_tokens_seen": 1024560,
      "step": 982,
      "train_runtime": 2373.2234,
      "train_tokens_per_second": 431.717
    },
    {
      "epoch": 0.42013035580724434,
      "grad_norm": 2.3002512454986572,
      "learning_rate": 1.5820320855614973e-05,
      "loss": 0.09414497017860413,
      "num_input_tokens_seen": 1025730,
      "step": 983,
      "train_runtime": 2375.3405,
      "train_tokens_per_second": 431.824
    },
    {
      "epoch": 0.42055775189657013,
      "grad_norm": 5.44786262512207,
      "learning_rate": 1.5816042780748662e-05,
      "loss": 0.2069467455148697,
      "num_input_tokens_seen": 1026550,
      "step": 984,
      "train_runtime": 2377.3645,
      "train_tokens_per_second": 431.802
    },
    {
      "epoch": 0.4209851479858959,
      "grad_norm": 4.028661727905273,
      "learning_rate": 1.5811764705882352e-05,
      "loss": 0.09830475598573685,
      "num_input_tokens_seen": 1027168,
      "step": 985,
      "train_runtime": 2379.3332,
      "train_tokens_per_second": 431.704
    },
    {
      "epoch": 0.4214125440752217,
      "grad_norm": 2.412536382675171,
      "learning_rate": 1.5807486631016045e-05,
      "loss": 0.1301649808883667,
      "num_input_tokens_seen": 1028250,
      "step": 986,
      "train_runtime": 2381.4668,
      "train_tokens_per_second": 431.772
    },
    {
      "epoch": 0.4218399401645475,
      "grad_norm": 2.8748650550842285,
      "learning_rate": 1.5803208556149734e-05,
      "loss": 0.10201525688171387,
      "num_input_tokens_seen": 1029064,
      "step": 987,
      "train_runtime": 2383.4946,
      "train_tokens_per_second": 431.746
    },
    {
      "epoch": 0.4222673362538733,
      "grad_norm": 3.6433022022247314,
      "learning_rate": 1.5798930481283424e-05,
      "loss": 0.1748078465461731,
      "num_input_tokens_seen": 1030018,
      "step": 988,
      "train_runtime": 2385.6021,
      "train_tokens_per_second": 431.764
    },
    {
      "epoch": 0.42269473234319904,
      "grad_norm": 3.15360164642334,
      "learning_rate": 1.5794652406417113e-05,
      "loss": 0.11847887933254242,
      "num_input_tokens_seen": 1030886,
      "step": 989,
      "train_runtime": 2387.6681,
      "train_tokens_per_second": 431.754
    },
    {
      "epoch": 0.42312212843252484,
      "grad_norm": 2.9097654819488525,
      "learning_rate": 1.5790374331550802e-05,
      "loss": 0.15133161842823029,
      "num_input_tokens_seen": 1032168,
      "step": 990,
      "train_runtime": 2389.7963,
      "train_tokens_per_second": 431.906
    },
    {
      "epoch": 0.42354952452185063,
      "grad_norm": 3.6200013160705566,
      "learning_rate": 1.5786096256684492e-05,
      "loss": 0.14536435902118683,
      "num_input_tokens_seen": 1032956,
      "step": 991,
      "train_runtime": 2398.4351,
      "train_tokens_per_second": 430.679
    },
    {
      "epoch": 0.4239769206111764,
      "grad_norm": 2.526433229446411,
      "learning_rate": 1.578181818181818e-05,
      "loss": 0.13084016740322113,
      "num_input_tokens_seen": 1033768,
      "step": 992,
      "train_runtime": 2400.4557,
      "train_tokens_per_second": 430.655
    },
    {
      "epoch": 0.4244043167005022,
      "grad_norm": 3.6554641723632812,
      "learning_rate": 1.577754010695187e-05,
      "loss": 0.14103198051452637,
      "num_input_tokens_seen": 1034492,
      "step": 993,
      "train_runtime": 2402.4641,
      "train_tokens_per_second": 430.596
    },
    {
      "epoch": 0.42483171278982795,
      "grad_norm": 3.3080387115478516,
      "learning_rate": 1.5773262032085564e-05,
      "loss": 0.12623611092567444,
      "num_input_tokens_seen": 1035644,
      "step": 994,
      "train_runtime": 2404.5783,
      "train_tokens_per_second": 430.697
    },
    {
      "epoch": 0.42525910887915375,
      "grad_norm": 2.4104950428009033,
      "learning_rate": 1.5768983957219253e-05,
      "loss": 0.11813364177942276,
      "num_input_tokens_seen": 1037086,
      "step": 995,
      "train_runtime": 2406.7565,
      "train_tokens_per_second": 430.906
    },
    {
      "epoch": 0.42568650496847954,
      "grad_norm": 3.64646053314209,
      "learning_rate": 1.5764705882352943e-05,
      "loss": 0.0654887780547142,
      "num_input_tokens_seen": 1037730,
      "step": 996,
      "train_runtime": 2408.705,
      "train_tokens_per_second": 430.825
    },
    {
      "epoch": 0.42611390105780533,
      "grad_norm": 2.640672206878662,
      "learning_rate": 1.5760427807486632e-05,
      "loss": 0.09134610742330551,
      "num_input_tokens_seen": 1038528,
      "step": 997,
      "train_runtime": 2410.7094,
      "train_tokens_per_second": 430.798
    },
    {
      "epoch": 0.4265412971471311,
      "grad_norm": 2.6327028274536133,
      "learning_rate": 1.575614973262032e-05,
      "loss": 0.11012723296880722,
      "num_input_tokens_seen": 1039386,
      "step": 998,
      "train_runtime": 2412.7598,
      "train_tokens_per_second": 430.787
    },
    {
      "epoch": 0.42696869323645686,
      "grad_norm": 3.207695722579956,
      "learning_rate": 1.575187165775401e-05,
      "loss": 0.1508389264345169,
      "num_input_tokens_seen": 1040338,
      "step": 999,
      "train_runtime": 2414.8065,
      "train_tokens_per_second": 430.816
    },
    {
      "epoch": 0.42739608932578266,
      "grad_norm": 4.508682727813721,
      "learning_rate": 1.57475935828877e-05,
      "loss": 0.24495993554592133,
      "num_input_tokens_seen": 1041176,
      "step": 1000,
      "train_runtime": 2416.8504,
      "train_tokens_per_second": 430.799
    },
    {
      "epoch": 0.42782348541510845,
      "grad_norm": 2.4895706176757812,
      "learning_rate": 1.574331550802139e-05,
      "loss": 0.11581746488809586,
      "num_input_tokens_seen": 1042066,
      "step": 1001,
      "train_runtime": 2418.9551,
      "train_tokens_per_second": 430.792
    },
    {
      "epoch": 0.42825088150443424,
      "grad_norm": 2.77655029296875,
      "learning_rate": 1.5739037433155083e-05,
      "loss": 0.21308037638664246,
      "num_input_tokens_seen": 1043372,
      "step": 1002,
      "train_runtime": 2421.1268,
      "train_tokens_per_second": 430.945
    },
    {
      "epoch": 0.42867827759376004,
      "grad_norm": 5.7979326248168945,
      "learning_rate": 1.5734759358288772e-05,
      "loss": 0.222947359085083,
      "num_input_tokens_seen": 1044332,
      "step": 1003,
      "train_runtime": 2423.17,
      "train_tokens_per_second": 430.978
    },
    {
      "epoch": 0.4291056736830858,
      "grad_norm": 2.3011908531188965,
      "learning_rate": 1.573048128342246e-05,
      "loss": 0.08945601433515549,
      "num_input_tokens_seen": 1045352,
      "step": 1004,
      "train_runtime": 2425.2196,
      "train_tokens_per_second": 431.034
    },
    {
      "epoch": 0.42953306977241157,
      "grad_norm": 3.217158794403076,
      "learning_rate": 1.572620320855615e-05,
      "loss": 0.18294289708137512,
      "num_input_tokens_seen": 1046092,
      "step": 1005,
      "train_runtime": 2427.3962,
      "train_tokens_per_second": 430.952
    },
    {
      "epoch": 0.42996046586173736,
      "grad_norm": 4.006203651428223,
      "learning_rate": 1.572192513368984e-05,
      "loss": 0.22173276543617249,
      "num_input_tokens_seen": 1046896,
      "step": 1006,
      "train_runtime": 2429.4375,
      "train_tokens_per_second": 430.921
    },
    {
      "epoch": 0.43038786195106316,
      "grad_norm": 3.962876319885254,
      "learning_rate": 1.571764705882353e-05,
      "loss": 0.18439872562885284,
      "num_input_tokens_seen": 1047566,
      "step": 1007,
      "train_runtime": 2431.4412,
      "train_tokens_per_second": 430.842
    },
    {
      "epoch": 0.43081525804038895,
      "grad_norm": 4.599574565887451,
      "learning_rate": 1.571336898395722e-05,
      "loss": 0.14483045041561127,
      "num_input_tokens_seen": 1048180,
      "step": 1008,
      "train_runtime": 2433.4211,
      "train_tokens_per_second": 430.743
    },
    {
      "epoch": 0.4312426541297147,
      "grad_norm": 4.2527570724487305,
      "learning_rate": 1.570909090909091e-05,
      "loss": 0.24055345356464386,
      "num_input_tokens_seen": 1049332,
      "step": 1009,
      "train_runtime": 2435.5341,
      "train_tokens_per_second": 430.843
    },
    {
      "epoch": 0.4316700502190405,
      "grad_norm": 4.76500129699707,
      "learning_rate": 1.57048128342246e-05,
      "loss": 0.13520678877830505,
      "num_input_tokens_seen": 1049910,
      "step": 1010,
      "train_runtime": 2437.4816,
      "train_tokens_per_second": 430.736
    },
    {
      "epoch": 0.4320974463083663,
      "grad_norm": 3.4047281742095947,
      "learning_rate": 1.570053475935829e-05,
      "loss": 0.1783725619316101,
      "num_input_tokens_seen": 1051190,
      "step": 1011,
      "train_runtime": 2439.6085,
      "train_tokens_per_second": 430.885
    },
    {
      "epoch": 0.43252484239769207,
      "grad_norm": 3.4894444942474365,
      "learning_rate": 1.569625668449198e-05,
      "loss": 0.22947156429290771,
      "num_input_tokens_seen": 1052216,
      "step": 1012,
      "train_runtime": 2441.7337,
      "train_tokens_per_second": 430.93
    },
    {
      "epoch": 0.43295223848701786,
      "grad_norm": 3.4204695224761963,
      "learning_rate": 1.569197860962567e-05,
      "loss": 0.15205669403076172,
      "num_input_tokens_seen": 1053108,
      "step": 1013,
      "train_runtime": 2443.7524,
      "train_tokens_per_second": 430.939
    },
    {
      "epoch": 0.43337963457634365,
      "grad_norm": 4.279384136199951,
      "learning_rate": 1.568770053475936e-05,
      "loss": 0.22293099761009216,
      "num_input_tokens_seen": 1053874,
      "step": 1014,
      "train_runtime": 2445.7765,
      "train_tokens_per_second": 430.895
    },
    {
      "epoch": 0.4338070306656694,
      "grad_norm": 1.7959322929382324,
      "learning_rate": 1.568342245989305e-05,
      "loss": 0.07155629992485046,
      "num_input_tokens_seen": 1055180,
      "step": 1015,
      "train_runtime": 2447.9586,
      "train_tokens_per_second": 431.045
    },
    {
      "epoch": 0.4342344267549952,
      "grad_norm": 3.1679439544677734,
      "learning_rate": 1.5679144385026738e-05,
      "loss": 0.10849887132644653,
      "num_input_tokens_seen": 1055904,
      "step": 1016,
      "train_runtime": 2449.9532,
      "train_tokens_per_second": 430.989
    },
    {
      "epoch": 0.434661822844321,
      "grad_norm": 2.921274423599243,
      "learning_rate": 1.5674866310160428e-05,
      "loss": 0.15974636375904083,
      "num_input_tokens_seen": 1056978,
      "step": 1017,
      "train_runtime": 2452.0714,
      "train_tokens_per_second": 431.055
    },
    {
      "epoch": 0.43508921893364677,
      "grad_norm": 4.978099346160889,
      "learning_rate": 1.567058823529412e-05,
      "loss": 0.21475082635879517,
      "num_input_tokens_seen": 1058052,
      "step": 1018,
      "train_runtime": 2454.1392,
      "train_tokens_per_second": 431.13
    },
    {
      "epoch": 0.43551661502297256,
      "grad_norm": 2.737354040145874,
      "learning_rate": 1.566631016042781e-05,
      "loss": 0.09027975797653198,
      "num_input_tokens_seen": 1058838,
      "step": 1019,
      "train_runtime": 2456.1161,
      "train_tokens_per_second": 431.103
    },
    {
      "epoch": 0.4359440111122983,
      "grad_norm": 3.2369279861450195,
      "learning_rate": 1.56620320855615e-05,
      "loss": 0.10335059463977814,
      "num_input_tokens_seen": 1059646,
      "step": 1020,
      "train_runtime": 2458.1705,
      "train_tokens_per_second": 431.071
    },
    {
      "epoch": 0.4363714072016241,
      "grad_norm": 2.6113617420196533,
      "learning_rate": 1.565775401069519e-05,
      "loss": 0.17476630210876465,
      "num_input_tokens_seen": 1060874,
      "step": 1021,
      "train_runtime": 2466.8843,
      "train_tokens_per_second": 430.046
    },
    {
      "epoch": 0.4367988032909499,
      "grad_norm": 6.517242908477783,
      "learning_rate": 1.5653475935828878e-05,
      "loss": 0.27782416343688965,
      "num_input_tokens_seen": 1061970,
      "step": 1022,
      "train_runtime": 2469.0171,
      "train_tokens_per_second": 430.119
    },
    {
      "epoch": 0.4372261993802757,
      "grad_norm": 2.8367512226104736,
      "learning_rate": 1.5649197860962568e-05,
      "loss": 0.2036033272743225,
      "num_input_tokens_seen": 1063252,
      "step": 1023,
      "train_runtime": 2471.1951,
      "train_tokens_per_second": 430.258
    },
    {
      "epoch": 0.4376535954696015,
      "grad_norm": 4.166869640350342,
      "learning_rate": 1.5644919786096257e-05,
      "loss": 0.1926107108592987,
      "num_input_tokens_seen": 1063910,
      "step": 1024,
      "train_runtime": 2473.1854,
      "train_tokens_per_second": 430.178
    },
    {
      "epoch": 0.4380809915589272,
      "grad_norm": 2.81599497795105,
      "learning_rate": 1.5640641711229947e-05,
      "loss": 0.11440451443195343,
      "num_input_tokens_seen": 1065248,
      "step": 1025,
      "train_runtime": 2475.3945,
      "train_tokens_per_second": 430.335
    },
    {
      "epoch": 0.438508387648253,
      "grad_norm": 2.8661296367645264,
      "learning_rate": 1.563636363636364e-05,
      "loss": 0.10879606008529663,
      "num_input_tokens_seen": 1065962,
      "step": 1026,
      "train_runtime": 2477.4197,
      "train_tokens_per_second": 430.271
    },
    {
      "epoch": 0.4389357837375788,
      "grad_norm": 2.863218069076538,
      "learning_rate": 1.563208556149733e-05,
      "loss": 0.1990147829055786,
      "num_input_tokens_seen": 1067372,
      "step": 1027,
      "train_runtime": 2479.6154,
      "train_tokens_per_second": 430.459
    },
    {
      "epoch": 0.4393631798269046,
      "grad_norm": 3.031972885131836,
      "learning_rate": 1.562780748663102e-05,
      "loss": 0.14785142242908478,
      "num_input_tokens_seen": 1068310,
      "step": 1028,
      "train_runtime": 2481.6769,
      "train_tokens_per_second": 430.479
    },
    {
      "epoch": 0.4397905759162304,
      "grad_norm": 2.5645766258239746,
      "learning_rate": 1.5623529411764708e-05,
      "loss": 0.11684894561767578,
      "num_input_tokens_seen": 1069292,
      "step": 1029,
      "train_runtime": 2483.8156,
      "train_tokens_per_second": 430.504
    },
    {
      "epoch": 0.4402179720055561,
      "grad_norm": 2.743408441543579,
      "learning_rate": 1.5619251336898397e-05,
      "loss": 0.09588809311389923,
      "num_input_tokens_seen": 1070098,
      "step": 1030,
      "train_runtime": 2485.8608,
      "train_tokens_per_second": 430.474
    },
    {
      "epoch": 0.4406453680948819,
      "grad_norm": 3.146395683288574,
      "learning_rate": 1.5614973262032087e-05,
      "loss": 0.15148451924324036,
      "num_input_tokens_seen": 1071068,
      "step": 1031,
      "train_runtime": 2487.9986,
      "train_tokens_per_second": 430.494
    },
    {
      "epoch": 0.4410727641842077,
      "grad_norm": 3.2485437393188477,
      "learning_rate": 1.5610695187165776e-05,
      "loss": 0.1452752947807312,
      "num_input_tokens_seen": 1072092,
      "step": 1032,
      "train_runtime": 2490.0908,
      "train_tokens_per_second": 430.543
    },
    {
      "epoch": 0.4415001602735335,
      "grad_norm": 3.1901021003723145,
      "learning_rate": 1.5606417112299466e-05,
      "loss": 0.11025238782167435,
      "num_input_tokens_seen": 1072790,
      "step": 1033,
      "train_runtime": 2492.1486,
      "train_tokens_per_second": 430.468
    },
    {
      "epoch": 0.4419275563628593,
      "grad_norm": 3.3421692848205566,
      "learning_rate": 1.560213903743316e-05,
      "loss": 0.14787673950195312,
      "num_input_tokens_seen": 1073782,
      "step": 1034,
      "train_runtime": 2494.2112,
      "train_tokens_per_second": 430.51
    },
    {
      "epoch": 0.44235495245218504,
      "grad_norm": 2.9150612354278564,
      "learning_rate": 1.5597860962566848e-05,
      "loss": 0.175185427069664,
      "num_input_tokens_seen": 1074934,
      "step": 1035,
      "train_runtime": 2496.3646,
      "train_tokens_per_second": 430.6
    },
    {
      "epoch": 0.44278234854151083,
      "grad_norm": 2.544585943222046,
      "learning_rate": 1.5593582887700537e-05,
      "loss": 0.1855897307395935,
      "num_input_tokens_seen": 1076546,
      "step": 1036,
      "train_runtime": 2498.6453,
      "train_tokens_per_second": 430.852
    },
    {
      "epoch": 0.4432097446308366,
      "grad_norm": 5.125703811645508,
      "learning_rate": 1.5589304812834227e-05,
      "loss": 0.18655644357204437,
      "num_input_tokens_seen": 1077670,
      "step": 1037,
      "train_runtime": 2500.7785,
      "train_tokens_per_second": 430.934
    },
    {
      "epoch": 0.4436371407201624,
      "grad_norm": 2.39841628074646,
      "learning_rate": 1.5585026737967916e-05,
      "loss": 0.11417564749717712,
      "num_input_tokens_seen": 1079310,
      "step": 1038,
      "train_runtime": 2503.0452,
      "train_tokens_per_second": 431.199
    },
    {
      "epoch": 0.4440645368094882,
      "grad_norm": 3.2729265689849854,
      "learning_rate": 1.5580748663101606e-05,
      "loss": 0.13362637162208557,
      "num_input_tokens_seen": 1080016,
      "step": 1039,
      "train_runtime": 2504.9995,
      "train_tokens_per_second": 431.144
    },
    {
      "epoch": 0.444491932898814,
      "grad_norm": 2.726480007171631,
      "learning_rate": 1.5576470588235295e-05,
      "loss": 0.14196565747261047,
      "num_input_tokens_seen": 1080900,
      "step": 1040,
      "train_runtime": 2507.0937,
      "train_tokens_per_second": 431.137
    },
    {
      "epoch": 0.44491932898813974,
      "grad_norm": 3.0665643215179443,
      "learning_rate": 1.5572192513368985e-05,
      "loss": 0.1690657138824463,
      "num_input_tokens_seen": 1082256,
      "step": 1041,
      "train_runtime": 2509.2713,
      "train_tokens_per_second": 431.303
    },
    {
      "epoch": 0.44534672507746553,
      "grad_norm": 2.6255617141723633,
      "learning_rate": 1.5567914438502677e-05,
      "loss": 0.11941884458065033,
      "num_input_tokens_seen": 1083402,
      "step": 1042,
      "train_runtime": 2511.3705,
      "train_tokens_per_second": 431.399
    },
    {
      "epoch": 0.4457741211667913,
      "grad_norm": 4.067086219787598,
      "learning_rate": 1.5563636363636367e-05,
      "loss": 0.1705426424741745,
      "num_input_tokens_seen": 1084648,
      "step": 1043,
      "train_runtime": 2513.4912,
      "train_tokens_per_second": 431.53
    },
    {
      "epoch": 0.4462015172561171,
      "grad_norm": 2.9490678310394287,
      "learning_rate": 1.5559358288770053e-05,
      "loss": 0.17178130149841309,
      "num_input_tokens_seen": 1085474,
      "step": 1044,
      "train_runtime": 2515.536,
      "train_tokens_per_second": 431.508
    },
    {
      "epoch": 0.4466289133454429,
      "grad_norm": 3.8613905906677246,
      "learning_rate": 1.5555080213903742e-05,
      "loss": 0.12446797639131546,
      "num_input_tokens_seen": 1086270,
      "step": 1045,
      "train_runtime": 2517.5803,
      "train_tokens_per_second": 431.474
    },
    {
      "epoch": 0.44705630943476865,
      "grad_norm": 3.081756830215454,
      "learning_rate": 1.5550802139037435e-05,
      "loss": 0.11352714896202087,
      "num_input_tokens_seen": 1086938,
      "step": 1046,
      "train_runtime": 2519.5362,
      "train_tokens_per_second": 431.404
    },
    {
      "epoch": 0.44748370552409444,
      "grad_norm": 5.03386926651001,
      "learning_rate": 1.5546524064171125e-05,
      "loss": 0.19863486289978027,
      "num_input_tokens_seen": 1087992,
      "step": 1047,
      "train_runtime": 2521.6272,
      "train_tokens_per_second": 431.464
    },
    {
      "epoch": 0.44791110161342024,
      "grad_norm": 2.281496286392212,
      "learning_rate": 1.5542245989304814e-05,
      "loss": 0.07257408648729324,
      "num_input_tokens_seen": 1088910,
      "step": 1048,
      "train_runtime": 2523.7002,
      "train_tokens_per_second": 431.474
    },
    {
      "epoch": 0.44833849770274603,
      "grad_norm": 2.7421557903289795,
      "learning_rate": 1.5537967914438503e-05,
      "loss": 0.11143845319747925,
      "num_input_tokens_seen": 1090138,
      "step": 1049,
      "train_runtime": 2525.8293,
      "train_tokens_per_second": 431.596
    },
    {
      "epoch": 0.4487658937920718,
      "grad_norm": 3.2605109214782715,
      "learning_rate": 1.5533689839572196e-05,
      "loss": 0.1240977793931961,
      "num_input_tokens_seen": 1091126,
      "step": 1050,
      "train_runtime": 2527.8936,
      "train_tokens_per_second": 431.634
    },
    {
      "epoch": 0.44919328988139756,
      "grad_norm": 4.209699630737305,
      "learning_rate": 1.5529411764705882e-05,
      "loss": 0.21886172890663147,
      "num_input_tokens_seen": 1092242,
      "step": 1051,
      "train_runtime": 2536.5567,
      "train_tokens_per_second": 430.6
    },
    {
      "epoch": 0.44962068597072335,
      "grad_norm": 2.4231152534484863,
      "learning_rate": 1.5525133689839572e-05,
      "loss": 0.12536178529262543,
      "num_input_tokens_seen": 1093496,
      "step": 1052,
      "train_runtime": 2538.6946,
      "train_tokens_per_second": 430.732
    },
    {
      "epoch": 0.45004808206004915,
      "grad_norm": 2.6356184482574463,
      "learning_rate": 1.552085561497326e-05,
      "loss": 0.1799614429473877,
      "num_input_tokens_seen": 1094654,
      "step": 1053,
      "train_runtime": 2540.8711,
      "train_tokens_per_second": 430.818
    },
    {
      "epoch": 0.45047547814937494,
      "grad_norm": 2.4675352573394775,
      "learning_rate": 1.5516577540106954e-05,
      "loss": 0.10288633406162262,
      "num_input_tokens_seen": 1095680,
      "step": 1054,
      "train_runtime": 2542.961,
      "train_tokens_per_second": 430.868
    },
    {
      "epoch": 0.45090287423870073,
      "grad_norm": 2.2310800552368164,
      "learning_rate": 1.5512299465240644e-05,
      "loss": 0.1007671132683754,
      "num_input_tokens_seen": 1097010,
      "step": 1055,
      "train_runtime": 2545.1425,
      "train_tokens_per_second": 431.021
    },
    {
      "epoch": 0.4513302703280265,
      "grad_norm": 3.2314248085021973,
      "learning_rate": 1.5508021390374333e-05,
      "loss": 0.2237885594367981,
      "num_input_tokens_seen": 1098134,
      "step": 1056,
      "train_runtime": 2547.2799,
      "train_tokens_per_second": 431.101
    },
    {
      "epoch": 0.45175766641735227,
      "grad_norm": 2.0864264965057373,
      "learning_rate": 1.5503743315508022e-05,
      "loss": 0.1292007714509964,
      "num_input_tokens_seen": 1099590,
      "step": 1057,
      "train_runtime": 2549.4666,
      "train_tokens_per_second": 431.302
    },
    {
      "epoch": 0.45218506250667806,
      "grad_norm": 1.9484848976135254,
      "learning_rate": 1.5499465240641712e-05,
      "loss": 0.11468847841024399,
      "num_input_tokens_seen": 1100842,
      "step": 1058,
      "train_runtime": 2551.6175,
      "train_tokens_per_second": 431.429
    },
    {
      "epoch": 0.45261245859600385,
      "grad_norm": 2.9358811378479004,
      "learning_rate": 1.54951871657754e-05,
      "loss": 0.2071388214826584,
      "num_input_tokens_seen": 1102114,
      "step": 1059,
      "train_runtime": 2553.7753,
      "train_tokens_per_second": 431.563
    },
    {
      "epoch": 0.45303985468532965,
      "grad_norm": 3.30251145362854,
      "learning_rate": 1.549090909090909e-05,
      "loss": 0.20812180638313293,
      "num_input_tokens_seen": 1103470,
      "step": 1060,
      "train_runtime": 2555.9323,
      "train_tokens_per_second": 431.729
    },
    {
      "epoch": 0.45346725077465544,
      "grad_norm": 2.4831864833831787,
      "learning_rate": 1.548663101604278e-05,
      "loss": 0.14195820689201355,
      "num_input_tokens_seen": 1104796,
      "step": 1061,
      "train_runtime": 2558.1064,
      "train_tokens_per_second": 431.88
    },
    {
      "epoch": 0.4538946468639812,
      "grad_norm": 3.6319377422332764,
      "learning_rate": 1.5482352941176473e-05,
      "loss": 0.13811862468719482,
      "num_input_tokens_seen": 1105546,
      "step": 1062,
      "train_runtime": 2560.0693,
      "train_tokens_per_second": 431.842
    },
    {
      "epoch": 0.45432204295330697,
      "grad_norm": 2.107161283493042,
      "learning_rate": 1.5478074866310162e-05,
      "loss": 0.10408823937177658,
      "num_input_tokens_seen": 1106758,
      "step": 1063,
      "train_runtime": 2562.2012,
      "train_tokens_per_second": 431.956
    },
    {
      "epoch": 0.45474943904263276,
      "grad_norm": 3.1638989448547363,
      "learning_rate": 1.5473796791443852e-05,
      "loss": 0.15363234281539917,
      "num_input_tokens_seen": 1107616,
      "step": 1064,
      "train_runtime": 2564.2235,
      "train_tokens_per_second": 431.95
    },
    {
      "epoch": 0.45517683513195856,
      "grad_norm": 7.830322265625,
      "learning_rate": 1.546951871657754e-05,
      "loss": 0.1365651786327362,
      "num_input_tokens_seen": 1108174,
      "step": 1065,
      "train_runtime": 2566.1901,
      "train_tokens_per_second": 431.836
    },
    {
      "epoch": 0.45560423122128435,
      "grad_norm": 3.503216505050659,
      "learning_rate": 1.546524064171123e-05,
      "loss": 0.2846662402153015,
      "num_input_tokens_seen": 1109014,
      "step": 1066,
      "train_runtime": 2568.2761,
      "train_tokens_per_second": 431.813
    },
    {
      "epoch": 0.4560316273106101,
      "grad_norm": 3.514633893966675,
      "learning_rate": 1.546096256684492e-05,
      "loss": 0.19525128602981567,
      "num_input_tokens_seen": 1110314,
      "step": 1067,
      "train_runtime": 2570.4397,
      "train_tokens_per_second": 431.955
    },
    {
      "epoch": 0.4564590233999359,
      "grad_norm": 2.669102191925049,
      "learning_rate": 1.545668449197861e-05,
      "loss": 0.14375221729278564,
      "num_input_tokens_seen": 1111590,
      "step": 1068,
      "train_runtime": 2572.591,
      "train_tokens_per_second": 432.09
    },
    {
      "epoch": 0.4568864194892617,
      "grad_norm": 3.464954137802124,
      "learning_rate": 1.54524064171123e-05,
      "loss": 0.13595668971538544,
      "num_input_tokens_seen": 1112282,
      "step": 1069,
      "train_runtime": 2574.593,
      "train_tokens_per_second": 432.022
    },
    {
      "epoch": 0.45731381557858747,
      "grad_norm": 1.8810628652572632,
      "learning_rate": 1.5448128342245992e-05,
      "loss": 0.08300885558128357,
      "num_input_tokens_seen": 1113702,
      "step": 1070,
      "train_runtime": 2576.7576,
      "train_tokens_per_second": 432.211
    },
    {
      "epoch": 0.45774121166791326,
      "grad_norm": 2.5656237602233887,
      "learning_rate": 1.544385026737968e-05,
      "loss": 0.11466950923204422,
      "num_input_tokens_seen": 1114542,
      "step": 1071,
      "train_runtime": 2578.8251,
      "train_tokens_per_second": 432.19
    },
    {
      "epoch": 0.458168607757239,
      "grad_norm": 3.370588779449463,
      "learning_rate": 1.543957219251337e-05,
      "loss": 0.18732677400112152,
      "num_input_tokens_seen": 1115466,
      "step": 1072,
      "train_runtime": 2580.9135,
      "train_tokens_per_second": 432.198
    },
    {
      "epoch": 0.4585960038465648,
      "grad_norm": 3.3462183475494385,
      "learning_rate": 1.543529411764706e-05,
      "loss": 0.14999401569366455,
      "num_input_tokens_seen": 1116862,
      "step": 1073,
      "train_runtime": 2583.0996,
      "train_tokens_per_second": 432.373
    },
    {
      "epoch": 0.4590233999358906,
      "grad_norm": 2.879754066467285,
      "learning_rate": 1.543101604278075e-05,
      "loss": 0.13676685094833374,
      "num_input_tokens_seen": 1117870,
      "step": 1074,
      "train_runtime": 2585.1904,
      "train_tokens_per_second": 432.413
    },
    {
      "epoch": 0.4594507960252164,
      "grad_norm": 3.713593006134033,
      "learning_rate": 1.542673796791444e-05,
      "loss": 0.18806308507919312,
      "num_input_tokens_seen": 1118678,
      "step": 1075,
      "train_runtime": 2587.2352,
      "train_tokens_per_second": 432.384
    },
    {
      "epoch": 0.4598781921145422,
      "grad_norm": 3.257934093475342,
      "learning_rate": 1.542245989304813e-05,
      "loss": 0.17128415405750275,
      "num_input_tokens_seen": 1119906,
      "step": 1076,
      "train_runtime": 2589.3376,
      "train_tokens_per_second": 432.507
    },
    {
      "epoch": 0.4603055882038679,
      "grad_norm": 2.9704718589782715,
      "learning_rate": 1.5418181818181818e-05,
      "loss": 0.1316772848367691,
      "num_input_tokens_seen": 1120874,
      "step": 1077,
      "train_runtime": 2591.4121,
      "train_tokens_per_second": 432.534
    },
    {
      "epoch": 0.4607329842931937,
      "grad_norm": 3.433804750442505,
      "learning_rate": 1.541390374331551e-05,
      "loss": 0.17128492891788483,
      "num_input_tokens_seen": 1122506,
      "step": 1078,
      "train_runtime": 2593.6091,
      "train_tokens_per_second": 432.797
    },
    {
      "epoch": 0.4611603803825195,
      "grad_norm": 2.7072598934173584,
      "learning_rate": 1.54096256684492e-05,
      "loss": 0.08850735425949097,
      "num_input_tokens_seen": 1123354,
      "step": 1079,
      "train_runtime": 2595.6782,
      "train_tokens_per_second": 432.779
    },
    {
      "epoch": 0.4615877764718453,
      "grad_norm": 2.266479253768921,
      "learning_rate": 1.540534759358289e-05,
      "loss": 0.10116958618164062,
      "num_input_tokens_seen": 1124486,
      "step": 1080,
      "train_runtime": 2597.7872,
      "train_tokens_per_second": 432.863
    },
    {
      "epoch": 0.4620151725611711,
      "grad_norm": 3.3960344791412354,
      "learning_rate": 1.540106951871658e-05,
      "loss": 0.1324056088924408,
      "num_input_tokens_seen": 1125286,
      "step": 1081,
      "train_runtime": 2606.5205,
      "train_tokens_per_second": 431.72
    },
    {
      "epoch": 0.4624425686504968,
      "grad_norm": 3.4813549518585205,
      "learning_rate": 1.539679144385027e-05,
      "loss": 0.13071005046367645,
      "num_input_tokens_seen": 1126018,
      "step": 1082,
      "train_runtime": 2608.5292,
      "train_tokens_per_second": 431.668
    },
    {
      "epoch": 0.4628699647398226,
      "grad_norm": 2.5402848720550537,
      "learning_rate": 1.5392513368983958e-05,
      "loss": 0.11979816854000092,
      "num_input_tokens_seen": 1127760,
      "step": 1083,
      "train_runtime": 2610.7841,
      "train_tokens_per_second": 431.962
    },
    {
      "epoch": 0.4632973608291484,
      "grad_norm": 3.59625244140625,
      "learning_rate": 1.5388235294117648e-05,
      "loss": 0.1375684142112732,
      "num_input_tokens_seen": 1128962,
      "step": 1084,
      "train_runtime": 2612.9241,
      "train_tokens_per_second": 432.068
    },
    {
      "epoch": 0.4637247569184742,
      "grad_norm": 3.143054246902466,
      "learning_rate": 1.5383957219251337e-05,
      "loss": 0.15120042860507965,
      "num_input_tokens_seen": 1129876,
      "step": 1085,
      "train_runtime": 2614.9657,
      "train_tokens_per_second": 432.081
    },
    {
      "epoch": 0.4641521530078,
      "grad_norm": 2.635488510131836,
      "learning_rate": 1.537967914438503e-05,
      "loss": 0.10741516947746277,
      "num_input_tokens_seen": 1130876,
      "step": 1086,
      "train_runtime": 2617.0489,
      "train_tokens_per_second": 432.119
    },
    {
      "epoch": 0.4645795490971258,
      "grad_norm": 3.1969196796417236,
      "learning_rate": 1.537540106951872e-05,
      "loss": 0.17901310324668884,
      "num_input_tokens_seen": 1131882,
      "step": 1087,
      "train_runtime": 2619.0932,
      "train_tokens_per_second": 432.166
    },
    {
      "epoch": 0.4650069451864515,
      "grad_norm": 4.070721626281738,
      "learning_rate": 1.537112299465241e-05,
      "loss": 0.16284756362438202,
      "num_input_tokens_seen": 1132664,
      "step": 1088,
      "train_runtime": 2621.146,
      "train_tokens_per_second": 432.125
    },
    {
      "epoch": 0.4654343412757773,
      "grad_norm": 2.8047432899475098,
      "learning_rate": 1.5366844919786098e-05,
      "loss": 0.1308150589466095,
      "num_input_tokens_seen": 1133980,
      "step": 1089,
      "train_runtime": 2623.3297,
      "train_tokens_per_second": 432.267
    },
    {
      "epoch": 0.4658617373651031,
      "grad_norm": 2.700371742248535,
      "learning_rate": 1.5362566844919788e-05,
      "loss": 0.060718096792697906,
      "num_input_tokens_seen": 1134990,
      "step": 1090,
      "train_runtime": 2625.3968,
      "train_tokens_per_second": 432.312
    },
    {
      "epoch": 0.4662891334544289,
      "grad_norm": 2.8845772743225098,
      "learning_rate": 1.5358288770053477e-05,
      "loss": 0.1135871410369873,
      "num_input_tokens_seen": 1135714,
      "step": 1091,
      "train_runtime": 2627.3711,
      "train_tokens_per_second": 432.263
    },
    {
      "epoch": 0.4667165295437547,
      "grad_norm": 3.114514112472534,
      "learning_rate": 1.5354010695187167e-05,
      "loss": 0.1570318341255188,
      "num_input_tokens_seen": 1136538,
      "step": 1092,
      "train_runtime": 2629.3806,
      "train_tokens_per_second": 432.246
    },
    {
      "epoch": 0.46714392563308044,
      "grad_norm": 3.0474343299865723,
      "learning_rate": 1.5349732620320856e-05,
      "loss": 0.147725448012352,
      "num_input_tokens_seen": 1137472,
      "step": 1093,
      "train_runtime": 2631.5313,
      "train_tokens_per_second": 432.247
    },
    {
      "epoch": 0.46757132172240623,
      "grad_norm": 4.396843910217285,
      "learning_rate": 1.5345454545454545e-05,
      "loss": 0.17232832312583923,
      "num_input_tokens_seen": 1138134,
      "step": 1094,
      "train_runtime": 2633.5612,
      "train_tokens_per_second": 432.165
    },
    {
      "epoch": 0.467998717811732,
      "grad_norm": 2.4748077392578125,
      "learning_rate": 1.5341176470588238e-05,
      "loss": 0.15381349623203278,
      "num_input_tokens_seen": 1139682,
      "step": 1095,
      "train_runtime": 2635.8125,
      "train_tokens_per_second": 432.384
    },
    {
      "epoch": 0.4684261139010578,
      "grad_norm": 3.068317413330078,
      "learning_rate": 1.5336898395721928e-05,
      "loss": 0.10474956035614014,
      "num_input_tokens_seen": 1140220,
      "step": 1096,
      "train_runtime": 2637.7737,
      "train_tokens_per_second": 432.266
    },
    {
      "epoch": 0.4688535099903836,
      "grad_norm": 3.164912223815918,
      "learning_rate": 1.5332620320855617e-05,
      "loss": 0.20970091223716736,
      "num_input_tokens_seen": 1141284,
      "step": 1097,
      "train_runtime": 2639.8805,
      "train_tokens_per_second": 432.324
    },
    {
      "epoch": 0.46928090607970935,
      "grad_norm": 2.8378591537475586,
      "learning_rate": 1.5328342245989307e-05,
      "loss": 0.16146932542324066,
      "num_input_tokens_seen": 1142276,
      "step": 1098,
      "train_runtime": 2641.9333,
      "train_tokens_per_second": 432.364
    },
    {
      "epoch": 0.46970830216903514,
      "grad_norm": 1.890136957168579,
      "learning_rate": 1.5324064171122996e-05,
      "loss": 0.1063518226146698,
      "num_input_tokens_seen": 1143250,
      "step": 1099,
      "train_runtime": 2644.0071,
      "train_tokens_per_second": 432.393
    },
    {
      "epoch": 0.47013569825836093,
      "grad_norm": 2.5111641883850098,
      "learning_rate": 1.5319786096256686e-05,
      "loss": 0.130143940448761,
      "num_input_tokens_seen": 1144518,
      "step": 1100,
      "train_runtime": 2646.1267,
      "train_tokens_per_second": 432.526
    },
    {
      "epoch": 0.47056309434768673,
      "grad_norm": 5.517326831817627,
      "learning_rate": 1.5315508021390375e-05,
      "loss": 0.1228143498301506,
      "num_input_tokens_seen": 1145362,
      "step": 1101,
      "train_runtime": 2648.1678,
      "train_tokens_per_second": 432.511
    },
    {
      "epoch": 0.4709904904370125,
      "grad_norm": 3.1789982318878174,
      "learning_rate": 1.5311229946524064e-05,
      "loss": 0.14044930040836334,
      "num_input_tokens_seen": 1146036,
      "step": 1102,
      "train_runtime": 2650.1478,
      "train_tokens_per_second": 432.442
    },
    {
      "epoch": 0.47141788652633826,
      "grad_norm": 3.9956135749816895,
      "learning_rate": 1.5306951871657757e-05,
      "loss": 0.20187225937843323,
      "num_input_tokens_seen": 1146846,
      "step": 1103,
      "train_runtime": 2652.1751,
      "train_tokens_per_second": 432.417
    },
    {
      "epoch": 0.47184528261566405,
      "grad_norm": 2.8002490997314453,
      "learning_rate": 1.5302673796791447e-05,
      "loss": 0.10556461662054062,
      "num_input_tokens_seen": 1147790,
      "step": 1104,
      "train_runtime": 2654.2198,
      "train_tokens_per_second": 432.44
    },
    {
      "epoch": 0.47227267870498985,
      "grad_norm": 4.445836067199707,
      "learning_rate": 1.5298395721925136e-05,
      "loss": 0.17272546887397766,
      "num_input_tokens_seen": 1148852,
      "step": 1105,
      "train_runtime": 2656.2914,
      "train_tokens_per_second": 432.502
    },
    {
      "epoch": 0.47270007479431564,
      "grad_norm": 2.8741371631622314,
      "learning_rate": 1.5294117647058822e-05,
      "loss": 0.1411958634853363,
      "num_input_tokens_seen": 1149814,
      "step": 1106,
      "train_runtime": 2658.3826,
      "train_tokens_per_second": 432.524
    },
    {
      "epoch": 0.47312747088364143,
      "grad_norm": 2.783771276473999,
      "learning_rate": 1.5289839572192515e-05,
      "loss": 0.18507781624794006,
      "num_input_tokens_seen": 1150958,
      "step": 1107,
      "train_runtime": 2660.5235,
      "train_tokens_per_second": 432.606
    },
    {
      "epoch": 0.47355486697296717,
      "grad_norm": 3.2462594509124756,
      "learning_rate": 1.5285561497326204e-05,
      "loss": 0.11716610938310623,
      "num_input_tokens_seen": 1152078,
      "step": 1108,
      "train_runtime": 2662.6088,
      "train_tokens_per_second": 432.688
    },
    {
      "epoch": 0.47398226306229296,
      "grad_norm": 3.969907283782959,
      "learning_rate": 1.5281283422459894e-05,
      "loss": 0.1353527307510376,
      "num_input_tokens_seen": 1153404,
      "step": 1109,
      "train_runtime": 2664.7814,
      "train_tokens_per_second": 432.832
    },
    {
      "epoch": 0.47440965915161876,
      "grad_norm": 3.9630134105682373,
      "learning_rate": 1.5277005347593583e-05,
      "loss": 0.2100336104631424,
      "num_input_tokens_seen": 1154246,
      "step": 1110,
      "train_runtime": 2666.823,
      "train_tokens_per_second": 432.817
    },
    {
      "epoch": 0.47483705524094455,
      "grad_norm": 4.457049369812012,
      "learning_rate": 1.5272727272727276e-05,
      "loss": 0.24514269828796387,
      "num_input_tokens_seen": 1155348,
      "step": 1111,
      "train_runtime": 2675.5169,
      "train_tokens_per_second": 431.822
    },
    {
      "epoch": 0.47526445133027034,
      "grad_norm": 2.624235153198242,
      "learning_rate": 1.5268449197860962e-05,
      "loss": 0.10719535499811172,
      "num_input_tokens_seen": 1156148,
      "step": 1112,
      "train_runtime": 2677.5034,
      "train_tokens_per_second": 431.801
    },
    {
      "epoch": 0.47569184741959614,
      "grad_norm": 3.359612226486206,
      "learning_rate": 1.5264171122994652e-05,
      "loss": 0.15921300649642944,
      "num_input_tokens_seen": 1156964,
      "step": 1113,
      "train_runtime": 2679.5169,
      "train_tokens_per_second": 431.781
    },
    {
      "epoch": 0.4761192435089219,
      "grad_norm": 2.3543784618377686,
      "learning_rate": 1.525989304812834e-05,
      "loss": 0.09240389615297318,
      "num_input_tokens_seen": 1157788,
      "step": 1114,
      "train_runtime": 2681.5775,
      "train_tokens_per_second": 431.756
    },
    {
      "epoch": 0.47654663959824767,
      "grad_norm": 2.728663206100464,
      "learning_rate": 1.5255614973262034e-05,
      "loss": 0.10826657712459564,
      "num_input_tokens_seen": 1159104,
      "step": 1115,
      "train_runtime": 2683.7424,
      "train_tokens_per_second": 431.898
    },
    {
      "epoch": 0.47697403568757346,
      "grad_norm": 2.6188294887542725,
      "learning_rate": 1.5251336898395723e-05,
      "loss": 0.0742989256978035,
      "num_input_tokens_seen": 1160388,
      "step": 1116,
      "train_runtime": 2685.9199,
      "train_tokens_per_second": 432.026
    },
    {
      "epoch": 0.47740143177689925,
      "grad_norm": 3.8709871768951416,
      "learning_rate": 1.5247058823529413e-05,
      "loss": 0.16443318128585815,
      "num_input_tokens_seen": 1161552,
      "step": 1117,
      "train_runtime": 2688.0382,
      "train_tokens_per_second": 432.119
    },
    {
      "epoch": 0.47782882786622505,
      "grad_norm": 4.712742328643799,
      "learning_rate": 1.5242780748663102e-05,
      "loss": 0.24033494293689728,
      "num_input_tokens_seen": 1162296,
      "step": 1118,
      "train_runtime": 2690.0499,
      "train_tokens_per_second": 432.072
    },
    {
      "epoch": 0.4782562239555508,
      "grad_norm": 2.0754475593566895,
      "learning_rate": 1.5238502673796793e-05,
      "loss": 0.07046375423669815,
      "num_input_tokens_seen": 1163414,
      "step": 1119,
      "train_runtime": 2692.1436,
      "train_tokens_per_second": 432.152
    },
    {
      "epoch": 0.4786836200448766,
      "grad_norm": 3.374993324279785,
      "learning_rate": 1.5234224598930483e-05,
      "loss": 0.1427287757396698,
      "num_input_tokens_seen": 1164268,
      "step": 1120,
      "train_runtime": 2694.1787,
      "train_tokens_per_second": 432.142
    },
    {
      "epoch": 0.47911101613420237,
      "grad_norm": 4.995404243469238,
      "learning_rate": 1.5229946524064172e-05,
      "loss": 0.1521342247724533,
      "num_input_tokens_seen": 1164864,
      "step": 1121,
      "train_runtime": 2696.1585,
      "train_tokens_per_second": 432.046
    },
    {
      "epoch": 0.47953841222352817,
      "grad_norm": 2.663304328918457,
      "learning_rate": 1.5225668449197862e-05,
      "loss": 0.12766656279563904,
      "num_input_tokens_seen": 1165722,
      "step": 1122,
      "train_runtime": 2698.2046,
      "train_tokens_per_second": 432.036
    },
    {
      "epoch": 0.47996580831285396,
      "grad_norm": 3.3409948348999023,
      "learning_rate": 1.5221390374331553e-05,
      "loss": 0.18630322813987732,
      "num_input_tokens_seen": 1166924,
      "step": 1123,
      "train_runtime": 2700.3715,
      "train_tokens_per_second": 432.135
    },
    {
      "epoch": 0.4803932044021797,
      "grad_norm": 3.103919267654419,
      "learning_rate": 1.5217112299465242e-05,
      "loss": 0.10242228955030441,
      "num_input_tokens_seen": 1168254,
      "step": 1124,
      "train_runtime": 2702.5499,
      "train_tokens_per_second": 432.278
    },
    {
      "epoch": 0.4808206004915055,
      "grad_norm": 4.313792705535889,
      "learning_rate": 1.5212834224598932e-05,
      "loss": 0.15389426052570343,
      "num_input_tokens_seen": 1168950,
      "step": 1125,
      "train_runtime": 2704.5885,
      "train_tokens_per_second": 432.21
    },
    {
      "epoch": 0.4812479965808313,
      "grad_norm": 2.292739152908325,
      "learning_rate": 1.5208556149732621e-05,
      "loss": 0.08169770240783691,
      "num_input_tokens_seen": 1169718,
      "step": 1126,
      "train_runtime": 2706.5988,
      "train_tokens_per_second": 432.173
    },
    {
      "epoch": 0.4816753926701571,
      "grad_norm": 2.8973989486694336,
      "learning_rate": 1.5204278074866312e-05,
      "loss": 0.10626824200153351,
      "num_input_tokens_seen": 1171080,
      "step": 1127,
      "train_runtime": 2708.7664,
      "train_tokens_per_second": 432.33
    },
    {
      "epoch": 0.48210278875948287,
      "grad_norm": 3.474763870239258,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.2161523699760437,
      "num_input_tokens_seen": 1171958,
      "step": 1128,
      "train_runtime": 2710.8134,
      "train_tokens_per_second": 432.327
    },
    {
      "epoch": 0.4825301848488086,
      "grad_norm": 3.202106475830078,
      "learning_rate": 1.5195721925133691e-05,
      "loss": 0.1991758793592453,
      "num_input_tokens_seen": 1173202,
      "step": 1129,
      "train_runtime": 2712.9734,
      "train_tokens_per_second": 432.441
    },
    {
      "epoch": 0.4829575809381344,
      "grad_norm": 4.2022809982299805,
      "learning_rate": 1.519144385026738e-05,
      "loss": 0.2232508510351181,
      "num_input_tokens_seen": 1173970,
      "step": 1130,
      "train_runtime": 2715.0167,
      "train_tokens_per_second": 432.399
    },
    {
      "epoch": 0.4833849770274602,
      "grad_norm": 2.7030551433563232,
      "learning_rate": 1.5187165775401072e-05,
      "loss": 0.08495761454105377,
      "num_input_tokens_seen": 1174890,
      "step": 1131,
      "train_runtime": 2717.114,
      "train_tokens_per_second": 432.404
    },
    {
      "epoch": 0.483812373116786,
      "grad_norm": 3.4785704612731934,
      "learning_rate": 1.5182887700534761e-05,
      "loss": 0.1914873868227005,
      "num_input_tokens_seen": 1175908,
      "step": 1132,
      "train_runtime": 2719.2131,
      "train_tokens_per_second": 432.444
    },
    {
      "epoch": 0.4842397692061118,
      "grad_norm": 3.721217632293701,
      "learning_rate": 1.517860962566845e-05,
      "loss": 0.16007255017757416,
      "num_input_tokens_seen": 1176818,
      "step": 1133,
      "train_runtime": 2721.2342,
      "train_tokens_per_second": 432.457
    },
    {
      "epoch": 0.4846671652954376,
      "grad_norm": 2.4107444286346436,
      "learning_rate": 1.517433155080214e-05,
      "loss": 0.12073090672492981,
      "num_input_tokens_seen": 1178280,
      "step": 1134,
      "train_runtime": 2723.4186,
      "train_tokens_per_second": 432.647
    },
    {
      "epoch": 0.4850945613847633,
      "grad_norm": 3.2907817363739014,
      "learning_rate": 1.5170053475935831e-05,
      "loss": 0.13351722061634064,
      "num_input_tokens_seen": 1179326,
      "step": 1135,
      "train_runtime": 2725.4962,
      "train_tokens_per_second": 432.701
    },
    {
      "epoch": 0.4855219574740891,
      "grad_norm": 3.914668560028076,
      "learning_rate": 1.516577540106952e-05,
      "loss": 0.1750204861164093,
      "num_input_tokens_seen": 1180014,
      "step": 1136,
      "train_runtime": 2727.4869,
      "train_tokens_per_second": 432.638
    },
    {
      "epoch": 0.4859493535634149,
      "grad_norm": 2.2116777896881104,
      "learning_rate": 1.516149732620321e-05,
      "loss": 0.12250398099422455,
      "num_input_tokens_seen": 1180960,
      "step": 1137,
      "train_runtime": 2729.5648,
      "train_tokens_per_second": 432.655
    },
    {
      "epoch": 0.4863767496527407,
      "grad_norm": 4.321115970611572,
      "learning_rate": 1.5157219251336898e-05,
      "loss": 0.21933779120445251,
      "num_input_tokens_seen": 1182068,
      "step": 1138,
      "train_runtime": 2731.6408,
      "train_tokens_per_second": 432.732
    },
    {
      "epoch": 0.4868041457420665,
      "grad_norm": 3.981973171234131,
      "learning_rate": 1.515294117647059e-05,
      "loss": 0.217737078666687,
      "num_input_tokens_seen": 1183120,
      "step": 1139,
      "train_runtime": 2733.7497,
      "train_tokens_per_second": 432.783
    },
    {
      "epoch": 0.4872315418313922,
      "grad_norm": 2.112104892730713,
      "learning_rate": 1.514866310160428e-05,
      "loss": 0.10334925353527069,
      "num_input_tokens_seen": 1183972,
      "step": 1140,
      "train_runtime": 2735.774,
      "train_tokens_per_second": 432.774
    },
    {
      "epoch": 0.487658937920718,
      "grad_norm": 2.6317176818847656,
      "learning_rate": 1.5144385026737968e-05,
      "loss": 0.13040843605995178,
      "num_input_tokens_seen": 1185140,
      "step": 1141,
      "train_runtime": 2744.1302,
      "train_tokens_per_second": 431.882
    },
    {
      "epoch": 0.4880863340100438,
      "grad_norm": 3.8773531913757324,
      "learning_rate": 1.5140106951871657e-05,
      "loss": 0.19817647337913513,
      "num_input_tokens_seen": 1185734,
      "step": 1142,
      "train_runtime": 2746.1518,
      "train_tokens_per_second": 431.78
    },
    {
      "epoch": 0.4885137300993696,
      "grad_norm": 2.8795385360717773,
      "learning_rate": 1.513582887700535e-05,
      "loss": 0.11026828736066818,
      "num_input_tokens_seen": 1186790,
      "step": 1143,
      "train_runtime": 2748.2465,
      "train_tokens_per_second": 431.835
    },
    {
      "epoch": 0.4889411261886954,
      "grad_norm": 3.0057780742645264,
      "learning_rate": 1.513155080213904e-05,
      "loss": 0.14564253389835358,
      "num_input_tokens_seen": 1188504,
      "step": 1144,
      "train_runtime": 2750.4822,
      "train_tokens_per_second": 432.107
    },
    {
      "epoch": 0.48936852227802113,
      "grad_norm": 3.8096261024475098,
      "learning_rate": 1.5127272727272728e-05,
      "loss": 0.1662902534008026,
      "num_input_tokens_seen": 1189264,
      "step": 1145,
      "train_runtime": 2752.5067,
      "train_tokens_per_second": 432.066
    },
    {
      "epoch": 0.4897959183673469,
      "grad_norm": 2.5535879135131836,
      "learning_rate": 1.5122994652406417e-05,
      "loss": 0.1270386278629303,
      "num_input_tokens_seen": 1190702,
      "step": 1146,
      "train_runtime": 2754.7312,
      "train_tokens_per_second": 432.239
    },
    {
      "epoch": 0.4902233144566727,
      "grad_norm": 2.360358238220215,
      "learning_rate": 1.511871657754011e-05,
      "loss": 0.08077384531497955,
      "num_input_tokens_seen": 1191380,
      "step": 1147,
      "train_runtime": 2756.7204,
      "train_tokens_per_second": 432.173
    },
    {
      "epoch": 0.4906507105459985,
      "grad_norm": 2.111539363861084,
      "learning_rate": 1.5114438502673798e-05,
      "loss": 0.09322850406169891,
      "num_input_tokens_seen": 1192598,
      "step": 1148,
      "train_runtime": 2758.8433,
      "train_tokens_per_second": 432.282
    },
    {
      "epoch": 0.4910781066353243,
      "grad_norm": 3.5963425636291504,
      "learning_rate": 1.5110160427807487e-05,
      "loss": 0.12126210331916809,
      "num_input_tokens_seen": 1193422,
      "step": 1149,
      "train_runtime": 2760.9023,
      "train_tokens_per_second": 432.258
    },
    {
      "epoch": 0.49150550272465005,
      "grad_norm": 2.635592222213745,
      "learning_rate": 1.5105882352941176e-05,
      "loss": 0.1305210292339325,
      "num_input_tokens_seen": 1194250,
      "step": 1150,
      "train_runtime": 2762.925,
      "train_tokens_per_second": 432.241
    },
    {
      "epoch": 0.49193289881397584,
      "grad_norm": 2.4423794746398926,
      "learning_rate": 1.5101604278074868e-05,
      "loss": 0.09712070971727371,
      "num_input_tokens_seen": 1195288,
      "step": 1151,
      "train_runtime": 2765.0006,
      "train_tokens_per_second": 432.292
    },
    {
      "epoch": 0.49236029490330163,
      "grad_norm": 2.4243600368499756,
      "learning_rate": 1.5097326203208557e-05,
      "loss": 0.13342250883579254,
      "num_input_tokens_seen": 1196658,
      "step": 1152,
      "train_runtime": 2767.1455,
      "train_tokens_per_second": 432.452
    },
    {
      "epoch": 0.4927876909926274,
      "grad_norm": 2.272446870803833,
      "learning_rate": 1.5093048128342246e-05,
      "loss": 0.1295868456363678,
      "num_input_tokens_seen": 1197858,
      "step": 1153,
      "train_runtime": 2769.2727,
      "train_tokens_per_second": 432.553
    },
    {
      "epoch": 0.4932150870819532,
      "grad_norm": 3.037247896194458,
      "learning_rate": 1.5088770053475936e-05,
      "loss": 0.10061538964509964,
      "num_input_tokens_seen": 1198460,
      "step": 1154,
      "train_runtime": 2771.4284,
      "train_tokens_per_second": 432.434
    },
    {
      "epoch": 0.49364248317127896,
      "grad_norm": 2.932354211807251,
      "learning_rate": 1.5084491978609627e-05,
      "loss": 0.14099803566932678,
      "num_input_tokens_seen": 1199602,
      "step": 1155,
      "train_runtime": 2773.5479,
      "train_tokens_per_second": 432.515
    },
    {
      "epoch": 0.49406987926060475,
      "grad_norm": 3.6686477661132812,
      "learning_rate": 1.5080213903743316e-05,
      "loss": 0.13770988583564758,
      "num_input_tokens_seen": 1200682,
      "step": 1156,
      "train_runtime": 2775.6376,
      "train_tokens_per_second": 432.579
    },
    {
      "epoch": 0.49449727534993054,
      "grad_norm": 2.9927093982696533,
      "learning_rate": 1.5075935828877006e-05,
      "loss": 0.1364351361989975,
      "num_input_tokens_seen": 1201562,
      "step": 1157,
      "train_runtime": 2777.6922,
      "train_tokens_per_second": 432.576
    },
    {
      "epoch": 0.49492467143925634,
      "grad_norm": 2.8519723415374756,
      "learning_rate": 1.5071657754010695e-05,
      "loss": 0.09951236099004745,
      "num_input_tokens_seen": 1202292,
      "step": 1158,
      "train_runtime": 2779.7434,
      "train_tokens_per_second": 432.519
    },
    {
      "epoch": 0.49535206752858213,
      "grad_norm": 4.1219964027404785,
      "learning_rate": 1.5067379679144387e-05,
      "loss": 0.24936291575431824,
      "num_input_tokens_seen": 1203078,
      "step": 1159,
      "train_runtime": 2781.7782,
      "train_tokens_per_second": 432.485
    },
    {
      "epoch": 0.4957794636179079,
      "grad_norm": 3.9951977729797363,
      "learning_rate": 1.5063101604278076e-05,
      "loss": 0.20271098613739014,
      "num_input_tokens_seen": 1204314,
      "step": 1160,
      "train_runtime": 2783.9284,
      "train_tokens_per_second": 432.595
    },
    {
      "epoch": 0.49620685970723366,
      "grad_norm": 2.3672826290130615,
      "learning_rate": 1.5058823529411765e-05,
      "loss": 0.14124973118305206,
      "num_input_tokens_seen": 1205414,
      "step": 1161,
      "train_runtime": 2786.0063,
      "train_tokens_per_second": 432.667
    },
    {
      "epoch": 0.49663425579655945,
      "grad_norm": 2.570695638656616,
      "learning_rate": 1.5054545454545455e-05,
      "loss": 0.10842475295066833,
      "num_input_tokens_seen": 1206124,
      "step": 1162,
      "train_runtime": 2788.0188,
      "train_tokens_per_second": 432.61
    },
    {
      "epoch": 0.49706165188588525,
      "grad_norm": 3.4628758430480957,
      "learning_rate": 1.5050267379679146e-05,
      "loss": 0.20646989345550537,
      "num_input_tokens_seen": 1207038,
      "step": 1163,
      "train_runtime": 2790.1003,
      "train_tokens_per_second": 432.615
    },
    {
      "epoch": 0.49748904797521104,
      "grad_norm": 3.9845736026763916,
      "learning_rate": 1.5045989304812835e-05,
      "loss": 0.18305808305740356,
      "num_input_tokens_seen": 1207746,
      "step": 1164,
      "train_runtime": 2792.079,
      "train_tokens_per_second": 432.562
    },
    {
      "epoch": 0.49791644406453683,
      "grad_norm": 2.352217197418213,
      "learning_rate": 1.5041711229946525e-05,
      "loss": 0.12100671231746674,
      "num_input_tokens_seen": 1208784,
      "step": 1165,
      "train_runtime": 2794.1257,
      "train_tokens_per_second": 432.616
    },
    {
      "epoch": 0.49834384015386257,
      "grad_norm": 2.5519676208496094,
      "learning_rate": 1.5037433155080214e-05,
      "loss": 0.1499929428100586,
      "num_input_tokens_seen": 1209892,
      "step": 1166,
      "train_runtime": 2796.2164,
      "train_tokens_per_second": 432.689
    },
    {
      "epoch": 0.49877123624318837,
      "grad_norm": 2.3918063640594482,
      "learning_rate": 1.5033155080213905e-05,
      "loss": 0.1183585524559021,
      "num_input_tokens_seen": 1210896,
      "step": 1167,
      "train_runtime": 2798.3089,
      "train_tokens_per_second": 432.724
    },
    {
      "epoch": 0.49919863233251416,
      "grad_norm": 3.267407178878784,
      "learning_rate": 1.5028877005347595e-05,
      "loss": 0.18828095495700836,
      "num_input_tokens_seen": 1211932,
      "step": 1168,
      "train_runtime": 2800.3926,
      "train_tokens_per_second": 432.772
    },
    {
      "epoch": 0.49962602842183995,
      "grad_norm": 1.88737952709198,
      "learning_rate": 1.5024598930481284e-05,
      "loss": 0.08741053938865662,
      "num_input_tokens_seen": 1213072,
      "step": 1169,
      "train_runtime": 2802.5118,
      "train_tokens_per_second": 432.852
    },
    {
      "epoch": 0.5000534245111657,
      "grad_norm": 2.6835086345672607,
      "learning_rate": 1.5020320855614974e-05,
      "loss": 0.12560200691223145,
      "num_input_tokens_seen": 1214044,
      "step": 1170,
      "train_runtime": 2804.6022,
      "train_tokens_per_second": 432.876
    },
    {
      "epoch": 0.5004808206004915,
      "grad_norm": 2.1941637992858887,
      "learning_rate": 1.5016042780748665e-05,
      "loss": 0.1008794829249382,
      "num_input_tokens_seen": 1215246,
      "step": 1171,
      "train_runtime": 2813.3537,
      "train_tokens_per_second": 431.956
    },
    {
      "epoch": 0.5009082166898173,
      "grad_norm": 3.1997156143188477,
      "learning_rate": 1.5011764705882354e-05,
      "loss": 0.20159912109375,
      "num_input_tokens_seen": 1216162,
      "step": 1172,
      "train_runtime": 2815.4394,
      "train_tokens_per_second": 431.962
    },
    {
      "epoch": 0.5013356127791431,
      "grad_norm": 2.143730640411377,
      "learning_rate": 1.5007486631016044e-05,
      "loss": 0.10188774019479752,
      "num_input_tokens_seen": 1218486,
      "step": 1173,
      "train_runtime": 2817.8827,
      "train_tokens_per_second": 432.412
    },
    {
      "epoch": 0.5017630088684688,
      "grad_norm": 2.7107653617858887,
      "learning_rate": 1.5003208556149733e-05,
      "loss": 0.17050811648368835,
      "num_input_tokens_seen": 1219380,
      "step": 1174,
      "train_runtime": 2819.9317,
      "train_tokens_per_second": 432.415
    },
    {
      "epoch": 0.5021904049577947,
      "grad_norm": 2.7906951904296875,
      "learning_rate": 1.4998930481283424e-05,
      "loss": 0.09755326062440872,
      "num_input_tokens_seen": 1219936,
      "step": 1175,
      "train_runtime": 2821.876,
      "train_tokens_per_second": 432.314
    },
    {
      "epoch": 0.5026178010471204,
      "grad_norm": 5.099658966064453,
      "learning_rate": 1.4994652406417114e-05,
      "loss": 0.24060800671577454,
      "num_input_tokens_seen": 1220910,
      "step": 1176,
      "train_runtime": 2823.9513,
      "train_tokens_per_second": 432.341
    },
    {
      "epoch": 0.5030451971364462,
      "grad_norm": 3.3546831607818604,
      "learning_rate": 1.4990374331550803e-05,
      "loss": 0.1906435191631317,
      "num_input_tokens_seen": 1221952,
      "step": 1177,
      "train_runtime": 2826.0428,
      "train_tokens_per_second": 432.39
    },
    {
      "epoch": 0.503472593225772,
      "grad_norm": 3.424024820327759,
      "learning_rate": 1.4986096256684493e-05,
      "loss": 0.2020358443260193,
      "num_input_tokens_seen": 1222912,
      "step": 1178,
      "train_runtime": 2828.1283,
      "train_tokens_per_second": 432.41
    },
    {
      "epoch": 0.5038999893150977,
      "grad_norm": 3.646613836288452,
      "learning_rate": 1.4981818181818184e-05,
      "loss": 0.2279026210308075,
      "num_input_tokens_seen": 1224156,
      "step": 1179,
      "train_runtime": 2830.2702,
      "train_tokens_per_second": 432.523
    },
    {
      "epoch": 0.5043273854044236,
      "grad_norm": 2.5153822898864746,
      "learning_rate": 1.4977540106951873e-05,
      "loss": 0.15767836570739746,
      "num_input_tokens_seen": 1225652,
      "step": 1180,
      "train_runtime": 2832.4461,
      "train_tokens_per_second": 432.719
    },
    {
      "epoch": 0.5047547814937493,
      "grad_norm": 4.689395427703857,
      "learning_rate": 1.4973262032085563e-05,
      "loss": 0.09848980605602264,
      "num_input_tokens_seen": 1227204,
      "step": 1181,
      "train_runtime": 2834.6759,
      "train_tokens_per_second": 432.926
    },
    {
      "epoch": 0.5051821775830752,
      "grad_norm": 2.905390977859497,
      "learning_rate": 1.4968983957219252e-05,
      "loss": 0.14566992223262787,
      "num_input_tokens_seen": 1228700,
      "step": 1182,
      "train_runtime": 2836.8944,
      "train_tokens_per_second": 433.114
    },
    {
      "epoch": 0.5056095736724009,
      "grad_norm": 2.8243939876556396,
      "learning_rate": 1.4964705882352943e-05,
      "loss": 0.12898750603199005,
      "num_input_tokens_seen": 1229644,
      "step": 1183,
      "train_runtime": 2838.9925,
      "train_tokens_per_second": 433.127
    },
    {
      "epoch": 0.5060369697617266,
      "grad_norm": 5.052365303039551,
      "learning_rate": 1.4960427807486633e-05,
      "loss": 0.22005364298820496,
      "num_input_tokens_seen": 1230508,
      "step": 1184,
      "train_runtime": 2841.046,
      "train_tokens_per_second": 433.118
    },
    {
      "epoch": 0.5064643658510525,
      "grad_norm": 2.562453269958496,
      "learning_rate": 1.4956149732620322e-05,
      "loss": 0.09847698360681534,
      "num_input_tokens_seen": 1231260,
      "step": 1185,
      "train_runtime": 2843.1161,
      "train_tokens_per_second": 433.067
    },
    {
      "epoch": 0.5068917619403782,
      "grad_norm": 3.255903720855713,
      "learning_rate": 1.4951871657754012e-05,
      "loss": 0.13542407751083374,
      "num_input_tokens_seen": 1232160,
      "step": 1186,
      "train_runtime": 2845.1369,
      "train_tokens_per_second": 433.076
    },
    {
      "epoch": 0.5073191580297041,
      "grad_norm": 3.189725875854492,
      "learning_rate": 1.4947593582887703e-05,
      "loss": 0.09219875186681747,
      "num_input_tokens_seen": 1233090,
      "step": 1187,
      "train_runtime": 2847.2089,
      "train_tokens_per_second": 433.087
    },
    {
      "epoch": 0.5077465541190298,
      "grad_norm": 2.5523719787597656,
      "learning_rate": 1.4943315508021392e-05,
      "loss": 0.1534418761730194,
      "num_input_tokens_seen": 1234300,
      "step": 1188,
      "train_runtime": 2849.3436,
      "train_tokens_per_second": 433.187
    },
    {
      "epoch": 0.5081739502083555,
      "grad_norm": 3.37165904045105,
      "learning_rate": 1.4939037433155082e-05,
      "loss": 0.12875303626060486,
      "num_input_tokens_seen": 1235186,
      "step": 1189,
      "train_runtime": 2851.397,
      "train_tokens_per_second": 433.186
    },
    {
      "epoch": 0.5086013462976814,
      "grad_norm": 2.39281964302063,
      "learning_rate": 1.4934759358288771e-05,
      "loss": 0.10504047572612762,
      "num_input_tokens_seen": 1236154,
      "step": 1190,
      "train_runtime": 2853.548,
      "train_tokens_per_second": 433.199
    },
    {
      "epoch": 0.5090287423870071,
      "grad_norm": 2.805588722229004,
      "learning_rate": 1.4930481283422462e-05,
      "loss": 0.16197718679904938,
      "num_input_tokens_seen": 1237918,
      "step": 1191,
      "train_runtime": 2855.8138,
      "train_tokens_per_second": 433.473
    },
    {
      "epoch": 0.509456138476333,
      "grad_norm": 4.094025611877441,
      "learning_rate": 1.4926203208556152e-05,
      "loss": 0.17036931216716766,
      "num_input_tokens_seen": 1238520,
      "step": 1192,
      "train_runtime": 2857.7931,
      "train_tokens_per_second": 433.383
    },
    {
      "epoch": 0.5098835345656587,
      "grad_norm": 4.176519870758057,
      "learning_rate": 1.4921925133689841e-05,
      "loss": 0.14350777864456177,
      "num_input_tokens_seen": 1239132,
      "step": 1193,
      "train_runtime": 2859.7843,
      "train_tokens_per_second": 433.296
    },
    {
      "epoch": 0.5103109306549845,
      "grad_norm": 5.641793727874756,
      "learning_rate": 1.491764705882353e-05,
      "loss": 0.24869948625564575,
      "num_input_tokens_seen": 1239858,
      "step": 1194,
      "train_runtime": 2861.8105,
      "train_tokens_per_second": 433.243
    },
    {
      "epoch": 0.5107383267443103,
      "grad_norm": 3.246270179748535,
      "learning_rate": 1.4913368983957222e-05,
      "loss": 0.14501456916332245,
      "num_input_tokens_seen": 1240696,
      "step": 1195,
      "train_runtime": 2863.8971,
      "train_tokens_per_second": 433.219
    },
    {
      "epoch": 0.511165722833636,
      "grad_norm": 2.838711738586426,
      "learning_rate": 1.4909090909090911e-05,
      "loss": 0.17654375731945038,
      "num_input_tokens_seen": 1242014,
      "step": 1196,
      "train_runtime": 2866.0444,
      "train_tokens_per_second": 433.355
    },
    {
      "epoch": 0.5115931189229619,
      "grad_norm": 2.7171852588653564,
      "learning_rate": 1.49048128342246e-05,
      "loss": 0.09624326229095459,
      "num_input_tokens_seen": 1242658,
      "step": 1197,
      "train_runtime": 2868.034,
      "train_tokens_per_second": 433.279
    },
    {
      "epoch": 0.5120205150122876,
      "grad_norm": 2.2829556465148926,
      "learning_rate": 1.490053475935829e-05,
      "loss": 0.14937518537044525,
      "num_input_tokens_seen": 1244118,
      "step": 1198,
      "train_runtime": 2870.2146,
      "train_tokens_per_second": 433.458
    },
    {
      "epoch": 0.5124479111016135,
      "grad_norm": 3.8092312812805176,
      "learning_rate": 1.4896256684491981e-05,
      "loss": 0.18598896265029907,
      "num_input_tokens_seen": 1244880,
      "step": 1199,
      "train_runtime": 2872.2545,
      "train_tokens_per_second": 433.416
    },
    {
      "epoch": 0.5128753071909392,
      "grad_norm": 4.290783405303955,
      "learning_rate": 1.489197860962567e-05,
      "loss": 0.2228667140007019,
      "num_input_tokens_seen": 1246146,
      "step": 1200,
      "train_runtime": 2874.378,
      "train_tokens_per_second": 433.536
    },
    {
      "epoch": 0.513302703280265,
      "grad_norm": 3.054795980453491,
      "learning_rate": 1.488770053475936e-05,
      "loss": 0.1417790800333023,
      "num_input_tokens_seen": 1246980,
      "step": 1201,
      "train_runtime": 2882.8688,
      "train_tokens_per_second": 432.548
    },
    {
      "epoch": 0.5137300993695908,
      "grad_norm": 2.4083077907562256,
      "learning_rate": 1.488342245989305e-05,
      "loss": 0.11850110441446304,
      "num_input_tokens_seen": 1247924,
      "step": 1202,
      "train_runtime": 2885.0891,
      "train_tokens_per_second": 432.543
    },
    {
      "epoch": 0.5141574954589165,
      "grad_norm": 2.284047842025757,
      "learning_rate": 1.4879144385026737e-05,
      "loss": 0.1140502467751503,
      "num_input_tokens_seen": 1248902,
      "step": 1203,
      "train_runtime": 2887.1313,
      "train_tokens_per_second": 432.575
    },
    {
      "epoch": 0.5145848915482424,
      "grad_norm": 4.041154861450195,
      "learning_rate": 1.487486631016043e-05,
      "loss": 0.11437393724918365,
      "num_input_tokens_seen": 1249520,
      "step": 1204,
      "train_runtime": 2889.1697,
      "train_tokens_per_second": 432.484
    },
    {
      "epoch": 0.5150122876375681,
      "grad_norm": 2.480656147003174,
      "learning_rate": 1.487058823529412e-05,
      "loss": 0.14181768894195557,
      "num_input_tokens_seen": 1250896,
      "step": 1205,
      "train_runtime": 2891.3688,
      "train_tokens_per_second": 432.631
    },
    {
      "epoch": 0.5154396837268939,
      "grad_norm": 4.123666286468506,
      "learning_rate": 1.4866310160427807e-05,
      "loss": 0.2572101652622223,
      "num_input_tokens_seen": 1252022,
      "step": 1206,
      "train_runtime": 2893.5111,
      "train_tokens_per_second": 432.7
    },
    {
      "epoch": 0.5158670798162197,
      "grad_norm": 2.35606050491333,
      "learning_rate": 1.4862032085561497e-05,
      "loss": 0.13891059160232544,
      "num_input_tokens_seen": 1253436,
      "step": 1207,
      "train_runtime": 2895.7185,
      "train_tokens_per_second": 432.858
    },
    {
      "epoch": 0.5162944759055454,
      "grad_norm": 2.973144292831421,
      "learning_rate": 1.485775401069519e-05,
      "loss": 0.1084960475564003,
      "num_input_tokens_seen": 1254522,
      "step": 1208,
      "train_runtime": 2897.8911,
      "train_tokens_per_second": 432.909
    },
    {
      "epoch": 0.5167218719948713,
      "grad_norm": 2.1476330757141113,
      "learning_rate": 1.4853475935828877e-05,
      "loss": 0.09674285352230072,
      "num_input_tokens_seen": 1256084,
      "step": 1209,
      "train_runtime": 2900.1252,
      "train_tokens_per_second": 433.114
    },
    {
      "epoch": 0.517149268084197,
      "grad_norm": 3.6345908641815186,
      "learning_rate": 1.4849197860962567e-05,
      "loss": 0.13477930426597595,
      "num_input_tokens_seen": 1257036,
      "step": 1210,
      "train_runtime": 2902.1925,
      "train_tokens_per_second": 433.133
    },
    {
      "epoch": 0.5175766641735228,
      "grad_norm": 3.2305147647857666,
      "learning_rate": 1.4844919786096256e-05,
      "loss": 0.1655772626399994,
      "num_input_tokens_seen": 1257922,
      "step": 1211,
      "train_runtime": 2904.2798,
      "train_tokens_per_second": 433.127
    },
    {
      "epoch": 0.5180040602628486,
      "grad_norm": 1.967262625694275,
      "learning_rate": 1.484064171122995e-05,
      "loss": 0.0806393250823021,
      "num_input_tokens_seen": 1259290,
      "step": 1212,
      "train_runtime": 2906.4589,
      "train_tokens_per_second": 433.273
    },
    {
      "epoch": 0.5184314563521744,
      "grad_norm": 2.1603286266326904,
      "learning_rate": 1.4836363636363637e-05,
      "loss": 0.1084219217300415,
      "num_input_tokens_seen": 1260326,
      "step": 1213,
      "train_runtime": 2908.5513,
      "train_tokens_per_second": 433.317
    },
    {
      "epoch": 0.5188588524415002,
      "grad_norm": 3.709182024002075,
      "learning_rate": 1.4832085561497326e-05,
      "loss": 0.13470333814620972,
      "num_input_tokens_seen": 1261044,
      "step": 1214,
      "train_runtime": 2910.5172,
      "train_tokens_per_second": 433.271
    },
    {
      "epoch": 0.519286248530826,
      "grad_norm": 2.3523428440093994,
      "learning_rate": 1.4827807486631016e-05,
      "loss": 0.10855934023857117,
      "num_input_tokens_seen": 1262616,
      "step": 1215,
      "train_runtime": 2912.7449,
      "train_tokens_per_second": 433.48
    },
    {
      "epoch": 0.5197136446201517,
      "grad_norm": 4.7110724449157715,
      "learning_rate": 1.4823529411764707e-05,
      "loss": 0.22755569219589233,
      "num_input_tokens_seen": 1263170,
      "step": 1216,
      "train_runtime": 2914.7034,
      "train_tokens_per_second": 433.379
    },
    {
      "epoch": 0.5201410407094775,
      "grad_norm": 2.7887420654296875,
      "learning_rate": 1.4819251336898396e-05,
      "loss": 0.12684106826782227,
      "num_input_tokens_seen": 1264336,
      "step": 1217,
      "train_runtime": 2916.8437,
      "train_tokens_per_second": 433.46
    },
    {
      "epoch": 0.5205684367988033,
      "grad_norm": 2.5725595951080322,
      "learning_rate": 1.4814973262032086e-05,
      "loss": 0.15530602633953094,
      "num_input_tokens_seen": 1265598,
      "step": 1218,
      "train_runtime": 2918.9847,
      "train_tokens_per_second": 433.575
    },
    {
      "epoch": 0.5209958328881291,
      "grad_norm": 2.8241114616394043,
      "learning_rate": 1.4810695187165775e-05,
      "loss": 0.18477758765220642,
      "num_input_tokens_seen": 1266874,
      "step": 1219,
      "train_runtime": 2921.1907,
      "train_tokens_per_second": 433.684
    },
    {
      "epoch": 0.5214232289774549,
      "grad_norm": 4.6259870529174805,
      "learning_rate": 1.4806417112299466e-05,
      "loss": 0.185531347990036,
      "num_input_tokens_seen": 1267652,
      "step": 1220,
      "train_runtime": 2923.2287,
      "train_tokens_per_second": 433.648
    },
    {
      "epoch": 0.5218506250667806,
      "grad_norm": 2.4858689308166504,
      "learning_rate": 1.4802139037433156e-05,
      "loss": 0.08196789026260376,
      "num_input_tokens_seen": 1269138,
      "step": 1221,
      "train_runtime": 2925.4713,
      "train_tokens_per_second": 433.823
    },
    {
      "epoch": 0.5222780211561064,
      "grad_norm": 2.866215467453003,
      "learning_rate": 1.4797860962566845e-05,
      "loss": 0.11320745944976807,
      "num_input_tokens_seen": 1269886,
      "step": 1222,
      "train_runtime": 2927.5289,
      "train_tokens_per_second": 433.774
    },
    {
      "epoch": 0.5227054172454322,
      "grad_norm": 3.460625648498535,
      "learning_rate": 1.4793582887700535e-05,
      "loss": 0.14805959165096283,
      "num_input_tokens_seen": 1270554,
      "step": 1223,
      "train_runtime": 2929.5692,
      "train_tokens_per_second": 433.7
    },
    {
      "epoch": 0.523132813334758,
      "grad_norm": 3.169494390487671,
      "learning_rate": 1.4789304812834226e-05,
      "loss": 0.11480510979890823,
      "num_input_tokens_seen": 1271226,
      "step": 1224,
      "train_runtime": 2931.596,
      "train_tokens_per_second": 433.629
    },
    {
      "epoch": 0.5235602094240838,
      "grad_norm": 3.1488611698150635,
      "learning_rate": 1.4785026737967915e-05,
      "loss": 0.08535999804735184,
      "num_input_tokens_seen": 1272126,
      "step": 1225,
      "train_runtime": 2933.6436,
      "train_tokens_per_second": 433.633
    },
    {
      "epoch": 0.5239876055134095,
      "grad_norm": 1.1961338520050049,
      "learning_rate": 1.4780748663101605e-05,
      "loss": 0.062140218913555145,
      "num_input_tokens_seen": 1275302,
      "step": 1226,
      "train_runtime": 2936.2875,
      "train_tokens_per_second": 434.325
    },
    {
      "epoch": 0.5244150016027354,
      "grad_norm": 3.8686330318450928,
      "learning_rate": 1.4776470588235294e-05,
      "loss": 0.1642768383026123,
      "num_input_tokens_seen": 1276538,
      "step": 1227,
      "train_runtime": 2938.4278,
      "train_tokens_per_second": 434.429
    },
    {
      "epoch": 0.5248423976920611,
      "grad_norm": 2.709268093109131,
      "learning_rate": 1.4772192513368985e-05,
      "loss": 0.15790528059005737,
      "num_input_tokens_seen": 1277534,
      "step": 1228,
      "train_runtime": 2940.5046,
      "train_tokens_per_second": 434.461
    },
    {
      "epoch": 0.5252697937813869,
      "grad_norm": 3.1894426345825195,
      "learning_rate": 1.4767914438502675e-05,
      "loss": 0.14597229659557343,
      "num_input_tokens_seen": 1278534,
      "step": 1229,
      "train_runtime": 2942.5887,
      "train_tokens_per_second": 434.493
    },
    {
      "epoch": 0.5256971898707127,
      "grad_norm": 3.300455331802368,
      "learning_rate": 1.4763636363636364e-05,
      "loss": 0.19414545595645905,
      "num_input_tokens_seen": 1279530,
      "step": 1230,
      "train_runtime": 2944.6592,
      "train_tokens_per_second": 434.526
    },
    {
      "epoch": 0.5261245859600384,
      "grad_norm": 2.5099072456359863,
      "learning_rate": 1.4759358288770054e-05,
      "loss": 0.10873032361268997,
      "num_input_tokens_seen": 1280870,
      "step": 1231,
      "train_runtime": 2953.638,
      "train_tokens_per_second": 433.658
    },
    {
      "epoch": 0.5265519820493643,
      "grad_norm": 2.3596484661102295,
      "learning_rate": 1.4755080213903745e-05,
      "loss": 0.10290868580341339,
      "num_input_tokens_seen": 1282036,
      "step": 1232,
      "train_runtime": 2955.7866,
      "train_tokens_per_second": 433.738
    },
    {
      "epoch": 0.52697937813869,
      "grad_norm": 2.6016011238098145,
      "learning_rate": 1.4750802139037434e-05,
      "loss": 0.09967711567878723,
      "num_input_tokens_seen": 1282954,
      "step": 1233,
      "train_runtime": 2957.8801,
      "train_tokens_per_second": 433.741
    },
    {
      "epoch": 0.5274067742280159,
      "grad_norm": 4.108160018920898,
      "learning_rate": 1.4746524064171124e-05,
      "loss": 0.2169618308544159,
      "num_input_tokens_seen": 1283920,
      "step": 1234,
      "train_runtime": 2959.9763,
      "train_tokens_per_second": 433.76
    },
    {
      "epoch": 0.5278341703173416,
      "grad_norm": 3.0595760345458984,
      "learning_rate": 1.4742245989304813e-05,
      "loss": 0.15109138190746307,
      "num_input_tokens_seen": 1284836,
      "step": 1235,
      "train_runtime": 2962.0189,
      "train_tokens_per_second": 433.77
    },
    {
      "epoch": 0.5282615664066673,
      "grad_norm": 2.926755666732788,
      "learning_rate": 1.4737967914438504e-05,
      "loss": 0.16534759104251862,
      "num_input_tokens_seen": 1285940,
      "step": 1236,
      "train_runtime": 2964.1334,
      "train_tokens_per_second": 433.833
    },
    {
      "epoch": 0.5286889624959932,
      "grad_norm": 3.5862061977386475,
      "learning_rate": 1.4733689839572194e-05,
      "loss": 0.24955226480960846,
      "num_input_tokens_seen": 1286944,
      "step": 1237,
      "train_runtime": 2966.2328,
      "train_tokens_per_second": 433.865
    },
    {
      "epoch": 0.5291163585853189,
      "grad_norm": 3.403506278991699,
      "learning_rate": 1.4729411764705883e-05,
      "loss": 0.10444283485412598,
      "num_input_tokens_seen": 1287744,
      "step": 1238,
      "train_runtime": 2968.2536,
      "train_tokens_per_second": 433.839
    },
    {
      "epoch": 0.5295437546746448,
      "grad_norm": 3.841616153717041,
      "learning_rate": 1.4725133689839573e-05,
      "loss": 0.21578004956245422,
      "num_input_tokens_seen": 1289008,
      "step": 1239,
      "train_runtime": 2970.4004,
      "train_tokens_per_second": 433.951
    },
    {
      "epoch": 0.5299711507639705,
      "grad_norm": 3.752838611602783,
      "learning_rate": 1.4720855614973264e-05,
      "loss": 0.16529375314712524,
      "num_input_tokens_seen": 1289992,
      "step": 1240,
      "train_runtime": 2972.4899,
      "train_tokens_per_second": 433.977
    },
    {
      "epoch": 0.5303985468532962,
      "grad_norm": 2.8671066761016846,
      "learning_rate": 1.4716577540106953e-05,
      "loss": 0.12689585983753204,
      "num_input_tokens_seen": 1291310,
      "step": 1241,
      "train_runtime": 2974.6648,
      "train_tokens_per_second": 434.103
    },
    {
      "epoch": 0.5308259429426221,
      "grad_norm": 3.026271343231201,
      "learning_rate": 1.4712299465240643e-05,
      "loss": 0.15771925449371338,
      "num_input_tokens_seen": 1292422,
      "step": 1242,
      "train_runtime": 2976.7651,
      "train_tokens_per_second": 434.17
    },
    {
      "epoch": 0.5312533390319478,
      "grad_norm": 4.134592056274414,
      "learning_rate": 1.4708021390374332e-05,
      "loss": 0.16895480453968048,
      "num_input_tokens_seen": 1293270,
      "step": 1243,
      "train_runtime": 2978.8477,
      "train_tokens_per_second": 434.151
    },
    {
      "epoch": 0.5316807351212737,
      "grad_norm": 3.4088222980499268,
      "learning_rate": 1.4703743315508023e-05,
      "loss": 0.10249975323677063,
      "num_input_tokens_seen": 1293884,
      "step": 1244,
      "train_runtime": 2980.9077,
      "train_tokens_per_second": 434.057
    },
    {
      "epoch": 0.5321081312105994,
      "grad_norm": 4.0659027099609375,
      "learning_rate": 1.4699465240641713e-05,
      "loss": 0.18310728669166565,
      "num_input_tokens_seen": 1294644,
      "step": 1245,
      "train_runtime": 2982.9946,
      "train_tokens_per_second": 434.008
    },
    {
      "epoch": 0.5325355272999253,
      "grad_norm": 4.350315570831299,
      "learning_rate": 1.4695187165775402e-05,
      "loss": 0.1722957342863083,
      "num_input_tokens_seen": 1295360,
      "step": 1246,
      "train_runtime": 2984.9821,
      "train_tokens_per_second": 433.959
    },
    {
      "epoch": 0.532962923389251,
      "grad_norm": 3.4019153118133545,
      "learning_rate": 1.4690909090909092e-05,
      "loss": 0.19986552000045776,
      "num_input_tokens_seen": 1296580,
      "step": 1247,
      "train_runtime": 2987.1504,
      "train_tokens_per_second": 434.052
    },
    {
      "epoch": 0.5333903194785767,
      "grad_norm": 3.4604334831237793,
      "learning_rate": 1.4686631016042783e-05,
      "loss": 0.13814860582351685,
      "num_input_tokens_seen": 1297590,
      "step": 1248,
      "train_runtime": 2989.2033,
      "train_tokens_per_second": 434.092
    },
    {
      "epoch": 0.5338177155679026,
      "grad_norm": 2.296515941619873,
      "learning_rate": 1.4682352941176472e-05,
      "loss": 0.1151202991604805,
      "num_input_tokens_seen": 1298560,
      "step": 1249,
      "train_runtime": 2991.2654,
      "train_tokens_per_second": 434.117
    },
    {
      "epoch": 0.5342451116572283,
      "grad_norm": 3.181269884109497,
      "learning_rate": 1.4678074866310162e-05,
      "loss": 0.19524168968200684,
      "num_input_tokens_seen": 1299314,
      "step": 1250,
      "train_runtime": 2993.3266,
      "train_tokens_per_second": 434.07
    },
    {
      "epoch": 0.5346725077465542,
      "grad_norm": 3.111470937728882,
      "learning_rate": 1.4673796791443851e-05,
      "loss": 0.09875819087028503,
      "num_input_tokens_seen": 1300070,
      "step": 1251,
      "train_runtime": 2995.423,
      "train_tokens_per_second": 434.019
    },
    {
      "epoch": 0.5350999038358799,
      "grad_norm": 5.305385589599609,
      "learning_rate": 1.4669518716577542e-05,
      "loss": 0.18836107850074768,
      "num_input_tokens_seen": 1300824,
      "step": 1252,
      "train_runtime": 2997.4636,
      "train_tokens_per_second": 433.975
    },
    {
      "epoch": 0.5355272999252056,
      "grad_norm": 3.39542555809021,
      "learning_rate": 1.4665240641711232e-05,
      "loss": 0.1877380907535553,
      "num_input_tokens_seen": 1301600,
      "step": 1253,
      "train_runtime": 2999.4933,
      "train_tokens_per_second": 433.94
    },
    {
      "epoch": 0.5359546960145315,
      "grad_norm": 2.797130584716797,
      "learning_rate": 1.4660962566844921e-05,
      "loss": 0.12603871524333954,
      "num_input_tokens_seen": 1302436,
      "step": 1254,
      "train_runtime": 3001.5592,
      "train_tokens_per_second": 433.92
    },
    {
      "epoch": 0.5363820921038572,
      "grad_norm": 2.6841440200805664,
      "learning_rate": 1.465668449197861e-05,
      "loss": 0.13795852661132812,
      "num_input_tokens_seen": 1303460,
      "step": 1255,
      "train_runtime": 3003.6618,
      "train_tokens_per_second": 433.957
    },
    {
      "epoch": 0.5368094881931831,
      "grad_norm": 3.3998546600341797,
      "learning_rate": 1.4652406417112302e-05,
      "loss": 0.12445757538080215,
      "num_input_tokens_seen": 1304480,
      "step": 1256,
      "train_runtime": 3005.7477,
      "train_tokens_per_second": 433.995
    },
    {
      "epoch": 0.5372368842825088,
      "grad_norm": 2.7438855171203613,
      "learning_rate": 1.4648128342245991e-05,
      "loss": 0.18717393279075623,
      "num_input_tokens_seen": 1305782,
      "step": 1257,
      "train_runtime": 3007.9045,
      "train_tokens_per_second": 434.117
    },
    {
      "epoch": 0.5376642803718346,
      "grad_norm": 3.3281304836273193,
      "learning_rate": 1.464385026737968e-05,
      "loss": 0.17011483013629913,
      "num_input_tokens_seen": 1306524,
      "step": 1258,
      "train_runtime": 3009.9567,
      "train_tokens_per_second": 434.067
    },
    {
      "epoch": 0.5380916764611604,
      "grad_norm": 2.7448432445526123,
      "learning_rate": 1.463957219251337e-05,
      "loss": 0.19452330470085144,
      "num_input_tokens_seen": 1307818,
      "step": 1259,
      "train_runtime": 3012.0997,
      "train_tokens_per_second": 434.188
    },
    {
      "epoch": 0.5385190725504861,
      "grad_norm": 2.0388052463531494,
      "learning_rate": 1.4635294117647061e-05,
      "loss": 0.1313883513212204,
      "num_input_tokens_seen": 1309754,
      "step": 1260,
      "train_runtime": 3014.404,
      "train_tokens_per_second": 434.498
    },
    {
      "epoch": 0.538946468639812,
      "grad_norm": 3.47153639793396,
      "learning_rate": 1.463101604278075e-05,
      "loss": 0.18669819831848145,
      "num_input_tokens_seen": 1310512,
      "step": 1261,
      "train_runtime": 3022.9912,
      "train_tokens_per_second": 433.515
    },
    {
      "epoch": 0.5393738647291377,
      "grad_norm": 3.1900274753570557,
      "learning_rate": 1.462673796791444e-05,
      "loss": 0.13144513964653015,
      "num_input_tokens_seen": 1311126,
      "step": 1262,
      "train_runtime": 3024.9515,
      "train_tokens_per_second": 433.437
    },
    {
      "epoch": 0.5398012608184635,
      "grad_norm": 3.0377893447875977,
      "learning_rate": 1.462245989304813e-05,
      "loss": 0.16956773400306702,
      "num_input_tokens_seen": 1312054,
      "step": 1263,
      "train_runtime": 3027.1519,
      "train_tokens_per_second": 433.429
    },
    {
      "epoch": 0.5402286569077893,
      "grad_norm": 3.7872700691223145,
      "learning_rate": 1.461818181818182e-05,
      "loss": 0.17255957424640656,
      "num_input_tokens_seen": 1313044,
      "step": 1264,
      "train_runtime": 3029.2521,
      "train_tokens_per_second": 433.455
    },
    {
      "epoch": 0.540656052997115,
      "grad_norm": 4.534908771514893,
      "learning_rate": 1.461390374331551e-05,
      "loss": 0.1506447196006775,
      "num_input_tokens_seen": 1313798,
      "step": 1265,
      "train_runtime": 3031.2379,
      "train_tokens_per_second": 433.42
    },
    {
      "epoch": 0.5410834490864409,
      "grad_norm": 4.061362266540527,
      "learning_rate": 1.46096256684492e-05,
      "loss": 0.2508237659931183,
      "num_input_tokens_seen": 1315170,
      "step": 1266,
      "train_runtime": 3033.4115,
      "train_tokens_per_second": 433.561
    },
    {
      "epoch": 0.5415108451757666,
      "grad_norm": 3.1945958137512207,
      "learning_rate": 1.4605347593582887e-05,
      "loss": 0.122700996696949,
      "num_input_tokens_seen": 1316050,
      "step": 1267,
      "train_runtime": 3035.5316,
      "train_tokens_per_second": 433.548
    },
    {
      "epoch": 0.5419382412650924,
      "grad_norm": 2.2648088932037354,
      "learning_rate": 1.460106951871658e-05,
      "loss": 0.07432137429714203,
      "num_input_tokens_seen": 1317004,
      "step": 1268,
      "train_runtime": 3037.6613,
      "train_tokens_per_second": 433.559
    },
    {
      "epoch": 0.5423656373544182,
      "grad_norm": 3.122262954711914,
      "learning_rate": 1.459679144385027e-05,
      "loss": 0.11405198276042938,
      "num_input_tokens_seen": 1318044,
      "step": 1269,
      "train_runtime": 3039.7727,
      "train_tokens_per_second": 433.6
    },
    {
      "epoch": 0.542793033443744,
      "grad_norm": 3.263747453689575,
      "learning_rate": 1.4592513368983959e-05,
      "loss": 0.12836995720863342,
      "num_input_tokens_seen": 1318916,
      "step": 1270,
      "train_runtime": 3041.8447,
      "train_tokens_per_second": 433.591
    },
    {
      "epoch": 0.5432204295330698,
      "grad_norm": 3.0479540824890137,
      "learning_rate": 1.4588235294117647e-05,
      "loss": 0.26140522956848145,
      "num_input_tokens_seen": 1320104,
      "step": 1271,
      "train_runtime": 3043.962,
      "train_tokens_per_second": 433.68
    },
    {
      "epoch": 0.5436478256223956,
      "grad_norm": 4.577423095703125,
      "learning_rate": 1.458395721925134e-05,
      "loss": 0.14573048055171967,
      "num_input_tokens_seen": 1320730,
      "step": 1272,
      "train_runtime": 3045.9948,
      "train_tokens_per_second": 433.596
    },
    {
      "epoch": 0.5440752217117213,
      "grad_norm": 3.4006383419036865,
      "learning_rate": 1.4579679144385029e-05,
      "loss": 0.10004554688930511,
      "num_input_tokens_seen": 1321634,
      "step": 1273,
      "train_runtime": 3048.075,
      "train_tokens_per_second": 433.596
    },
    {
      "epoch": 0.5445026178010471,
      "grad_norm": 2.7472174167633057,
      "learning_rate": 1.4575401069518717e-05,
      "loss": 0.18967130780220032,
      "num_input_tokens_seen": 1322748,
      "step": 1274,
      "train_runtime": 3050.199,
      "train_tokens_per_second": 433.66
    },
    {
      "epoch": 0.5449300138903729,
      "grad_norm": 2.7711236476898193,
      "learning_rate": 1.4571122994652406e-05,
      "loss": 0.187877357006073,
      "num_input_tokens_seen": 1323836,
      "step": 1275,
      "train_runtime": 3052.3368,
      "train_tokens_per_second": 433.712
    },
    {
      "epoch": 0.5453574099796987,
      "grad_norm": 1.9772099256515503,
      "learning_rate": 1.4566844919786099e-05,
      "loss": 0.10886204242706299,
      "num_input_tokens_seen": 1325094,
      "step": 1276,
      "train_runtime": 3054.5087,
      "train_tokens_per_second": 433.816
    },
    {
      "epoch": 0.5457848060690245,
      "grad_norm": 3.4452152252197266,
      "learning_rate": 1.4562566844919787e-05,
      "loss": 0.12986767292022705,
      "num_input_tokens_seen": 1326388,
      "step": 1277,
      "train_runtime": 3056.6539,
      "train_tokens_per_second": 433.935
    },
    {
      "epoch": 0.5462122021583502,
      "grad_norm": 3.2720425128936768,
      "learning_rate": 1.4558288770053476e-05,
      "loss": 0.1310553252696991,
      "num_input_tokens_seen": 1327346,
      "step": 1278,
      "train_runtime": 3058.7511,
      "train_tokens_per_second": 433.95
    },
    {
      "epoch": 0.546639598247676,
      "grad_norm": 3.2607574462890625,
      "learning_rate": 1.4554010695187166e-05,
      "loss": 0.17354868352413177,
      "num_input_tokens_seen": 1328328,
      "step": 1279,
      "train_runtime": 3060.8379,
      "train_tokens_per_second": 433.975
    },
    {
      "epoch": 0.5470669943370018,
      "grad_norm": 3.38954758644104,
      "learning_rate": 1.4549732620320859e-05,
      "loss": 0.15475034713745117,
      "num_input_tokens_seen": 1329264,
      "step": 1280,
      "train_runtime": 3062.8619,
      "train_tokens_per_second": 433.994
    },
    {
      "epoch": 0.5474943904263276,
      "grad_norm": 3.611175537109375,
      "learning_rate": 1.4545454545454546e-05,
      "loss": 0.09882596880197525,
      "num_input_tokens_seen": 1330392,
      "step": 1281,
      "train_runtime": 3064.9539,
      "train_tokens_per_second": 434.066
    },
    {
      "epoch": 0.5479217865156534,
      "grad_norm": 2.675280809402466,
      "learning_rate": 1.4541176470588236e-05,
      "loss": 0.11324083060026169,
      "num_input_tokens_seen": 1331500,
      "step": 1282,
      "train_runtime": 3067.0639,
      "train_tokens_per_second": 434.129
    },
    {
      "epoch": 0.5483491826049791,
      "grad_norm": 3.086268424987793,
      "learning_rate": 1.4536898395721925e-05,
      "loss": 0.1743987500667572,
      "num_input_tokens_seen": 1332680,
      "step": 1283,
      "train_runtime": 3069.1876,
      "train_tokens_per_second": 434.213
    },
    {
      "epoch": 0.548776578694305,
      "grad_norm": 2.768984794616699,
      "learning_rate": 1.4532620320855616e-05,
      "loss": 0.16291919350624084,
      "num_input_tokens_seen": 1333844,
      "step": 1284,
      "train_runtime": 3071.2953,
      "train_tokens_per_second": 434.294
    },
    {
      "epoch": 0.5492039747836307,
      "grad_norm": 2.387202501296997,
      "learning_rate": 1.4528342245989306e-05,
      "loss": 0.1337834596633911,
      "num_input_tokens_seen": 1335014,
      "step": 1285,
      "train_runtime": 3073.4412,
      "train_tokens_per_second": 434.371
    },
    {
      "epoch": 0.5496313708729565,
      "grad_norm": 2.024073839187622,
      "learning_rate": 1.4524064171122995e-05,
      "loss": 0.13841214776039124,
      "num_input_tokens_seen": 1337410,
      "step": 1286,
      "train_runtime": 3075.8339,
      "train_tokens_per_second": 434.812
    },
    {
      "epoch": 0.5500587669622823,
      "grad_norm": 3.5594098567962646,
      "learning_rate": 1.4519786096256685e-05,
      "loss": 0.13000838458538055,
      "num_input_tokens_seen": 1338056,
      "step": 1287,
      "train_runtime": 3077.7845,
      "train_tokens_per_second": 434.746
    },
    {
      "epoch": 0.550486163051608,
      "grad_norm": 4.240809917449951,
      "learning_rate": 1.4515508021390376e-05,
      "loss": 0.12927718460559845,
      "num_input_tokens_seen": 1338814,
      "step": 1288,
      "train_runtime": 3079.8386,
      "train_tokens_per_second": 434.703
    },
    {
      "epoch": 0.5509135591409339,
      "grad_norm": 1.9870355129241943,
      "learning_rate": 1.4511229946524065e-05,
      "loss": 0.09284672141075134,
      "num_input_tokens_seen": 1340582,
      "step": 1289,
      "train_runtime": 3082.2007,
      "train_tokens_per_second": 434.943
    },
    {
      "epoch": 0.5513409552302596,
      "grad_norm": 4.011283874511719,
      "learning_rate": 1.4506951871657755e-05,
      "loss": 0.08587135374546051,
      "num_input_tokens_seen": 1341116,
      "step": 1290,
      "train_runtime": 3084.1558,
      "train_tokens_per_second": 434.841
    },
    {
      "epoch": 0.5517683513195855,
      "grad_norm": 2.6522579193115234,
      "learning_rate": 1.4502673796791444e-05,
      "loss": 0.1723092496395111,
      "num_input_tokens_seen": 1342370,
      "step": 1291,
      "train_runtime": 3092.9473,
      "train_tokens_per_second": 434.01
    },
    {
      "epoch": 0.5521957474089112,
      "grad_norm": 3.418346881866455,
      "learning_rate": 1.4498395721925135e-05,
      "loss": 0.22418902814388275,
      "num_input_tokens_seen": 1343356,
      "step": 1292,
      "train_runtime": 3095.0061,
      "train_tokens_per_second": 434.04
    },
    {
      "epoch": 0.552623143498237,
      "grad_norm": 2.2360384464263916,
      "learning_rate": 1.4494117647058825e-05,
      "loss": 0.09782730042934418,
      "num_input_tokens_seen": 1344480,
      "step": 1293,
      "train_runtime": 3097.3633,
      "train_tokens_per_second": 434.072
    },
    {
      "epoch": 0.5530505395875628,
      "grad_norm": 2.2155373096466064,
      "learning_rate": 1.4489839572192514e-05,
      "loss": 0.09152060002088547,
      "num_input_tokens_seen": 1345570,
      "step": 1294,
      "train_runtime": 3099.5012,
      "train_tokens_per_second": 434.125
    },
    {
      "epoch": 0.5534779356768885,
      "grad_norm": 2.8570311069488525,
      "learning_rate": 1.4485561497326204e-05,
      "loss": 0.12493006885051727,
      "num_input_tokens_seen": 1346498,
      "step": 1295,
      "train_runtime": 3101.5673,
      "train_tokens_per_second": 434.135
    },
    {
      "epoch": 0.5539053317662144,
      "grad_norm": 3.504819869995117,
      "learning_rate": 1.4481283422459895e-05,
      "loss": 0.15750575065612793,
      "num_input_tokens_seen": 1347320,
      "step": 1296,
      "train_runtime": 3103.621,
      "train_tokens_per_second": 434.112
    },
    {
      "epoch": 0.5543327278555401,
      "grad_norm": 2.975438117980957,
      "learning_rate": 1.4477005347593584e-05,
      "loss": 0.15112178027629852,
      "num_input_tokens_seen": 1348416,
      "step": 1297,
      "train_runtime": 3105.7492,
      "train_tokens_per_second": 434.168
    },
    {
      "epoch": 0.554760123944866,
      "grad_norm": 2.8611741065979004,
      "learning_rate": 1.4472727272727274e-05,
      "loss": 0.08872049301862717,
      "num_input_tokens_seen": 1349308,
      "step": 1298,
      "train_runtime": 3107.833,
      "train_tokens_per_second": 434.164
    },
    {
      "epoch": 0.5551875200341917,
      "grad_norm": 2.685610294342041,
      "learning_rate": 1.4468449197860963e-05,
      "loss": 0.11205846071243286,
      "num_input_tokens_seen": 1350388,
      "step": 1299,
      "train_runtime": 3109.9127,
      "train_tokens_per_second": 434.221
    },
    {
      "epoch": 0.5556149161235174,
      "grad_norm": 2.436103105545044,
      "learning_rate": 1.4464171122994654e-05,
      "loss": 0.04306722432374954,
      "num_input_tokens_seen": 1350990,
      "step": 1300,
      "train_runtime": 3111.9037,
      "train_tokens_per_second": 434.136
    },
    {
      "epoch": 0.5560423122128433,
      "grad_norm": 5.107798099517822,
      "learning_rate": 1.4459893048128344e-05,
      "loss": 0.2502478063106537,
      "num_input_tokens_seen": 1351776,
      "step": 1301,
      "train_runtime": 3113.8952,
      "train_tokens_per_second": 434.111
    },
    {
      "epoch": 0.556469708302169,
      "grad_norm": 5.479470252990723,
      "learning_rate": 1.4455614973262033e-05,
      "loss": 0.19407883286476135,
      "num_input_tokens_seen": 1352406,
      "step": 1302,
      "train_runtime": 3115.8991,
      "train_tokens_per_second": 434.034
    },
    {
      "epoch": 0.5568971043914949,
      "grad_norm": 2.941948413848877,
      "learning_rate": 1.4451336898395723e-05,
      "loss": 0.10889407992362976,
      "num_input_tokens_seen": 1353458,
      "step": 1303,
      "train_runtime": 3118.0275,
      "train_tokens_per_second": 434.075
    },
    {
      "epoch": 0.5573245004808206,
      "grad_norm": 4.489953994750977,
      "learning_rate": 1.4447058823529414e-05,
      "loss": 0.15718862414360046,
      "num_input_tokens_seen": 1354264,
      "step": 1304,
      "train_runtime": 3120.0442,
      "train_tokens_per_second": 434.053
    },
    {
      "epoch": 0.5577518965701463,
      "grad_norm": 2.7625503540039062,
      "learning_rate": 1.4442780748663103e-05,
      "loss": 0.16560781002044678,
      "num_input_tokens_seen": 1355288,
      "step": 1305,
      "train_runtime": 3122.1,
      "train_tokens_per_second": 434.095
    },
    {
      "epoch": 0.5581792926594722,
      "grad_norm": 3.410099506378174,
      "learning_rate": 1.4438502673796793e-05,
      "loss": 0.1142624020576477,
      "num_input_tokens_seen": 1356086,
      "step": 1306,
      "train_runtime": 3124.1233,
      "train_tokens_per_second": 434.069
    },
    {
      "epoch": 0.5586066887487979,
      "grad_norm": 2.389314889907837,
      "learning_rate": 1.4434224598930482e-05,
      "loss": 0.14941425621509552,
      "num_input_tokens_seen": 1357528,
      "step": 1307,
      "train_runtime": 3126.3436,
      "train_tokens_per_second": 434.222
    },
    {
      "epoch": 0.5590340848381238,
      "grad_norm": 2.9370124340057373,
      "learning_rate": 1.4429946524064173e-05,
      "loss": 0.20251935720443726,
      "num_input_tokens_seen": 1358710,
      "step": 1308,
      "train_runtime": 3128.4627,
      "train_tokens_per_second": 434.306
    },
    {
      "epoch": 0.5594614809274495,
      "grad_norm": 5.125398635864258,
      "learning_rate": 1.4425668449197863e-05,
      "loss": 0.1502501219511032,
      "num_input_tokens_seen": 1359730,
      "step": 1309,
      "train_runtime": 3130.5562,
      "train_tokens_per_second": 434.341
    },
    {
      "epoch": 0.5598888770167753,
      "grad_norm": 2.7636914253234863,
      "learning_rate": 1.4421390374331552e-05,
      "loss": 0.17712441086769104,
      "num_input_tokens_seen": 1360966,
      "step": 1310,
      "train_runtime": 3132.7133,
      "train_tokens_per_second": 434.437
    },
    {
      "epoch": 0.5603162731061011,
      "grad_norm": 4.530420780181885,
      "learning_rate": 1.4417112299465242e-05,
      "loss": 0.18464383482933044,
      "num_input_tokens_seen": 1361962,
      "step": 1311,
      "train_runtime": 3134.8087,
      "train_tokens_per_second": 434.464
    },
    {
      "epoch": 0.5607436691954268,
      "grad_norm": 3.288115978240967,
      "learning_rate": 1.4412834224598931e-05,
      "loss": 0.15711161494255066,
      "num_input_tokens_seen": 1362752,
      "step": 1312,
      "train_runtime": 3136.8402,
      "train_tokens_per_second": 434.435
    },
    {
      "epoch": 0.5611710652847527,
      "grad_norm": 2.5233519077301025,
      "learning_rate": 1.4408556149732622e-05,
      "loss": 0.14466112852096558,
      "num_input_tokens_seen": 1363708,
      "step": 1313,
      "train_runtime": 3138.9263,
      "train_tokens_per_second": 434.45
    },
    {
      "epoch": 0.5615984613740784,
      "grad_norm": 2.770968198776245,
      "learning_rate": 1.4404278074866312e-05,
      "loss": 0.1307995617389679,
      "num_input_tokens_seen": 1364932,
      "step": 1314,
      "train_runtime": 3141.0926,
      "train_tokens_per_second": 434.541
    },
    {
      "epoch": 0.5620258574634042,
      "grad_norm": 2.6532740592956543,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.15271973609924316,
      "num_input_tokens_seen": 1365832,
      "step": 1315,
      "train_runtime": 3143.2687,
      "train_tokens_per_second": 434.526
    },
    {
      "epoch": 0.56245325355273,
      "grad_norm": 2.3266897201538086,
      "learning_rate": 1.439572192513369e-05,
      "loss": 0.11991928517818451,
      "num_input_tokens_seen": 1367160,
      "step": 1316,
      "train_runtime": 3145.3845,
      "train_tokens_per_second": 434.656
    },
    {
      "epoch": 0.5628806496420558,
      "grad_norm": 3.4439101219177246,
      "learning_rate": 1.4391443850267382e-05,
      "loss": 0.23416438698768616,
      "num_input_tokens_seen": 1368060,
      "step": 1317,
      "train_runtime": 3147.4299,
      "train_tokens_per_second": 434.659
    },
    {
      "epoch": 0.5633080457313816,
      "grad_norm": 3.6046524047851562,
      "learning_rate": 1.4387165775401071e-05,
      "loss": 0.1598702222108841,
      "num_input_tokens_seen": 1368762,
      "step": 1318,
      "train_runtime": 3149.4738,
      "train_tokens_per_second": 434.6
    },
    {
      "epoch": 0.5637354418207073,
      "grad_norm": 2.938603401184082,
      "learning_rate": 1.438288770053476e-05,
      "loss": 0.1904858946800232,
      "num_input_tokens_seen": 1369696,
      "step": 1319,
      "train_runtime": 3151.5342,
      "train_tokens_per_second": 434.612
    },
    {
      "epoch": 0.5641628379100331,
      "grad_norm": 2.813359260559082,
      "learning_rate": 1.437860962566845e-05,
      "loss": 0.10506963729858398,
      "num_input_tokens_seen": 1370364,
      "step": 1320,
      "train_runtime": 3153.514,
      "train_tokens_per_second": 434.551
    },
    {
      "epoch": 0.5645902339993589,
      "grad_norm": 4.190011024475098,
      "learning_rate": 1.4374331550802141e-05,
      "loss": 0.1983892172574997,
      "num_input_tokens_seen": 1371314,
      "step": 1321,
      "train_runtime": 3162.523,
      "train_tokens_per_second": 433.614
    },
    {
      "epoch": 0.5650176300886847,
      "grad_norm": 2.8000965118408203,
      "learning_rate": 1.437005347593583e-05,
      "loss": 0.13952013850212097,
      "num_input_tokens_seen": 1372200,
      "step": 1322,
      "train_runtime": 3164.5991,
      "train_tokens_per_second": 433.609
    },
    {
      "epoch": 0.5654450261780105,
      "grad_norm": 2.927912950515747,
      "learning_rate": 1.436577540106952e-05,
      "loss": 0.16799136996269226,
      "num_input_tokens_seen": 1373306,
      "step": 1323,
      "train_runtime": 3166.7036,
      "train_tokens_per_second": 433.671
    },
    {
      "epoch": 0.5658724222673363,
      "grad_norm": 2.517874002456665,
      "learning_rate": 1.436149732620321e-05,
      "loss": 0.11278734356164932,
      "num_input_tokens_seen": 1374362,
      "step": 1324,
      "train_runtime": 3168.783,
      "train_tokens_per_second": 433.719
    },
    {
      "epoch": 0.566299818356662,
      "grad_norm": 3.439088821411133,
      "learning_rate": 1.43572192513369e-05,
      "loss": 0.16937866806983948,
      "num_input_tokens_seen": 1375172,
      "step": 1325,
      "train_runtime": 3170.8002,
      "train_tokens_per_second": 433.699
    },
    {
      "epoch": 0.5667272144459878,
      "grad_norm": 3.08162522315979,
      "learning_rate": 1.435294117647059e-05,
      "loss": 0.1832624077796936,
      "num_input_tokens_seen": 1376126,
      "step": 1326,
      "train_runtime": 3172.8748,
      "train_tokens_per_second": 433.716
    },
    {
      "epoch": 0.5671546105353136,
      "grad_norm": 2.31919002532959,
      "learning_rate": 1.434866310160428e-05,
      "loss": 0.12628880143165588,
      "num_input_tokens_seen": 1377132,
      "step": 1327,
      "train_runtime": 3174.9454,
      "train_tokens_per_second": 433.75
    },
    {
      "epoch": 0.5675820066246394,
      "grad_norm": 2.693835735321045,
      "learning_rate": 1.4344385026737967e-05,
      "loss": 0.13304206728935242,
      "num_input_tokens_seen": 1378052,
      "step": 1328,
      "train_runtime": 3176.987,
      "train_tokens_per_second": 433.761
    },
    {
      "epoch": 0.5680094027139652,
      "grad_norm": 2.582542657852173,
      "learning_rate": 1.434010695187166e-05,
      "loss": 0.12485780566930771,
      "num_input_tokens_seen": 1379138,
      "step": 1329,
      "train_runtime": 3179.0924,
      "train_tokens_per_second": 433.815
    },
    {
      "epoch": 0.5684367988032909,
      "grad_norm": 2.1392982006073,
      "learning_rate": 1.433582887700535e-05,
      "loss": 0.09820333868265152,
      "num_input_tokens_seen": 1380224,
      "step": 1330,
      "train_runtime": 3181.1739,
      "train_tokens_per_second": 433.873
    },
    {
      "epoch": 0.5688641948926167,
      "grad_norm": 2.1656980514526367,
      "learning_rate": 1.4331550802139039e-05,
      "loss": 0.08080549538135529,
      "num_input_tokens_seen": 1380956,
      "step": 1331,
      "train_runtime": 3183.1887,
      "train_tokens_per_second": 433.828
    },
    {
      "epoch": 0.5692915909819425,
      "grad_norm": 3.219681978225708,
      "learning_rate": 1.4327272727272727e-05,
      "loss": 0.0961499810218811,
      "num_input_tokens_seen": 1381664,
      "step": 1332,
      "train_runtime": 3185.1647,
      "train_tokens_per_second": 433.781
    },
    {
      "epoch": 0.5697189870712683,
      "grad_norm": 3.8177578449249268,
      "learning_rate": 1.432299465240642e-05,
      "loss": 0.2069641500711441,
      "num_input_tokens_seen": 1382478,
      "step": 1333,
      "train_runtime": 3187.1867,
      "train_tokens_per_second": 433.761
    },
    {
      "epoch": 0.5701463831605941,
      "grad_norm": 2.5426411628723145,
      "learning_rate": 1.4318716577540109e-05,
      "loss": 0.1197671890258789,
      "num_input_tokens_seen": 1383428,
      "step": 1334,
      "train_runtime": 3189.3526,
      "train_tokens_per_second": 433.765
    },
    {
      "epoch": 0.5705737792499198,
      "grad_norm": 3.00923752784729,
      "learning_rate": 1.4314438502673797e-05,
      "loss": 0.14582058787345886,
      "num_input_tokens_seen": 1384562,
      "step": 1335,
      "train_runtime": 3191.4769,
      "train_tokens_per_second": 433.831
    },
    {
      "epoch": 0.5710011753392457,
      "grad_norm": 3.2409698963165283,
      "learning_rate": 1.4310160427807486e-05,
      "loss": 0.1074337512254715,
      "num_input_tokens_seen": 1385582,
      "step": 1336,
      "train_runtime": 3193.6214,
      "train_tokens_per_second": 433.859
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 2.674619674682617,
      "learning_rate": 1.4305882352941179e-05,
      "loss": 0.17064478993415833,
      "num_input_tokens_seen": 1387014,
      "step": 1337,
      "train_runtime": 3195.7833,
      "train_tokens_per_second": 434.014
    },
    {
      "epoch": 0.5718559675178972,
      "grad_norm": 2.36714768409729,
      "learning_rate": 1.4301604278074868e-05,
      "loss": 0.07887314260005951,
      "num_input_tokens_seen": 1387936,
      "step": 1338,
      "train_runtime": 3197.8804,
      "train_tokens_per_second": 434.017
    },
    {
      "epoch": 0.572283363607223,
      "grad_norm": 3.2736432552337646,
      "learning_rate": 1.4297326203208556e-05,
      "loss": 0.11346343904733658,
      "num_input_tokens_seen": 1388608,
      "step": 1339,
      "train_runtime": 3199.8956,
      "train_tokens_per_second": 433.954
    },
    {
      "epoch": 0.5727107596965487,
      "grad_norm": 4.206210136413574,
      "learning_rate": 1.4293048128342246e-05,
      "loss": 0.18226617574691772,
      "num_input_tokens_seen": 1389596,
      "step": 1340,
      "train_runtime": 3201.9641,
      "train_tokens_per_second": 433.982
    },
    {
      "epoch": 0.5731381557858746,
      "grad_norm": 2.84797739982605,
      "learning_rate": 1.4288770053475938e-05,
      "loss": 0.12390857934951782,
      "num_input_tokens_seen": 1391016,
      "step": 1341,
      "train_runtime": 3204.1733,
      "train_tokens_per_second": 434.126
    },
    {
      "epoch": 0.5735655518752003,
      "grad_norm": 3.604992151260376,
      "learning_rate": 1.4284491978609626e-05,
      "loss": 0.1389472484588623,
      "num_input_tokens_seen": 1392974,
      "step": 1342,
      "train_runtime": 3206.4591,
      "train_tokens_per_second": 434.428
    },
    {
      "epoch": 0.5739929479645262,
      "grad_norm": 3.0878748893737793,
      "learning_rate": 1.4280213903743316e-05,
      "loss": 0.1665935218334198,
      "num_input_tokens_seen": 1394476,
      "step": 1343,
      "train_runtime": 3208.7286,
      "train_tokens_per_second": 434.588
    },
    {
      "epoch": 0.5744203440538519,
      "grad_norm": 3.2702009677886963,
      "learning_rate": 1.4275935828877005e-05,
      "loss": 0.1423080712556839,
      "num_input_tokens_seen": 1395278,
      "step": 1344,
      "train_runtime": 3210.7495,
      "train_tokens_per_second": 434.565
    },
    {
      "epoch": 0.5748477401431777,
      "grad_norm": 3.2000083923339844,
      "learning_rate": 1.4271657754010696e-05,
      "loss": 0.17459425330162048,
      "num_input_tokens_seen": 1396106,
      "step": 1345,
      "train_runtime": 3212.7956,
      "train_tokens_per_second": 434.546
    },
    {
      "epoch": 0.5752751362325035,
      "grad_norm": 3.0125555992126465,
      "learning_rate": 1.4267379679144386e-05,
      "loss": 0.1326783001422882,
      "num_input_tokens_seen": 1396912,
      "step": 1346,
      "train_runtime": 3214.8143,
      "train_tokens_per_second": 434.523
    },
    {
      "epoch": 0.5757025323218292,
      "grad_norm": 3.782444715499878,
      "learning_rate": 1.4263101604278075e-05,
      "loss": 0.17368285357952118,
      "num_input_tokens_seen": 1397604,
      "step": 1347,
      "train_runtime": 3216.8295,
      "train_tokens_per_second": 434.466
    },
    {
      "epoch": 0.5761299284111551,
      "grad_norm": 2.9148635864257812,
      "learning_rate": 1.4258823529411765e-05,
      "loss": 0.16578298807144165,
      "num_input_tokens_seen": 1398700,
      "step": 1348,
      "train_runtime": 3218.9106,
      "train_tokens_per_second": 434.526
    },
    {
      "epoch": 0.5765573245004808,
      "grad_norm": 2.904982805252075,
      "learning_rate": 1.4254545454545456e-05,
      "loss": 0.17295828461647034,
      "num_input_tokens_seen": 1399822,
      "step": 1349,
      "train_runtime": 3221.0225,
      "train_tokens_per_second": 434.589
    },
    {
      "epoch": 0.5769847205898067,
      "grad_norm": 2.682614803314209,
      "learning_rate": 1.4250267379679145e-05,
      "loss": 0.12830506265163422,
      "num_input_tokens_seen": 1401042,
      "step": 1350,
      "train_runtime": 3223.1815,
      "train_tokens_per_second": 434.677
    },
    {
      "epoch": 0.5774121166791324,
      "grad_norm": 3.2086193561553955,
      "learning_rate": 1.4245989304812835e-05,
      "loss": 0.15853194892406464,
      "num_input_tokens_seen": 1401878,
      "step": 1351,
      "train_runtime": 3231.5951,
      "train_tokens_per_second": 433.804
    },
    {
      "epoch": 0.5778395127684581,
      "grad_norm": 2.0538580417633057,
      "learning_rate": 1.4241711229946524e-05,
      "loss": 0.12270934134721756,
      "num_input_tokens_seen": 1402938,
      "step": 1352,
      "train_runtime": 3233.6739,
      "train_tokens_per_second": 433.853
    },
    {
      "epoch": 0.578266908857784,
      "grad_norm": 3.130448579788208,
      "learning_rate": 1.4237433155080215e-05,
      "loss": 0.12257538735866547,
      "num_input_tokens_seen": 1403754,
      "step": 1353,
      "train_runtime": 3235.6999,
      "train_tokens_per_second": 433.833
    },
    {
      "epoch": 0.5786943049471097,
      "grad_norm": 2.407013416290283,
      "learning_rate": 1.4233155080213905e-05,
      "loss": 0.12908770143985748,
      "num_input_tokens_seen": 1405168,
      "step": 1354,
      "train_runtime": 3237.9098,
      "train_tokens_per_second": 433.974
    },
    {
      "epoch": 0.5791217010364356,
      "grad_norm": 2.1559484004974365,
      "learning_rate": 1.4228877005347594e-05,
      "loss": 0.1556459665298462,
      "num_input_tokens_seen": 1406528,
      "step": 1355,
      "train_runtime": 3240.0631,
      "train_tokens_per_second": 434.105
    },
    {
      "epoch": 0.5795490971257613,
      "grad_norm": 2.7706427574157715,
      "learning_rate": 1.4224598930481284e-05,
      "loss": 0.19304051995277405,
      "num_input_tokens_seen": 1407702,
      "step": 1356,
      "train_runtime": 3242.1866,
      "train_tokens_per_second": 434.183
    },
    {
      "epoch": 0.579976493215087,
      "grad_norm": 1.9366533756256104,
      "learning_rate": 1.4220320855614975e-05,
      "loss": 0.07077852636575699,
      "num_input_tokens_seen": 1409130,
      "step": 1357,
      "train_runtime": 3244.3929,
      "train_tokens_per_second": 434.328
    },
    {
      "epoch": 0.5804038893044129,
      "grad_norm": 4.998035430908203,
      "learning_rate": 1.4216042780748664e-05,
      "loss": 0.13400226831436157,
      "num_input_tokens_seen": 1410038,
      "step": 1358,
      "train_runtime": 3246.436,
      "train_tokens_per_second": 434.334
    },
    {
      "epoch": 0.5808312853937386,
      "grad_norm": 2.978074789047241,
      "learning_rate": 1.4211764705882354e-05,
      "loss": 0.15089115500450134,
      "num_input_tokens_seen": 1410926,
      "step": 1359,
      "train_runtime": 3248.4758,
      "train_tokens_per_second": 434.335
    },
    {
      "epoch": 0.5812586814830645,
      "grad_norm": 3.8650741577148438,
      "learning_rate": 1.4207486631016043e-05,
      "loss": 0.14304523169994354,
      "num_input_tokens_seen": 1411922,
      "step": 1360,
      "train_runtime": 3250.587,
      "train_tokens_per_second": 434.359
    },
    {
      "epoch": 0.5816860775723902,
      "grad_norm": 2.8047916889190674,
      "learning_rate": 1.4203208556149734e-05,
      "loss": 0.14779841899871826,
      "num_input_tokens_seen": 1412952,
      "step": 1361,
      "train_runtime": 3252.6364,
      "train_tokens_per_second": 434.402
    },
    {
      "epoch": 0.582113473661716,
      "grad_norm": 3.7442920207977295,
      "learning_rate": 1.4198930481283424e-05,
      "loss": 0.19863316416740417,
      "num_input_tokens_seen": 1414016,
      "step": 1362,
      "train_runtime": 3254.7119,
      "train_tokens_per_second": 434.452
    },
    {
      "epoch": 0.5825408697510418,
      "grad_norm": 2.6118645668029785,
      "learning_rate": 1.4194652406417113e-05,
      "loss": 0.1354249268770218,
      "num_input_tokens_seen": 1415272,
      "step": 1363,
      "train_runtime": 3256.816,
      "train_tokens_per_second": 434.557
    },
    {
      "epoch": 0.5829682658403675,
      "grad_norm": 2.1527950763702393,
      "learning_rate": 1.4190374331550803e-05,
      "loss": 0.07591502368450165,
      "num_input_tokens_seen": 1416104,
      "step": 1364,
      "train_runtime": 3258.9519,
      "train_tokens_per_second": 434.527
    },
    {
      "epoch": 0.5833956619296934,
      "grad_norm": 3.659360647201538,
      "learning_rate": 1.4186096256684494e-05,
      "loss": 0.14368534088134766,
      "num_input_tokens_seen": 1417086,
      "step": 1365,
      "train_runtime": 3261.026,
      "train_tokens_per_second": 434.552
    },
    {
      "epoch": 0.5838230580190191,
      "grad_norm": 3.592390298843384,
      "learning_rate": 1.4181818181818183e-05,
      "loss": 0.18800120055675507,
      "num_input_tokens_seen": 1417778,
      "step": 1366,
      "train_runtime": 3262.9897,
      "train_tokens_per_second": 434.503
    },
    {
      "epoch": 0.5842504541083449,
      "grad_norm": 2.112363576889038,
      "learning_rate": 1.4177540106951873e-05,
      "loss": 0.10255702584981918,
      "num_input_tokens_seen": 1418892,
      "step": 1367,
      "train_runtime": 3265.1122,
      "train_tokens_per_second": 434.561
    },
    {
      "epoch": 0.5846778501976707,
      "grad_norm": 2.676422119140625,
      "learning_rate": 1.4173262032085562e-05,
      "loss": 0.10220567137002945,
      "num_input_tokens_seen": 1419800,
      "step": 1368,
      "train_runtime": 3267.1567,
      "train_tokens_per_second": 434.567
    },
    {
      "epoch": 0.5851052462869964,
      "grad_norm": 2.871175527572632,
      "learning_rate": 1.4168983957219253e-05,
      "loss": 0.14551648497581482,
      "num_input_tokens_seen": 1420856,
      "step": 1369,
      "train_runtime": 3269.2633,
      "train_tokens_per_second": 434.61
    },
    {
      "epoch": 0.5855326423763223,
      "grad_norm": 2.1074609756469727,
      "learning_rate": 1.4164705882352943e-05,
      "loss": 0.0820218026638031,
      "num_input_tokens_seen": 1421690,
      "step": 1370,
      "train_runtime": 3271.3095,
      "train_tokens_per_second": 434.594
    },
    {
      "epoch": 0.585960038465648,
      "grad_norm": 3.1287553310394287,
      "learning_rate": 1.4160427807486632e-05,
      "loss": 0.16858786344528198,
      "num_input_tokens_seen": 1422638,
      "step": 1371,
      "train_runtime": 3273.3897,
      "train_tokens_per_second": 434.607
    },
    {
      "epoch": 0.5863874345549738,
      "grad_norm": 3.2365715503692627,
      "learning_rate": 1.4156149732620321e-05,
      "loss": 0.15625795722007751,
      "num_input_tokens_seen": 1423712,
      "step": 1372,
      "train_runtime": 3275.462,
      "train_tokens_per_second": 434.66
    },
    {
      "epoch": 0.5868148306442996,
      "grad_norm": 2.434537649154663,
      "learning_rate": 1.4151871657754013e-05,
      "loss": 0.12010855227708817,
      "num_input_tokens_seen": 1424810,
      "step": 1373,
      "train_runtime": 3277.6151,
      "train_tokens_per_second": 434.709
    },
    {
      "epoch": 0.5872422267336254,
      "grad_norm": 2.9297800064086914,
      "learning_rate": 1.4147593582887702e-05,
      "loss": 0.15987205505371094,
      "num_input_tokens_seen": 1425684,
      "step": 1374,
      "train_runtime": 3279.6771,
      "train_tokens_per_second": 434.703
    },
    {
      "epoch": 0.5876696228229512,
      "grad_norm": 2.5956802368164062,
      "learning_rate": 1.4143315508021392e-05,
      "loss": 0.14205943048000336,
      "num_input_tokens_seen": 1427042,
      "step": 1375,
      "train_runtime": 3281.8156,
      "train_tokens_per_second": 434.833
    },
    {
      "epoch": 0.588097018912277,
      "grad_norm": 2.2184929847717285,
      "learning_rate": 1.4139037433155081e-05,
      "loss": 0.13644662499427795,
      "num_input_tokens_seen": 1428144,
      "step": 1376,
      "train_runtime": 3283.9122,
      "train_tokens_per_second": 434.891
    },
    {
      "epoch": 0.5885244150016027,
      "grad_norm": 4.882390022277832,
      "learning_rate": 1.4134759358288772e-05,
      "loss": 0.20036539435386658,
      "num_input_tokens_seen": 1428934,
      "step": 1377,
      "train_runtime": 3285.9482,
      "train_tokens_per_second": 434.862
    },
    {
      "epoch": 0.5889518110909285,
      "grad_norm": 2.24263858795166,
      "learning_rate": 1.4130481283422462e-05,
      "loss": 0.12761899828910828,
      "num_input_tokens_seen": 1430552,
      "step": 1378,
      "train_runtime": 3288.1843,
      "train_tokens_per_second": 435.058
    },
    {
      "epoch": 0.5893792071802543,
      "grad_norm": 2.6268224716186523,
      "learning_rate": 1.4126203208556151e-05,
      "loss": 0.13226649165153503,
      "num_input_tokens_seen": 1431452,
      "step": 1379,
      "train_runtime": 3290.2762,
      "train_tokens_per_second": 435.055
    },
    {
      "epoch": 0.5898066032695801,
      "grad_norm": 4.0447678565979,
      "learning_rate": 1.412192513368984e-05,
      "loss": 0.11312992870807648,
      "num_input_tokens_seen": 1432088,
      "step": 1380,
      "train_runtime": 3292.2512,
      "train_tokens_per_second": 434.987
    },
    {
      "epoch": 0.5902339993589059,
      "grad_norm": 2.023045778274536,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 0.07444490492343903,
      "num_input_tokens_seen": 1433322,
      "step": 1381,
      "train_runtime": 3301.1164,
      "train_tokens_per_second": 434.193
    },
    {
      "epoch": 0.5906613954482316,
      "grad_norm": 3.653201103210449,
      "learning_rate": 1.4113368983957221e-05,
      "loss": 0.20476751029491425,
      "num_input_tokens_seen": 1434046,
      "step": 1382,
      "train_runtime": 3303.1286,
      "train_tokens_per_second": 434.148
    },
    {
      "epoch": 0.5910887915375574,
      "grad_norm": 3.6361706256866455,
      "learning_rate": 1.410909090909091e-05,
      "loss": 0.16467516124248505,
      "num_input_tokens_seen": 1434882,
      "step": 1383,
      "train_runtime": 3305.316,
      "train_tokens_per_second": 434.113
    },
    {
      "epoch": 0.5915161876268832,
      "grad_norm": 2.283289909362793,
      "learning_rate": 1.41048128342246e-05,
      "loss": 0.14137494564056396,
      "num_input_tokens_seen": 1436088,
      "step": 1384,
      "train_runtime": 3307.6527,
      "train_tokens_per_second": 434.171
    },
    {
      "epoch": 0.591943583716209,
      "grad_norm": 2.8381502628326416,
      "learning_rate": 1.4100534759358291e-05,
      "loss": 0.14940938353538513,
      "num_input_tokens_seen": 1437124,
      "step": 1385,
      "train_runtime": 3309.7282,
      "train_tokens_per_second": 434.212
    },
    {
      "epoch": 0.5923709798055348,
      "grad_norm": 2.9985997676849365,
      "learning_rate": 1.409625668449198e-05,
      "loss": 0.1341642588376999,
      "num_input_tokens_seen": 1438190,
      "step": 1386,
      "train_runtime": 3311.8006,
      "train_tokens_per_second": 434.262
    },
    {
      "epoch": 0.5927983758948605,
      "grad_norm": 2.535414218902588,
      "learning_rate": 1.409197860962567e-05,
      "loss": 0.13133931159973145,
      "num_input_tokens_seen": 1439252,
      "step": 1387,
      "train_runtime": 3313.8776,
      "train_tokens_per_second": 434.311
    },
    {
      "epoch": 0.5932257719841864,
      "grad_norm": 3.5592312812805176,
      "learning_rate": 1.408770053475936e-05,
      "loss": 0.10969515144824982,
      "num_input_tokens_seen": 1439840,
      "step": 1388,
      "train_runtime": 3315.8,
      "train_tokens_per_second": 434.236
    },
    {
      "epoch": 0.5936531680735121,
      "grad_norm": 3.807300329208374,
      "learning_rate": 1.408342245989305e-05,
      "loss": 0.19459311664104462,
      "num_input_tokens_seen": 1440750,
      "step": 1389,
      "train_runtime": 3317.8803,
      "train_tokens_per_second": 434.238
    },
    {
      "epoch": 0.5940805641628379,
      "grad_norm": 2.7336909770965576,
      "learning_rate": 1.407914438502674e-05,
      "loss": 0.14076997339725494,
      "num_input_tokens_seen": 1441662,
      "step": 1390,
      "train_runtime": 3319.9232,
      "train_tokens_per_second": 434.246
    },
    {
      "epoch": 0.5945079602521637,
      "grad_norm": 3.067060947418213,
      "learning_rate": 1.407486631016043e-05,
      "loss": 0.17666979134082794,
      "num_input_tokens_seen": 1442694,
      "step": 1391,
      "train_runtime": 3322.0132,
      "train_tokens_per_second": 434.283
    },
    {
      "epoch": 0.5949353563414895,
      "grad_norm": 2.2158451080322266,
      "learning_rate": 1.4070588235294119e-05,
      "loss": 0.1015886515378952,
      "num_input_tokens_seen": 1443642,
      "step": 1392,
      "train_runtime": 3324.0748,
      "train_tokens_per_second": 434.299
    },
    {
      "epoch": 0.5953627524308153,
      "grad_norm": 4.7660746574401855,
      "learning_rate": 1.406631016042781e-05,
      "loss": 0.16630813479423523,
      "num_input_tokens_seen": 1444738,
      "step": 1393,
      "train_runtime": 3326.2039,
      "train_tokens_per_second": 434.35
    },
    {
      "epoch": 0.595790148520141,
      "grad_norm": 4.274450778961182,
      "learning_rate": 1.40620320855615e-05,
      "loss": 0.25991204380989075,
      "num_input_tokens_seen": 1445596,
      "step": 1394,
      "train_runtime": 3328.251,
      "train_tokens_per_second": 434.341
    },
    {
      "epoch": 0.5962175446094669,
      "grad_norm": 3.8187432289123535,
      "learning_rate": 1.4057754010695189e-05,
      "loss": 0.17650507390499115,
      "num_input_tokens_seen": 1446378,
      "step": 1395,
      "train_runtime": 3330.2601,
      "train_tokens_per_second": 434.314
    },
    {
      "epoch": 0.5966449406987926,
      "grad_norm": 3.882115364074707,
      "learning_rate": 1.4053475935828877e-05,
      "loss": 0.20205016434192657,
      "num_input_tokens_seen": 1447422,
      "step": 1396,
      "train_runtime": 3332.3464,
      "train_tokens_per_second": 434.355
    },
    {
      "epoch": 0.5970723367881184,
      "grad_norm": 3.42087721824646,
      "learning_rate": 1.404919786096257e-05,
      "loss": 0.2658466398715973,
      "num_input_tokens_seen": 1449220,
      "step": 1397,
      "train_runtime": 3334.5762,
      "train_tokens_per_second": 434.604
    },
    {
      "epoch": 0.5974997328774442,
      "grad_norm": 3.7142629623413086,
      "learning_rate": 1.4044919786096259e-05,
      "loss": 0.23548118770122528,
      "num_input_tokens_seen": 1449952,
      "step": 1398,
      "train_runtime": 3336.5859,
      "train_tokens_per_second": 434.562
    },
    {
      "epoch": 0.5979271289667699,
      "grad_norm": 2.3331501483917236,
      "learning_rate": 1.4040641711229948e-05,
      "loss": 0.05423790216445923,
      "num_input_tokens_seen": 1450596,
      "step": 1399,
      "train_runtime": 3338.5411,
      "train_tokens_per_second": 434.5
    },
    {
      "epoch": 0.5983545250560958,
      "grad_norm": 2.72188401222229,
      "learning_rate": 1.4036363636363636e-05,
      "loss": 0.12585221230983734,
      "num_input_tokens_seen": 1451548,
      "step": 1400,
      "train_runtime": 3340.6263,
      "train_tokens_per_second": 434.514
    },
    {
      "epoch": 0.5987819211454215,
      "grad_norm": 2.5179193019866943,
      "learning_rate": 1.4032085561497329e-05,
      "loss": 0.15464454889297485,
      "num_input_tokens_seen": 1452764,
      "step": 1401,
      "train_runtime": 3342.7816,
      "train_tokens_per_second": 434.597
    },
    {
      "epoch": 0.5992093172347474,
      "grad_norm": 2.6366610527038574,
      "learning_rate": 1.4027807486631018e-05,
      "loss": 0.11611218750476837,
      "num_input_tokens_seen": 1453610,
      "step": 1402,
      "train_runtime": 3344.8676,
      "train_tokens_per_second": 434.579
    },
    {
      "epoch": 0.5996367133240731,
      "grad_norm": 2.4527528285980225,
      "learning_rate": 1.4023529411764706e-05,
      "loss": 0.10760036110877991,
      "num_input_tokens_seen": 1454470,
      "step": 1403,
      "train_runtime": 3346.9549,
      "train_tokens_per_second": 434.565
    },
    {
      "epoch": 0.6000641094133988,
      "grad_norm": 3.5098700523376465,
      "learning_rate": 1.4019251336898396e-05,
      "loss": 0.12446847558021545,
      "num_input_tokens_seen": 1455998,
      "step": 1404,
      "train_runtime": 3349.2072,
      "train_tokens_per_second": 434.729
    },
    {
      "epoch": 0.6000641094133988,
      "eval_loss": 0.5446999669075012,
      "eval_runtime": 58.011,
      "eval_samples_per_second": 17.204,
      "eval_steps_per_second": 8.602,
      "num_input_tokens_seen": 1455998,
      "step": 1404
    },
    {
      "epoch": 0.6004915055027247,
      "grad_norm": 3.699281692504883,
      "learning_rate": 1.4014973262032088e-05,
      "loss": 0.19538527727127075,
      "num_input_tokens_seen": 1456830,
      "step": 1405,
      "train_runtime": 3409.262,
      "train_tokens_per_second": 427.315
    },
    {
      "epoch": 0.6009189015920504,
      "grad_norm": 3.037367105484009,
      "learning_rate": 1.4010695187165778e-05,
      "loss": 0.19640500843524933,
      "num_input_tokens_seen": 1457788,
      "step": 1406,
      "train_runtime": 3411.3672,
      "train_tokens_per_second": 427.332
    },
    {
      "epoch": 0.6013462976813763,
      "grad_norm": 3.306971788406372,
      "learning_rate": 1.4006417112299466e-05,
      "loss": 0.18082109093666077,
      "num_input_tokens_seen": 1458838,
      "step": 1407,
      "train_runtime": 3413.4472,
      "train_tokens_per_second": 427.38
    },
    {
      "epoch": 0.601773693770702,
      "grad_norm": 2.507986545562744,
      "learning_rate": 1.4002139037433155e-05,
      "loss": 0.11207126826047897,
      "num_input_tokens_seen": 1459840,
      "step": 1408,
      "train_runtime": 3415.5138,
      "train_tokens_per_second": 427.414
    },
    {
      "epoch": 0.6022010898600277,
      "grad_norm": 2.9219322204589844,
      "learning_rate": 1.3997860962566848e-05,
      "loss": 0.19506515562534332,
      "num_input_tokens_seen": 1460590,
      "step": 1409,
      "train_runtime": 3417.5371,
      "train_tokens_per_second": 427.381
    },
    {
      "epoch": 0.6026284859493536,
      "grad_norm": 2.4391496181488037,
      "learning_rate": 1.3993582887700536e-05,
      "loss": 0.15730175375938416,
      "num_input_tokens_seen": 1462106,
      "step": 1410,
      "train_runtime": 3419.7508,
      "train_tokens_per_second": 427.548
    },
    {
      "epoch": 0.6030558820386793,
      "grad_norm": 2.110074996948242,
      "learning_rate": 1.3989304812834225e-05,
      "loss": 0.10840006917715073,
      "num_input_tokens_seen": 1463456,
      "step": 1411,
      "train_runtime": 3428.6259,
      "train_tokens_per_second": 426.835
    },
    {
      "epoch": 0.6034832781280052,
      "grad_norm": 2.5467004776000977,
      "learning_rate": 1.3985026737967915e-05,
      "loss": 0.13251107931137085,
      "num_input_tokens_seen": 1464550,
      "step": 1412,
      "train_runtime": 3430.7299,
      "train_tokens_per_second": 426.892
    },
    {
      "epoch": 0.6039106742173309,
      "grad_norm": 3.6813597679138184,
      "learning_rate": 1.3980748663101606e-05,
      "loss": 0.17481614649295807,
      "num_input_tokens_seen": 1465682,
      "step": 1413,
      "train_runtime": 3433.2347,
      "train_tokens_per_second": 426.91
    },
    {
      "epoch": 0.6043380703066566,
      "grad_norm": 2.6558637619018555,
      "learning_rate": 1.3976470588235295e-05,
      "loss": 0.17672036588191986,
      "num_input_tokens_seen": 1466848,
      "step": 1414,
      "train_runtime": 3435.4078,
      "train_tokens_per_second": 426.979
    },
    {
      "epoch": 0.6047654663959825,
      "grad_norm": 2.469191789627075,
      "learning_rate": 1.3972192513368985e-05,
      "loss": 0.08628380298614502,
      "num_input_tokens_seen": 1467706,
      "step": 1415,
      "train_runtime": 3437.4489,
      "train_tokens_per_second": 426.975
    },
    {
      "epoch": 0.6051928624853082,
      "grad_norm": 3.496084690093994,
      "learning_rate": 1.3967914438502674e-05,
      "loss": 0.22254379093647003,
      "num_input_tokens_seen": 1468522,
      "step": 1416,
      "train_runtime": 3439.4674,
      "train_tokens_per_second": 426.962
    },
    {
      "epoch": 0.6056202585746341,
      "grad_norm": 2.5808959007263184,
      "learning_rate": 1.3963636363636365e-05,
      "loss": 0.12229837477207184,
      "num_input_tokens_seen": 1469476,
      "step": 1417,
      "train_runtime": 3441.5663,
      "train_tokens_per_second": 426.979
    },
    {
      "epoch": 0.6060476546639598,
      "grad_norm": 2.2333199977874756,
      "learning_rate": 1.3959358288770055e-05,
      "loss": 0.1297299712896347,
      "num_input_tokens_seen": 1470466,
      "step": 1418,
      "train_runtime": 3443.6979,
      "train_tokens_per_second": 427.002
    },
    {
      "epoch": 0.6064750507532856,
      "grad_norm": 2.1581215858459473,
      "learning_rate": 1.3955080213903744e-05,
      "loss": 0.08850215375423431,
      "num_input_tokens_seen": 1471740,
      "step": 1419,
      "train_runtime": 3445.8413,
      "train_tokens_per_second": 427.106
    },
    {
      "epoch": 0.6069024468426114,
      "grad_norm": 2.688570976257324,
      "learning_rate": 1.3950802139037433e-05,
      "loss": 0.10546266287565231,
      "num_input_tokens_seen": 1472736,
      "step": 1420,
      "train_runtime": 3447.9303,
      "train_tokens_per_second": 427.136
    },
    {
      "epoch": 0.6073298429319371,
      "grad_norm": 3.4864931106567383,
      "learning_rate": 1.3946524064171123e-05,
      "loss": 0.17743486166000366,
      "num_input_tokens_seen": 1473460,
      "step": 1421,
      "train_runtime": 3449.9209,
      "train_tokens_per_second": 427.1
    },
    {
      "epoch": 0.607757239021263,
      "grad_norm": 2.444497585296631,
      "learning_rate": 1.3942245989304814e-05,
      "loss": 0.09064759314060211,
      "num_input_tokens_seen": 1474266,
      "step": 1422,
      "train_runtime": 3451.9487,
      "train_tokens_per_second": 427.082
    },
    {
      "epoch": 0.6081846351105887,
      "grad_norm": 4.110256195068359,
      "learning_rate": 1.3937967914438504e-05,
      "loss": 0.1383572667837143,
      "num_input_tokens_seen": 1475044,
      "step": 1423,
      "train_runtime": 3453.9683,
      "train_tokens_per_second": 427.058
    },
    {
      "epoch": 0.6086120311999145,
      "grad_norm": 4.379847049713135,
      "learning_rate": 1.3933689839572193e-05,
      "loss": 0.16037997603416443,
      "num_input_tokens_seen": 1475826,
      "step": 1424,
      "train_runtime": 3455.9833,
      "train_tokens_per_second": 427.035
    },
    {
      "epoch": 0.6090394272892403,
      "grad_norm": 3.4214606285095215,
      "learning_rate": 1.3929411764705882e-05,
      "loss": 0.14068466424942017,
      "num_input_tokens_seen": 1476594,
      "step": 1425,
      "train_runtime": 3458.0036,
      "train_tokens_per_second": 427.008
    },
    {
      "epoch": 0.6094668233785661,
      "grad_norm": 2.551879644393921,
      "learning_rate": 1.3925133689839574e-05,
      "loss": 0.10875682532787323,
      "num_input_tokens_seen": 1477446,
      "step": 1426,
      "train_runtime": 3460.0547,
      "train_tokens_per_second": 427.001
    },
    {
      "epoch": 0.6098942194678919,
      "grad_norm": 3.4368040561676025,
      "learning_rate": 1.3920855614973263e-05,
      "loss": 0.18043841421604156,
      "num_input_tokens_seen": 1478308,
      "step": 1427,
      "train_runtime": 3462.1006,
      "train_tokens_per_second": 426.997
    },
    {
      "epoch": 0.6103216155572176,
      "grad_norm": 3.0184576511383057,
      "learning_rate": 1.3916577540106952e-05,
      "loss": 0.09612550586462021,
      "num_input_tokens_seen": 1479076,
      "step": 1428,
      "train_runtime": 3464.142,
      "train_tokens_per_second": 426.967
    },
    {
      "epoch": 0.6107490116465434,
      "grad_norm": 2.936882734298706,
      "learning_rate": 1.3912299465240642e-05,
      "loss": 0.08449357748031616,
      "num_input_tokens_seen": 1479708,
      "step": 1429,
      "train_runtime": 3466.0802,
      "train_tokens_per_second": 426.911
    },
    {
      "epoch": 0.6111764077358692,
      "grad_norm": 2.52876615524292,
      "learning_rate": 1.3908021390374333e-05,
      "loss": 0.16039231419563293,
      "num_input_tokens_seen": 1480992,
      "step": 1430,
      "train_runtime": 3468.2199,
      "train_tokens_per_second": 427.018
    },
    {
      "epoch": 0.611603803825195,
      "grad_norm": 3.0217013359069824,
      "learning_rate": 1.3903743315508022e-05,
      "loss": 0.11687301099300385,
      "num_input_tokens_seen": 1482360,
      "step": 1431,
      "train_runtime": 3470.3997,
      "train_tokens_per_second": 427.144
    },
    {
      "epoch": 0.6120311999145208,
      "grad_norm": 2.031693696975708,
      "learning_rate": 1.3899465240641712e-05,
      "loss": 0.09202948212623596,
      "num_input_tokens_seen": 1483492,
      "step": 1432,
      "train_runtime": 3472.4846,
      "train_tokens_per_second": 427.213
    },
    {
      "epoch": 0.6124585960038466,
      "grad_norm": 2.821016788482666,
      "learning_rate": 1.3895187165775401e-05,
      "loss": 0.17000722885131836,
      "num_input_tokens_seen": 1484432,
      "step": 1433,
      "train_runtime": 3474.5355,
      "train_tokens_per_second": 427.232
    },
    {
      "epoch": 0.6128859920931723,
      "grad_norm": 2.8982903957366943,
      "learning_rate": 1.3890909090909093e-05,
      "loss": 0.17944198846817017,
      "num_input_tokens_seen": 1485592,
      "step": 1434,
      "train_runtime": 3476.654,
      "train_tokens_per_second": 427.305
    },
    {
      "epoch": 0.6133133881824981,
      "grad_norm": 3.3795251846313477,
      "learning_rate": 1.3886631016042782e-05,
      "loss": 0.178505539894104,
      "num_input_tokens_seen": 1486452,
      "step": 1435,
      "train_runtime": 3478.7327,
      "train_tokens_per_second": 427.297
    },
    {
      "epoch": 0.6137407842718239,
      "grad_norm": 3.649630546569824,
      "learning_rate": 1.3882352941176471e-05,
      "loss": 0.17836499214172363,
      "num_input_tokens_seen": 1487382,
      "step": 1436,
      "train_runtime": 3480.8015,
      "train_tokens_per_second": 427.31
    },
    {
      "epoch": 0.6141681803611497,
      "grad_norm": 3.3058955669403076,
      "learning_rate": 1.3878074866310161e-05,
      "loss": 0.22165974974632263,
      "num_input_tokens_seen": 1488180,
      "step": 1437,
      "train_runtime": 3482.8414,
      "train_tokens_per_second": 427.289
    },
    {
      "epoch": 0.6145955764504755,
      "grad_norm": 2.4731411933898926,
      "learning_rate": 1.3873796791443852e-05,
      "loss": 0.09999997913837433,
      "num_input_tokens_seen": 1489138,
      "step": 1438,
      "train_runtime": 3484.9521,
      "train_tokens_per_second": 427.305
    },
    {
      "epoch": 0.6150229725398013,
      "grad_norm": 3.4230947494506836,
      "learning_rate": 1.3869518716577541e-05,
      "loss": 0.0926116406917572,
      "num_input_tokens_seen": 1490098,
      "step": 1439,
      "train_runtime": 3487.0557,
      "train_tokens_per_second": 427.323
    },
    {
      "epoch": 0.615450368629127,
      "grad_norm": 2.837399959564209,
      "learning_rate": 1.3865240641711231e-05,
      "loss": 0.13646970689296722,
      "num_input_tokens_seen": 1490940,
      "step": 1440,
      "train_runtime": 3489.0748,
      "train_tokens_per_second": 427.317
    },
    {
      "epoch": 0.6158777647184528,
      "grad_norm": 5.884909152984619,
      "learning_rate": 1.386096256684492e-05,
      "loss": 0.11018891632556915,
      "num_input_tokens_seen": 1491526,
      "step": 1441,
      "train_runtime": 3498.0344,
      "train_tokens_per_second": 426.39
    },
    {
      "epoch": 0.6163051608077786,
      "grad_norm": 3.2749173641204834,
      "learning_rate": 1.3856684491978611e-05,
      "loss": 0.16875940561294556,
      "num_input_tokens_seen": 1492750,
      "step": 1442,
      "train_runtime": 3500.1556,
      "train_tokens_per_second": 426.481
    },
    {
      "epoch": 0.6167325568971044,
      "grad_norm": 2.897189140319824,
      "learning_rate": 1.3852406417112301e-05,
      "loss": 0.11121410876512527,
      "num_input_tokens_seen": 1493648,
      "step": 1443,
      "train_runtime": 3502.2555,
      "train_tokens_per_second": 426.482
    },
    {
      "epoch": 0.6171599529864302,
      "grad_norm": 3.547976493835449,
      "learning_rate": 1.384812834224599e-05,
      "loss": 0.1049327403306961,
      "num_input_tokens_seen": 1494386,
      "step": 1444,
      "train_runtime": 3504.2846,
      "train_tokens_per_second": 426.445
    },
    {
      "epoch": 0.617587349075756,
      "grad_norm": 2.83156156539917,
      "learning_rate": 1.384385026737968e-05,
      "loss": 0.13763849437236786,
      "num_input_tokens_seen": 1495750,
      "step": 1445,
      "train_runtime": 3506.4312,
      "train_tokens_per_second": 426.573
    },
    {
      "epoch": 0.6180147451650817,
      "grad_norm": 3.80690336227417,
      "learning_rate": 1.3839572192513371e-05,
      "loss": 0.21457195281982422,
      "num_input_tokens_seen": 1496632,
      "step": 1446,
      "train_runtime": 3508.4691,
      "train_tokens_per_second": 426.577
    },
    {
      "epoch": 0.6184421412544076,
      "grad_norm": 3.0698084831237793,
      "learning_rate": 1.383529411764706e-05,
      "loss": 0.13854141533374786,
      "num_input_tokens_seen": 1497298,
      "step": 1447,
      "train_runtime": 3510.4857,
      "train_tokens_per_second": 426.522
    },
    {
      "epoch": 0.6188695373437333,
      "grad_norm": 3.7672619819641113,
      "learning_rate": 1.383101604278075e-05,
      "loss": 0.21939069032669067,
      "num_input_tokens_seen": 1497988,
      "step": 1448,
      "train_runtime": 3512.4619,
      "train_tokens_per_second": 426.478
    },
    {
      "epoch": 0.6192969334330591,
      "grad_norm": 2.3915205001831055,
      "learning_rate": 1.382673796791444e-05,
      "loss": 0.08274801075458527,
      "num_input_tokens_seen": 1498954,
      "step": 1449,
      "train_runtime": 3514.5473,
      "train_tokens_per_second": 426.5
    },
    {
      "epoch": 0.6197243295223849,
      "grad_norm": 2.8864173889160156,
      "learning_rate": 1.382245989304813e-05,
      "loss": 0.09981369227170944,
      "num_input_tokens_seen": 1499586,
      "step": 1450,
      "train_runtime": 3516.4958,
      "train_tokens_per_second": 426.443
    },
    {
      "epoch": 0.6201517256117106,
      "grad_norm": 3.0598063468933105,
      "learning_rate": 1.381818181818182e-05,
      "loss": 0.14454570412635803,
      "num_input_tokens_seen": 1500474,
      "step": 1451,
      "train_runtime": 3518.5357,
      "train_tokens_per_second": 426.448
    },
    {
      "epoch": 0.6205791217010365,
      "grad_norm": 2.9035451412200928,
      "learning_rate": 1.381390374331551e-05,
      "loss": 0.1534239947795868,
      "num_input_tokens_seen": 1501270,
      "step": 1452,
      "train_runtime": 3520.6057,
      "train_tokens_per_second": 426.424
    },
    {
      "epoch": 0.6210065177903622,
      "grad_norm": 3.257843255996704,
      "learning_rate": 1.3809625668449199e-05,
      "loss": 0.11779599636793137,
      "num_input_tokens_seen": 1502062,
      "step": 1453,
      "train_runtime": 3522.6247,
      "train_tokens_per_second": 426.404
    },
    {
      "epoch": 0.621433913879688,
      "grad_norm": 2.6349258422851562,
      "learning_rate": 1.380534759358289e-05,
      "loss": 0.10472257435321808,
      "num_input_tokens_seen": 1502904,
      "step": 1454,
      "train_runtime": 3524.6418,
      "train_tokens_per_second": 426.399
    },
    {
      "epoch": 0.6218613099690138,
      "grad_norm": 3.440901756286621,
      "learning_rate": 1.380106951871658e-05,
      "loss": 0.12523989379405975,
      "num_input_tokens_seen": 1503628,
      "step": 1455,
      "train_runtime": 3526.6536,
      "train_tokens_per_second": 426.361
    },
    {
      "epoch": 0.6222887060583395,
      "grad_norm": 4.631787300109863,
      "learning_rate": 1.3796791443850269e-05,
      "loss": 0.14164529740810394,
      "num_input_tokens_seen": 1504378,
      "step": 1456,
      "train_runtime": 3528.7365,
      "train_tokens_per_second": 426.322
    },
    {
      "epoch": 0.6227161021476654,
      "grad_norm": 3.1214423179626465,
      "learning_rate": 1.3792513368983958e-05,
      "loss": 0.16745080053806305,
      "num_input_tokens_seen": 1505316,
      "step": 1457,
      "train_runtime": 3530.809,
      "train_tokens_per_second": 426.337
    },
    {
      "epoch": 0.6231434982369911,
      "grad_norm": 3.417099952697754,
      "learning_rate": 1.378823529411765e-05,
      "loss": 0.22412510216236115,
      "num_input_tokens_seen": 1506384,
      "step": 1458,
      "train_runtime": 3532.9084,
      "train_tokens_per_second": 426.386
    },
    {
      "epoch": 0.623570894326317,
      "grad_norm": 2.4870553016662598,
      "learning_rate": 1.3783957219251339e-05,
      "loss": 0.12703093886375427,
      "num_input_tokens_seen": 1507456,
      "step": 1459,
      "train_runtime": 3534.9855,
      "train_tokens_per_second": 426.439
    },
    {
      "epoch": 0.6239982904156427,
      "grad_norm": 3.1759095191955566,
      "learning_rate": 1.3779679144385028e-05,
      "loss": 0.11046059429645538,
      "num_input_tokens_seen": 1508336,
      "step": 1460,
      "train_runtime": 3537.0315,
      "train_tokens_per_second": 426.441
    },
    {
      "epoch": 0.6244256865049684,
      "grad_norm": 2.964648485183716,
      "learning_rate": 1.3775401069518716e-05,
      "loss": 0.1784340739250183,
      "num_input_tokens_seen": 1509728,
      "step": 1461,
      "train_runtime": 3539.1928,
      "train_tokens_per_second": 426.574
    },
    {
      "epoch": 0.6248530825942943,
      "grad_norm": 2.6775224208831787,
      "learning_rate": 1.3771122994652409e-05,
      "loss": 0.11438018828630447,
      "num_input_tokens_seen": 1510484,
      "step": 1462,
      "train_runtime": 3541.1968,
      "train_tokens_per_second": 426.546
    },
    {
      "epoch": 0.62528047868362,
      "grad_norm": 3.12679386138916,
      "learning_rate": 1.3766844919786098e-05,
      "loss": 0.21539583802223206,
      "num_input_tokens_seen": 1511506,
      "step": 1463,
      "train_runtime": 3543.2609,
      "train_tokens_per_second": 426.586
    },
    {
      "epoch": 0.6257078747729459,
      "grad_norm": 2.8939411640167236,
      "learning_rate": 1.3762566844919786e-05,
      "loss": 0.12157222628593445,
      "num_input_tokens_seen": 1512420,
      "step": 1464,
      "train_runtime": 3545.2975,
      "train_tokens_per_second": 426.599
    },
    {
      "epoch": 0.6261352708622716,
      "grad_norm": 2.128225564956665,
      "learning_rate": 1.3758288770053475e-05,
      "loss": 0.10394015908241272,
      "num_input_tokens_seen": 1513738,
      "step": 1465,
      "train_runtime": 3547.4778,
      "train_tokens_per_second": 426.708
    },
    {
      "epoch": 0.6265626669515973,
      "grad_norm": 4.000294208526611,
      "learning_rate": 1.3754010695187168e-05,
      "loss": 0.1444389969110489,
      "num_input_tokens_seen": 1514442,
      "step": 1466,
      "train_runtime": 3549.4848,
      "train_tokens_per_second": 426.665
    },
    {
      "epoch": 0.6269900630409232,
      "grad_norm": 4.2455291748046875,
      "learning_rate": 1.3749732620320858e-05,
      "loss": 0.1115453839302063,
      "num_input_tokens_seen": 1515064,
      "step": 1467,
      "train_runtime": 3551.4289,
      "train_tokens_per_second": 426.607
    },
    {
      "epoch": 0.6274174591302489,
      "grad_norm": 9.257857322692871,
      "learning_rate": 1.3745454545454546e-05,
      "loss": 0.22937318682670593,
      "num_input_tokens_seen": 1516422,
      "step": 1468,
      "train_runtime": 3553.5886,
      "train_tokens_per_second": 426.73
    },
    {
      "epoch": 0.6278448552195748,
      "grad_norm": 2.7044317722320557,
      "learning_rate": 1.3741176470588235e-05,
      "loss": 0.13837867975234985,
      "num_input_tokens_seen": 1517430,
      "step": 1469,
      "train_runtime": 3555.6592,
      "train_tokens_per_second": 426.765
    },
    {
      "epoch": 0.6282722513089005,
      "grad_norm": 3.1235721111297607,
      "learning_rate": 1.3736898395721928e-05,
      "loss": 0.09635145962238312,
      "num_input_tokens_seen": 1518076,
      "step": 1470,
      "train_runtime": 3557.63,
      "train_tokens_per_second": 426.71
    },
    {
      "epoch": 0.6286996473982263,
      "grad_norm": 2.6795527935028076,
      "learning_rate": 1.3732620320855616e-05,
      "loss": 0.11742737144231796,
      "num_input_tokens_seen": 1519230,
      "step": 1471,
      "train_runtime": 3566.3589,
      "train_tokens_per_second": 425.989
    },
    {
      "epoch": 0.6291270434875521,
      "grad_norm": 2.943836212158203,
      "learning_rate": 1.3728342245989305e-05,
      "loss": 0.09473095834255219,
      "num_input_tokens_seen": 1519822,
      "step": 1472,
      "train_runtime": 3568.3578,
      "train_tokens_per_second": 425.916
    },
    {
      "epoch": 0.6295544395768778,
      "grad_norm": 2.960627317428589,
      "learning_rate": 1.3724064171122994e-05,
      "loss": 0.08305860310792923,
      "num_input_tokens_seen": 1520542,
      "step": 1473,
      "train_runtime": 3570.4332,
      "train_tokens_per_second": 425.87
    },
    {
      "epoch": 0.6299818356662037,
      "grad_norm": 2.246575355529785,
      "learning_rate": 1.3719786096256687e-05,
      "loss": 0.10312852263450623,
      "num_input_tokens_seen": 1521506,
      "step": 1474,
      "train_runtime": 3572.588,
      "train_tokens_per_second": 425.883
    },
    {
      "epoch": 0.6304092317555294,
      "grad_norm": 2.5485050678253174,
      "learning_rate": 1.3715508021390375e-05,
      "loss": 0.09694022685289383,
      "num_input_tokens_seen": 1522358,
      "step": 1475,
      "train_runtime": 3574.6459,
      "train_tokens_per_second": 425.877
    },
    {
      "epoch": 0.6308366278448552,
      "grad_norm": 2.582261085510254,
      "learning_rate": 1.3711229946524064e-05,
      "loss": 0.13314642012119293,
      "num_input_tokens_seen": 1523224,
      "step": 1476,
      "train_runtime": 3576.8789,
      "train_tokens_per_second": 425.853
    },
    {
      "epoch": 0.631264023934181,
      "grad_norm": 3.18725323677063,
      "learning_rate": 1.3706951871657754e-05,
      "loss": 0.14739792048931122,
      "num_input_tokens_seen": 1524170,
      "step": 1477,
      "train_runtime": 3578.9434,
      "train_tokens_per_second": 425.872
    },
    {
      "epoch": 0.6316914200235068,
      "grad_norm": 3.830138683319092,
      "learning_rate": 1.3702673796791445e-05,
      "loss": 0.10739552974700928,
      "num_input_tokens_seen": 1524894,
      "step": 1478,
      "train_runtime": 3580.9907,
      "train_tokens_per_second": 425.83
    },
    {
      "epoch": 0.6321188161128326,
      "grad_norm": 2.9134812355041504,
      "learning_rate": 1.3698395721925135e-05,
      "loss": 0.10831707715988159,
      "num_input_tokens_seen": 1525752,
      "step": 1479,
      "train_runtime": 3583.0466,
      "train_tokens_per_second": 425.825
    },
    {
      "epoch": 0.6325462122021583,
      "grad_norm": 2.2162744998931885,
      "learning_rate": 1.3694117647058824e-05,
      "loss": 0.10464231669902802,
      "num_input_tokens_seen": 1526760,
      "step": 1480,
      "train_runtime": 3585.1445,
      "train_tokens_per_second": 425.857
    },
    {
      "epoch": 0.6329736082914841,
      "grad_norm": 3.623624324798584,
      "learning_rate": 1.3689839572192513e-05,
      "loss": 0.13597632944583893,
      "num_input_tokens_seen": 1527564,
      "step": 1481,
      "train_runtime": 3587.1526,
      "train_tokens_per_second": 425.843
    },
    {
      "epoch": 0.6334010043808099,
      "grad_norm": 1.9764485359191895,
      "learning_rate": 1.3685561497326205e-05,
      "loss": 0.09632347524166107,
      "num_input_tokens_seen": 1528844,
      "step": 1482,
      "train_runtime": 3589.335,
      "train_tokens_per_second": 425.941
    },
    {
      "epoch": 0.6338284004701357,
      "grad_norm": 2.834679126739502,
      "learning_rate": 1.3681283422459894e-05,
      "loss": 0.14813737571239471,
      "num_input_tokens_seen": 1529794,
      "step": 1483,
      "train_runtime": 3591.4052,
      "train_tokens_per_second": 425.96
    },
    {
      "epoch": 0.6342557965594615,
      "grad_norm": 3.9278035163879395,
      "learning_rate": 1.3677005347593583e-05,
      "loss": 0.14398543536663055,
      "num_input_tokens_seen": 1530606,
      "step": 1484,
      "train_runtime": 3593.4261,
      "train_tokens_per_second": 425.946
    },
    {
      "epoch": 0.6346831926487873,
      "grad_norm": 2.2874929904937744,
      "learning_rate": 1.3672727272727273e-05,
      "loss": 0.11771330237388611,
      "num_input_tokens_seen": 1531714,
      "step": 1485,
      "train_runtime": 3595.5683,
      "train_tokens_per_second": 426.001
    },
    {
      "epoch": 0.635110588738113,
      "grad_norm": 3.821604013442993,
      "learning_rate": 1.3668449197860964e-05,
      "loss": 0.2711103856563568,
      "num_input_tokens_seen": 1532570,
      "step": 1486,
      "train_runtime": 3597.6115,
      "train_tokens_per_second": 425.997
    },
    {
      "epoch": 0.6355379848274388,
      "grad_norm": 3.055563449859619,
      "learning_rate": 1.3664171122994653e-05,
      "loss": 0.1839410960674286,
      "num_input_tokens_seen": 1533636,
      "step": 1487,
      "train_runtime": 3599.6752,
      "train_tokens_per_second": 426.048
    },
    {
      "epoch": 0.6359653809167646,
      "grad_norm": 3.095349073410034,
      "learning_rate": 1.3659893048128343e-05,
      "loss": 0.17613187432289124,
      "num_input_tokens_seen": 1534644,
      "step": 1488,
      "train_runtime": 3601.7773,
      "train_tokens_per_second": 426.08
    },
    {
      "epoch": 0.6363927770060904,
      "grad_norm": 2.953834295272827,
      "learning_rate": 1.3655614973262032e-05,
      "loss": 0.11626684665679932,
      "num_input_tokens_seen": 1535460,
      "step": 1489,
      "train_runtime": 3603.8341,
      "train_tokens_per_second": 426.063
    },
    {
      "epoch": 0.6368201730954162,
      "grad_norm": 2.8666796684265137,
      "learning_rate": 1.3651336898395723e-05,
      "loss": 0.11040861159563065,
      "num_input_tokens_seen": 1536424,
      "step": 1490,
      "train_runtime": 3605.9671,
      "train_tokens_per_second": 426.078
    },
    {
      "epoch": 0.637247569184742,
      "grad_norm": 2.5996861457824707,
      "learning_rate": 1.3647058823529413e-05,
      "loss": 0.18704327940940857,
      "num_input_tokens_seen": 1538244,
      "step": 1491,
      "train_runtime": 3608.2906,
      "train_tokens_per_second": 426.308
    },
    {
      "epoch": 0.6376749652740678,
      "grad_norm": 2.9655489921569824,
      "learning_rate": 1.3642780748663102e-05,
      "loss": 0.2014021873474121,
      "num_input_tokens_seen": 1539572,
      "step": 1492,
      "train_runtime": 3610.4583,
      "train_tokens_per_second": 426.42
    },
    {
      "epoch": 0.6381023613633935,
      "grad_norm": 4.0340142250061035,
      "learning_rate": 1.3638502673796792e-05,
      "loss": 0.18054339289665222,
      "num_input_tokens_seen": 1540412,
      "step": 1493,
      "train_runtime": 3612.4996,
      "train_tokens_per_second": 426.412
    },
    {
      "epoch": 0.6385297574527193,
      "grad_norm": 4.150583744049072,
      "learning_rate": 1.3634224598930483e-05,
      "loss": 0.11605513095855713,
      "num_input_tokens_seen": 1541012,
      "step": 1494,
      "train_runtime": 3614.4356,
      "train_tokens_per_second": 426.349
    },
    {
      "epoch": 0.6389571535420451,
      "grad_norm": 1.9334378242492676,
      "learning_rate": 1.3629946524064172e-05,
      "loss": 0.08048471808433533,
      "num_input_tokens_seen": 1541764,
      "step": 1495,
      "train_runtime": 3616.4707,
      "train_tokens_per_second": 426.317
    },
    {
      "epoch": 0.6393845496313709,
      "grad_norm": 2.171452045440674,
      "learning_rate": 1.3625668449197862e-05,
      "loss": 0.08667883276939392,
      "num_input_tokens_seen": 1543226,
      "step": 1496,
      "train_runtime": 3618.6214,
      "train_tokens_per_second": 426.468
    },
    {
      "epoch": 0.6398119457206967,
      "grad_norm": 3.2747793197631836,
      "learning_rate": 1.3621390374331551e-05,
      "loss": 0.13285577297210693,
      "num_input_tokens_seen": 1543984,
      "step": 1497,
      "train_runtime": 3620.6189,
      "train_tokens_per_second": 426.442
    },
    {
      "epoch": 0.6402393418100224,
      "grad_norm": 2.9296987056732178,
      "learning_rate": 1.3617112299465242e-05,
      "loss": 0.18483984470367432,
      "num_input_tokens_seen": 1545230,
      "step": 1498,
      "train_runtime": 3622.7598,
      "train_tokens_per_second": 426.534
    },
    {
      "epoch": 0.6406667378993482,
      "grad_norm": 2.39042592048645,
      "learning_rate": 1.3612834224598932e-05,
      "loss": 0.11871865391731262,
      "num_input_tokens_seen": 1546236,
      "step": 1499,
      "train_runtime": 3624.8432,
      "train_tokens_per_second": 426.566
    },
    {
      "epoch": 0.641094133988674,
      "grad_norm": 3.414876937866211,
      "learning_rate": 1.3608556149732621e-05,
      "loss": 0.11641217768192291,
      "num_input_tokens_seen": 1546874,
      "step": 1500,
      "train_runtime": 3626.8248,
      "train_tokens_per_second": 426.509
    },
    {
      "epoch": 0.6415215300779998,
      "grad_norm": 2.3719992637634277,
      "learning_rate": 1.360427807486631e-05,
      "loss": 0.10685586929321289,
      "num_input_tokens_seen": 1547780,
      "step": 1501,
      "train_runtime": 3635.5739,
      "train_tokens_per_second": 425.732
    },
    {
      "epoch": 0.6419489261673256,
      "grad_norm": 2.7630510330200195,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.11462070047855377,
      "num_input_tokens_seen": 1548564,
      "step": 1502,
      "train_runtime": 3637.6234,
      "train_tokens_per_second": 425.708
    },
    {
      "epoch": 0.6423763222566513,
      "grad_norm": 4.8656744956970215,
      "learning_rate": 1.3595721925133691e-05,
      "loss": 0.212483748793602,
      "num_input_tokens_seen": 1549224,
      "step": 1503,
      "train_runtime": 3639.7043,
      "train_tokens_per_second": 425.646
    },
    {
      "epoch": 0.6428037183459772,
      "grad_norm": 3.504530668258667,
      "learning_rate": 1.359144385026738e-05,
      "loss": 0.1726633906364441,
      "num_input_tokens_seen": 1550292,
      "step": 1504,
      "train_runtime": 3641.8245,
      "train_tokens_per_second": 425.691
    },
    {
      "epoch": 0.6432311144353029,
      "grad_norm": 3.975522518157959,
      "learning_rate": 1.358716577540107e-05,
      "loss": 0.11641143262386322,
      "num_input_tokens_seen": 1551226,
      "step": 1505,
      "train_runtime": 3643.8712,
      "train_tokens_per_second": 425.708
    },
    {
      "epoch": 0.6436585105246287,
      "grad_norm": 3.188652276992798,
      "learning_rate": 1.3582887700534761e-05,
      "loss": 0.1587495058774948,
      "num_input_tokens_seen": 1552042,
      "step": 1506,
      "train_runtime": 3645.967,
      "train_tokens_per_second": 425.687
    },
    {
      "epoch": 0.6440859066139545,
      "grad_norm": 2.3829288482666016,
      "learning_rate": 1.3578609625668451e-05,
      "loss": 0.10228777676820755,
      "num_input_tokens_seen": 1553214,
      "step": 1507,
      "train_runtime": 3648.0782,
      "train_tokens_per_second": 425.762
    },
    {
      "epoch": 0.6445133027032802,
      "grad_norm": 3.0020737648010254,
      "learning_rate": 1.357433155080214e-05,
      "loss": 0.24458348751068115,
      "num_input_tokens_seen": 1554276,
      "step": 1508,
      "train_runtime": 3650.1715,
      "train_tokens_per_second": 425.809
    },
    {
      "epoch": 0.6449406987926061,
      "grad_norm": 2.494694232940674,
      "learning_rate": 1.357005347593583e-05,
      "loss": 0.09424354135990143,
      "num_input_tokens_seen": 1554968,
      "step": 1509,
      "train_runtime": 3652.1636,
      "train_tokens_per_second": 425.766
    },
    {
      "epoch": 0.6453680948819318,
      "grad_norm": 2.695605754852295,
      "learning_rate": 1.3565775401069521e-05,
      "loss": 0.12960883975028992,
      "num_input_tokens_seen": 1556068,
      "step": 1510,
      "train_runtime": 3654.2804,
      "train_tokens_per_second": 425.821
    },
    {
      "epoch": 0.6457954909712577,
      "grad_norm": 3.6422171592712402,
      "learning_rate": 1.356149732620321e-05,
      "loss": 0.22197654843330383,
      "num_input_tokens_seen": 1556930,
      "step": 1511,
      "train_runtime": 3656.3265,
      "train_tokens_per_second": 425.818
    },
    {
      "epoch": 0.6462228870605834,
      "grad_norm": 2.917511463165283,
      "learning_rate": 1.35572192513369e-05,
      "loss": 0.11555804312229156,
      "num_input_tokens_seen": 1558272,
      "step": 1512,
      "train_runtime": 3658.5094,
      "train_tokens_per_second": 425.931
    },
    {
      "epoch": 0.6466502831499091,
      "grad_norm": 3.0762832164764404,
      "learning_rate": 1.355294117647059e-05,
      "loss": 0.09562921524047852,
      "num_input_tokens_seen": 1558930,
      "step": 1513,
      "train_runtime": 3660.4957,
      "train_tokens_per_second": 425.879
    },
    {
      "epoch": 0.647077679239235,
      "grad_norm": 2.5248517990112305,
      "learning_rate": 1.354866310160428e-05,
      "loss": 0.08613090217113495,
      "num_input_tokens_seen": 1559754,
      "step": 1514,
      "train_runtime": 3662.5137,
      "train_tokens_per_second": 425.87
    },
    {
      "epoch": 0.6475050753285607,
      "grad_norm": 3.964998960494995,
      "learning_rate": 1.354438502673797e-05,
      "loss": 0.204328715801239,
      "num_input_tokens_seen": 1560548,
      "step": 1515,
      "train_runtime": 3664.5412,
      "train_tokens_per_second": 425.851
    },
    {
      "epoch": 0.6479324714178866,
      "grad_norm": 2.447803258895874,
      "learning_rate": 1.354010695187166e-05,
      "loss": 0.14509670436382294,
      "num_input_tokens_seen": 1561562,
      "step": 1516,
      "train_runtime": 3666.5968,
      "train_tokens_per_second": 425.889
    },
    {
      "epoch": 0.6483598675072123,
      "grad_norm": 3.774406909942627,
      "learning_rate": 1.3535828877005349e-05,
      "loss": 0.15953212976455688,
      "num_input_tokens_seen": 1562250,
      "step": 1517,
      "train_runtime": 3668.572,
      "train_tokens_per_second": 425.847
    },
    {
      "epoch": 0.648787263596538,
      "grad_norm": 3.978740930557251,
      "learning_rate": 1.353155080213904e-05,
      "loss": 0.2058103382587433,
      "num_input_tokens_seen": 1563270,
      "step": 1518,
      "train_runtime": 3670.6249,
      "train_tokens_per_second": 425.887
    },
    {
      "epoch": 0.6492146596858639,
      "grad_norm": 3.0934062004089355,
      "learning_rate": 1.352727272727273e-05,
      "loss": 0.14622631669044495,
      "num_input_tokens_seen": 1564114,
      "step": 1519,
      "train_runtime": 3672.7534,
      "train_tokens_per_second": 425.87
    },
    {
      "epoch": 0.6496420557751896,
      "grad_norm": 3.5222365856170654,
      "learning_rate": 1.3522994652406419e-05,
      "loss": 0.21718433499336243,
      "num_input_tokens_seen": 1565672,
      "step": 1520,
      "train_runtime": 3674.9575,
      "train_tokens_per_second": 426.038
    },
    {
      "epoch": 0.6500694518645155,
      "grad_norm": 3.4182534217834473,
      "learning_rate": 1.3518716577540108e-05,
      "loss": 0.13513398170471191,
      "num_input_tokens_seen": 1566850,
      "step": 1521,
      "train_runtime": 3677.0353,
      "train_tokens_per_second": 426.118
    },
    {
      "epoch": 0.6504968479538412,
      "grad_norm": 5.0255842208862305,
      "learning_rate": 1.35144385026738e-05,
      "loss": 0.21971845626831055,
      "num_input_tokens_seen": 1567950,
      "step": 1522,
      "train_runtime": 3679.1193,
      "train_tokens_per_second": 426.175
    },
    {
      "epoch": 0.650924244043167,
      "grad_norm": 3.1422319412231445,
      "learning_rate": 1.3510160427807489e-05,
      "loss": 0.13466057181358337,
      "num_input_tokens_seen": 1568690,
      "step": 1523,
      "train_runtime": 3681.1443,
      "train_tokens_per_second": 426.142
    },
    {
      "epoch": 0.6513516401324928,
      "grad_norm": 3.234555959701538,
      "learning_rate": 1.3505882352941178e-05,
      "loss": 0.2144131064414978,
      "num_input_tokens_seen": 1569822,
      "step": 1524,
      "train_runtime": 3683.2819,
      "train_tokens_per_second": 426.202
    },
    {
      "epoch": 0.6517790362218185,
      "grad_norm": 3.9785099029541016,
      "learning_rate": 1.3501604278074868e-05,
      "loss": 0.15267924964427948,
      "num_input_tokens_seen": 1570408,
      "step": 1525,
      "train_runtime": 3685.3198,
      "train_tokens_per_second": 426.125
    },
    {
      "epoch": 0.6522064323111444,
      "grad_norm": 3.465639591217041,
      "learning_rate": 1.3497326203208559e-05,
      "loss": 0.19032518565654755,
      "num_input_tokens_seen": 1571396,
      "step": 1526,
      "train_runtime": 3687.3866,
      "train_tokens_per_second": 426.154
    },
    {
      "epoch": 0.6526338284004701,
      "grad_norm": 2.8069920539855957,
      "learning_rate": 1.3493048128342248e-05,
      "loss": 0.14490953087806702,
      "num_input_tokens_seen": 1572208,
      "step": 1527,
      "train_runtime": 3689.4538,
      "train_tokens_per_second": 426.136
    },
    {
      "epoch": 0.6530612244897959,
      "grad_norm": 2.129127264022827,
      "learning_rate": 1.3488770053475938e-05,
      "loss": 0.11624617874622345,
      "num_input_tokens_seen": 1573508,
      "step": 1528,
      "train_runtime": 3691.6414,
      "train_tokens_per_second": 426.235
    },
    {
      "epoch": 0.6534886205791217,
      "grad_norm": 1.8786301612854004,
      "learning_rate": 1.3484491978609625e-05,
      "loss": 0.10244782269001007,
      "num_input_tokens_seen": 1574602,
      "step": 1529,
      "train_runtime": 3693.8163,
      "train_tokens_per_second": 426.281
    },
    {
      "epoch": 0.6539160166684475,
      "grad_norm": 4.182097911834717,
      "learning_rate": 1.3480213903743315e-05,
      "loss": 0.17187145352363586,
      "num_input_tokens_seen": 1575186,
      "step": 1530,
      "train_runtime": 3695.8232,
      "train_tokens_per_second": 426.207
    },
    {
      "epoch": 0.6543434127577733,
      "grad_norm": 1.8183155059814453,
      "learning_rate": 1.3475935828877008e-05,
      "loss": 0.0856410339474678,
      "num_input_tokens_seen": 1576610,
      "step": 1531,
      "train_runtime": 3704.6141,
      "train_tokens_per_second": 425.58
    },
    {
      "epoch": 0.654770808847099,
      "grad_norm": 2.8884012699127197,
      "learning_rate": 1.3471657754010697e-05,
      "loss": 0.14600466191768646,
      "num_input_tokens_seen": 1577596,
      "step": 1532,
      "train_runtime": 3706.6629,
      "train_tokens_per_second": 425.611
    },
    {
      "epoch": 0.6551982049364248,
      "grad_norm": 3.003814458847046,
      "learning_rate": 1.3467379679144385e-05,
      "loss": 0.1828392744064331,
      "num_input_tokens_seen": 1578504,
      "step": 1533,
      "train_runtime": 3708.9029,
      "train_tokens_per_second": 425.599
    },
    {
      "epoch": 0.6556256010257506,
      "grad_norm": 2.1931958198547363,
      "learning_rate": 1.3463101604278074e-05,
      "loss": 0.13615339994430542,
      "num_input_tokens_seen": 1579710,
      "step": 1534,
      "train_runtime": 3711.0566,
      "train_tokens_per_second": 425.677
    },
    {
      "epoch": 0.6560529971150764,
      "grad_norm": 2.2456908226013184,
      "learning_rate": 1.3458823529411767e-05,
      "loss": 0.08627387136220932,
      "num_input_tokens_seen": 1580692,
      "step": 1535,
      "train_runtime": 3713.1495,
      "train_tokens_per_second": 425.701
    },
    {
      "epoch": 0.6564803932044022,
      "grad_norm": 3.724116563796997,
      "learning_rate": 1.3454545454545455e-05,
      "loss": 0.15534056723117828,
      "num_input_tokens_seen": 1581530,
      "step": 1536,
      "train_runtime": 3715.1728,
      "train_tokens_per_second": 425.695
    },
    {
      "epoch": 0.656907789293728,
      "grad_norm": 2.5453593730926514,
      "learning_rate": 1.3450267379679144e-05,
      "loss": 0.12748880684375763,
      "num_input_tokens_seen": 1582596,
      "step": 1537,
      "train_runtime": 3717.2833,
      "train_tokens_per_second": 425.74
    },
    {
      "epoch": 0.6573351853830538,
      "grad_norm": 5.269556999206543,
      "learning_rate": 1.3445989304812834e-05,
      "loss": 0.14962272346019745,
      "num_input_tokens_seen": 1583350,
      "step": 1538,
      "train_runtime": 3719.3436,
      "train_tokens_per_second": 425.707
    },
    {
      "epoch": 0.6577625814723795,
      "grad_norm": 4.323054313659668,
      "learning_rate": 1.3441711229946525e-05,
      "loss": 0.21885833144187927,
      "num_input_tokens_seen": 1584178,
      "step": 1539,
      "train_runtime": 3721.4368,
      "train_tokens_per_second": 425.69
    },
    {
      "epoch": 0.6581899775617053,
      "grad_norm": 2.9094793796539307,
      "learning_rate": 1.3437433155080214e-05,
      "loss": 0.13586312532424927,
      "num_input_tokens_seen": 1585064,
      "step": 1540,
      "train_runtime": 3723.4866,
      "train_tokens_per_second": 425.693
    },
    {
      "epoch": 0.6586173736510311,
      "grad_norm": 3.597817897796631,
      "learning_rate": 1.3433155080213904e-05,
      "loss": 0.14414404332637787,
      "num_input_tokens_seen": 1585794,
      "step": 1541,
      "train_runtime": 3725.5101,
      "train_tokens_per_second": 425.658
    },
    {
      "epoch": 0.6590447697403569,
      "grad_norm": 4.029960632324219,
      "learning_rate": 1.3428877005347593e-05,
      "loss": 0.2180078774690628,
      "num_input_tokens_seen": 1587408,
      "step": 1542,
      "train_runtime": 3727.7169,
      "train_tokens_per_second": 425.839
    },
    {
      "epoch": 0.6594721658296827,
      "grad_norm": 4.100533962249756,
      "learning_rate": 1.3424598930481284e-05,
      "loss": 0.17998196184635162,
      "num_input_tokens_seen": 1588142,
      "step": 1543,
      "train_runtime": 3729.7397,
      "train_tokens_per_second": 425.805
    },
    {
      "epoch": 0.6598995619190084,
      "grad_norm": 4.7648515701293945,
      "learning_rate": 1.3420320855614974e-05,
      "loss": 0.23594880104064941,
      "num_input_tokens_seen": 1588860,
      "step": 1544,
      "train_runtime": 3731.7605,
      "train_tokens_per_second": 425.767
    },
    {
      "epoch": 0.6603269580083342,
      "grad_norm": 4.623931884765625,
      "learning_rate": 1.3416042780748663e-05,
      "loss": 0.14606264233589172,
      "num_input_tokens_seen": 1589590,
      "step": 1545,
      "train_runtime": 3733.7727,
      "train_tokens_per_second": 425.733
    },
    {
      "epoch": 0.66075435409766,
      "grad_norm": 2.120661973953247,
      "learning_rate": 1.3411764705882353e-05,
      "loss": 0.07236392796039581,
      "num_input_tokens_seen": 1590230,
      "step": 1546,
      "train_runtime": 3735.7574,
      "train_tokens_per_second": 425.678
    },
    {
      "epoch": 0.6611817501869858,
      "grad_norm": 2.899981737136841,
      "learning_rate": 1.3407486631016044e-05,
      "loss": 0.09665043652057648,
      "num_input_tokens_seen": 1591156,
      "step": 1547,
      "train_runtime": 3737.7945,
      "train_tokens_per_second": 425.694
    },
    {
      "epoch": 0.6616091462763116,
      "grad_norm": 2.396935224533081,
      "learning_rate": 1.3403208556149733e-05,
      "loss": 0.07971307635307312,
      "num_input_tokens_seen": 1592442,
      "step": 1548,
      "train_runtime": 3740.0039,
      "train_tokens_per_second": 425.786
    },
    {
      "epoch": 0.6620365423656374,
      "grad_norm": 5.512351036071777,
      "learning_rate": 1.3398930481283423e-05,
      "loss": 0.08992433547973633,
      "num_input_tokens_seen": 1593202,
      "step": 1549,
      "train_runtime": 3742.0402,
      "train_tokens_per_second": 425.758
    },
    {
      "epoch": 0.6624639384549631,
      "grad_norm": 2.763360023498535,
      "learning_rate": 1.3394652406417112e-05,
      "loss": 0.17814138531684875,
      "num_input_tokens_seen": 1594400,
      "step": 1550,
      "train_runtime": 3744.2789,
      "train_tokens_per_second": 425.823
    },
    {
      "epoch": 0.662891334544289,
      "grad_norm": 3.826706886291504,
      "learning_rate": 1.3390374331550803e-05,
      "loss": 0.09510751068592072,
      "num_input_tokens_seen": 1595116,
      "step": 1551,
      "train_runtime": 3746.299,
      "train_tokens_per_second": 425.784
    },
    {
      "epoch": 0.6633187306336147,
      "grad_norm": 2.5190634727478027,
      "learning_rate": 1.3386096256684493e-05,
      "loss": 0.14573213458061218,
      "num_input_tokens_seen": 1596084,
      "step": 1552,
      "train_runtime": 3748.3903,
      "train_tokens_per_second": 425.805
    },
    {
      "epoch": 0.6637461267229405,
      "grad_norm": 2.3098104000091553,
      "learning_rate": 1.3381818181818182e-05,
      "loss": 0.1424594223499298,
      "num_input_tokens_seen": 1597872,
      "step": 1553,
      "train_runtime": 3750.6652,
      "train_tokens_per_second": 426.024
    },
    {
      "epoch": 0.6641735228122663,
      "grad_norm": 2.7591264247894287,
      "learning_rate": 1.3377540106951872e-05,
      "loss": 0.11576004326343536,
      "num_input_tokens_seen": 1598962,
      "step": 1554,
      "train_runtime": 3752.7653,
      "train_tokens_per_second": 426.076
    },
    {
      "epoch": 0.664600918901592,
      "grad_norm": 2.9065213203430176,
      "learning_rate": 1.3373262032085563e-05,
      "loss": 0.1090204119682312,
      "num_input_tokens_seen": 1599828,
      "step": 1555,
      "train_runtime": 3754.8307,
      "train_tokens_per_second": 426.072
    },
    {
      "epoch": 0.6650283149909179,
      "grad_norm": 2.754087209701538,
      "learning_rate": 1.3368983957219252e-05,
      "loss": 0.17998015880584717,
      "num_input_tokens_seen": 1600918,
      "step": 1556,
      "train_runtime": 3756.9799,
      "train_tokens_per_second": 426.118
    },
    {
      "epoch": 0.6654557110802436,
      "grad_norm": 3.2658700942993164,
      "learning_rate": 1.3364705882352942e-05,
      "loss": 0.19984865188598633,
      "num_input_tokens_seen": 1601938,
      "step": 1557,
      "train_runtime": 3759.1082,
      "train_tokens_per_second": 426.148
    },
    {
      "epoch": 0.6658831071695694,
      "grad_norm": 3.116774082183838,
      "learning_rate": 1.3360427807486631e-05,
      "loss": 0.13655301928520203,
      "num_input_tokens_seen": 1602692,
      "step": 1558,
      "train_runtime": 3761.0848,
      "train_tokens_per_second": 426.125
    },
    {
      "epoch": 0.6663105032588952,
      "grad_norm": 3.1976685523986816,
      "learning_rate": 1.3356149732620322e-05,
      "loss": 0.10624785721302032,
      "num_input_tokens_seen": 1603628,
      "step": 1559,
      "train_runtime": 3763.1232,
      "train_tokens_per_second": 426.143
    },
    {
      "epoch": 0.6667378993482209,
      "grad_norm": 2.3742599487304688,
      "learning_rate": 1.3351871657754012e-05,
      "loss": 0.1465083211660385,
      "num_input_tokens_seen": 1604734,
      "step": 1560,
      "train_runtime": 3765.2301,
      "train_tokens_per_second": 426.198
    },
    {
      "epoch": 0.6671652954375468,
      "grad_norm": 3.4235386848449707,
      "learning_rate": 1.3347593582887701e-05,
      "loss": 0.18542777001857758,
      "num_input_tokens_seen": 1605706,
      "step": 1561,
      "train_runtime": 3773.9514,
      "train_tokens_per_second": 425.471
    },
    {
      "epoch": 0.6675926915268725,
      "grad_norm": 2.669586181640625,
      "learning_rate": 1.334331550802139e-05,
      "loss": 0.1491749882698059,
      "num_input_tokens_seen": 1606732,
      "step": 1562,
      "train_runtime": 3776.0445,
      "train_tokens_per_second": 425.507
    },
    {
      "epoch": 0.6680200876161984,
      "grad_norm": 2.283149242401123,
      "learning_rate": 1.3339037433155082e-05,
      "loss": 0.11235065758228302,
      "num_input_tokens_seen": 1607630,
      "step": 1563,
      "train_runtime": 3778.1366,
      "train_tokens_per_second": 425.509
    },
    {
      "epoch": 0.6684474837055241,
      "grad_norm": 2.0375988483428955,
      "learning_rate": 1.3334759358288771e-05,
      "loss": 0.09467926621437073,
      "num_input_tokens_seen": 1608904,
      "step": 1564,
      "train_runtime": 3780.2712,
      "train_tokens_per_second": 425.605
    },
    {
      "epoch": 0.6688748797948498,
      "grad_norm": 3.068815231323242,
      "learning_rate": 1.333048128342246e-05,
      "loss": 0.16510456800460815,
      "num_input_tokens_seen": 1609904,
      "step": 1565,
      "train_runtime": 3782.349,
      "train_tokens_per_second": 425.636
    },
    {
      "epoch": 0.6693022758841757,
      "grad_norm": 2.582709789276123,
      "learning_rate": 1.332620320855615e-05,
      "loss": 0.14242790639400482,
      "num_input_tokens_seen": 1610686,
      "step": 1566,
      "train_runtime": 3784.3429,
      "train_tokens_per_second": 425.618
    },
    {
      "epoch": 0.6697296719735014,
      "grad_norm": 2.302830457687378,
      "learning_rate": 1.3321925133689841e-05,
      "loss": 0.08061867207288742,
      "num_input_tokens_seen": 1611608,
      "step": 1567,
      "train_runtime": 3786.3874,
      "train_tokens_per_second": 425.632
    },
    {
      "epoch": 0.6701570680628273,
      "grad_norm": 2.428049087524414,
      "learning_rate": 1.331764705882353e-05,
      "loss": 0.08166418969631195,
      "num_input_tokens_seen": 1612824,
      "step": 1568,
      "train_runtime": 3788.5362,
      "train_tokens_per_second": 425.712
    },
    {
      "epoch": 0.670584464152153,
      "grad_norm": 3.02074933052063,
      "learning_rate": 1.331336898395722e-05,
      "loss": 0.13414841890335083,
      "num_input_tokens_seen": 1613618,
      "step": 1569,
      "train_runtime": 3790.6202,
      "train_tokens_per_second": 425.687
    },
    {
      "epoch": 0.6710118602414787,
      "grad_norm": 2.169861316680908,
      "learning_rate": 1.330909090909091e-05,
      "loss": 0.12823854386806488,
      "num_input_tokens_seen": 1614690,
      "step": 1570,
      "train_runtime": 3792.7021,
      "train_tokens_per_second": 425.736
    },
    {
      "epoch": 0.6714392563308046,
      "grad_norm": 2.4305543899536133,
      "learning_rate": 1.33048128342246e-05,
      "loss": 0.11599336564540863,
      "num_input_tokens_seen": 1615700,
      "step": 1571,
      "train_runtime": 3794.7872,
      "train_tokens_per_second": 425.768
    },
    {
      "epoch": 0.6718666524201303,
      "grad_norm": 2.0328714847564697,
      "learning_rate": 1.330053475935829e-05,
      "loss": 0.16121914982795715,
      "num_input_tokens_seen": 1617158,
      "step": 1572,
      "train_runtime": 3796.9905,
      "train_tokens_per_second": 425.905
    },
    {
      "epoch": 0.6722940485094562,
      "grad_norm": 3.055515766143799,
      "learning_rate": 1.329625668449198e-05,
      "loss": 0.12334509193897247,
      "num_input_tokens_seen": 1617860,
      "step": 1573,
      "train_runtime": 3799.0205,
      "train_tokens_per_second": 425.862
    },
    {
      "epoch": 0.6727214445987819,
      "grad_norm": 2.4867000579833984,
      "learning_rate": 1.3291978609625669e-05,
      "loss": 0.11558125168085098,
      "num_input_tokens_seen": 1618652,
      "step": 1574,
      "train_runtime": 3801.0557,
      "train_tokens_per_second": 425.843
    },
    {
      "epoch": 0.6731488406881077,
      "grad_norm": 2.851879119873047,
      "learning_rate": 1.328770053475936e-05,
      "loss": 0.14694936573505402,
      "num_input_tokens_seen": 1619922,
      "step": 1575,
      "train_runtime": 3803.224,
      "train_tokens_per_second": 425.934
    },
    {
      "epoch": 0.6735762367774335,
      "grad_norm": 3.393662691116333,
      "learning_rate": 1.328342245989305e-05,
      "loss": 0.17909544706344604,
      "num_input_tokens_seen": 1620640,
      "step": 1576,
      "train_runtime": 3805.2492,
      "train_tokens_per_second": 425.896
    },
    {
      "epoch": 0.6740036328667592,
      "grad_norm": 2.5993282794952393,
      "learning_rate": 1.3279144385026739e-05,
      "loss": 0.1457187980413437,
      "num_input_tokens_seen": 1621670,
      "step": 1577,
      "train_runtime": 3807.4052,
      "train_tokens_per_second": 425.925
    },
    {
      "epoch": 0.6744310289560851,
      "grad_norm": 2.9386048316955566,
      "learning_rate": 1.3274866310160429e-05,
      "loss": 0.11855722963809967,
      "num_input_tokens_seen": 1622578,
      "step": 1578,
      "train_runtime": 3809.4534,
      "train_tokens_per_second": 425.935
    },
    {
      "epoch": 0.6748584250454108,
      "grad_norm": 3.5167593955993652,
      "learning_rate": 1.327058823529412e-05,
      "loss": 0.20887939631938934,
      "num_input_tokens_seen": 1623332,
      "step": 1579,
      "train_runtime": 3811.4953,
      "train_tokens_per_second": 425.904
    },
    {
      "epoch": 0.6752858211347366,
      "grad_norm": 3.0010759830474854,
      "learning_rate": 1.326631016042781e-05,
      "loss": 0.18665343523025513,
      "num_input_tokens_seen": 1624418,
      "step": 1580,
      "train_runtime": 3813.6063,
      "train_tokens_per_second": 425.953
    },
    {
      "epoch": 0.6757132172240624,
      "grad_norm": 3.057358980178833,
      "learning_rate": 1.3262032085561499e-05,
      "loss": 0.10085485875606537,
      "num_input_tokens_seen": 1625200,
      "step": 1581,
      "train_runtime": 3815.6262,
      "train_tokens_per_second": 425.933
    },
    {
      "epoch": 0.6761406133133882,
      "grad_norm": 2.774298667907715,
      "learning_rate": 1.3257754010695188e-05,
      "loss": 0.14732888340950012,
      "num_input_tokens_seen": 1626426,
      "step": 1582,
      "train_runtime": 3817.7774,
      "train_tokens_per_second": 426.014
    },
    {
      "epoch": 0.676568009402714,
      "grad_norm": 2.9863998889923096,
      "learning_rate": 1.325347593582888e-05,
      "loss": 0.22564369440078735,
      "num_input_tokens_seen": 1627304,
      "step": 1583,
      "train_runtime": 3819.8628,
      "train_tokens_per_second": 426.011
    },
    {
      "epoch": 0.6769954054920397,
      "grad_norm": 3.1854777336120605,
      "learning_rate": 1.3249197860962569e-05,
      "loss": 0.15117281675338745,
      "num_input_tokens_seen": 1628384,
      "step": 1584,
      "train_runtime": 3821.9792,
      "train_tokens_per_second": 426.058
    },
    {
      "epoch": 0.6774228015813656,
      "grad_norm": 2.56819748878479,
      "learning_rate": 1.3244919786096258e-05,
      "loss": 0.15319907665252686,
      "num_input_tokens_seen": 1629448,
      "step": 1585,
      "train_runtime": 3824.0864,
      "train_tokens_per_second": 426.101
    },
    {
      "epoch": 0.6778501976706913,
      "grad_norm": 3.392935276031494,
      "learning_rate": 1.3240641711229948e-05,
      "loss": 0.15829937160015106,
      "num_input_tokens_seen": 1630412,
      "step": 1586,
      "train_runtime": 3826.0936,
      "train_tokens_per_second": 426.13
    },
    {
      "epoch": 0.6782775937600171,
      "grad_norm": 2.1995556354522705,
      "learning_rate": 1.3236363636363639e-05,
      "loss": 0.12209456413984299,
      "num_input_tokens_seen": 1631754,
      "step": 1587,
      "train_runtime": 3828.2619,
      "train_tokens_per_second": 426.239
    },
    {
      "epoch": 0.6787049898493429,
      "grad_norm": 2.5163869857788086,
      "learning_rate": 1.3232085561497328e-05,
      "loss": 0.09759660065174103,
      "num_input_tokens_seen": 1633172,
      "step": 1588,
      "train_runtime": 3830.4421,
      "train_tokens_per_second": 426.366
    },
    {
      "epoch": 0.6791323859386686,
      "grad_norm": 2.2898318767547607,
      "learning_rate": 1.3227807486631018e-05,
      "loss": 0.11964927613735199,
      "num_input_tokens_seen": 1634354,
      "step": 1589,
      "train_runtime": 3832.6316,
      "train_tokens_per_second": 426.431
    },
    {
      "epoch": 0.6795597820279945,
      "grad_norm": 4.861701488494873,
      "learning_rate": 1.3223529411764705e-05,
      "loss": 0.1981499046087265,
      "num_input_tokens_seen": 1635210,
      "step": 1590,
      "train_runtime": 3834.6696,
      "train_tokens_per_second": 426.428
    },
    {
      "epoch": 0.6799871781173202,
      "grad_norm": 2.4189250469207764,
      "learning_rate": 1.3219251336898398e-05,
      "loss": 0.16792207956314087,
      "num_input_tokens_seen": 1636860,
      "step": 1591,
      "train_runtime": 3843.5762,
      "train_tokens_per_second": 425.869
    },
    {
      "epoch": 0.680414574206646,
      "grad_norm": 3.1255605220794678,
      "learning_rate": 1.3214973262032088e-05,
      "loss": 0.14919006824493408,
      "num_input_tokens_seen": 1637730,
      "step": 1592,
      "train_runtime": 3845.7644,
      "train_tokens_per_second": 425.853
    },
    {
      "epoch": 0.6808419702959718,
      "grad_norm": 2.6487908363342285,
      "learning_rate": 1.3210695187165777e-05,
      "loss": 0.19162176549434662,
      "num_input_tokens_seen": 1639292,
      "step": 1593,
      "train_runtime": 3847.9444,
      "train_tokens_per_second": 426.018
    },
    {
      "epoch": 0.6812693663852976,
      "grad_norm": 2.6720197200775146,
      "learning_rate": 1.3206417112299465e-05,
      "loss": 0.11653536558151245,
      "num_input_tokens_seen": 1640188,
      "step": 1594,
      "train_runtime": 3850.0154,
      "train_tokens_per_second": 426.021
    },
    {
      "epoch": 0.6816967624746234,
      "grad_norm": 4.183128833770752,
      "learning_rate": 1.3202139037433158e-05,
      "loss": 0.20200294256210327,
      "num_input_tokens_seen": 1640806,
      "step": 1595,
      "train_runtime": 3852.0142,
      "train_tokens_per_second": 425.961
    },
    {
      "epoch": 0.6821241585639491,
      "grad_norm": 3.8790345191955566,
      "learning_rate": 1.3197860962566847e-05,
      "loss": 0.14347238838672638,
      "num_input_tokens_seen": 1642176,
      "step": 1596,
      "train_runtime": 3854.2747,
      "train_tokens_per_second": 426.066
    },
    {
      "epoch": 0.6825515546532749,
      "grad_norm": 2.6697440147399902,
      "learning_rate": 1.3193582887700535e-05,
      "loss": 0.20666377246379852,
      "num_input_tokens_seen": 1643650,
      "step": 1597,
      "train_runtime": 3856.4888,
      "train_tokens_per_second": 426.204
    },
    {
      "epoch": 0.6829789507426007,
      "grad_norm": 3.147174835205078,
      "learning_rate": 1.3189304812834224e-05,
      "loss": 0.10880240052938461,
      "num_input_tokens_seen": 1644390,
      "step": 1598,
      "train_runtime": 3858.5296,
      "train_tokens_per_second": 426.17
    },
    {
      "epoch": 0.6834063468319265,
      "grad_norm": 3.357785224914551,
      "learning_rate": 1.3185026737967917e-05,
      "loss": 0.19138175249099731,
      "num_input_tokens_seen": 1645308,
      "step": 1599,
      "train_runtime": 3860.612,
      "train_tokens_per_second": 426.178
    },
    {
      "epoch": 0.6838337429212523,
      "grad_norm": 3.105679988861084,
      "learning_rate": 1.3180748663101607e-05,
      "loss": 0.18142981827259064,
      "num_input_tokens_seen": 1646310,
      "step": 1600,
      "train_runtime": 3862.651,
      "train_tokens_per_second": 426.212
    },
    {
      "epoch": 0.6842611390105781,
      "grad_norm": 1.9695712327957153,
      "learning_rate": 1.3176470588235294e-05,
      "loss": 0.0836038589477539,
      "num_input_tokens_seen": 1647374,
      "step": 1601,
      "train_runtime": 3864.685,
      "train_tokens_per_second": 426.263
    },
    {
      "epoch": 0.6846885350999038,
      "grad_norm": 3.3684911727905273,
      "learning_rate": 1.3172192513368984e-05,
      "loss": 0.18620504438877106,
      "num_input_tokens_seen": 1648536,
      "step": 1602,
      "train_runtime": 3866.7931,
      "train_tokens_per_second": 426.332
    },
    {
      "epoch": 0.6851159311892296,
      "grad_norm": 3.576066493988037,
      "learning_rate": 1.3167914438502677e-05,
      "loss": 0.18863776326179504,
      "num_input_tokens_seen": 1649318,
      "step": 1603,
      "train_runtime": 3868.8053,
      "train_tokens_per_second": 426.312
    },
    {
      "epoch": 0.6855433272785554,
      "grad_norm": 3.392570972442627,
      "learning_rate": 1.3163636363636364e-05,
      "loss": 0.13371005654335022,
      "num_input_tokens_seen": 1650252,
      "step": 1604,
      "train_runtime": 3870.8864,
      "train_tokens_per_second": 426.324
    },
    {
      "epoch": 0.6859707233678812,
      "grad_norm": 1.8862580060958862,
      "learning_rate": 1.3159358288770054e-05,
      "loss": 0.07925024628639221,
      "num_input_tokens_seen": 1652300,
      "step": 1605,
      "train_runtime": 3873.2108,
      "train_tokens_per_second": 426.597
    },
    {
      "epoch": 0.686398119457207,
      "grad_norm": 3.5801241397857666,
      "learning_rate": 1.3155080213903743e-05,
      "loss": 0.16683319211006165,
      "num_input_tokens_seen": 1653100,
      "step": 1606,
      "train_runtime": 3875.2585,
      "train_tokens_per_second": 426.578
    },
    {
      "epoch": 0.6868255155465327,
      "grad_norm": 2.4567997455596924,
      "learning_rate": 1.3150802139037434e-05,
      "loss": 0.12541936337947845,
      "num_input_tokens_seen": 1654178,
      "step": 1607,
      "train_runtime": 3877.3414,
      "train_tokens_per_second": 426.627
    },
    {
      "epoch": 0.6872529116358586,
      "grad_norm": 2.737053394317627,
      "learning_rate": 1.3146524064171124e-05,
      "loss": 0.12982957065105438,
      "num_input_tokens_seen": 1655230,
      "step": 1608,
      "train_runtime": 3879.4203,
      "train_tokens_per_second": 426.669
    },
    {
      "epoch": 0.6876803077251843,
      "grad_norm": 3.2906999588012695,
      "learning_rate": 1.3142245989304813e-05,
      "loss": 0.24963447451591492,
      "num_input_tokens_seen": 1657018,
      "step": 1609,
      "train_runtime": 3881.7743,
      "train_tokens_per_second": 426.871
    },
    {
      "epoch": 0.6881077038145101,
      "grad_norm": 3.597764253616333,
      "learning_rate": 1.3137967914438503e-05,
      "loss": 0.2881978154182434,
      "num_input_tokens_seen": 1657826,
      "step": 1610,
      "train_runtime": 3883.7548,
      "train_tokens_per_second": 426.862
    },
    {
      "epoch": 0.6885350999038359,
      "grad_norm": 2.057811737060547,
      "learning_rate": 1.3133689839572194e-05,
      "loss": 0.08862215280532837,
      "num_input_tokens_seen": 1659216,
      "step": 1611,
      "train_runtime": 3885.9402,
      "train_tokens_per_second": 426.979
    },
    {
      "epoch": 0.6889624959931616,
      "grad_norm": 2.7082576751708984,
      "learning_rate": 1.3129411764705883e-05,
      "loss": 0.09579111635684967,
      "num_input_tokens_seen": 1660136,
      "step": 1612,
      "train_runtime": 3887.9689,
      "train_tokens_per_second": 426.993
    },
    {
      "epoch": 0.6893898920824875,
      "grad_norm": 2.8410701751708984,
      "learning_rate": 1.3125133689839573e-05,
      "loss": 0.18607118725776672,
      "num_input_tokens_seen": 1661920,
      "step": 1613,
      "train_runtime": 3890.1898,
      "train_tokens_per_second": 427.208
    },
    {
      "epoch": 0.6898172881718132,
      "grad_norm": 2.691599130630493,
      "learning_rate": 1.3120855614973262e-05,
      "loss": 0.1239597424864769,
      "num_input_tokens_seen": 1662820,
      "step": 1614,
      "train_runtime": 3892.2341,
      "train_tokens_per_second": 427.215
    },
    {
      "epoch": 0.690244684261139,
      "grad_norm": 2.18497633934021,
      "learning_rate": 1.3116577540106953e-05,
      "loss": 0.11108903586864471,
      "num_input_tokens_seen": 1664108,
      "step": 1615,
      "train_runtime": 3894.3359,
      "train_tokens_per_second": 427.315
    },
    {
      "epoch": 0.6906720803504648,
      "grad_norm": 2.9395556449890137,
      "learning_rate": 1.3112299465240643e-05,
      "loss": 0.18891631066799164,
      "num_input_tokens_seen": 1665014,
      "step": 1616,
      "train_runtime": 3896.4118,
      "train_tokens_per_second": 427.32
    },
    {
      "epoch": 0.6910994764397905,
      "grad_norm": 2.8265609741210938,
      "learning_rate": 1.3108021390374332e-05,
      "loss": 0.0613570436835289,
      "num_input_tokens_seen": 1666400,
      "step": 1617,
      "train_runtime": 3898.5977,
      "train_tokens_per_second": 427.436
    },
    {
      "epoch": 0.6915268725291164,
      "grad_norm": 2.4524476528167725,
      "learning_rate": 1.3103743315508022e-05,
      "loss": 0.1273629367351532,
      "num_input_tokens_seen": 1667142,
      "step": 1618,
      "train_runtime": 3900.6362,
      "train_tokens_per_second": 427.403
    },
    {
      "epoch": 0.6919542686184421,
      "grad_norm": 2.3937675952911377,
      "learning_rate": 1.3099465240641713e-05,
      "loss": 0.09456370770931244,
      "num_input_tokens_seen": 1668150,
      "step": 1619,
      "train_runtime": 3902.7099,
      "train_tokens_per_second": 427.434
    },
    {
      "epoch": 0.692381664707768,
      "grad_norm": 2.7047691345214844,
      "learning_rate": 1.3095187165775402e-05,
      "loss": 0.13802814483642578,
      "num_input_tokens_seen": 1669652,
      "step": 1620,
      "train_runtime": 3904.9061,
      "train_tokens_per_second": 427.578
    },
    {
      "epoch": 0.6928090607970937,
      "grad_norm": 3.2640371322631836,
      "learning_rate": 1.3090909090909092e-05,
      "loss": 0.11681882292032242,
      "num_input_tokens_seen": 1670674,
      "step": 1621,
      "train_runtime": 3913.2306,
      "train_tokens_per_second": 426.93
    },
    {
      "epoch": 0.6932364568864194,
      "grad_norm": 3.0372519493103027,
      "learning_rate": 1.3086631016042781e-05,
      "loss": 0.14249123632907867,
      "num_input_tokens_seen": 1671534,
      "step": 1622,
      "train_runtime": 3915.2407,
      "train_tokens_per_second": 426.93
    },
    {
      "epoch": 0.6936638529757453,
      "grad_norm": 4.135215759277344,
      "learning_rate": 1.3082352941176472e-05,
      "loss": 0.2676665782928467,
      "num_input_tokens_seen": 1672412,
      "step": 1623,
      "train_runtime": 3917.3077,
      "train_tokens_per_second": 426.929
    },
    {
      "epoch": 0.694091249065071,
      "grad_norm": 3.9823038578033447,
      "learning_rate": 1.3078074866310162e-05,
      "loss": 0.4981217682361603,
      "num_input_tokens_seen": 1673666,
      "step": 1624,
      "train_runtime": 3919.4495,
      "train_tokens_per_second": 427.016
    },
    {
      "epoch": 0.6945186451543969,
      "grad_norm": 2.845492124557495,
      "learning_rate": 1.3073796791443851e-05,
      "loss": 0.12419921159744263,
      "num_input_tokens_seen": 1674652,
      "step": 1625,
      "train_runtime": 3921.495,
      "train_tokens_per_second": 427.044
    },
    {
      "epoch": 0.6949460412437226,
      "grad_norm": 2.095703601837158,
      "learning_rate": 1.306951871657754e-05,
      "loss": 0.09531547129154205,
      "num_input_tokens_seen": 1675698,
      "step": 1626,
      "train_runtime": 3923.579,
      "train_tokens_per_second": 427.084
    },
    {
      "epoch": 0.6953734373330483,
      "grad_norm": 2.989895820617676,
      "learning_rate": 1.3065240641711232e-05,
      "loss": 0.09823733568191528,
      "num_input_tokens_seen": 1676306,
      "step": 1627,
      "train_runtime": 3925.5613,
      "train_tokens_per_second": 427.023
    },
    {
      "epoch": 0.6958008334223742,
      "grad_norm": 2.459953546524048,
      "learning_rate": 1.3060962566844921e-05,
      "loss": 0.1193925142288208,
      "num_input_tokens_seen": 1677460,
      "step": 1628,
      "train_runtime": 3927.7076,
      "train_tokens_per_second": 427.084
    },
    {
      "epoch": 0.6962282295116999,
      "grad_norm": 3.317596197128296,
      "learning_rate": 1.305668449197861e-05,
      "loss": 0.0970437079668045,
      "num_input_tokens_seen": 1678144,
      "step": 1629,
      "train_runtime": 3929.7129,
      "train_tokens_per_second": 427.04
    },
    {
      "epoch": 0.6966556256010258,
      "grad_norm": 2.53511118888855,
      "learning_rate": 1.30524064171123e-05,
      "loss": 0.12688416242599487,
      "num_input_tokens_seen": 1679080,
      "step": 1630,
      "train_runtime": 3931.7921,
      "train_tokens_per_second": 427.052
    },
    {
      "epoch": 0.6970830216903515,
      "grad_norm": 2.4305365085601807,
      "learning_rate": 1.3048128342245991e-05,
      "loss": 0.07128015160560608,
      "num_input_tokens_seen": 1680132,
      "step": 1631,
      "train_runtime": 3933.9185,
      "train_tokens_per_second": 427.089
    },
    {
      "epoch": 0.6975104177796773,
      "grad_norm": 3.819319248199463,
      "learning_rate": 1.304385026737968e-05,
      "loss": 0.09876695275306702,
      "num_input_tokens_seen": 1680726,
      "step": 1632,
      "train_runtime": 3935.9616,
      "train_tokens_per_second": 427.018
    },
    {
      "epoch": 0.6979378138690031,
      "grad_norm": 3.7172605991363525,
      "learning_rate": 1.303957219251337e-05,
      "loss": 0.22836706042289734,
      "num_input_tokens_seen": 1681756,
      "step": 1633,
      "train_runtime": 3938.1041,
      "train_tokens_per_second": 427.047
    },
    {
      "epoch": 0.6983652099583288,
      "grad_norm": 3.323859930038452,
      "learning_rate": 1.303529411764706e-05,
      "loss": 0.1308099925518036,
      "num_input_tokens_seen": 1682650,
      "step": 1634,
      "train_runtime": 3940.3052,
      "train_tokens_per_second": 427.035
    },
    {
      "epoch": 0.6987926060476547,
      "grad_norm": 3.0783824920654297,
      "learning_rate": 1.303101604278075e-05,
      "loss": 0.21143673360347748,
      "num_input_tokens_seen": 1684078,
      "step": 1635,
      "train_runtime": 3942.473,
      "train_tokens_per_second": 427.163
    },
    {
      "epoch": 0.6992200021369804,
      "grad_norm": 2.3745343685150146,
      "learning_rate": 1.302673796791444e-05,
      "loss": 0.14765599370002747,
      "num_input_tokens_seen": 1685224,
      "step": 1636,
      "train_runtime": 3944.5702,
      "train_tokens_per_second": 427.226
    },
    {
      "epoch": 0.6996473982263063,
      "grad_norm": 3.0867974758148193,
      "learning_rate": 1.302245989304813e-05,
      "loss": 0.1804279088973999,
      "num_input_tokens_seen": 1686050,
      "step": 1637,
      "train_runtime": 3946.6139,
      "train_tokens_per_second": 427.214
    },
    {
      "epoch": 0.700074794315632,
      "grad_norm": 2.6826188564300537,
      "learning_rate": 1.3018181818181819e-05,
      "loss": 0.08859211206436157,
      "num_input_tokens_seen": 1686956,
      "step": 1638,
      "train_runtime": 3948.6891,
      "train_tokens_per_second": 427.219
    },
    {
      "epoch": 0.7005021904049578,
      "grad_norm": 3.395402669906616,
      "learning_rate": 1.3013903743315509e-05,
      "loss": 0.15626779198646545,
      "num_input_tokens_seen": 1687798,
      "step": 1639,
      "train_runtime": 3950.6905,
      "train_tokens_per_second": 427.216
    },
    {
      "epoch": 0.7009295864942836,
      "grad_norm": 2.332399368286133,
      "learning_rate": 1.30096256684492e-05,
      "loss": 0.10119998455047607,
      "num_input_tokens_seen": 1689004,
      "step": 1640,
      "train_runtime": 3952.7852,
      "train_tokens_per_second": 427.295
    },
    {
      "epoch": 0.7013569825836093,
      "grad_norm": 3.0781655311584473,
      "learning_rate": 1.3005347593582889e-05,
      "loss": 0.10670844465494156,
      "num_input_tokens_seen": 1689856,
      "step": 1641,
      "train_runtime": 3954.8167,
      "train_tokens_per_second": 427.291
    },
    {
      "epoch": 0.7017843786729352,
      "grad_norm": 2.5547902584075928,
      "learning_rate": 1.3001069518716579e-05,
      "loss": 0.06811079382896423,
      "num_input_tokens_seen": 1690670,
      "step": 1642,
      "train_runtime": 3956.8163,
      "train_tokens_per_second": 427.28
    },
    {
      "epoch": 0.7022117747622609,
      "grad_norm": 3.101245164871216,
      "learning_rate": 1.2996791443850268e-05,
      "loss": 0.1471184641122818,
      "num_input_tokens_seen": 1691480,
      "step": 1643,
      "train_runtime": 3958.8592,
      "train_tokens_per_second": 427.264
    },
    {
      "epoch": 0.7026391708515867,
      "grad_norm": 3.2463717460632324,
      "learning_rate": 1.2992513368983959e-05,
      "loss": 0.10457038879394531,
      "num_input_tokens_seen": 1692250,
      "step": 1644,
      "train_runtime": 3960.8557,
      "train_tokens_per_second": 427.244
    },
    {
      "epoch": 0.7030665669409125,
      "grad_norm": 2.880979537963867,
      "learning_rate": 1.2988235294117649e-05,
      "loss": 0.1903861165046692,
      "num_input_tokens_seen": 1693420,
      "step": 1645,
      "train_runtime": 3962.9732,
      "train_tokens_per_second": 427.31
    },
    {
      "epoch": 0.7034939630302383,
      "grad_norm": 4.939394950866699,
      "learning_rate": 1.2983957219251338e-05,
      "loss": 0.20906499028205872,
      "num_input_tokens_seen": 1694210,
      "step": 1646,
      "train_runtime": 3964.9752,
      "train_tokens_per_second": 427.294
    },
    {
      "epoch": 0.7039213591195641,
      "grad_norm": 2.6525001525878906,
      "learning_rate": 1.2979679144385027e-05,
      "loss": 0.11395084112882614,
      "num_input_tokens_seen": 1695012,
      "step": 1647,
      "train_runtime": 3967.0044,
      "train_tokens_per_second": 427.278
    },
    {
      "epoch": 0.7043487552088898,
      "grad_norm": 2.86409854888916,
      "learning_rate": 1.2975401069518719e-05,
      "loss": 0.1560884714126587,
      "num_input_tokens_seen": 1696218,
      "step": 1648,
      "train_runtime": 3969.1362,
      "train_tokens_per_second": 427.352
    },
    {
      "epoch": 0.7047761512982156,
      "grad_norm": 2.9886152744293213,
      "learning_rate": 1.2971122994652408e-05,
      "loss": 0.12156147509813309,
      "num_input_tokens_seen": 1697258,
      "step": 1649,
      "train_runtime": 3971.2053,
      "train_tokens_per_second": 427.391
    },
    {
      "epoch": 0.7052035473875414,
      "grad_norm": 3.0794358253479004,
      "learning_rate": 1.2966844919786097e-05,
      "loss": 0.10936135798692703,
      "num_input_tokens_seen": 1698094,
      "step": 1650,
      "train_runtime": 3973.2434,
      "train_tokens_per_second": 427.382
    },
    {
      "epoch": 0.7056309434768672,
      "grad_norm": 3.7119407653808594,
      "learning_rate": 1.2962566844919787e-05,
      "loss": 0.11862250417470932,
      "num_input_tokens_seen": 1698594,
      "step": 1651,
      "train_runtime": 3981.421,
      "train_tokens_per_second": 426.63
    },
    {
      "epoch": 0.706058339566193,
      "grad_norm": 3.069287061691284,
      "learning_rate": 1.2958288770053478e-05,
      "loss": 0.18450075387954712,
      "num_input_tokens_seen": 1699658,
      "step": 1652,
      "train_runtime": 3983.501,
      "train_tokens_per_second": 426.674
    },
    {
      "epoch": 0.7064857356555188,
      "grad_norm": 2.3116512298583984,
      "learning_rate": 1.2954010695187168e-05,
      "loss": 0.13053485751152039,
      "num_input_tokens_seen": 1700770,
      "step": 1653,
      "train_runtime": 3985.6066,
      "train_tokens_per_second": 426.728
    },
    {
      "epoch": 0.7069131317448445,
      "grad_norm": 2.779745101928711,
      "learning_rate": 1.2949732620320857e-05,
      "loss": 0.16588063538074493,
      "num_input_tokens_seen": 1701816,
      "step": 1654,
      "train_runtime": 3987.6714,
      "train_tokens_per_second": 426.769
    },
    {
      "epoch": 0.7073405278341703,
      "grad_norm": 2.9462406635284424,
      "learning_rate": 1.2945454545454545e-05,
      "loss": 0.20384477078914642,
      "num_input_tokens_seen": 1703090,
      "step": 1655,
      "train_runtime": 3989.7657,
      "train_tokens_per_second": 426.865
    },
    {
      "epoch": 0.7077679239234961,
      "grad_norm": 2.217874526977539,
      "learning_rate": 1.2941176470588238e-05,
      "loss": 0.10553063452243805,
      "num_input_tokens_seen": 1704284,
      "step": 1656,
      "train_runtime": 3991.8634,
      "train_tokens_per_second": 426.939
    },
    {
      "epoch": 0.7081953200128219,
      "grad_norm": 3.3472745418548584,
      "learning_rate": 1.2936898395721927e-05,
      "loss": 0.17429442703723907,
      "num_input_tokens_seen": 1705218,
      "step": 1657,
      "train_runtime": 3993.9054,
      "train_tokens_per_second": 426.955
    },
    {
      "epoch": 0.7086227161021477,
      "grad_norm": 3.8400657176971436,
      "learning_rate": 1.2932620320855615e-05,
      "loss": 0.1665058434009552,
      "num_input_tokens_seen": 1706048,
      "step": 1658,
      "train_runtime": 3995.9538,
      "train_tokens_per_second": 426.944
    },
    {
      "epoch": 0.7090501121914734,
      "grad_norm": 2.0666720867156982,
      "learning_rate": 1.2928342245989304e-05,
      "loss": 0.0790431946516037,
      "num_input_tokens_seen": 1707348,
      "step": 1659,
      "train_runtime": 3998.0856,
      "train_tokens_per_second": 427.041
    },
    {
      "epoch": 0.7094775082807993,
      "grad_norm": 2.7541933059692383,
      "learning_rate": 1.2924064171122997e-05,
      "loss": 0.12006007879972458,
      "num_input_tokens_seen": 1708384,
      "step": 1660,
      "train_runtime": 4000.1165,
      "train_tokens_per_second": 427.084
    },
    {
      "epoch": 0.709904904370125,
      "grad_norm": 2.458078145980835,
      "learning_rate": 1.2919786096256686e-05,
      "loss": 0.09136639535427094,
      "num_input_tokens_seen": 1709360,
      "step": 1661,
      "train_runtime": 4002.1602,
      "train_tokens_per_second": 427.109
    },
    {
      "epoch": 0.7103323004594508,
      "grad_norm": 2.8389453887939453,
      "learning_rate": 1.2915508021390374e-05,
      "loss": 0.16383452713489532,
      "num_input_tokens_seen": 1710508,
      "step": 1662,
      "train_runtime": 4004.2777,
      "train_tokens_per_second": 427.17
    },
    {
      "epoch": 0.7107596965487766,
      "grad_norm": 3.5363619327545166,
      "learning_rate": 1.2911229946524064e-05,
      "loss": 0.22730599343776703,
      "num_input_tokens_seen": 1712138,
      "step": 1663,
      "train_runtime": 4006.6545,
      "train_tokens_per_second": 427.324
    },
    {
      "epoch": 0.7111870926381023,
      "grad_norm": 3.8893401622772217,
      "learning_rate": 1.2906951871657757e-05,
      "loss": 0.19639477133750916,
      "num_input_tokens_seen": 1712860,
      "step": 1664,
      "train_runtime": 4008.7035,
      "train_tokens_per_second": 427.285
    },
    {
      "epoch": 0.7116144887274282,
      "grad_norm": 3.728217124938965,
      "learning_rate": 1.2902673796791444e-05,
      "loss": 0.11914101988077164,
      "num_input_tokens_seen": 1713704,
      "step": 1665,
      "train_runtime": 4010.7475,
      "train_tokens_per_second": 427.278
    },
    {
      "epoch": 0.7120418848167539,
      "grad_norm": 3.5283544063568115,
      "learning_rate": 1.2898395721925134e-05,
      "loss": 0.1672677844762802,
      "num_input_tokens_seen": 1714516,
      "step": 1666,
      "train_runtime": 4012.7993,
      "train_tokens_per_second": 427.262
    },
    {
      "epoch": 0.7124692809060797,
      "grad_norm": 2.4075896739959717,
      "learning_rate": 1.2894117647058823e-05,
      "loss": 0.12488129734992981,
      "num_input_tokens_seen": 1715660,
      "step": 1667,
      "train_runtime": 4014.8872,
      "train_tokens_per_second": 427.325
    },
    {
      "epoch": 0.7128966769954055,
      "grad_norm": 3.5406494140625,
      "learning_rate": 1.2889839572192516e-05,
      "loss": 0.1721637099981308,
      "num_input_tokens_seen": 1716730,
      "step": 1668,
      "train_runtime": 4016.9938,
      "train_tokens_per_second": 427.367
    },
    {
      "epoch": 0.7133240730847312,
      "grad_norm": 2.7573654651641846,
      "learning_rate": 1.2885561497326204e-05,
      "loss": 0.11639227718114853,
      "num_input_tokens_seen": 1717640,
      "step": 1669,
      "train_runtime": 4019.0426,
      "train_tokens_per_second": 427.375
    },
    {
      "epoch": 0.7137514691740571,
      "grad_norm": 4.804013729095459,
      "learning_rate": 1.2881283422459893e-05,
      "loss": 0.21367575228214264,
      "num_input_tokens_seen": 1718460,
      "step": 1670,
      "train_runtime": 4021.0944,
      "train_tokens_per_second": 427.361
    },
    {
      "epoch": 0.7141788652633828,
      "grad_norm": 3.78324294090271,
      "learning_rate": 1.2877005347593583e-05,
      "loss": 0.14412060379981995,
      "num_input_tokens_seen": 1719306,
      "step": 1671,
      "train_runtime": 4023.1503,
      "train_tokens_per_second": 427.353
    },
    {
      "epoch": 0.7146062613527087,
      "grad_norm": 2.279505968093872,
      "learning_rate": 1.2872727272727274e-05,
      "loss": 0.12978610396385193,
      "num_input_tokens_seen": 1720296,
      "step": 1672,
      "train_runtime": 4025.2769,
      "train_tokens_per_second": 427.373
    },
    {
      "epoch": 0.7150336574420344,
      "grad_norm": 4.358551979064941,
      "learning_rate": 1.2868449197860963e-05,
      "loss": 0.1565861701965332,
      "num_input_tokens_seen": 1720946,
      "step": 1673,
      "train_runtime": 4027.2598,
      "train_tokens_per_second": 427.324
    },
    {
      "epoch": 0.7154610535313601,
      "grad_norm": 3.662733316421509,
      "learning_rate": 1.2864171122994653e-05,
      "loss": 0.18193666636943817,
      "num_input_tokens_seen": 1722042,
      "step": 1674,
      "train_runtime": 4029.393,
      "train_tokens_per_second": 427.37
    },
    {
      "epoch": 0.715888449620686,
      "grad_norm": 2.6407856941223145,
      "learning_rate": 1.2859893048128342e-05,
      "loss": 0.13663464784622192,
      "num_input_tokens_seen": 1723140,
      "step": 1675,
      "train_runtime": 4031.4938,
      "train_tokens_per_second": 427.42
    },
    {
      "epoch": 0.7163158457100117,
      "grad_norm": 2.2533209323883057,
      "learning_rate": 1.2855614973262033e-05,
      "loss": 0.09304480999708176,
      "num_input_tokens_seen": 1724224,
      "step": 1676,
      "train_runtime": 4033.6004,
      "train_tokens_per_second": 427.465
    },
    {
      "epoch": 0.7167432417993376,
      "grad_norm": 2.527470111846924,
      "learning_rate": 1.2851336898395723e-05,
      "loss": 0.09156791120767593,
      "num_input_tokens_seen": 1725548,
      "step": 1677,
      "train_runtime": 4035.7495,
      "train_tokens_per_second": 427.566
    },
    {
      "epoch": 0.7171706378886633,
      "grad_norm": 3.8725194931030273,
      "learning_rate": 1.2847058823529412e-05,
      "loss": 0.13674314320087433,
      "num_input_tokens_seen": 1726676,
      "step": 1678,
      "train_runtime": 4037.8574,
      "train_tokens_per_second": 427.622
    },
    {
      "epoch": 0.717598033977989,
      "grad_norm": 3.6760237216949463,
      "learning_rate": 1.2842780748663102e-05,
      "loss": 0.17109842598438263,
      "num_input_tokens_seen": 1727428,
      "step": 1679,
      "train_runtime": 4039.8438,
      "train_tokens_per_second": 427.598
    },
    {
      "epoch": 0.7180254300673149,
      "grad_norm": 2.799755334854126,
      "learning_rate": 1.2838502673796793e-05,
      "loss": 0.16299965977668762,
      "num_input_tokens_seen": 1728446,
      "step": 1680,
      "train_runtime": 4041.9223,
      "train_tokens_per_second": 427.63
    },
    {
      "epoch": 0.7184528261566406,
      "grad_norm": 3.8620645999908447,
      "learning_rate": 1.2834224598930482e-05,
      "loss": 0.10270003974437714,
      "num_input_tokens_seen": 1729056,
      "step": 1681,
      "train_runtime": 4050.3046,
      "train_tokens_per_second": 426.895
    },
    {
      "epoch": 0.7188802222459665,
      "grad_norm": 3.610529899597168,
      "learning_rate": 1.2829946524064172e-05,
      "loss": 0.18845506012439728,
      "num_input_tokens_seen": 1730126,
      "step": 1682,
      "train_runtime": 4052.3803,
      "train_tokens_per_second": 426.941
    },
    {
      "epoch": 0.7193076183352922,
      "grad_norm": 3.0155282020568848,
      "learning_rate": 1.2825668449197861e-05,
      "loss": 0.11069008708000183,
      "num_input_tokens_seen": 1730996,
      "step": 1683,
      "train_runtime": 4054.4851,
      "train_tokens_per_second": 426.934
    },
    {
      "epoch": 0.7197350144246181,
      "grad_norm": 3.03389310836792,
      "learning_rate": 1.2821390374331552e-05,
      "loss": 0.16556021571159363,
      "num_input_tokens_seen": 1732152,
      "step": 1684,
      "train_runtime": 4056.5803,
      "train_tokens_per_second": 426.998
    },
    {
      "epoch": 0.7201624105139438,
      "grad_norm": 2.5943901538848877,
      "learning_rate": 1.2817112299465242e-05,
      "loss": 0.12813101708889008,
      "num_input_tokens_seen": 1733204,
      "step": 1685,
      "train_runtime": 4058.6533,
      "train_tokens_per_second": 427.039
    },
    {
      "epoch": 0.7205898066032695,
      "grad_norm": 2.7205615043640137,
      "learning_rate": 1.2812834224598931e-05,
      "loss": 0.07578977197408676,
      "num_input_tokens_seen": 1734046,
      "step": 1686,
      "train_runtime": 4060.6929,
      "train_tokens_per_second": 427.032
    },
    {
      "epoch": 0.7210172026925954,
      "grad_norm": 2.8328237533569336,
      "learning_rate": 1.280855614973262e-05,
      "loss": 0.1297164112329483,
      "num_input_tokens_seen": 1735042,
      "step": 1687,
      "train_runtime": 4062.8256,
      "train_tokens_per_second": 427.053
    },
    {
      "epoch": 0.7214445987819211,
      "grad_norm": 3.1164746284484863,
      "learning_rate": 1.2804278074866312e-05,
      "loss": 0.168115496635437,
      "num_input_tokens_seen": 1736068,
      "step": 1688,
      "train_runtime": 4064.9025,
      "train_tokens_per_second": 427.087
    },
    {
      "epoch": 0.721871994871247,
      "grad_norm": 3.218972682952881,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.1437757909297943,
      "num_input_tokens_seen": 1737184,
      "step": 1689,
      "train_runtime": 4067.0197,
      "train_tokens_per_second": 427.139
    },
    {
      "epoch": 0.7222993909605727,
      "grad_norm": 3.383817195892334,
      "learning_rate": 1.279572192513369e-05,
      "loss": 0.20265325903892517,
      "num_input_tokens_seen": 1738208,
      "step": 1690,
      "train_runtime": 4069.0634,
      "train_tokens_per_second": 427.176
    },
    {
      "epoch": 0.7227267870498985,
      "grad_norm": 2.3254318237304688,
      "learning_rate": 1.279144385026738e-05,
      "loss": 0.12910565733909607,
      "num_input_tokens_seen": 1739152,
      "step": 1691,
      "train_runtime": 4071.1144,
      "train_tokens_per_second": 427.193
    },
    {
      "epoch": 0.7231541831392243,
      "grad_norm": 3.0438852310180664,
      "learning_rate": 1.2787165775401071e-05,
      "loss": 0.12765420973300934,
      "num_input_tokens_seen": 1740346,
      "step": 1692,
      "train_runtime": 4073.2656,
      "train_tokens_per_second": 427.261
    },
    {
      "epoch": 0.72358157922855,
      "grad_norm": 3.4133188724517822,
      "learning_rate": 1.278288770053476e-05,
      "loss": 0.2313060462474823,
      "num_input_tokens_seen": 1741248,
      "step": 1693,
      "train_runtime": 4075.469,
      "train_tokens_per_second": 427.251
    },
    {
      "epoch": 0.7240089753178759,
      "grad_norm": 3.4037530422210693,
      "learning_rate": 1.277860962566845e-05,
      "loss": 0.15263766050338745,
      "num_input_tokens_seen": 1742122,
      "step": 1694,
      "train_runtime": 4077.5272,
      "train_tokens_per_second": 427.25
    },
    {
      "epoch": 0.7244363714072016,
      "grad_norm": 2.177854299545288,
      "learning_rate": 1.277433155080214e-05,
      "loss": 0.11862416565418243,
      "num_input_tokens_seen": 1743328,
      "step": 1695,
      "train_runtime": 4079.6776,
      "train_tokens_per_second": 427.32
    },
    {
      "epoch": 0.7248637674965274,
      "grad_norm": 3.185788869857788,
      "learning_rate": 1.277005347593583e-05,
      "loss": 0.1639426201581955,
      "num_input_tokens_seen": 1744486,
      "step": 1696,
      "train_runtime": 4081.8278,
      "train_tokens_per_second": 427.379
    },
    {
      "epoch": 0.7252911635858532,
      "grad_norm": 3.2895989418029785,
      "learning_rate": 1.276577540106952e-05,
      "loss": 0.1792648732662201,
      "num_input_tokens_seen": 1745474,
      "step": 1697,
      "train_runtime": 4083.9907,
      "train_tokens_per_second": 427.394
    },
    {
      "epoch": 0.725718559675179,
      "grad_norm": 4.185188293457031,
      "learning_rate": 1.276149732620321e-05,
      "loss": 0.17716136574745178,
      "num_input_tokens_seen": 1746166,
      "step": 1698,
      "train_runtime": 4085.9399,
      "train_tokens_per_second": 427.36
    },
    {
      "epoch": 0.7261459557645048,
      "grad_norm": 3.3039119243621826,
      "learning_rate": 1.2757219251336899e-05,
      "loss": 0.20853915810585022,
      "num_input_tokens_seen": 1747250,
      "step": 1699,
      "train_runtime": 4088.011,
      "train_tokens_per_second": 427.408
    },
    {
      "epoch": 0.7265733518538305,
      "grad_norm": 2.205228090286255,
      "learning_rate": 1.275294117647059e-05,
      "loss": 0.11350136995315552,
      "num_input_tokens_seen": 1748350,
      "step": 1700,
      "train_runtime": 4090.125,
      "train_tokens_per_second": 427.456
    },
    {
      "epoch": 0.7270007479431563,
      "grad_norm": 3.152604341506958,
      "learning_rate": 1.274866310160428e-05,
      "loss": 0.1099860817193985,
      "num_input_tokens_seen": 1749094,
      "step": 1701,
      "train_runtime": 4092.1684,
      "train_tokens_per_second": 427.425
    },
    {
      "epoch": 0.7274281440324821,
      "grad_norm": 3.303657054901123,
      "learning_rate": 1.2744385026737969e-05,
      "loss": 0.1371927261352539,
      "num_input_tokens_seen": 1749952,
      "step": 1702,
      "train_runtime": 4094.2086,
      "train_tokens_per_second": 427.421
    },
    {
      "epoch": 0.7278555401218079,
      "grad_norm": 2.9780490398406982,
      "learning_rate": 1.2740106951871658e-05,
      "loss": 0.1036418080329895,
      "num_input_tokens_seen": 1750582,
      "step": 1703,
      "train_runtime": 4096.195,
      "train_tokens_per_second": 427.368
    },
    {
      "epoch": 0.7282829362111337,
      "grad_norm": 3.7338707447052,
      "learning_rate": 1.273582887700535e-05,
      "loss": 0.11675436794757843,
      "num_input_tokens_seen": 1751290,
      "step": 1704,
      "train_runtime": 4098.1705,
      "train_tokens_per_second": 427.335
    },
    {
      "epoch": 0.7287103323004595,
      "grad_norm": 4.176034450531006,
      "learning_rate": 1.2731550802139039e-05,
      "loss": 0.23184752464294434,
      "num_input_tokens_seen": 1752338,
      "step": 1705,
      "train_runtime": 4100.2506,
      "train_tokens_per_second": 427.373
    },
    {
      "epoch": 0.7291377283897852,
      "grad_norm": 2.412473201751709,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 0.10220357030630112,
      "num_input_tokens_seen": 1753384,
      "step": 1706,
      "train_runtime": 4102.3417,
      "train_tokens_per_second": 427.411
    },
    {
      "epoch": 0.729565124479111,
      "grad_norm": 3.10713267326355,
      "learning_rate": 1.2722994652406418e-05,
      "loss": 0.1686929315328598,
      "num_input_tokens_seen": 1754138,
      "step": 1707,
      "train_runtime": 4104.3655,
      "train_tokens_per_second": 427.383
    },
    {
      "epoch": 0.7299925205684368,
      "grad_norm": 3.179652214050293,
      "learning_rate": 1.2718716577540109e-05,
      "loss": 0.19970983266830444,
      "num_input_tokens_seen": 1755136,
      "step": 1708,
      "train_runtime": 4106.4543,
      "train_tokens_per_second": 427.409
    },
    {
      "epoch": 0.7304199166577626,
      "grad_norm": 3.3354885578155518,
      "learning_rate": 1.2714438502673798e-05,
      "loss": 0.18996183574199677,
      "num_input_tokens_seen": 1756140,
      "step": 1709,
      "train_runtime": 4108.5043,
      "train_tokens_per_second": 427.44
    },
    {
      "epoch": 0.7308473127470884,
      "grad_norm": 2.7286245822906494,
      "learning_rate": 1.2710160427807488e-05,
      "loss": 0.12833094596862793,
      "num_input_tokens_seen": 1757116,
      "step": 1710,
      "train_runtime": 4110.5429,
      "train_tokens_per_second": 427.466
    },
    {
      "epoch": 0.7312747088364141,
      "grad_norm": 2.2786917686462402,
      "learning_rate": 1.2705882352941177e-05,
      "loss": 0.05347301810979843,
      "num_input_tokens_seen": 1757922,
      "step": 1711,
      "train_runtime": 4118.8575,
      "train_tokens_per_second": 426.798
    },
    {
      "epoch": 0.73170210492574,
      "grad_norm": 2.3057162761688232,
      "learning_rate": 1.2701604278074869e-05,
      "loss": 0.13636146485805511,
      "num_input_tokens_seen": 1759588,
      "step": 1712,
      "train_runtime": 4121.0813,
      "train_tokens_per_second": 426.972
    },
    {
      "epoch": 0.7321295010150657,
      "grad_norm": 3.88169264793396,
      "learning_rate": 1.2697326203208558e-05,
      "loss": 0.11356302350759506,
      "num_input_tokens_seen": 1760308,
      "step": 1713,
      "train_runtime": 4123.1221,
      "train_tokens_per_second": 426.936
    },
    {
      "epoch": 0.7325568971043915,
      "grad_norm": 3.0245745182037354,
      "learning_rate": 1.2693048128342247e-05,
      "loss": 0.23742002248764038,
      "num_input_tokens_seen": 1762078,
      "step": 1714,
      "train_runtime": 4125.3936,
      "train_tokens_per_second": 427.13
    },
    {
      "epoch": 0.7329842931937173,
      "grad_norm": 1.959012508392334,
      "learning_rate": 1.2688770053475937e-05,
      "loss": 0.0950913280248642,
      "num_input_tokens_seen": 1766162,
      "step": 1715,
      "train_runtime": 4128.1949,
      "train_tokens_per_second": 427.829
    },
    {
      "epoch": 0.733411689283043,
      "grad_norm": 3.113219738006592,
      "learning_rate": 1.2684491978609628e-05,
      "loss": 0.08972372114658356,
      "num_input_tokens_seen": 1766872,
      "step": 1716,
      "train_runtime": 4130.1733,
      "train_tokens_per_second": 427.796
    },
    {
      "epoch": 0.7338390853723689,
      "grad_norm": 2.6438539028167725,
      "learning_rate": 1.2680213903743317e-05,
      "loss": 0.13002948462963104,
      "num_input_tokens_seen": 1767812,
      "step": 1717,
      "train_runtime": 4132.2657,
      "train_tokens_per_second": 427.807
    },
    {
      "epoch": 0.7342664814616946,
      "grad_norm": 3.5801172256469727,
      "learning_rate": 1.2675935828877007e-05,
      "loss": 0.12898202240467072,
      "num_input_tokens_seen": 1768594,
      "step": 1718,
      "train_runtime": 4134.2679,
      "train_tokens_per_second": 427.789
    },
    {
      "epoch": 0.7346938775510204,
      "grad_norm": 3.45713472366333,
      "learning_rate": 1.2671657754010696e-05,
      "loss": 0.16828308999538422,
      "num_input_tokens_seen": 1769324,
      "step": 1719,
      "train_runtime": 4136.2092,
      "train_tokens_per_second": 427.765
    },
    {
      "epoch": 0.7351212736403462,
      "grad_norm": 4.62410306930542,
      "learning_rate": 1.2667379679144387e-05,
      "loss": 0.1622631996870041,
      "num_input_tokens_seen": 1770244,
      "step": 1720,
      "train_runtime": 4138.3029,
      "train_tokens_per_second": 427.771
    },
    {
      "epoch": 0.7355486697296719,
      "grad_norm": 3.0335936546325684,
      "learning_rate": 1.2663101604278077e-05,
      "loss": 0.16483764350414276,
      "num_input_tokens_seen": 1771034,
      "step": 1721,
      "train_runtime": 4140.2914,
      "train_tokens_per_second": 427.756
    },
    {
      "epoch": 0.7359760658189978,
      "grad_norm": 3.9681930541992188,
      "learning_rate": 1.2658823529411766e-05,
      "loss": 0.11987143754959106,
      "num_input_tokens_seen": 1771848,
      "step": 1722,
      "train_runtime": 4142.422,
      "train_tokens_per_second": 427.732
    },
    {
      "epoch": 0.7364034619083235,
      "grad_norm": 3.1155405044555664,
      "learning_rate": 1.2654545454545454e-05,
      "loss": 0.10921688377857208,
      "num_input_tokens_seen": 1772830,
      "step": 1723,
      "train_runtime": 4144.4785,
      "train_tokens_per_second": 427.757
    },
    {
      "epoch": 0.7368308579976494,
      "grad_norm": 4.516935348510742,
      "learning_rate": 1.2650267379679147e-05,
      "loss": 0.14564579725265503,
      "num_input_tokens_seen": 1773516,
      "step": 1724,
      "train_runtime": 4146.4903,
      "train_tokens_per_second": 427.715
    },
    {
      "epoch": 0.7372582540869751,
      "grad_norm": 3.113145351409912,
      "learning_rate": 1.2645989304812836e-05,
      "loss": 0.12443186342716217,
      "num_input_tokens_seen": 1774302,
      "step": 1725,
      "train_runtime": 4148.6709,
      "train_tokens_per_second": 427.68
    },
    {
      "epoch": 0.7376856501763008,
      "grad_norm": 3.4054698944091797,
      "learning_rate": 1.2641711229946524e-05,
      "loss": 0.19764456152915955,
      "num_input_tokens_seen": 1775332,
      "step": 1726,
      "train_runtime": 4150.708,
      "train_tokens_per_second": 427.718
    },
    {
      "epoch": 0.7381130462656267,
      "grad_norm": 1.7858580350875854,
      "learning_rate": 1.2637433155080214e-05,
      "loss": 0.0921870619058609,
      "num_input_tokens_seen": 1777012,
      "step": 1727,
      "train_runtime": 4152.9714,
      "train_tokens_per_second": 427.889
    },
    {
      "epoch": 0.7385404423549524,
      "grad_norm": 2.897024631500244,
      "learning_rate": 1.2633155080213906e-05,
      "loss": 0.1305638700723648,
      "num_input_tokens_seen": 1778124,
      "step": 1728,
      "train_runtime": 4155.1162,
      "train_tokens_per_second": 427.936
    },
    {
      "epoch": 0.7389678384442783,
      "grad_norm": 2.877156972885132,
      "learning_rate": 1.2628877005347596e-05,
      "loss": 0.15893763303756714,
      "num_input_tokens_seen": 1778986,
      "step": 1729,
      "train_runtime": 4157.1675,
      "train_tokens_per_second": 427.932
    },
    {
      "epoch": 0.739395234533604,
      "grad_norm": 2.7703163623809814,
      "learning_rate": 1.2624598930481284e-05,
      "loss": 0.16486668586730957,
      "num_input_tokens_seen": 1780028,
      "step": 1730,
      "train_runtime": 4159.2708,
      "train_tokens_per_second": 427.966
    },
    {
      "epoch": 0.7398226306229299,
      "grad_norm": 2.6259467601776123,
      "learning_rate": 1.2620320855614973e-05,
      "loss": 0.0976109430193901,
      "num_input_tokens_seen": 1781036,
      "step": 1731,
      "train_runtime": 4161.3548,
      "train_tokens_per_second": 427.994
    },
    {
      "epoch": 0.7402500267122556,
      "grad_norm": 2.6563796997070312,
      "learning_rate": 1.2616042780748666e-05,
      "loss": 0.10229581594467163,
      "num_input_tokens_seen": 1782164,
      "step": 1732,
      "train_runtime": 4163.4706,
      "train_tokens_per_second": 428.048
    },
    {
      "epoch": 0.7406774228015813,
      "grad_norm": 2.7999024391174316,
      "learning_rate": 1.2611764705882354e-05,
      "loss": 0.10329177975654602,
      "num_input_tokens_seen": 1783190,
      "step": 1733,
      "train_runtime": 4165.5409,
      "train_tokens_per_second": 428.081
    },
    {
      "epoch": 0.7411048188909072,
      "grad_norm": 2.784428119659424,
      "learning_rate": 1.2607486631016043e-05,
      "loss": 0.15808075666427612,
      "num_input_tokens_seen": 1784506,
      "step": 1734,
      "train_runtime": 4167.7083,
      "train_tokens_per_second": 428.174
    },
    {
      "epoch": 0.7415322149802329,
      "grad_norm": 3.747777223587036,
      "learning_rate": 1.2603208556149733e-05,
      "loss": 0.12574762105941772,
      "num_input_tokens_seen": 1785292,
      "step": 1735,
      "train_runtime": 4169.7539,
      "train_tokens_per_second": 428.153
    },
    {
      "epoch": 0.7419596110695588,
      "grad_norm": 3.335115432739258,
      "learning_rate": 1.2598930481283425e-05,
      "loss": 0.14596965909004211,
      "num_input_tokens_seen": 1786170,
      "step": 1736,
      "train_runtime": 4171.8037,
      "train_tokens_per_second": 428.153
    },
    {
      "epoch": 0.7423870071588845,
      "grad_norm": 5.01472806930542,
      "learning_rate": 1.2594652406417113e-05,
      "loss": 0.16974447667598724,
      "num_input_tokens_seen": 1786842,
      "step": 1737,
      "train_runtime": 4173.7808,
      "train_tokens_per_second": 428.111
    },
    {
      "epoch": 0.7428144032482102,
      "grad_norm": 3.320187568664551,
      "learning_rate": 1.2590374331550803e-05,
      "loss": 0.17597053945064545,
      "num_input_tokens_seen": 1788128,
      "step": 1738,
      "train_runtime": 4175.944,
      "train_tokens_per_second": 428.197
    },
    {
      "epoch": 0.7432417993375361,
      "grad_norm": 3.2284138202667236,
      "learning_rate": 1.2586096256684492e-05,
      "loss": 0.15552473068237305,
      "num_input_tokens_seen": 1789048,
      "step": 1739,
      "train_runtime": 4178.02,
      "train_tokens_per_second": 428.205
    },
    {
      "epoch": 0.7436691954268618,
      "grad_norm": 3.238884687423706,
      "learning_rate": 1.2581818181818183e-05,
      "loss": 0.18473714590072632,
      "num_input_tokens_seen": 1790130,
      "step": 1740,
      "train_runtime": 4180.1213,
      "train_tokens_per_second": 428.248
    },
    {
      "epoch": 0.7440965915161877,
      "grad_norm": 2.707418918609619,
      "learning_rate": 1.2577540106951873e-05,
      "loss": 0.1684379130601883,
      "num_input_tokens_seen": 1791316,
      "step": 1741,
      "train_runtime": 4188.4932,
      "train_tokens_per_second": 427.676
    },
    {
      "epoch": 0.7445239876055134,
      "grad_norm": 3.8198659420013428,
      "learning_rate": 1.2573262032085562e-05,
      "loss": 0.10199625045061111,
      "num_input_tokens_seen": 1792168,
      "step": 1742,
      "train_runtime": 4190.4965,
      "train_tokens_per_second": 427.674
    },
    {
      "epoch": 0.7449513836948392,
      "grad_norm": 4.6382012367248535,
      "learning_rate": 1.2568983957219252e-05,
      "loss": 0.10959888249635696,
      "num_input_tokens_seen": 1792770,
      "step": 1743,
      "train_runtime": 4192.4783,
      "train_tokens_per_second": 427.616
    },
    {
      "epoch": 0.745378779784165,
      "grad_norm": 2.788638114929199,
      "learning_rate": 1.2564705882352943e-05,
      "loss": 0.1129089891910553,
      "num_input_tokens_seen": 1793856,
      "step": 1744,
      "train_runtime": 4194.5809,
      "train_tokens_per_second": 427.66
    },
    {
      "epoch": 0.7458061758734907,
      "grad_norm": 2.915764570236206,
      "learning_rate": 1.2560427807486632e-05,
      "loss": 0.16279761493206024,
      "num_input_tokens_seen": 1794662,
      "step": 1745,
      "train_runtime": 4196.5897,
      "train_tokens_per_second": 427.648
    },
    {
      "epoch": 0.7462335719628166,
      "grad_norm": 2.4101195335388184,
      "learning_rate": 1.2556149732620322e-05,
      "loss": 0.10009346902370453,
      "num_input_tokens_seen": 1795566,
      "step": 1746,
      "train_runtime": 4198.6187,
      "train_tokens_per_second": 427.656
    },
    {
      "epoch": 0.7466609680521423,
      "grad_norm": 3.4743504524230957,
      "learning_rate": 1.2551871657754011e-05,
      "loss": 0.12854984402656555,
      "num_input_tokens_seen": 1796186,
      "step": 1747,
      "train_runtime": 4200.561,
      "train_tokens_per_second": 427.606
    },
    {
      "epoch": 0.7470883641414681,
      "grad_norm": 2.804262638092041,
      "learning_rate": 1.25475935828877e-05,
      "loss": 0.18147476017475128,
      "num_input_tokens_seen": 1798050,
      "step": 1748,
      "train_runtime": 4202.8551,
      "train_tokens_per_second": 427.816
    },
    {
      "epoch": 0.7475157602307939,
      "grad_norm": 2.84769344329834,
      "learning_rate": 1.2543315508021392e-05,
      "loss": 0.1201382502913475,
      "num_input_tokens_seen": 1799004,
      "step": 1749,
      "train_runtime": 4204.9327,
      "train_tokens_per_second": 427.832
    },
    {
      "epoch": 0.7479431563201197,
      "grad_norm": 3.483933448791504,
      "learning_rate": 1.2539037433155081e-05,
      "loss": 0.10916777700185776,
      "num_input_tokens_seen": 1799964,
      "step": 1750,
      "train_runtime": 4207.006,
      "train_tokens_per_second": 427.849
    },
    {
      "epoch": 0.7483705524094455,
      "grad_norm": 3.2232701778411865,
      "learning_rate": 1.253475935828877e-05,
      "loss": 0.12337847799062729,
      "num_input_tokens_seen": 1800854,
      "step": 1751,
      "train_runtime": 4209.0698,
      "train_tokens_per_second": 427.851
    },
    {
      "epoch": 0.7487979484987712,
      "grad_norm": 2.846959352493286,
      "learning_rate": 1.253048128342246e-05,
      "loss": 0.16990630328655243,
      "num_input_tokens_seen": 1801794,
      "step": 1752,
      "train_runtime": 4211.1746,
      "train_tokens_per_second": 427.86
    },
    {
      "epoch": 0.749225344588097,
      "grad_norm": 2.67598032951355,
      "learning_rate": 1.2526203208556151e-05,
      "loss": 0.10821860283613205,
      "num_input_tokens_seen": 1802644,
      "step": 1753,
      "train_runtime": 4213.2295,
      "train_tokens_per_second": 427.853
    },
    {
      "epoch": 0.7496527406774228,
      "grad_norm": 3.1944963932037354,
      "learning_rate": 1.252192513368984e-05,
      "loss": 0.24526366591453552,
      "num_input_tokens_seen": 1803724,
      "step": 1754,
      "train_runtime": 4215.3604,
      "train_tokens_per_second": 427.893
    },
    {
      "epoch": 0.7500801367667486,
      "grad_norm": 2.943840742111206,
      "learning_rate": 1.251764705882353e-05,
      "loss": 0.15970027446746826,
      "num_input_tokens_seen": 1804638,
      "step": 1755,
      "train_runtime": 4217.5597,
      "train_tokens_per_second": 427.887
    },
    {
      "epoch": 0.7505075328560744,
      "grad_norm": 3.1619248390197754,
      "learning_rate": 1.251336898395722e-05,
      "loss": 0.16502642631530762,
      "num_input_tokens_seen": 1805340,
      "step": 1756,
      "train_runtime": 4219.538,
      "train_tokens_per_second": 427.853
    },
    {
      "epoch": 0.7509349289454001,
      "grad_norm": 3.274695634841919,
      "learning_rate": 1.250909090909091e-05,
      "loss": 0.15791615843772888,
      "num_input_tokens_seen": 1806728,
      "step": 1757,
      "train_runtime": 4221.7075,
      "train_tokens_per_second": 427.961
    },
    {
      "epoch": 0.7513623250347259,
      "grad_norm": 2.935295581817627,
      "learning_rate": 1.25048128342246e-05,
      "loss": 0.13088148832321167,
      "num_input_tokens_seen": 1807926,
      "step": 1758,
      "train_runtime": 4223.7775,
      "train_tokens_per_second": 428.035
    },
    {
      "epoch": 0.7517897211240517,
      "grad_norm": 3.229001045227051,
      "learning_rate": 1.250053475935829e-05,
      "loss": 0.19611865282058716,
      "num_input_tokens_seen": 1808960,
      "step": 1759,
      "train_runtime": 4225.852,
      "train_tokens_per_second": 428.07
    },
    {
      "epoch": 0.7522171172133775,
      "grad_norm": 2.4109506607055664,
      "learning_rate": 1.2496256684491979e-05,
      "loss": 0.11808124929666519,
      "num_input_tokens_seen": 1810242,
      "step": 1760,
      "train_runtime": 4227.9918,
      "train_tokens_per_second": 428.156
    },
    {
      "epoch": 0.7526445133027033,
      "grad_norm": 2.914564371109009,
      "learning_rate": 1.249197860962567e-05,
      "loss": 0.14882615208625793,
      "num_input_tokens_seen": 1811086,
      "step": 1761,
      "train_runtime": 4230.042,
      "train_tokens_per_second": 428.148
    },
    {
      "epoch": 0.7530719093920291,
      "grad_norm": 2.611783742904663,
      "learning_rate": 1.248770053475936e-05,
      "loss": 0.2134699523448944,
      "num_input_tokens_seen": 1812102,
      "step": 1762,
      "train_runtime": 4232.173,
      "train_tokens_per_second": 428.173
    },
    {
      "epoch": 0.7534993054813548,
      "grad_norm": 3.4775607585906982,
      "learning_rate": 1.2483422459893049e-05,
      "loss": 0.17539821565151215,
      "num_input_tokens_seen": 1812822,
      "step": 1763,
      "train_runtime": 4234.1844,
      "train_tokens_per_second": 428.14
    },
    {
      "epoch": 0.7539267015706806,
      "grad_norm": 3.1305854320526123,
      "learning_rate": 1.2479144385026738e-05,
      "loss": 0.18104904890060425,
      "num_input_tokens_seen": 1813886,
      "step": 1764,
      "train_runtime": 4236.2635,
      "train_tokens_per_second": 428.181
    },
    {
      "epoch": 0.7543540976600064,
      "grad_norm": 2.763216257095337,
      "learning_rate": 1.247486631016043e-05,
      "loss": 0.12206174433231354,
      "num_input_tokens_seen": 1815006,
      "step": 1765,
      "train_runtime": 4238.3948,
      "train_tokens_per_second": 428.23
    },
    {
      "epoch": 0.7547814937493322,
      "grad_norm": 2.1365344524383545,
      "learning_rate": 1.2470588235294119e-05,
      "loss": 0.10479989647865295,
      "num_input_tokens_seen": 1815906,
      "step": 1766,
      "train_runtime": 4240.4826,
      "train_tokens_per_second": 428.231
    },
    {
      "epoch": 0.755208889838658,
      "grad_norm": 2.2437663078308105,
      "learning_rate": 1.2466310160427808e-05,
      "loss": 0.07789380103349686,
      "num_input_tokens_seen": 1816980,
      "step": 1767,
      "train_runtime": 4242.5519,
      "train_tokens_per_second": 428.275
    },
    {
      "epoch": 0.7556362859279837,
      "grad_norm": 2.27329158782959,
      "learning_rate": 1.2462032085561498e-05,
      "loss": 0.11260631680488586,
      "num_input_tokens_seen": 1817988,
      "step": 1768,
      "train_runtime": 4244.591,
      "train_tokens_per_second": 428.307
    },
    {
      "epoch": 0.7560636820173096,
      "grad_norm": 2.837125778198242,
      "learning_rate": 1.2457754010695189e-05,
      "loss": 0.16431836783885956,
      "num_input_tokens_seen": 1819128,
      "step": 1769,
      "train_runtime": 4246.6892,
      "train_tokens_per_second": 428.364
    },
    {
      "epoch": 0.7564910781066353,
      "grad_norm": 1.7793630361557007,
      "learning_rate": 1.2453475935828878e-05,
      "loss": 0.08038732409477234,
      "num_input_tokens_seen": 1820096,
      "step": 1770,
      "train_runtime": 4248.7316,
      "train_tokens_per_second": 428.386
    },
    {
      "epoch": 0.7569184741959611,
      "grad_norm": 2.8599798679351807,
      "learning_rate": 1.2449197860962568e-05,
      "loss": 0.1291077733039856,
      "num_input_tokens_seen": 1820958,
      "step": 1771,
      "train_runtime": 4257.3071,
      "train_tokens_per_second": 427.725
    },
    {
      "epoch": 0.7573458702852869,
      "grad_norm": 3.425774335861206,
      "learning_rate": 1.2444919786096257e-05,
      "loss": 0.1876506507396698,
      "num_input_tokens_seen": 1821880,
      "step": 1772,
      "train_runtime": 4259.3967,
      "train_tokens_per_second": 427.732
    },
    {
      "epoch": 0.7577732663746126,
      "grad_norm": 2.5635290145874023,
      "learning_rate": 1.2440641711229948e-05,
      "loss": 0.14740212261676788,
      "num_input_tokens_seen": 1822866,
      "step": 1773,
      "train_runtime": 4261.4621,
      "train_tokens_per_second": 427.756
    },
    {
      "epoch": 0.7582006624639385,
      "grad_norm": 3.358508825302124,
      "learning_rate": 1.2436363636363638e-05,
      "loss": 0.2512654662132263,
      "num_input_tokens_seen": 1823646,
      "step": 1774,
      "train_runtime": 4263.4883,
      "train_tokens_per_second": 427.736
    },
    {
      "epoch": 0.7586280585532642,
      "grad_norm": 2.407524347305298,
      "learning_rate": 1.2432085561497327e-05,
      "loss": 0.14994464814662933,
      "num_input_tokens_seen": 1825138,
      "step": 1775,
      "train_runtime": 4265.7137,
      "train_tokens_per_second": 427.862
    },
    {
      "epoch": 0.75905545464259,
      "grad_norm": 2.9009687900543213,
      "learning_rate": 1.2427807486631017e-05,
      "loss": 0.1654253900051117,
      "num_input_tokens_seen": 1825992,
      "step": 1776,
      "train_runtime": 4267.7444,
      "train_tokens_per_second": 427.859
    },
    {
      "epoch": 0.7594828507319158,
      "grad_norm": 1.869413137435913,
      "learning_rate": 1.2423529411764708e-05,
      "loss": 0.0809844508767128,
      "num_input_tokens_seen": 1827504,
      "step": 1777,
      "train_runtime": 4269.96,
      "train_tokens_per_second": 427.991
    },
    {
      "epoch": 0.7599102468212415,
      "grad_norm": 1.9973881244659424,
      "learning_rate": 1.2419251336898397e-05,
      "loss": 0.07713896036148071,
      "num_input_tokens_seen": 1828998,
      "step": 1778,
      "train_runtime": 4272.1232,
      "train_tokens_per_second": 428.124
    },
    {
      "epoch": 0.7603376429105674,
      "grad_norm": 2.606424331665039,
      "learning_rate": 1.2414973262032087e-05,
      "loss": 0.09174610674381256,
      "num_input_tokens_seen": 1829948,
      "step": 1779,
      "train_runtime": 4274.2064,
      "train_tokens_per_second": 428.137
    },
    {
      "epoch": 0.7607650389998931,
      "grad_norm": 3.9461727142333984,
      "learning_rate": 1.2410695187165776e-05,
      "loss": 0.1827625185251236,
      "num_input_tokens_seen": 1831222,
      "step": 1780,
      "train_runtime": 4276.3501,
      "train_tokens_per_second": 428.221
    },
    {
      "epoch": 0.761192435089219,
      "grad_norm": 3.3259847164154053,
      "learning_rate": 1.2406417112299467e-05,
      "loss": 0.17484451830387115,
      "num_input_tokens_seen": 1832244,
      "step": 1781,
      "train_runtime": 4278.4585,
      "train_tokens_per_second": 428.249
    },
    {
      "epoch": 0.7616198311785447,
      "grad_norm": 1.900813341140747,
      "learning_rate": 1.2402139037433157e-05,
      "loss": 0.08792635053396225,
      "num_input_tokens_seen": 1833752,
      "step": 1782,
      "train_runtime": 4280.6523,
      "train_tokens_per_second": 428.381
    },
    {
      "epoch": 0.7620472272678706,
      "grad_norm": 3.819176435470581,
      "learning_rate": 1.2397860962566846e-05,
      "loss": 0.2154502421617508,
      "num_input_tokens_seen": 1834596,
      "step": 1783,
      "train_runtime": 4282.6576,
      "train_tokens_per_second": 428.378
    },
    {
      "epoch": 0.7624746233571963,
      "grad_norm": 2.9362082481384277,
      "learning_rate": 1.2393582887700534e-05,
      "loss": 0.18833956122398376,
      "num_input_tokens_seen": 1835838,
      "step": 1784,
      "train_runtime": 4284.8036,
      "train_tokens_per_second": 428.453
    },
    {
      "epoch": 0.762902019446522,
      "grad_norm": 4.101992607116699,
      "learning_rate": 1.2389304812834227e-05,
      "loss": 0.1190357506275177,
      "num_input_tokens_seen": 1836712,
      "step": 1785,
      "train_runtime": 4286.8422,
      "train_tokens_per_second": 428.453
    },
    {
      "epoch": 0.7633294155358479,
      "grad_norm": 4.283321380615234,
      "learning_rate": 1.2385026737967916e-05,
      "loss": 0.2087101936340332,
      "num_input_tokens_seen": 1837436,
      "step": 1786,
      "train_runtime": 4288.8695,
      "train_tokens_per_second": 428.42
    },
    {
      "epoch": 0.7637568116251736,
      "grad_norm": 2.6511285305023193,
      "learning_rate": 1.2380748663101606e-05,
      "loss": 0.09836596250534058,
      "num_input_tokens_seen": 1838206,
      "step": 1787,
      "train_runtime": 4290.9125,
      "train_tokens_per_second": 428.395
    },
    {
      "epoch": 0.7641842077144995,
      "grad_norm": 4.136745452880859,
      "learning_rate": 1.2376470588235294e-05,
      "loss": 0.1616525650024414,
      "num_input_tokens_seen": 1839590,
      "step": 1788,
      "train_runtime": 4293.0719,
      "train_tokens_per_second": 428.502
    },
    {
      "epoch": 0.7646116038038252,
      "grad_norm": 2.8372857570648193,
      "learning_rate": 1.2372192513368986e-05,
      "loss": 0.09618731588125229,
      "num_input_tokens_seen": 1840380,
      "step": 1789,
      "train_runtime": 4295.0617,
      "train_tokens_per_second": 428.487
    },
    {
      "epoch": 0.7650389998931509,
      "grad_norm": 2.310497283935547,
      "learning_rate": 1.2367914438502676e-05,
      "loss": 0.12419398128986359,
      "num_input_tokens_seen": 1841326,
      "step": 1790,
      "train_runtime": 4297.1207,
      "train_tokens_per_second": 428.502
    },
    {
      "epoch": 0.7654663959824768,
      "grad_norm": 3.164872884750366,
      "learning_rate": 1.2363636363636364e-05,
      "loss": 0.11873511970043182,
      "num_input_tokens_seen": 1842096,
      "step": 1791,
      "train_runtime": 4299.1351,
      "train_tokens_per_second": 428.481
    },
    {
      "epoch": 0.7658937920718025,
      "grad_norm": 3.534221887588501,
      "learning_rate": 1.2359358288770053e-05,
      "loss": 0.15829111635684967,
      "num_input_tokens_seen": 1842650,
      "step": 1792,
      "train_runtime": 4301.0667,
      "train_tokens_per_second": 428.417
    },
    {
      "epoch": 0.7663211881611284,
      "grad_norm": 2.8869614601135254,
      "learning_rate": 1.2355080213903746e-05,
      "loss": 0.16093742847442627,
      "num_input_tokens_seen": 1843962,
      "step": 1793,
      "train_runtime": 4303.2007,
      "train_tokens_per_second": 428.509
    },
    {
      "epoch": 0.7667485842504541,
      "grad_norm": 4.311562538146973,
      "learning_rate": 1.2350802139037434e-05,
      "loss": 0.2588258981704712,
      "num_input_tokens_seen": 1844878,
      "step": 1794,
      "train_runtime": 4305.2642,
      "train_tokens_per_second": 428.517
    },
    {
      "epoch": 0.7671759803397799,
      "grad_norm": 4.153341293334961,
      "learning_rate": 1.2346524064171123e-05,
      "loss": 0.14938843250274658,
      "num_input_tokens_seen": 1845536,
      "step": 1795,
      "train_runtime": 4307.2342,
      "train_tokens_per_second": 428.474
    },
    {
      "epoch": 0.7676033764291057,
      "grad_norm": 2.732607126235962,
      "learning_rate": 1.2342245989304812e-05,
      "loss": 0.15332266688346863,
      "num_input_tokens_seen": 1846996,
      "step": 1796,
      "train_runtime": 4309.4255,
      "train_tokens_per_second": 428.594
    },
    {
      "epoch": 0.7680307725184314,
      "grad_norm": 3.60335111618042,
      "learning_rate": 1.2337967914438505e-05,
      "loss": 0.15083912014961243,
      "num_input_tokens_seen": 1848206,
      "step": 1797,
      "train_runtime": 4311.5582,
      "train_tokens_per_second": 428.663
    },
    {
      "epoch": 0.7684581686077573,
      "grad_norm": 3.4597063064575195,
      "learning_rate": 1.2333689839572193e-05,
      "loss": 0.17319080233573914,
      "num_input_tokens_seen": 1849206,
      "step": 1798,
      "train_runtime": 4313.6339,
      "train_tokens_per_second": 428.689
    },
    {
      "epoch": 0.768885564697083,
      "grad_norm": 2.7628331184387207,
      "learning_rate": 1.2329411764705882e-05,
      "loss": 0.11759724467992783,
      "num_input_tokens_seen": 1849852,
      "step": 1799,
      "train_runtime": 4315.6069,
      "train_tokens_per_second": 428.642
    },
    {
      "epoch": 0.7693129607864088,
      "grad_norm": 3.115293502807617,
      "learning_rate": 1.2325133689839572e-05,
      "loss": 0.14337000250816345,
      "num_input_tokens_seen": 1851192,
      "step": 1800,
      "train_runtime": 4317.738,
      "train_tokens_per_second": 428.741
    },
    {
      "epoch": 0.7697403568757346,
      "grad_norm": 3.182490587234497,
      "learning_rate": 1.2320855614973263e-05,
      "loss": 0.1292513906955719,
      "num_input_tokens_seen": 1852074,
      "step": 1801,
      "train_runtime": 4326.4491,
      "train_tokens_per_second": 428.082
    },
    {
      "epoch": 0.7701677529650603,
      "grad_norm": 2.756492853164673,
      "learning_rate": 1.2316577540106953e-05,
      "loss": 0.16810327768325806,
      "num_input_tokens_seen": 1853086,
      "step": 1802,
      "train_runtime": 4328.5524,
      "train_tokens_per_second": 428.108
    },
    {
      "epoch": 0.7705951490543862,
      "grad_norm": 2.266127109527588,
      "learning_rate": 1.2312299465240642e-05,
      "loss": 0.134847030043602,
      "num_input_tokens_seen": 1854112,
      "step": 1803,
      "train_runtime": 4330.6585,
      "train_tokens_per_second": 428.136
    },
    {
      "epoch": 0.7710225451437119,
      "grad_norm": 3.6650660037994385,
      "learning_rate": 1.2308021390374331e-05,
      "loss": 0.12369846552610397,
      "num_input_tokens_seen": 1855120,
      "step": 1804,
      "train_runtime": 4332.7248,
      "train_tokens_per_second": 428.165
    },
    {
      "epoch": 0.7714499412330377,
      "grad_norm": 2.8042409420013428,
      "learning_rate": 1.2303743315508023e-05,
      "loss": 0.09973473846912384,
      "num_input_tokens_seen": 1855818,
      "step": 1805,
      "train_runtime": 4334.7173,
      "train_tokens_per_second": 428.129
    },
    {
      "epoch": 0.7718773373223635,
      "grad_norm": 2.951108932495117,
      "learning_rate": 1.2299465240641712e-05,
      "loss": 0.110304094851017,
      "num_input_tokens_seen": 1856408,
      "step": 1806,
      "train_runtime": 4336.6975,
      "train_tokens_per_second": 428.07
    },
    {
      "epoch": 0.7723047334116893,
      "grad_norm": 3.008371353149414,
      "learning_rate": 1.2295187165775401e-05,
      "loss": 0.1220649927854538,
      "num_input_tokens_seen": 1857328,
      "step": 1807,
      "train_runtime": 4338.8552,
      "train_tokens_per_second": 428.069
    },
    {
      "epoch": 0.7727321295010151,
      "grad_norm": 3.2356770038604736,
      "learning_rate": 1.2290909090909091e-05,
      "loss": 0.17200100421905518,
      "num_input_tokens_seen": 1858462,
      "step": 1808,
      "train_runtime": 4340.9454,
      "train_tokens_per_second": 428.124
    },
    {
      "epoch": 0.7731595255903408,
      "grad_norm": 3.228485584259033,
      "learning_rate": 1.2286631016042782e-05,
      "loss": 0.1288929283618927,
      "num_input_tokens_seen": 1859326,
      "step": 1809,
      "train_runtime": 4343.0364,
      "train_tokens_per_second": 428.117
    },
    {
      "epoch": 0.7735869216796666,
      "grad_norm": 2.7139246463775635,
      "learning_rate": 1.2282352941176471e-05,
      "loss": 0.12399152666330338,
      "num_input_tokens_seen": 1860078,
      "step": 1810,
      "train_runtime": 4345.2019,
      "train_tokens_per_second": 428.076
    },
    {
      "epoch": 0.7740143177689924,
      "grad_norm": 2.4700589179992676,
      "learning_rate": 1.2278074866310161e-05,
      "loss": 0.12367812544107437,
      "num_input_tokens_seen": 1861088,
      "step": 1811,
      "train_runtime": 4347.3237,
      "train_tokens_per_second": 428.1
    },
    {
      "epoch": 0.7744417138583182,
      "grad_norm": 3.568169355392456,
      "learning_rate": 1.227379679144385e-05,
      "loss": 0.13981027901172638,
      "num_input_tokens_seen": 1861916,
      "step": 1812,
      "train_runtime": 4349.4386,
      "train_tokens_per_second": 428.082
    },
    {
      "epoch": 0.774869109947644,
      "grad_norm": 2.8587687015533447,
      "learning_rate": 1.2269518716577542e-05,
      "loss": 0.12921200692653656,
      "num_input_tokens_seen": 1862624,
      "step": 1813,
      "train_runtime": 4351.5854,
      "train_tokens_per_second": 428.033
    },
    {
      "epoch": 0.7752965060369698,
      "grad_norm": 3.8217129707336426,
      "learning_rate": 1.2265240641711231e-05,
      "loss": 0.13139057159423828,
      "num_input_tokens_seen": 1863660,
      "step": 1814,
      "train_runtime": 4353.731,
      "train_tokens_per_second": 428.06
    },
    {
      "epoch": 0.7757239021262955,
      "grad_norm": 2.0437350273132324,
      "learning_rate": 1.226096256684492e-05,
      "loss": 0.10422183573246002,
      "num_input_tokens_seen": 1864714,
      "step": 1815,
      "train_runtime": 4355.8017,
      "train_tokens_per_second": 428.099
    },
    {
      "epoch": 0.7761512982156213,
      "grad_norm": 2.9613823890686035,
      "learning_rate": 1.225668449197861e-05,
      "loss": 0.16858945786952972,
      "num_input_tokens_seen": 1865414,
      "step": 1816,
      "train_runtime": 4357.8182,
      "train_tokens_per_second": 428.061
    },
    {
      "epoch": 0.7765786943049471,
      "grad_norm": 3.295197010040283,
      "learning_rate": 1.2252406417112301e-05,
      "loss": 0.17331922054290771,
      "num_input_tokens_seen": 1866346,
      "step": 1817,
      "train_runtime": 4359.8926,
      "train_tokens_per_second": 428.072
    },
    {
      "epoch": 0.7770060903942729,
      "grad_norm": 5.003626823425293,
      "learning_rate": 1.224812834224599e-05,
      "loss": 0.2478761374950409,
      "num_input_tokens_seen": 1867286,
      "step": 1818,
      "train_runtime": 4361.9827,
      "train_tokens_per_second": 428.082
    },
    {
      "epoch": 0.7774334864835987,
      "grad_norm": 3.4966726303100586,
      "learning_rate": 1.224385026737968e-05,
      "loss": 0.14820632338523865,
      "num_input_tokens_seen": 1868052,
      "step": 1819,
      "train_runtime": 4364.0258,
      "train_tokens_per_second": 428.057
    },
    {
      "epoch": 0.7778608825729244,
      "grad_norm": 2.869807720184326,
      "learning_rate": 1.223957219251337e-05,
      "loss": 0.1422082781791687,
      "num_input_tokens_seen": 1869264,
      "step": 1820,
      "train_runtime": 4366.1735,
      "train_tokens_per_second": 428.124
    },
    {
      "epoch": 0.7782882786622503,
      "grad_norm": 2.4492299556732178,
      "learning_rate": 1.223529411764706e-05,
      "loss": 0.12338537722826004,
      "num_input_tokens_seen": 1870456,
      "step": 1821,
      "train_runtime": 4368.3101,
      "train_tokens_per_second": 428.188
    },
    {
      "epoch": 0.778715674751576,
      "grad_norm": 2.3549654483795166,
      "learning_rate": 1.223101604278075e-05,
      "loss": 0.15092220902442932,
      "num_input_tokens_seen": 1871972,
      "step": 1822,
      "train_runtime": 4370.5129,
      "train_tokens_per_second": 428.319
    },
    {
      "epoch": 0.7791430708409018,
      "grad_norm": 4.360873222351074,
      "learning_rate": 1.222673796791444e-05,
      "loss": 0.13382264971733093,
      "num_input_tokens_seen": 1872810,
      "step": 1823,
      "train_runtime": 4372.5505,
      "train_tokens_per_second": 428.311
    },
    {
      "epoch": 0.7795704669302276,
      "grad_norm": 2.985707998275757,
      "learning_rate": 1.2222459893048129e-05,
      "loss": 0.11476238071918488,
      "num_input_tokens_seen": 1873992,
      "step": 1824,
      "train_runtime": 4374.6884,
      "train_tokens_per_second": 428.372
    },
    {
      "epoch": 0.7799978630195533,
      "grad_norm": 3.015141487121582,
      "learning_rate": 1.221818181818182e-05,
      "loss": 0.18612177670001984,
      "num_input_tokens_seen": 1874830,
      "step": 1825,
      "train_runtime": 4376.7349,
      "train_tokens_per_second": 428.363
    },
    {
      "epoch": 0.7804252591088792,
      "grad_norm": 3.2259037494659424,
      "learning_rate": 1.221390374331551e-05,
      "loss": 0.14476238191127777,
      "num_input_tokens_seen": 1875602,
      "step": 1826,
      "train_runtime": 4378.7863,
      "train_tokens_per_second": 428.338
    },
    {
      "epoch": 0.7808526551982049,
      "grad_norm": 2.8509538173675537,
      "learning_rate": 1.2209625668449199e-05,
      "loss": 0.1654292196035385,
      "num_input_tokens_seen": 1876560,
      "step": 1827,
      "train_runtime": 4380.9107,
      "train_tokens_per_second": 428.349
    },
    {
      "epoch": 0.7812800512875308,
      "grad_norm": 2.0071630477905273,
      "learning_rate": 1.2205347593582888e-05,
      "loss": 0.12034884095191956,
      "num_input_tokens_seen": 1877702,
      "step": 1828,
      "train_runtime": 4383.028,
      "train_tokens_per_second": 428.403
    },
    {
      "epoch": 0.7817074473768565,
      "grad_norm": 2.5182299613952637,
      "learning_rate": 1.220106951871658e-05,
      "loss": 0.24549561738967896,
      "num_input_tokens_seen": 1878878,
      "step": 1829,
      "train_runtime": 4385.1451,
      "train_tokens_per_second": 428.464
    },
    {
      "epoch": 0.7821348434661823,
      "grad_norm": 4.174243927001953,
      "learning_rate": 1.2196791443850269e-05,
      "loss": 0.08729416131973267,
      "num_input_tokens_seen": 1879378,
      "step": 1830,
      "train_runtime": 4387.095,
      "train_tokens_per_second": 428.388
    },
    {
      "epoch": 0.7825622395555081,
      "grad_norm": 3.676351308822632,
      "learning_rate": 1.2192513368983958e-05,
      "loss": 0.19944506883621216,
      "num_input_tokens_seen": 1880362,
      "step": 1831,
      "train_runtime": 4395.8122,
      "train_tokens_per_second": 427.762
    },
    {
      "epoch": 0.7829896356448338,
      "grad_norm": 5.0852766036987305,
      "learning_rate": 1.2188235294117648e-05,
      "loss": 0.26022642850875854,
      "num_input_tokens_seen": 1881118,
      "step": 1832,
      "train_runtime": 4397.8645,
      "train_tokens_per_second": 427.734
    },
    {
      "epoch": 0.7834170317341597,
      "grad_norm": 3.649362564086914,
      "learning_rate": 1.2183957219251339e-05,
      "loss": 0.1976720690727234,
      "num_input_tokens_seen": 1882114,
      "step": 1833,
      "train_runtime": 4400.0081,
      "train_tokens_per_second": 427.752
    },
    {
      "epoch": 0.7838444278234854,
      "grad_norm": 2.1792664527893066,
      "learning_rate": 1.2179679144385028e-05,
      "loss": 0.07211606204509735,
      "num_input_tokens_seen": 1883070,
      "step": 1834,
      "train_runtime": 4402.0918,
      "train_tokens_per_second": 427.767
    },
    {
      "epoch": 0.7842718239128113,
      "grad_norm": 3.1566128730773926,
      "learning_rate": 1.2175401069518718e-05,
      "loss": 0.16733992099761963,
      "num_input_tokens_seen": 1883892,
      "step": 1835,
      "train_runtime": 4404.1334,
      "train_tokens_per_second": 427.755
    },
    {
      "epoch": 0.784699220002137,
      "grad_norm": 1.6701500415802002,
      "learning_rate": 1.2171122994652407e-05,
      "loss": 0.07423573732376099,
      "num_input_tokens_seen": 1885090,
      "step": 1836,
      "train_runtime": 4406.2833,
      "train_tokens_per_second": 427.819
    },
    {
      "epoch": 0.7851266160914627,
      "grad_norm": 2.973598003387451,
      "learning_rate": 1.2166844919786098e-05,
      "loss": 0.1275968849658966,
      "num_input_tokens_seen": 1886034,
      "step": 1837,
      "train_runtime": 4408.3542,
      "train_tokens_per_second": 427.832
    },
    {
      "epoch": 0.7855540121807886,
      "grad_norm": 3.8972911834716797,
      "learning_rate": 1.2162566844919788e-05,
      "loss": 0.2016734480857849,
      "num_input_tokens_seen": 1887506,
      "step": 1838,
      "train_runtime": 4410.5179,
      "train_tokens_per_second": 427.956
    },
    {
      "epoch": 0.7859814082701143,
      "grad_norm": 2.3188345432281494,
      "learning_rate": 1.2158288770053477e-05,
      "loss": 0.1339975893497467,
      "num_input_tokens_seen": 1889108,
      "step": 1839,
      "train_runtime": 4412.7334,
      "train_tokens_per_second": 428.104
    },
    {
      "epoch": 0.7864088043594402,
      "grad_norm": 1.981019377708435,
      "learning_rate": 1.2154010695187167e-05,
      "loss": 0.11797478049993515,
      "num_input_tokens_seen": 1890818,
      "step": 1840,
      "train_runtime": 4414.9385,
      "train_tokens_per_second": 428.277
    },
    {
      "epoch": 0.7868362004487659,
      "grad_norm": 2.261542558670044,
      "learning_rate": 1.2149732620320858e-05,
      "loss": 0.07785042375326157,
      "num_input_tokens_seen": 1891494,
      "step": 1841,
      "train_runtime": 4416.9177,
      "train_tokens_per_second": 428.238
    },
    {
      "epoch": 0.7872635965380916,
      "grad_norm": 2.9760546684265137,
      "learning_rate": 1.2145454545454547e-05,
      "loss": 0.12564532458782196,
      "num_input_tokens_seen": 1892488,
      "step": 1842,
      "train_runtime": 4419.0074,
      "train_tokens_per_second": 428.261
    },
    {
      "epoch": 0.7876909926274175,
      "grad_norm": 1.758131504058838,
      "learning_rate": 1.2141176470588237e-05,
      "loss": 0.0910901129245758,
      "num_input_tokens_seen": 1893790,
      "step": 1843,
      "train_runtime": 4421.2268,
      "train_tokens_per_second": 428.34
    },
    {
      "epoch": 0.7881183887167432,
      "grad_norm": 2.690927505493164,
      "learning_rate": 1.2136898395721926e-05,
      "loss": 0.14538080990314484,
      "num_input_tokens_seen": 1894892,
      "step": 1844,
      "train_runtime": 4423.3054,
      "train_tokens_per_second": 428.388
    },
    {
      "epoch": 0.7885457848060691,
      "grad_norm": 4.087973117828369,
      "learning_rate": 1.2132620320855617e-05,
      "loss": 0.19727779924869537,
      "num_input_tokens_seen": 1895542,
      "step": 1845,
      "train_runtime": 4425.5222,
      "train_tokens_per_second": 428.321
    },
    {
      "epoch": 0.7889731808953948,
      "grad_norm": 3.1194968223571777,
      "learning_rate": 1.2128342245989307e-05,
      "loss": 0.2249072641134262,
      "num_input_tokens_seen": 1896596,
      "step": 1846,
      "train_runtime": 4427.576,
      "train_tokens_per_second": 428.36
    },
    {
      "epoch": 0.7894005769847205,
      "grad_norm": 3.3519327640533447,
      "learning_rate": 1.2124064171122996e-05,
      "loss": 0.1892613023519516,
      "num_input_tokens_seen": 1897578,
      "step": 1847,
      "train_runtime": 4429.6196,
      "train_tokens_per_second": 428.384
    },
    {
      "epoch": 0.7898279730740464,
      "grad_norm": 4.134048938751221,
      "learning_rate": 1.2119786096256686e-05,
      "loss": 0.21346129477024078,
      "num_input_tokens_seen": 1898646,
      "step": 1848,
      "train_runtime": 4431.6929,
      "train_tokens_per_second": 428.425
    },
    {
      "epoch": 0.7902553691633721,
      "grad_norm": 2.2150657176971436,
      "learning_rate": 1.2115508021390377e-05,
      "loss": 0.0927320048213005,
      "num_input_tokens_seen": 1899718,
      "step": 1849,
      "train_runtime": 4433.7996,
      "train_tokens_per_second": 428.463
    },
    {
      "epoch": 0.790682765252698,
      "grad_norm": 1.9254262447357178,
      "learning_rate": 1.2111229946524066e-05,
      "loss": 0.0757264494895935,
      "num_input_tokens_seen": 1900998,
      "step": 1850,
      "train_runtime": 4435.9456,
      "train_tokens_per_second": 428.544
    },
    {
      "epoch": 0.7911101613420237,
      "grad_norm": 2.189628839492798,
      "learning_rate": 1.2106951871657756e-05,
      "loss": 0.09491493552923203,
      "num_input_tokens_seen": 1901850,
      "step": 1851,
      "train_runtime": 4437.9879,
      "train_tokens_per_second": 428.539
    },
    {
      "epoch": 0.7915375574313495,
      "grad_norm": 2.282789468765259,
      "learning_rate": 1.2102673796791443e-05,
      "loss": 0.13587215542793274,
      "num_input_tokens_seen": 1902918,
      "step": 1852,
      "train_runtime": 4440.0609,
      "train_tokens_per_second": 428.579
    },
    {
      "epoch": 0.7919649535206753,
      "grad_norm": 3.310476541519165,
      "learning_rate": 1.2098395721925136e-05,
      "loss": 0.14387257397174835,
      "num_input_tokens_seen": 1903776,
      "step": 1853,
      "train_runtime": 4442.0678,
      "train_tokens_per_second": 428.579
    },
    {
      "epoch": 0.792392349610001,
      "grad_norm": 3.180478811264038,
      "learning_rate": 1.2094117647058826e-05,
      "loss": 0.19241297245025635,
      "num_input_tokens_seen": 1904982,
      "step": 1854,
      "train_runtime": 4444.1946,
      "train_tokens_per_second": 428.645
    },
    {
      "epoch": 0.7928197456993269,
      "grad_norm": 4.283200740814209,
      "learning_rate": 1.2089839572192515e-05,
      "loss": 0.18826483190059662,
      "num_input_tokens_seen": 1905624,
      "step": 1855,
      "train_runtime": 4446.2009,
      "train_tokens_per_second": 428.596
    },
    {
      "epoch": 0.7932471417886526,
      "grad_norm": 4.868692874908447,
      "learning_rate": 1.2085561497326203e-05,
      "loss": 0.1696545034646988,
      "num_input_tokens_seen": 1906454,
      "step": 1856,
      "train_runtime": 4448.2548,
      "train_tokens_per_second": 428.585
    },
    {
      "epoch": 0.7936745378779784,
      "grad_norm": 2.9834084510803223,
      "learning_rate": 1.2081283422459892e-05,
      "loss": 0.1212657168507576,
      "num_input_tokens_seen": 1907816,
      "step": 1857,
      "train_runtime": 4450.3991,
      "train_tokens_per_second": 428.684
    },
    {
      "epoch": 0.7941019339673042,
      "grad_norm": 2.3191118240356445,
      "learning_rate": 1.2077005347593585e-05,
      "loss": 0.10848109424114227,
      "num_input_tokens_seen": 1909478,
      "step": 1858,
      "train_runtime": 4452.6289,
      "train_tokens_per_second": 428.843
    },
    {
      "epoch": 0.79452933005663,
      "grad_norm": 3.261073350906372,
      "learning_rate": 1.2072727272727273e-05,
      "loss": 0.122526153922081,
      "num_input_tokens_seen": 1910396,
      "step": 1859,
      "train_runtime": 4454.6878,
      "train_tokens_per_second": 428.851
    },
    {
      "epoch": 0.7949567261459558,
      "grad_norm": 2.9008889198303223,
      "learning_rate": 1.2068449197860962e-05,
      "loss": 0.08777370303869247,
      "num_input_tokens_seen": 1911126,
      "step": 1860,
      "train_runtime": 4456.7003,
      "train_tokens_per_second": 428.821
    },
    {
      "epoch": 0.7953841222352815,
      "grad_norm": 3.3785629272460938,
      "learning_rate": 1.2064171122994652e-05,
      "loss": 0.10935518145561218,
      "num_input_tokens_seen": 1911846,
      "step": 1861,
      "train_runtime": 4465.2868,
      "train_tokens_per_second": 428.157
    },
    {
      "epoch": 0.7958115183246073,
      "grad_norm": 2.525773763656616,
      "learning_rate": 1.2059893048128343e-05,
      "loss": 0.12208137661218643,
      "num_input_tokens_seen": 1913048,
      "step": 1862,
      "train_runtime": 4467.4048,
      "train_tokens_per_second": 428.224
    },
    {
      "epoch": 0.7962389144139331,
      "grad_norm": 2.3992831707000732,
      "learning_rate": 1.2055614973262032e-05,
      "loss": 0.10732989013195038,
      "num_input_tokens_seen": 1914338,
      "step": 1863,
      "train_runtime": 4469.5447,
      "train_tokens_per_second": 428.307
    },
    {
      "epoch": 0.7966663105032589,
      "grad_norm": 3.6788501739501953,
      "learning_rate": 1.2051336898395722e-05,
      "loss": 0.18382295966148376,
      "num_input_tokens_seen": 1915156,
      "step": 1864,
      "train_runtime": 4471.5983,
      "train_tokens_per_second": 428.293
    },
    {
      "epoch": 0.7970937065925847,
      "grad_norm": 2.580742120742798,
      "learning_rate": 1.2047058823529411e-05,
      "loss": 0.2171119600534439,
      "num_input_tokens_seen": 1916360,
      "step": 1865,
      "train_runtime": 4473.714,
      "train_tokens_per_second": 428.36
    },
    {
      "epoch": 0.7975211026819105,
      "grad_norm": 2.459043502807617,
      "learning_rate": 1.2042780748663102e-05,
      "loss": 0.14649857580661774,
      "num_input_tokens_seen": 1917328,
      "step": 1866,
      "train_runtime": 4475.7737,
      "train_tokens_per_second": 428.379
    },
    {
      "epoch": 0.7979484987712362,
      "grad_norm": 2.722545862197876,
      "learning_rate": 1.2038502673796792e-05,
      "loss": 0.19640393555164337,
      "num_input_tokens_seen": 1918666,
      "step": 1867,
      "train_runtime": 4477.9542,
      "train_tokens_per_second": 428.469
    },
    {
      "epoch": 0.798375894860562,
      "grad_norm": 2.932243824005127,
      "learning_rate": 1.2034224598930481e-05,
      "loss": 0.17902055382728577,
      "num_input_tokens_seen": 1919910,
      "step": 1868,
      "train_runtime": 4480.0987,
      "train_tokens_per_second": 428.542
    },
    {
      "epoch": 0.7988032909498878,
      "grad_norm": 2.163515567779541,
      "learning_rate": 1.202994652406417e-05,
      "loss": 0.12526603043079376,
      "num_input_tokens_seen": 1921020,
      "step": 1869,
      "train_runtime": 4482.1928,
      "train_tokens_per_second": 428.589
    },
    {
      "epoch": 0.7992306870392136,
      "grad_norm": 2.9243991374969482,
      "learning_rate": 1.2025668449197862e-05,
      "loss": 0.12168437987565994,
      "num_input_tokens_seen": 1922298,
      "step": 1870,
      "train_runtime": 4484.3249,
      "train_tokens_per_second": 428.671
    },
    {
      "epoch": 0.7996580831285394,
      "grad_norm": 3.3715198040008545,
      "learning_rate": 1.2021390374331551e-05,
      "loss": 0.19191522896289825,
      "num_input_tokens_seen": 1923092,
      "step": 1871,
      "train_runtime": 4486.3351,
      "train_tokens_per_second": 428.655
    },
    {
      "epoch": 0.8000854792178651,
      "grad_norm": 4.176147937774658,
      "learning_rate": 1.201711229946524e-05,
      "loss": 0.10397354513406754,
      "num_input_tokens_seen": 1923818,
      "step": 1872,
      "train_runtime": 4488.3279,
      "train_tokens_per_second": 428.627
    },
    {
      "epoch": 0.8000854792178651,
      "eval_loss": 0.5246274471282959,
      "eval_runtime": 58.6395,
      "eval_samples_per_second": 17.019,
      "eval_steps_per_second": 8.51,
      "num_input_tokens_seen": 1923818,
      "step": 1872
    },
    {
      "epoch": 0.800512875307191,
      "grad_norm": 2.4083259105682373,
      "learning_rate": 1.201283422459893e-05,
      "loss": 0.11328636109828949,
      "num_input_tokens_seen": 1924722,
      "step": 1873,
      "train_runtime": 4549.0126,
      "train_tokens_per_second": 423.108
    },
    {
      "epoch": 0.8009402713965167,
      "grad_norm": 3.093846559524536,
      "learning_rate": 1.2008556149732621e-05,
      "loss": 0.11408796906471252,
      "num_input_tokens_seen": 1925368,
      "step": 1874,
      "train_runtime": 4551.0795,
      "train_tokens_per_second": 423.057
    },
    {
      "epoch": 0.8013676674858425,
      "grad_norm": 4.824093818664551,
      "learning_rate": 1.2004278074866311e-05,
      "loss": 0.20181873440742493,
      "num_input_tokens_seen": 1925972,
      "step": 1875,
      "train_runtime": 4553.0886,
      "train_tokens_per_second": 423.003
    },
    {
      "epoch": 0.8017950635751683,
      "grad_norm": 2.236607789993286,
      "learning_rate": 1.2e-05,
      "loss": 0.09922008216381073,
      "num_input_tokens_seen": 1927012,
      "step": 1876,
      "train_runtime": 4555.1625,
      "train_tokens_per_second": 423.039
    },
    {
      "epoch": 0.8022224596644941,
      "grad_norm": 2.522822618484497,
      "learning_rate": 1.199572192513369e-05,
      "loss": 0.1273057609796524,
      "num_input_tokens_seen": 1928134,
      "step": 1877,
      "train_runtime": 4557.2659,
      "train_tokens_per_second": 423.09
    },
    {
      "epoch": 0.8026498557538199,
      "grad_norm": 2.0603935718536377,
      "learning_rate": 1.1991443850267381e-05,
      "loss": 0.08036594837903976,
      "num_input_tokens_seen": 1928898,
      "step": 1878,
      "train_runtime": 4559.277,
      "train_tokens_per_second": 423.071
    },
    {
      "epoch": 0.8030772518431456,
      "grad_norm": 1.789493203163147,
      "learning_rate": 1.198716577540107e-05,
      "loss": 0.05749205872416496,
      "num_input_tokens_seen": 1930108,
      "step": 1879,
      "train_runtime": 4561.3809,
      "train_tokens_per_second": 423.141
    },
    {
      "epoch": 0.8035046479324714,
      "grad_norm": 2.3337302207946777,
      "learning_rate": 1.198288770053476e-05,
      "loss": 0.10587199032306671,
      "num_input_tokens_seen": 1931350,
      "step": 1880,
      "train_runtime": 4563.5158,
      "train_tokens_per_second": 423.215
    },
    {
      "epoch": 0.8039320440217972,
      "grad_norm": 2.002426862716675,
      "learning_rate": 1.197860962566845e-05,
      "loss": 0.0680728554725647,
      "num_input_tokens_seen": 1932190,
      "step": 1881,
      "train_runtime": 4565.5545,
      "train_tokens_per_second": 423.21
    },
    {
      "epoch": 0.804359440111123,
      "grad_norm": 3.0435385704040527,
      "learning_rate": 1.197433155080214e-05,
      "loss": 0.2086118906736374,
      "num_input_tokens_seen": 1933088,
      "step": 1882,
      "train_runtime": 4567.6067,
      "train_tokens_per_second": 423.217
    },
    {
      "epoch": 0.8047868362004488,
      "grad_norm": 2.4492571353912354,
      "learning_rate": 1.197005347593583e-05,
      "loss": 0.2437245100736618,
      "num_input_tokens_seen": 1935796,
      "step": 1883,
      "train_runtime": 4570.0887,
      "train_tokens_per_second": 423.58
    },
    {
      "epoch": 0.8052142322897745,
      "grad_norm": 3.086801767349243,
      "learning_rate": 1.196577540106952e-05,
      "loss": 0.15454979240894318,
      "num_input_tokens_seen": 1936758,
      "step": 1884,
      "train_runtime": 4572.1648,
      "train_tokens_per_second": 423.598
    },
    {
      "epoch": 0.8056416283791004,
      "grad_norm": 5.520110607147217,
      "learning_rate": 1.1961497326203209e-05,
      "loss": 0.206394761800766,
      "num_input_tokens_seen": 1937432,
      "step": 1885,
      "train_runtime": 4574.1513,
      "train_tokens_per_second": 423.561
    },
    {
      "epoch": 0.8060690244684261,
      "grad_norm": 3.722134590148926,
      "learning_rate": 1.19572192513369e-05,
      "loss": 0.1293383240699768,
      "num_input_tokens_seen": 1938202,
      "step": 1886,
      "train_runtime": 4576.1935,
      "train_tokens_per_second": 423.54
    },
    {
      "epoch": 0.806496420557752,
      "grad_norm": 2.78294038772583,
      "learning_rate": 1.195294117647059e-05,
      "loss": 0.12537311017513275,
      "num_input_tokens_seen": 1939054,
      "step": 1887,
      "train_runtime": 4578.2133,
      "train_tokens_per_second": 423.539
    },
    {
      "epoch": 0.8069238166470777,
      "grad_norm": 2.07055926322937,
      "learning_rate": 1.1948663101604279e-05,
      "loss": 0.10585374385118484,
      "num_input_tokens_seen": 1940562,
      "step": 1888,
      "train_runtime": 4580.4274,
      "train_tokens_per_second": 423.664
    },
    {
      "epoch": 0.8073512127364034,
      "grad_norm": 2.1547939777374268,
      "learning_rate": 1.1944385026737968e-05,
      "loss": 0.10604503750801086,
      "num_input_tokens_seen": 1941642,
      "step": 1889,
      "train_runtime": 4582.5527,
      "train_tokens_per_second": 423.703
    },
    {
      "epoch": 0.8077786088257293,
      "grad_norm": 2.444694757461548,
      "learning_rate": 1.194010695187166e-05,
      "loss": 0.12567241489887238,
      "num_input_tokens_seen": 1942750,
      "step": 1890,
      "train_runtime": 4584.6356,
      "train_tokens_per_second": 423.752
    },
    {
      "epoch": 0.808206004915055,
      "grad_norm": 2.3113696575164795,
      "learning_rate": 1.1935828877005349e-05,
      "loss": 0.09493012726306915,
      "num_input_tokens_seen": 1943888,
      "step": 1891,
      "train_runtime": 4593.283,
      "train_tokens_per_second": 423.202
    },
    {
      "epoch": 0.8086334010043809,
      "grad_norm": 3.0792691707611084,
      "learning_rate": 1.1931550802139038e-05,
      "loss": 0.09735411405563354,
      "num_input_tokens_seen": 1944556,
      "step": 1892,
      "train_runtime": 4595.3465,
      "train_tokens_per_second": 423.158
    },
    {
      "epoch": 0.8090607970937066,
      "grad_norm": 4.6262993812561035,
      "learning_rate": 1.1927272727272728e-05,
      "loss": 0.11058700829744339,
      "num_input_tokens_seen": 1945294,
      "step": 1893,
      "train_runtime": 4597.3755,
      "train_tokens_per_second": 423.131
    },
    {
      "epoch": 0.8094881931830323,
      "grad_norm": 3.428767442703247,
      "learning_rate": 1.1922994652406419e-05,
      "loss": 0.18326550722122192,
      "num_input_tokens_seen": 1946162,
      "step": 1894,
      "train_runtime": 4599.4886,
      "train_tokens_per_second": 423.126
    },
    {
      "epoch": 0.8099155892723582,
      "grad_norm": 3.5264241695404053,
      "learning_rate": 1.1918716577540108e-05,
      "loss": 0.1430865377187729,
      "num_input_tokens_seen": 1946952,
      "step": 1895,
      "train_runtime": 4601.5438,
      "train_tokens_per_second": 423.108
    },
    {
      "epoch": 0.8103429853616839,
      "grad_norm": 2.65175199508667,
      "learning_rate": 1.1914438502673798e-05,
      "loss": 0.11213567852973938,
      "num_input_tokens_seen": 1947842,
      "step": 1896,
      "train_runtime": 4603.6595,
      "train_tokens_per_second": 423.107
    },
    {
      "epoch": 0.8107703814510098,
      "grad_norm": 4.137303829193115,
      "learning_rate": 1.1910160427807487e-05,
      "loss": 0.2591896951198578,
      "num_input_tokens_seen": 1948972,
      "step": 1897,
      "train_runtime": 4605.8265,
      "train_tokens_per_second": 423.154
    },
    {
      "epoch": 0.8111977775403355,
      "grad_norm": 3.0623655319213867,
      "learning_rate": 1.1905882352941178e-05,
      "loss": 0.09301309287548065,
      "num_input_tokens_seen": 1949854,
      "step": 1898,
      "train_runtime": 4607.8824,
      "train_tokens_per_second": 423.156
    },
    {
      "epoch": 0.8116251736296612,
      "grad_norm": 3.4153971672058105,
      "learning_rate": 1.1901604278074868e-05,
      "loss": 0.15868785977363586,
      "num_input_tokens_seen": 1950624,
      "step": 1899,
      "train_runtime": 4609.9038,
      "train_tokens_per_second": 423.138
    },
    {
      "epoch": 0.8120525697189871,
      "grad_norm": 2.8564441204071045,
      "learning_rate": 1.1897326203208557e-05,
      "loss": 0.13919787108898163,
      "num_input_tokens_seen": 1951642,
      "step": 1900,
      "train_runtime": 4612.017,
      "train_tokens_per_second": 423.165
    },
    {
      "epoch": 0.8124799658083128,
      "grad_norm": 2.3190674781799316,
      "learning_rate": 1.1893048128342247e-05,
      "loss": 0.08593349158763885,
      "num_input_tokens_seen": 1952558,
      "step": 1901,
      "train_runtime": 4614.1222,
      "train_tokens_per_second": 423.17
    },
    {
      "epoch": 0.8129073618976387,
      "grad_norm": 2.6398873329162598,
      "learning_rate": 1.1888770053475938e-05,
      "loss": 0.16586089134216309,
      "num_input_tokens_seen": 1953702,
      "step": 1902,
      "train_runtime": 4616.3225,
      "train_tokens_per_second": 423.216
    },
    {
      "epoch": 0.8133347579869644,
      "grad_norm": 4.039870738983154,
      "learning_rate": 1.1884491978609627e-05,
      "loss": 0.11316657066345215,
      "num_input_tokens_seen": 1954534,
      "step": 1903,
      "train_runtime": 4618.3678,
      "train_tokens_per_second": 423.209
    },
    {
      "epoch": 0.8137621540762902,
      "grad_norm": 4.335970401763916,
      "learning_rate": 1.1880213903743317e-05,
      "loss": 0.11746223270893097,
      "num_input_tokens_seen": 1955774,
      "step": 1904,
      "train_runtime": 4620.5094,
      "train_tokens_per_second": 423.281
    },
    {
      "epoch": 0.814189550165616,
      "grad_norm": 2.930342674255371,
      "learning_rate": 1.1875935828877006e-05,
      "loss": 0.11023575812578201,
      "num_input_tokens_seen": 1956976,
      "step": 1905,
      "train_runtime": 4622.6954,
      "train_tokens_per_second": 423.341
    },
    {
      "epoch": 0.8146169462549417,
      "grad_norm": 3.6363377571105957,
      "learning_rate": 1.1871657754010697e-05,
      "loss": 0.1983751803636551,
      "num_input_tokens_seen": 1957994,
      "step": 1906,
      "train_runtime": 4624.7768,
      "train_tokens_per_second": 423.37
    },
    {
      "epoch": 0.8150443423442676,
      "grad_norm": 3.64961314201355,
      "learning_rate": 1.1867379679144387e-05,
      "loss": 0.17176346480846405,
      "num_input_tokens_seen": 1958802,
      "step": 1907,
      "train_runtime": 4626.8284,
      "train_tokens_per_second": 423.357
    },
    {
      "epoch": 0.8154717384335933,
      "grad_norm": 2.234658718109131,
      "learning_rate": 1.1863101604278076e-05,
      "loss": 0.10892599821090698,
      "num_input_tokens_seen": 1960204,
      "step": 1908,
      "train_runtime": 4629.0211,
      "train_tokens_per_second": 423.46
    },
    {
      "epoch": 0.8158991345229191,
      "grad_norm": 2.126518487930298,
      "learning_rate": 1.1858823529411766e-05,
      "loss": 0.15093612670898438,
      "num_input_tokens_seen": 1961966,
      "step": 1909,
      "train_runtime": 4631.2558,
      "train_tokens_per_second": 423.636
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 2.4514105319976807,
      "learning_rate": 1.1854545454545457e-05,
      "loss": 0.1282510757446289,
      "num_input_tokens_seen": 1963052,
      "step": 1910,
      "train_runtime": 4633.373,
      "train_tokens_per_second": 423.677
    },
    {
      "epoch": 0.8167539267015707,
      "grad_norm": 4.3745646476745605,
      "learning_rate": 1.1850267379679146e-05,
      "loss": 0.2642311751842499,
      "num_input_tokens_seen": 1963950,
      "step": 1911,
      "train_runtime": 4635.4425,
      "train_tokens_per_second": 423.681
    },
    {
      "epoch": 0.8171813227908965,
      "grad_norm": 2.5366644859313965,
      "learning_rate": 1.1845989304812836e-05,
      "loss": 0.11905498802661896,
      "num_input_tokens_seen": 1964986,
      "step": 1912,
      "train_runtime": 4637.5606,
      "train_tokens_per_second": 423.711
    },
    {
      "epoch": 0.8176087188802222,
      "grad_norm": 3.320986032485962,
      "learning_rate": 1.1841711229946525e-05,
      "loss": 0.1812538504600525,
      "num_input_tokens_seen": 1966226,
      "step": 1913,
      "train_runtime": 4639.718,
      "train_tokens_per_second": 423.781
    },
    {
      "epoch": 0.818036114969548,
      "grad_norm": 4.496185302734375,
      "learning_rate": 1.1837433155080216e-05,
      "loss": 0.21242013573646545,
      "num_input_tokens_seen": 1967028,
      "step": 1914,
      "train_runtime": 4641.7069,
      "train_tokens_per_second": 423.773
    },
    {
      "epoch": 0.8184635110588738,
      "grad_norm": 3.0694077014923096,
      "learning_rate": 1.1833155080213906e-05,
      "loss": 0.15557079017162323,
      "num_input_tokens_seen": 1967974,
      "step": 1915,
      "train_runtime": 4643.8124,
      "train_tokens_per_second": 423.784
    },
    {
      "epoch": 0.8188909071481996,
      "grad_norm": 4.743383407592773,
      "learning_rate": 1.1828877005347595e-05,
      "loss": 0.18137595057487488,
      "num_input_tokens_seen": 1968914,
      "step": 1916,
      "train_runtime": 4645.8921,
      "train_tokens_per_second": 423.797
    },
    {
      "epoch": 0.8193183032375254,
      "grad_norm": 3.5697715282440186,
      "learning_rate": 1.1824598930481283e-05,
      "loss": 0.14916926622390747,
      "num_input_tokens_seen": 1969620,
      "step": 1917,
      "train_runtime": 4647.8906,
      "train_tokens_per_second": 423.766
    },
    {
      "epoch": 0.8197456993268512,
      "grad_norm": 2.30627703666687,
      "learning_rate": 1.1820320855614976e-05,
      "loss": 0.14872755110263824,
      "num_input_tokens_seen": 1970726,
      "step": 1918,
      "train_runtime": 4650.0242,
      "train_tokens_per_second": 423.81
    },
    {
      "epoch": 0.8201730954161769,
      "grad_norm": 2.686779022216797,
      "learning_rate": 1.1816042780748665e-05,
      "loss": 0.08790980279445648,
      "num_input_tokens_seen": 1971684,
      "step": 1919,
      "train_runtime": 4652.1086,
      "train_tokens_per_second": 423.826
    },
    {
      "epoch": 0.8206004915055027,
      "grad_norm": 2.892587423324585,
      "learning_rate": 1.1811764705882353e-05,
      "loss": 0.15730991959571838,
      "num_input_tokens_seen": 1972462,
      "step": 1920,
      "train_runtime": 4654.1287,
      "train_tokens_per_second": 423.809
    },
    {
      "epoch": 0.8210278875948285,
      "grad_norm": 3.127499580383301,
      "learning_rate": 1.1807486631016042e-05,
      "loss": 0.13750679790973663,
      "num_input_tokens_seen": 1973296,
      "step": 1921,
      "train_runtime": 4662.8092,
      "train_tokens_per_second": 423.199
    },
    {
      "epoch": 0.8214552836841543,
      "grad_norm": 4.397209167480469,
      "learning_rate": 1.1803208556149735e-05,
      "loss": 0.2044524848461151,
      "num_input_tokens_seen": 1974076,
      "step": 1922,
      "train_runtime": 4665.0253,
      "train_tokens_per_second": 423.165
    },
    {
      "epoch": 0.8218826797734801,
      "grad_norm": 2.791097640991211,
      "learning_rate": 1.1798930481283425e-05,
      "loss": 0.09547609090805054,
      "num_input_tokens_seen": 1975028,
      "step": 1923,
      "train_runtime": 4667.0743,
      "train_tokens_per_second": 423.183
    },
    {
      "epoch": 0.8223100758628058,
      "grad_norm": 3.1614551544189453,
      "learning_rate": 1.1794652406417112e-05,
      "loss": 0.18947777152061462,
      "num_input_tokens_seen": 1975864,
      "step": 1924,
      "train_runtime": 4669.1105,
      "train_tokens_per_second": 423.178
    },
    {
      "epoch": 0.8227374719521316,
      "grad_norm": 3.162884473800659,
      "learning_rate": 1.1790374331550802e-05,
      "loss": 0.13253599405288696,
      "num_input_tokens_seen": 1976524,
      "step": 1925,
      "train_runtime": 4671.1638,
      "train_tokens_per_second": 423.133
    },
    {
      "epoch": 0.8231648680414574,
      "grad_norm": 4.1589884757995605,
      "learning_rate": 1.1786096256684495e-05,
      "loss": 0.22754980623722076,
      "num_input_tokens_seen": 1977354,
      "step": 1926,
      "train_runtime": 4673.3025,
      "train_tokens_per_second": 423.117
    },
    {
      "epoch": 0.8235922641307832,
      "grad_norm": 2.2708146572113037,
      "learning_rate": 1.1781818181818182e-05,
      "loss": 0.14771157503128052,
      "num_input_tokens_seen": 1978830,
      "step": 1927,
      "train_runtime": 4675.5408,
      "train_tokens_per_second": 423.23
    },
    {
      "epoch": 0.824019660220109,
      "grad_norm": 3.8743412494659424,
      "learning_rate": 1.1777540106951872e-05,
      "loss": 0.12557849287986755,
      "num_input_tokens_seen": 1979576,
      "step": 1928,
      "train_runtime": 4677.563,
      "train_tokens_per_second": 423.207
    },
    {
      "epoch": 0.8244470563094348,
      "grad_norm": 2.3909263610839844,
      "learning_rate": 1.1773262032085561e-05,
      "loss": 0.16569218039512634,
      "num_input_tokens_seen": 1980932,
      "step": 1929,
      "train_runtime": 4679.7242,
      "train_tokens_per_second": 423.301
    },
    {
      "epoch": 0.8248744523987606,
      "grad_norm": 1.8543733358383179,
      "learning_rate": 1.1768983957219254e-05,
      "loss": 0.10120807588100433,
      "num_input_tokens_seen": 1982492,
      "step": 1930,
      "train_runtime": 4681.9595,
      "train_tokens_per_second": 423.432
    },
    {
      "epoch": 0.8253018484880863,
      "grad_norm": 3.636075258255005,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.22500087320804596,
      "num_input_tokens_seen": 1983864,
      "step": 1931,
      "train_runtime": 4684.1361,
      "train_tokens_per_second": 423.528
    },
    {
      "epoch": 0.8257292445774121,
      "grad_norm": 2.86460018157959,
      "learning_rate": 1.1760427807486631e-05,
      "loss": 0.157918319106102,
      "num_input_tokens_seen": 1985584,
      "step": 1932,
      "train_runtime": 4686.3985,
      "train_tokens_per_second": 423.691
    },
    {
      "epoch": 0.8261566406667379,
      "grad_norm": 3.078482151031494,
      "learning_rate": 1.175614973262032e-05,
      "loss": 0.15079180896282196,
      "num_input_tokens_seen": 1986526,
      "step": 1933,
      "train_runtime": 4688.447,
      "train_tokens_per_second": 423.707
    },
    {
      "epoch": 0.8265840367560637,
      "grad_norm": 2.689268112182617,
      "learning_rate": 1.1751871657754012e-05,
      "loss": 0.10432767122983932,
      "num_input_tokens_seen": 1987252,
      "step": 1934,
      "train_runtime": 4690.445,
      "train_tokens_per_second": 423.681
    },
    {
      "epoch": 0.8270114328453895,
      "grad_norm": 1.8064614534378052,
      "learning_rate": 1.1747593582887701e-05,
      "loss": 0.04677347093820572,
      "num_input_tokens_seen": 1988562,
      "step": 1935,
      "train_runtime": 4692.5866,
      "train_tokens_per_second": 423.767
    },
    {
      "epoch": 0.8274388289347152,
      "grad_norm": 3.0575437545776367,
      "learning_rate": 1.174331550802139e-05,
      "loss": 0.21622297167778015,
      "num_input_tokens_seen": 1989806,
      "step": 1936,
      "train_runtime": 4694.7407,
      "train_tokens_per_second": 423.837
    },
    {
      "epoch": 0.8278662250240411,
      "grad_norm": 2.43487811088562,
      "learning_rate": 1.173903743315508e-05,
      "loss": 0.10695505142211914,
      "num_input_tokens_seen": 1990762,
      "step": 1937,
      "train_runtime": 4696.8247,
      "train_tokens_per_second": 423.853
    },
    {
      "epoch": 0.8282936211133668,
      "grad_norm": 2.327991247177124,
      "learning_rate": 1.1734759358288771e-05,
      "loss": 0.13072600960731506,
      "num_input_tokens_seen": 1991992,
      "step": 1938,
      "train_runtime": 4698.9745,
      "train_tokens_per_second": 423.921
    },
    {
      "epoch": 0.8287210172026926,
      "grad_norm": 3.2440786361694336,
      "learning_rate": 1.173048128342246e-05,
      "loss": 0.2048959732055664,
      "num_input_tokens_seen": 1993564,
      "step": 1939,
      "train_runtime": 4701.183,
      "train_tokens_per_second": 424.056
    },
    {
      "epoch": 0.8291484132920184,
      "grad_norm": 3.3145008087158203,
      "learning_rate": 1.172620320855615e-05,
      "loss": 0.1536371111869812,
      "num_input_tokens_seen": 1994270,
      "step": 1940,
      "train_runtime": 4703.1762,
      "train_tokens_per_second": 424.026
    },
    {
      "epoch": 0.8295758093813441,
      "grad_norm": 4.623020648956299,
      "learning_rate": 1.172192513368984e-05,
      "loss": 0.19404183328151703,
      "num_input_tokens_seen": 1995102,
      "step": 1941,
      "train_runtime": 4705.2492,
      "train_tokens_per_second": 424.016
    },
    {
      "epoch": 0.83000320547067,
      "grad_norm": 2.123108148574829,
      "learning_rate": 1.171764705882353e-05,
      "loss": 0.06709779798984528,
      "num_input_tokens_seen": 1996254,
      "step": 1942,
      "train_runtime": 4707.3668,
      "train_tokens_per_second": 424.07
    },
    {
      "epoch": 0.8304306015599957,
      "grad_norm": 3.1344902515411377,
      "learning_rate": 1.171336898395722e-05,
      "loss": 0.15405476093292236,
      "num_input_tokens_seen": 1997454,
      "step": 1943,
      "train_runtime": 4709.4949,
      "train_tokens_per_second": 424.133
    },
    {
      "epoch": 0.8308579976493216,
      "grad_norm": 2.6773366928100586,
      "learning_rate": 1.170909090909091e-05,
      "loss": 0.08066906034946442,
      "num_input_tokens_seen": 1998452,
      "step": 1944,
      "train_runtime": 4711.595,
      "train_tokens_per_second": 424.156
    },
    {
      "epoch": 0.8312853937386473,
      "grad_norm": 2.34963059425354,
      "learning_rate": 1.17048128342246e-05,
      "loss": 0.11968742311000824,
      "num_input_tokens_seen": 1999622,
      "step": 1945,
      "train_runtime": 4713.7296,
      "train_tokens_per_second": 424.212
    },
    {
      "epoch": 0.831712789827973,
      "grad_norm": 3.0777735710144043,
      "learning_rate": 1.170053475935829e-05,
      "loss": 0.13329118490219116,
      "num_input_tokens_seen": 2000404,
      "step": 1946,
      "train_runtime": 4715.748,
      "train_tokens_per_second": 424.197
    },
    {
      "epoch": 0.8321401859172989,
      "grad_norm": 2.622476100921631,
      "learning_rate": 1.169625668449198e-05,
      "loss": 0.14818327128887177,
      "num_input_tokens_seen": 2001258,
      "step": 1947,
      "train_runtime": 4717.8392,
      "train_tokens_per_second": 424.19
    },
    {
      "epoch": 0.8325675820066246,
      "grad_norm": 3.720508098602295,
      "learning_rate": 1.169197860962567e-05,
      "loss": 0.273777037858963,
      "num_input_tokens_seen": 2002180,
      "step": 1948,
      "train_runtime": 4719.8883,
      "train_tokens_per_second": 424.201
    },
    {
      "epoch": 0.8329949780959505,
      "grad_norm": 2.41963529586792,
      "learning_rate": 1.1687700534759359e-05,
      "loss": 0.13610133528709412,
      "num_input_tokens_seen": 2003156,
      "step": 1949,
      "train_runtime": 4721.9779,
      "train_tokens_per_second": 424.22
    },
    {
      "epoch": 0.8334223741852762,
      "grad_norm": 3.142697334289551,
      "learning_rate": 1.168342245989305e-05,
      "loss": 0.09877709299325943,
      "num_input_tokens_seen": 2003902,
      "step": 1950,
      "train_runtime": 4723.9611,
      "train_tokens_per_second": 424.2
    },
    {
      "epoch": 0.8338497702746019,
      "grad_norm": 2.2059855461120605,
      "learning_rate": 1.167914438502674e-05,
      "loss": 0.09511911869049072,
      "num_input_tokens_seen": 2004806,
      "step": 1951,
      "train_runtime": 4732.2731,
      "train_tokens_per_second": 423.645
    },
    {
      "epoch": 0.8342771663639278,
      "grad_norm": 3.3337655067443848,
      "learning_rate": 1.1674866310160429e-05,
      "loss": 0.16686686873435974,
      "num_input_tokens_seen": 2005616,
      "step": 1952,
      "train_runtime": 4734.3338,
      "train_tokens_per_second": 423.632
    },
    {
      "epoch": 0.8347045624532535,
      "grad_norm": 2.2215652465820312,
      "learning_rate": 1.1670588235294118e-05,
      "loss": 0.08955597877502441,
      "num_input_tokens_seen": 2006428,
      "step": 1953,
      "train_runtime": 4736.5735,
      "train_tokens_per_second": 423.603
    },
    {
      "epoch": 0.8351319585425794,
      "grad_norm": 2.384791135787964,
      "learning_rate": 1.166631016042781e-05,
      "loss": 0.11258506029844284,
      "num_input_tokens_seen": 2007744,
      "step": 1954,
      "train_runtime": 4738.7149,
      "train_tokens_per_second": 423.69
    },
    {
      "epoch": 0.8355593546319051,
      "grad_norm": 2.0092225074768066,
      "learning_rate": 1.1662032085561499e-05,
      "loss": 0.07927387207746506,
      "num_input_tokens_seen": 2008740,
      "step": 1955,
      "train_runtime": 4740.7597,
      "train_tokens_per_second": 423.717
    },
    {
      "epoch": 0.8359867507212309,
      "grad_norm": 4.950786113739014,
      "learning_rate": 1.1657754010695188e-05,
      "loss": 0.32466092705726624,
      "num_input_tokens_seen": 2009954,
      "step": 1956,
      "train_runtime": 4742.8656,
      "train_tokens_per_second": 423.785
    },
    {
      "epoch": 0.8364141468105567,
      "grad_norm": 2.8864147663116455,
      "learning_rate": 1.1653475935828878e-05,
      "loss": 0.12924933433532715,
      "num_input_tokens_seen": 2010976,
      "step": 1957,
      "train_runtime": 4744.9719,
      "train_tokens_per_second": 423.812
    },
    {
      "epoch": 0.8368415428998824,
      "grad_norm": 3.988420248031616,
      "learning_rate": 1.1649197860962569e-05,
      "loss": 0.21299467980861664,
      "num_input_tokens_seen": 2011758,
      "step": 1958,
      "train_runtime": 4746.9923,
      "train_tokens_per_second": 423.796
    },
    {
      "epoch": 0.8372689389892083,
      "grad_norm": 2.931626558303833,
      "learning_rate": 1.1644919786096258e-05,
      "loss": 0.14606085419654846,
      "num_input_tokens_seen": 2012714,
      "step": 1959,
      "train_runtime": 4749.0372,
      "train_tokens_per_second": 423.815
    },
    {
      "epoch": 0.837696335078534,
      "grad_norm": 2.99975323677063,
      "learning_rate": 1.1640641711229948e-05,
      "loss": 0.15143752098083496,
      "num_input_tokens_seen": 2013502,
      "step": 1960,
      "train_runtime": 4751.1021,
      "train_tokens_per_second": 423.797
    },
    {
      "epoch": 0.8381237311678598,
      "grad_norm": 5.287333965301514,
      "learning_rate": 1.1636363636363637e-05,
      "loss": 0.26975440979003906,
      "num_input_tokens_seen": 2014448,
      "step": 1961,
      "train_runtime": 4753.2108,
      "train_tokens_per_second": 423.808
    },
    {
      "epoch": 0.8385511272571856,
      "grad_norm": 1.9712953567504883,
      "learning_rate": 1.1632085561497328e-05,
      "loss": 0.11496014893054962,
      "num_input_tokens_seen": 2016488,
      "step": 1962,
      "train_runtime": 4755.568,
      "train_tokens_per_second": 424.027
    },
    {
      "epoch": 0.8389785233465114,
      "grad_norm": 2.6900217533111572,
      "learning_rate": 1.1627807486631018e-05,
      "loss": 0.1549610048532486,
      "num_input_tokens_seen": 2017248,
      "step": 1963,
      "train_runtime": 4757.6226,
      "train_tokens_per_second": 424.003
    },
    {
      "epoch": 0.8394059194358372,
      "grad_norm": 3.142636299133301,
      "learning_rate": 1.1623529411764707e-05,
      "loss": 0.07481295615434647,
      "num_input_tokens_seen": 2018118,
      "step": 1964,
      "train_runtime": 4759.6754,
      "train_tokens_per_second": 424.003
    },
    {
      "epoch": 0.8398333155251629,
      "grad_norm": 3.3055038452148438,
      "learning_rate": 1.1619251336898397e-05,
      "loss": 0.14716248214244843,
      "num_input_tokens_seen": 2019514,
      "step": 1965,
      "train_runtime": 4761.8196,
      "train_tokens_per_second": 424.106
    },
    {
      "epoch": 0.8402607116144887,
      "grad_norm": 2.9134085178375244,
      "learning_rate": 1.1614973262032086e-05,
      "loss": 0.1696421504020691,
      "num_input_tokens_seen": 2020222,
      "step": 1966,
      "train_runtime": 4763.8251,
      "train_tokens_per_second": 424.076
    },
    {
      "epoch": 0.8406881077038145,
      "grad_norm": 3.4489259719848633,
      "learning_rate": 1.1610695187165777e-05,
      "loss": 0.20588867366313934,
      "num_input_tokens_seen": 2021242,
      "step": 1967,
      "train_runtime": 4765.9112,
      "train_tokens_per_second": 424.104
    },
    {
      "epoch": 0.8411155037931403,
      "grad_norm": 2.7396275997161865,
      "learning_rate": 1.1606417112299467e-05,
      "loss": 0.1684694141149521,
      "num_input_tokens_seen": 2022708,
      "step": 1968,
      "train_runtime": 4768.1883,
      "train_tokens_per_second": 424.209
    },
    {
      "epoch": 0.8415428998824661,
      "grad_norm": 2.6376407146453857,
      "learning_rate": 1.1602139037433156e-05,
      "loss": 0.09553737938404083,
      "num_input_tokens_seen": 2023718,
      "step": 1969,
      "train_runtime": 4770.3403,
      "train_tokens_per_second": 424.229
    },
    {
      "epoch": 0.8419702959717918,
      "grad_norm": 2.150681257247925,
      "learning_rate": 1.1597860962566845e-05,
      "loss": 0.1336728185415268,
      "num_input_tokens_seen": 2025154,
      "step": 1970,
      "train_runtime": 4772.5238,
      "train_tokens_per_second": 424.336
    },
    {
      "epoch": 0.8423976920611176,
      "grad_norm": 1.6997156143188477,
      "learning_rate": 1.1593582887700537e-05,
      "loss": 0.1043979823589325,
      "num_input_tokens_seen": 2026996,
      "step": 1971,
      "train_runtime": 4774.8418,
      "train_tokens_per_second": 424.516
    },
    {
      "epoch": 0.8428250881504434,
      "grad_norm": 3.709462881088257,
      "learning_rate": 1.1589304812834226e-05,
      "loss": 0.14273390173912048,
      "num_input_tokens_seen": 2027706,
      "step": 1972,
      "train_runtime": 4776.8233,
      "train_tokens_per_second": 424.488
    },
    {
      "epoch": 0.8432524842397692,
      "grad_norm": 2.8449738025665283,
      "learning_rate": 1.1585026737967915e-05,
      "loss": 0.1415601223707199,
      "num_input_tokens_seen": 2028652,
      "step": 1973,
      "train_runtime": 4778.9036,
      "train_tokens_per_second": 424.502
    },
    {
      "epoch": 0.843679880329095,
      "grad_norm": 2.695927619934082,
      "learning_rate": 1.1580748663101605e-05,
      "loss": 0.12807713449001312,
      "num_input_tokens_seen": 2029382,
      "step": 1974,
      "train_runtime": 4780.9296,
      "train_tokens_per_second": 424.474
    },
    {
      "epoch": 0.8441072764184208,
      "grad_norm": 3.0551578998565674,
      "learning_rate": 1.1576470588235296e-05,
      "loss": 0.11282316595315933,
      "num_input_tokens_seen": 2030350,
      "step": 1975,
      "train_runtime": 4783.0003,
      "train_tokens_per_second": 424.493
    },
    {
      "epoch": 0.8445346725077466,
      "grad_norm": 3.949636697769165,
      "learning_rate": 1.1572192513368986e-05,
      "loss": 0.19138802587985992,
      "num_input_tokens_seen": 2031440,
      "step": 1976,
      "train_runtime": 4785.1854,
      "train_tokens_per_second": 424.527
    },
    {
      "epoch": 0.8449620685970723,
      "grad_norm": 2.253563642501831,
      "learning_rate": 1.1567914438502675e-05,
      "loss": 0.10519030690193176,
      "num_input_tokens_seen": 2032804,
      "step": 1977,
      "train_runtime": 4787.339,
      "train_tokens_per_second": 424.621
    },
    {
      "epoch": 0.8453894646863981,
      "grad_norm": 2.4796438217163086,
      "learning_rate": 1.1563636363636363e-05,
      "loss": 0.07710590213537216,
      "num_input_tokens_seen": 2033780,
      "step": 1978,
      "train_runtime": 4789.4032,
      "train_tokens_per_second": 424.642
    },
    {
      "epoch": 0.8458168607757239,
      "grad_norm": 2.4658453464508057,
      "learning_rate": 1.1559358288770056e-05,
      "loss": 0.10524742305278778,
      "num_input_tokens_seen": 2034834,
      "step": 1979,
      "train_runtime": 4791.4836,
      "train_tokens_per_second": 424.677
    },
    {
      "epoch": 0.8462442568650497,
      "grad_norm": 1.6197055578231812,
      "learning_rate": 1.1555080213903745e-05,
      "loss": 0.08397623896598816,
      "num_input_tokens_seen": 2036680,
      "step": 1980,
      "train_runtime": 4793.7819,
      "train_tokens_per_second": 424.859
    },
    {
      "epoch": 0.8466716529543755,
      "grad_norm": 1.9097598791122437,
      "learning_rate": 1.1550802139037434e-05,
      "loss": 0.10334040224552155,
      "num_input_tokens_seen": 2039644,
      "step": 1981,
      "train_runtime": 4802.9741,
      "train_tokens_per_second": 424.663
    },
    {
      "epoch": 0.8470990490437013,
      "grad_norm": 2.65543794631958,
      "learning_rate": 1.1546524064171122e-05,
      "loss": 0.12254716455936432,
      "num_input_tokens_seen": 2040578,
      "step": 1982,
      "train_runtime": 4805.2388,
      "train_tokens_per_second": 424.657
    },
    {
      "epoch": 0.847526445133027,
      "grad_norm": 3.56805682182312,
      "learning_rate": 1.1542245989304815e-05,
      "loss": 0.21757085621356964,
      "num_input_tokens_seen": 2041474,
      "step": 1983,
      "train_runtime": 4807.4609,
      "train_tokens_per_second": 424.647
    },
    {
      "epoch": 0.8479538412223528,
      "grad_norm": 2.963594913482666,
      "learning_rate": 1.1537967914438504e-05,
      "loss": 0.10216448456048965,
      "num_input_tokens_seen": 2042394,
      "step": 1984,
      "train_runtime": 4809.5267,
      "train_tokens_per_second": 424.656
    },
    {
      "epoch": 0.8483812373116786,
      "grad_norm": 2.03479266166687,
      "learning_rate": 1.1533689839572192e-05,
      "loss": 0.07313203811645508,
      "num_input_tokens_seen": 2043502,
      "step": 1985,
      "train_runtime": 4811.6572,
      "train_tokens_per_second": 424.698
    },
    {
      "epoch": 0.8488086334010044,
      "grad_norm": 1.762342929840088,
      "learning_rate": 1.1529411764705882e-05,
      "loss": 0.0892074853181839,
      "num_input_tokens_seen": 2044812,
      "step": 1986,
      "train_runtime": 4813.7836,
      "train_tokens_per_second": 424.783
    },
    {
      "epoch": 0.8492360294903302,
      "grad_norm": 2.834467887878418,
      "learning_rate": 1.1525133689839575e-05,
      "loss": 0.1746521145105362,
      "num_input_tokens_seen": 2046186,
      "step": 1987,
      "train_runtime": 4816.0122,
      "train_tokens_per_second": 424.871
    },
    {
      "epoch": 0.8496634255796559,
      "grad_norm": 2.2896041870117188,
      "learning_rate": 1.1520855614973262e-05,
      "loss": 0.08996517211198807,
      "num_input_tokens_seen": 2047098,
      "step": 1988,
      "train_runtime": 4818.0914,
      "train_tokens_per_second": 424.877
    },
    {
      "epoch": 0.8500908216689818,
      "grad_norm": 3.9072623252868652,
      "learning_rate": 1.1516577540106952e-05,
      "loss": 0.15439929068088531,
      "num_input_tokens_seen": 2047874,
      "step": 1989,
      "train_runtime": 4820.1573,
      "train_tokens_per_second": 424.856
    },
    {
      "epoch": 0.8505182177583075,
      "grad_norm": 3.1842758655548096,
      "learning_rate": 1.1512299465240641e-05,
      "loss": 0.1472446620464325,
      "num_input_tokens_seen": 2048806,
      "step": 1990,
      "train_runtime": 4822.2376,
      "train_tokens_per_second": 424.866
    },
    {
      "epoch": 0.8509456138476333,
      "grad_norm": 3.044966697692871,
      "learning_rate": 1.1508021390374334e-05,
      "loss": 0.15248648822307587,
      "num_input_tokens_seen": 2049882,
      "step": 1991,
      "train_runtime": 4824.3746,
      "train_tokens_per_second": 424.901
    },
    {
      "epoch": 0.8513730099369591,
      "grad_norm": 4.340991497039795,
      "learning_rate": 1.1503743315508022e-05,
      "loss": 0.13964655995368958,
      "num_input_tokens_seen": 2050772,
      "step": 1992,
      "train_runtime": 4826.4903,
      "train_tokens_per_second": 424.899
    },
    {
      "epoch": 0.8518004060262848,
      "grad_norm": 2.7062253952026367,
      "learning_rate": 1.1499465240641711e-05,
      "loss": 0.13101620972156525,
      "num_input_tokens_seen": 2051666,
      "step": 1993,
      "train_runtime": 4828.5504,
      "train_tokens_per_second": 424.903
    },
    {
      "epoch": 0.8522278021156107,
      "grad_norm": 3.3317129611968994,
      "learning_rate": 1.14951871657754e-05,
      "loss": 0.09891689568758011,
      "num_input_tokens_seen": 2052266,
      "step": 1994,
      "train_runtime": 4830.5044,
      "train_tokens_per_second": 424.855
    },
    {
      "epoch": 0.8526551982049364,
      "grad_norm": 2.544534921646118,
      "learning_rate": 1.1490909090909092e-05,
      "loss": 0.14052549004554749,
      "num_input_tokens_seen": 2053472,
      "step": 1995,
      "train_runtime": 4832.6697,
      "train_tokens_per_second": 424.915
    },
    {
      "epoch": 0.8530825942942623,
      "grad_norm": 3.747269630432129,
      "learning_rate": 1.1486631016042781e-05,
      "loss": 0.16506534814834595,
      "num_input_tokens_seen": 2054212,
      "step": 1996,
      "train_runtime": 4834.7681,
      "train_tokens_per_second": 424.883
    },
    {
      "epoch": 0.853509990383588,
      "grad_norm": 3.537022113800049,
      "learning_rate": 1.148235294117647e-05,
      "loss": 0.19596555829048157,
      "num_input_tokens_seen": 2055748,
      "step": 1997,
      "train_runtime": 4837.0151,
      "train_tokens_per_second": 425.003
    },
    {
      "epoch": 0.8539373864729137,
      "grad_norm": 2.9901909828186035,
      "learning_rate": 1.147807486631016e-05,
      "loss": 0.10402274876832962,
      "num_input_tokens_seen": 2056428,
      "step": 1998,
      "train_runtime": 4839.0355,
      "train_tokens_per_second": 424.967
    },
    {
      "epoch": 0.8543647825622396,
      "grad_norm": 3.015026092529297,
      "learning_rate": 1.1473796791443851e-05,
      "loss": 0.12196237593889236,
      "num_input_tokens_seen": 2057386,
      "step": 1999,
      "train_runtime": 4841.1324,
      "train_tokens_per_second": 424.98
    },
    {
      "epoch": 0.8547921786515653,
      "grad_norm": 1.937814712524414,
      "learning_rate": 1.146951871657754e-05,
      "loss": 0.09265303611755371,
      "num_input_tokens_seen": 2058680,
      "step": 2000,
      "train_runtime": 4843.2601,
      "train_tokens_per_second": 425.061
    },
    {
      "epoch": 0.8552195747408912,
      "grad_norm": 1.2096171379089355,
      "learning_rate": 1.146524064171123e-05,
      "loss": 0.04078340530395508,
      "num_input_tokens_seen": 2060918,
      "step": 2001,
      "train_runtime": 4845.7047,
      "train_tokens_per_second": 425.308
    },
    {
      "epoch": 0.8556469708302169,
      "grad_norm": 2.6704554557800293,
      "learning_rate": 1.146096256684492e-05,
      "loss": 0.049375370144844055,
      "num_input_tokens_seen": 2061456,
      "step": 2002,
      "train_runtime": 4847.6653,
      "train_tokens_per_second": 425.247
    },
    {
      "epoch": 0.8560743669195426,
      "grad_norm": 3.931211233139038,
      "learning_rate": 1.145668449197861e-05,
      "loss": 0.13609907031059265,
      "num_input_tokens_seen": 2062118,
      "step": 2003,
      "train_runtime": 4849.6926,
      "train_tokens_per_second": 425.206
    },
    {
      "epoch": 0.8565017630088685,
      "grad_norm": 2.8013854026794434,
      "learning_rate": 1.14524064171123e-05,
      "loss": 0.19170740246772766,
      "num_input_tokens_seen": 2063172,
      "step": 2004,
      "train_runtime": 4851.7945,
      "train_tokens_per_second": 425.239
    },
    {
      "epoch": 0.8569291590981942,
      "grad_norm": 2.5239267349243164,
      "learning_rate": 1.144812834224599e-05,
      "loss": 0.10953666269779205,
      "num_input_tokens_seen": 2064226,
      "step": 2005,
      "train_runtime": 4853.9267,
      "train_tokens_per_second": 425.269
    },
    {
      "epoch": 0.8573565551875201,
      "grad_norm": 2.671372652053833,
      "learning_rate": 1.1443850267379679e-05,
      "loss": 0.09042859822511673,
      "num_input_tokens_seen": 2064842,
      "step": 2006,
      "train_runtime": 4855.9363,
      "train_tokens_per_second": 425.22
    },
    {
      "epoch": 0.8577839512768458,
      "grad_norm": 2.8296258449554443,
      "learning_rate": 1.143957219251337e-05,
      "loss": 0.13653765618801117,
      "num_input_tokens_seen": 2065954,
      "step": 2007,
      "train_runtime": 4858.1143,
      "train_tokens_per_second": 425.258
    },
    {
      "epoch": 0.8582113473661716,
      "grad_norm": 2.457226037979126,
      "learning_rate": 1.143529411764706e-05,
      "loss": 0.14975999295711517,
      "num_input_tokens_seen": 2067154,
      "step": 2008,
      "train_runtime": 4860.2742,
      "train_tokens_per_second": 425.316
    },
    {
      "epoch": 0.8586387434554974,
      "grad_norm": 3.6422884464263916,
      "learning_rate": 1.1431016042780749e-05,
      "loss": 0.15724913775920868,
      "num_input_tokens_seen": 2067810,
      "step": 2009,
      "train_runtime": 4862.3395,
      "train_tokens_per_second": 425.271
    },
    {
      "epoch": 0.8590661395448231,
      "grad_norm": 2.1221868991851807,
      "learning_rate": 1.1426737967914439e-05,
      "loss": 0.09180396795272827,
      "num_input_tokens_seen": 2068922,
      "step": 2010,
      "train_runtime": 4864.4736,
      "train_tokens_per_second": 425.313
    },
    {
      "epoch": 0.859493535634149,
      "grad_norm": 3.1511943340301514,
      "learning_rate": 1.142245989304813e-05,
      "loss": 0.11383674293756485,
      "num_input_tokens_seen": 2069696,
      "step": 2011,
      "train_runtime": 4873.1578,
      "train_tokens_per_second": 424.714
    },
    {
      "epoch": 0.8599209317234747,
      "grad_norm": 3.0906403064727783,
      "learning_rate": 1.1418181818181819e-05,
      "loss": 0.18607240915298462,
      "num_input_tokens_seen": 2070610,
      "step": 2012,
      "train_runtime": 4875.4026,
      "train_tokens_per_second": 424.705
    },
    {
      "epoch": 0.8603483278128005,
      "grad_norm": 2.714343547821045,
      "learning_rate": 1.1413903743315509e-05,
      "loss": 0.10701340436935425,
      "num_input_tokens_seen": 2071972,
      "step": 2013,
      "train_runtime": 4877.601,
      "train_tokens_per_second": 424.793
    },
    {
      "epoch": 0.8607757239021263,
      "grad_norm": 3.5791478157043457,
      "learning_rate": 1.1409625668449198e-05,
      "loss": 0.18957960605621338,
      "num_input_tokens_seen": 2072962,
      "step": 2014,
      "train_runtime": 4879.8113,
      "train_tokens_per_second": 424.804
    },
    {
      "epoch": 0.861203119991452,
      "grad_norm": 5.472559928894043,
      "learning_rate": 1.140534759358289e-05,
      "loss": 0.14212319254875183,
      "num_input_tokens_seen": 2073506,
      "step": 2015,
      "train_runtime": 4881.8228,
      "train_tokens_per_second": 424.74
    },
    {
      "epoch": 0.8616305160807779,
      "grad_norm": 3.226473331451416,
      "learning_rate": 1.1401069518716579e-05,
      "loss": 0.10729537904262543,
      "num_input_tokens_seen": 2074234,
      "step": 2016,
      "train_runtime": 4883.8674,
      "train_tokens_per_second": 424.711
    },
    {
      "epoch": 0.8620579121701036,
      "grad_norm": 2.1194236278533936,
      "learning_rate": 1.1396791443850268e-05,
      "loss": 0.09174731373786926,
      "num_input_tokens_seen": 2075424,
      "step": 2017,
      "train_runtime": 4886.0529,
      "train_tokens_per_second": 424.765
    },
    {
      "epoch": 0.8624853082594294,
      "grad_norm": 2.6893062591552734,
      "learning_rate": 1.1392513368983957e-05,
      "loss": 0.23518908023834229,
      "num_input_tokens_seen": 2076470,
      "step": 2018,
      "train_runtime": 4888.147,
      "train_tokens_per_second": 424.797
    },
    {
      "epoch": 0.8629127043487552,
      "grad_norm": 3.5554754734039307,
      "learning_rate": 1.1388235294117649e-05,
      "loss": 0.21239030361175537,
      "num_input_tokens_seen": 2077754,
      "step": 2019,
      "train_runtime": 4890.3025,
      "train_tokens_per_second": 424.872
    },
    {
      "epoch": 0.863340100438081,
      "grad_norm": 2.337817668914795,
      "learning_rate": 1.1383957219251338e-05,
      "loss": 0.07666034251451492,
      "num_input_tokens_seen": 2078574,
      "step": 2020,
      "train_runtime": 4892.3816,
      "train_tokens_per_second": 424.859
    },
    {
      "epoch": 0.8637674965274068,
      "grad_norm": 3.9617319107055664,
      "learning_rate": 1.1379679144385028e-05,
      "loss": 0.1682778149843216,
      "num_input_tokens_seen": 2079704,
      "step": 2021,
      "train_runtime": 4894.527,
      "train_tokens_per_second": 424.904
    },
    {
      "epoch": 0.8641948926167325,
      "grad_norm": 2.1049556732177734,
      "learning_rate": 1.1375401069518717e-05,
      "loss": 0.1025129109621048,
      "num_input_tokens_seen": 2081392,
      "step": 2022,
      "train_runtime": 4896.813,
      "train_tokens_per_second": 425.05
    },
    {
      "epoch": 0.8646222887060584,
      "grad_norm": 3.863140821456909,
      "learning_rate": 1.1371122994652408e-05,
      "loss": 0.19519008696079254,
      "num_input_tokens_seen": 2082108,
      "step": 2023,
      "train_runtime": 4898.8537,
      "train_tokens_per_second": 425.019
    },
    {
      "epoch": 0.8650496847953841,
      "grad_norm": 2.683444023132324,
      "learning_rate": 1.1366844919786098e-05,
      "loss": 0.19205322861671448,
      "num_input_tokens_seen": 2083052,
      "step": 2024,
      "train_runtime": 4900.9514,
      "train_tokens_per_second": 425.03
    },
    {
      "epoch": 0.8654770808847099,
      "grad_norm": 2.886895179748535,
      "learning_rate": 1.1362566844919787e-05,
      "loss": 0.20484893023967743,
      "num_input_tokens_seen": 2084296,
      "step": 2025,
      "train_runtime": 4903.1165,
      "train_tokens_per_second": 425.096
    },
    {
      "epoch": 0.8659044769740357,
      "grad_norm": 2.769684314727783,
      "learning_rate": 1.1358288770053476e-05,
      "loss": 0.12484301626682281,
      "num_input_tokens_seen": 2085072,
      "step": 2026,
      "train_runtime": 4905.1531,
      "train_tokens_per_second": 425.078
    },
    {
      "epoch": 0.8663318730633615,
      "grad_norm": 1.8872156143188477,
      "learning_rate": 1.1354010695187168e-05,
      "loss": 0.0790603756904602,
      "num_input_tokens_seen": 2086082,
      "step": 2027,
      "train_runtime": 4907.2669,
      "train_tokens_per_second": 425.101
    },
    {
      "epoch": 0.8667592691526873,
      "grad_norm": 2.3702504634857178,
      "learning_rate": 1.1349732620320857e-05,
      "loss": 0.12505261600017548,
      "num_input_tokens_seen": 2087164,
      "step": 2028,
      "train_runtime": 4909.4236,
      "train_tokens_per_second": 425.134
    },
    {
      "epoch": 0.867186665242013,
      "grad_norm": 3.0951671600341797,
      "learning_rate": 1.1345454545454546e-05,
      "loss": 0.140278160572052,
      "num_input_tokens_seen": 2087896,
      "step": 2029,
      "train_runtime": 4911.4723,
      "train_tokens_per_second": 425.106
    },
    {
      "epoch": 0.8676140613313388,
      "grad_norm": 3.121394634246826,
      "learning_rate": 1.1341176470588236e-05,
      "loss": 0.13748930394649506,
      "num_input_tokens_seen": 2088562,
      "step": 2030,
      "train_runtime": 4913.4898,
      "train_tokens_per_second": 425.067
    },
    {
      "epoch": 0.8680414574206646,
      "grad_norm": 3.702622890472412,
      "learning_rate": 1.1336898395721927e-05,
      "loss": 0.1778581738471985,
      "num_input_tokens_seen": 2089798,
      "step": 2031,
      "train_runtime": 4915.6616,
      "train_tokens_per_second": 425.131
    },
    {
      "epoch": 0.8684688535099904,
      "grad_norm": 3.2283809185028076,
      "learning_rate": 1.1332620320855617e-05,
      "loss": 0.20485949516296387,
      "num_input_tokens_seen": 2090866,
      "step": 2032,
      "train_runtime": 4917.7714,
      "train_tokens_per_second": 425.165
    },
    {
      "epoch": 0.8688962495993162,
      "grad_norm": 2.1973516941070557,
      "learning_rate": 1.1328342245989306e-05,
      "loss": 0.11748316138982773,
      "num_input_tokens_seen": 2092152,
      "step": 2033,
      "train_runtime": 4919.9254,
      "train_tokens_per_second": 425.241
    },
    {
      "epoch": 0.869323645688642,
      "grad_norm": 4.244319915771484,
      "learning_rate": 1.1324064171122995e-05,
      "loss": 0.08979244530200958,
      "num_input_tokens_seen": 2092912,
      "step": 2034,
      "train_runtime": 4922.0096,
      "train_tokens_per_second": 425.215
    },
    {
      "epoch": 0.8697510417779677,
      "grad_norm": 3.4045071601867676,
      "learning_rate": 1.1319786096256687e-05,
      "loss": 0.17678019404411316,
      "num_input_tokens_seen": 2093722,
      "step": 2035,
      "train_runtime": 4924.0688,
      "train_tokens_per_second": 425.202
    },
    {
      "epoch": 0.8701784378672935,
      "grad_norm": 2.3533201217651367,
      "learning_rate": 1.1315508021390376e-05,
      "loss": 0.10691918432712555,
      "num_input_tokens_seen": 2094804,
      "step": 2036,
      "train_runtime": 4926.1652,
      "train_tokens_per_second": 425.24
    },
    {
      "epoch": 0.8706058339566193,
      "grad_norm": 3.7705631256103516,
      "learning_rate": 1.1311229946524065e-05,
      "loss": 0.12243242561817169,
      "num_input_tokens_seen": 2095692,
      "step": 2037,
      "train_runtime": 4928.2301,
      "train_tokens_per_second": 425.242
    },
    {
      "epoch": 0.8710332300459451,
      "grad_norm": 3.048866033554077,
      "learning_rate": 1.1306951871657755e-05,
      "loss": 0.1400669664144516,
      "num_input_tokens_seen": 2096448,
      "step": 2038,
      "train_runtime": 4930.2566,
      "train_tokens_per_second": 425.221
    },
    {
      "epoch": 0.8714606261352709,
      "grad_norm": 2.9495797157287598,
      "learning_rate": 1.1302673796791446e-05,
      "loss": 0.1668057143688202,
      "num_input_tokens_seen": 2097430,
      "step": 2039,
      "train_runtime": 4932.3492,
      "train_tokens_per_second": 425.24
    },
    {
      "epoch": 0.8718880222245966,
      "grad_norm": 2.709987163543701,
      "learning_rate": 1.1298395721925135e-05,
      "loss": 0.09622637927532196,
      "num_input_tokens_seen": 2098438,
      "step": 2040,
      "train_runtime": 4934.4223,
      "train_tokens_per_second": 425.265
    },
    {
      "epoch": 0.8723154183139225,
      "grad_norm": 3.0652785301208496,
      "learning_rate": 1.1294117647058825e-05,
      "loss": 0.20651371777057648,
      "num_input_tokens_seen": 2099334,
      "step": 2041,
      "train_runtime": 4943.1704,
      "train_tokens_per_second": 424.694
    },
    {
      "epoch": 0.8727428144032482,
      "grad_norm": 2.630715847015381,
      "learning_rate": 1.1289839572192514e-05,
      "loss": 0.11181443184614182,
      "num_input_tokens_seen": 2100336,
      "step": 2042,
      "train_runtime": 4945.2355,
      "train_tokens_per_second": 424.719
    },
    {
      "epoch": 0.873170210492574,
      "grad_norm": 3.4142298698425293,
      "learning_rate": 1.1285561497326205e-05,
      "loss": 0.1708744764328003,
      "num_input_tokens_seen": 2101246,
      "step": 2043,
      "train_runtime": 4947.2835,
      "train_tokens_per_second": 424.727
    },
    {
      "epoch": 0.8735976065818998,
      "grad_norm": 2.5575437545776367,
      "learning_rate": 1.1281283422459895e-05,
      "loss": 0.11476299911737442,
      "num_input_tokens_seen": 2102154,
      "step": 2044,
      "train_runtime": 4949.3579,
      "train_tokens_per_second": 424.733
    },
    {
      "epoch": 0.8740250026712255,
      "grad_norm": 2.3537535667419434,
      "learning_rate": 1.1277005347593584e-05,
      "loss": 0.14269980788230896,
      "num_input_tokens_seen": 2103916,
      "step": 2045,
      "train_runtime": 4951.622,
      "train_tokens_per_second": 424.894
    },
    {
      "epoch": 0.8744523987605514,
      "grad_norm": 2.941002130508423,
      "learning_rate": 1.1272727272727272e-05,
      "loss": 0.15500710904598236,
      "num_input_tokens_seen": 2104866,
      "step": 2046,
      "train_runtime": 4953.6685,
      "train_tokens_per_second": 424.911
    },
    {
      "epoch": 0.8748797948498771,
      "grad_norm": 2.7163405418395996,
      "learning_rate": 1.1268449197860965e-05,
      "loss": 0.17654132843017578,
      "num_input_tokens_seen": 2106214,
      "step": 2047,
      "train_runtime": 4955.8101,
      "train_tokens_per_second": 424.999
    },
    {
      "epoch": 0.875307190939203,
      "grad_norm": 4.698025703430176,
      "learning_rate": 1.1264171122994654e-05,
      "loss": 0.1784188598394394,
      "num_input_tokens_seen": 2107024,
      "step": 2048,
      "train_runtime": 4957.8575,
      "train_tokens_per_second": 424.987
    },
    {
      "epoch": 0.8757345870285287,
      "grad_norm": 2.271402359008789,
      "learning_rate": 1.1259893048128344e-05,
      "loss": 0.07826799154281616,
      "num_input_tokens_seen": 2107916,
      "step": 2049,
      "train_runtime": 4959.9411,
      "train_tokens_per_second": 424.988
    },
    {
      "epoch": 0.8761619831178544,
      "grad_norm": 2.663201093673706,
      "learning_rate": 1.1255614973262032e-05,
      "loss": 0.15355853736400604,
      "num_input_tokens_seen": 2109138,
      "step": 2050,
      "train_runtime": 4962.0428,
      "train_tokens_per_second": 425.054
    },
    {
      "epoch": 0.8765893792071803,
      "grad_norm": 2.6469366550445557,
      "learning_rate": 1.1251336898395724e-05,
      "loss": 0.16244632005691528,
      "num_input_tokens_seen": 2110444,
      "step": 2051,
      "train_runtime": 4964.1769,
      "train_tokens_per_second": 425.135
    },
    {
      "epoch": 0.877016775296506,
      "grad_norm": 3.0999233722686768,
      "learning_rate": 1.1247058823529414e-05,
      "loss": 0.1348665952682495,
      "num_input_tokens_seen": 2111194,
      "step": 2052,
      "train_runtime": 4966.1547,
      "train_tokens_per_second": 425.116
    },
    {
      "epoch": 0.8774441713858319,
      "grad_norm": 2.589658737182617,
      "learning_rate": 1.1242780748663102e-05,
      "loss": 0.11222142726182938,
      "num_input_tokens_seen": 2112414,
      "step": 2053,
      "train_runtime": 4968.2804,
      "train_tokens_per_second": 425.18
    },
    {
      "epoch": 0.8778715674751576,
      "grad_norm": 2.7313668727874756,
      "learning_rate": 1.1238502673796791e-05,
      "loss": 0.16139838099479675,
      "num_input_tokens_seen": 2113914,
      "step": 2054,
      "train_runtime": 4970.6352,
      "train_tokens_per_second": 425.28
    },
    {
      "epoch": 0.8782989635644833,
      "grad_norm": 3.9004340171813965,
      "learning_rate": 1.1234224598930484e-05,
      "loss": 0.1549561470746994,
      "num_input_tokens_seen": 2114710,
      "step": 2055,
      "train_runtime": 4972.6224,
      "train_tokens_per_second": 425.271
    },
    {
      "epoch": 0.8787263596538092,
      "grad_norm": 2.565876007080078,
      "learning_rate": 1.1229946524064172e-05,
      "loss": 0.17598728835582733,
      "num_input_tokens_seen": 2116414,
      "step": 2056,
      "train_runtime": 4974.9039,
      "train_tokens_per_second": 425.418
    },
    {
      "epoch": 0.8791537557431349,
      "grad_norm": 4.031659126281738,
      "learning_rate": 1.1225668449197861e-05,
      "loss": 0.1332961618900299,
      "num_input_tokens_seen": 2117190,
      "step": 2057,
      "train_runtime": 4976.9204,
      "train_tokens_per_second": 425.402
    },
    {
      "epoch": 0.8795811518324608,
      "grad_norm": 3.3271517753601074,
      "learning_rate": 1.122139037433155e-05,
      "loss": 0.19120244681835175,
      "num_input_tokens_seen": 2118212,
      "step": 2058,
      "train_runtime": 4979.0201,
      "train_tokens_per_second": 425.427
    },
    {
      "epoch": 0.8800085479217865,
      "grad_norm": 2.249606132507324,
      "learning_rate": 1.1217112299465243e-05,
      "loss": 0.10734176635742188,
      "num_input_tokens_seen": 2119170,
      "step": 2059,
      "train_runtime": 4981.1166,
      "train_tokens_per_second": 425.441
    },
    {
      "epoch": 0.8804359440111122,
      "grad_norm": 3.041447401046753,
      "learning_rate": 1.1212834224598931e-05,
      "loss": 0.10975335538387299,
      "num_input_tokens_seen": 2120156,
      "step": 2060,
      "train_runtime": 4983.1986,
      "train_tokens_per_second": 425.461
    },
    {
      "epoch": 0.8808633401004381,
      "grad_norm": 2.3502914905548096,
      "learning_rate": 1.120855614973262e-05,
      "loss": 0.15373003482818604,
      "num_input_tokens_seen": 2121670,
      "step": 2061,
      "train_runtime": 4985.4009,
      "train_tokens_per_second": 425.577
    },
    {
      "epoch": 0.8812907361897638,
      "grad_norm": 2.9987335205078125,
      "learning_rate": 1.120427807486631e-05,
      "loss": 0.14460055530071259,
      "num_input_tokens_seen": 2122614,
      "step": 2062,
      "train_runtime": 4987.49,
      "train_tokens_per_second": 425.588
    },
    {
      "epoch": 0.8817181322790897,
      "grad_norm": 3.730349540710449,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.1504986733198166,
      "num_input_tokens_seen": 2123342,
      "step": 2063,
      "train_runtime": 4989.5179,
      "train_tokens_per_second": 425.561
    },
    {
      "epoch": 0.8821455283684154,
      "grad_norm": 2.5371320247650146,
      "learning_rate": 1.119572192513369e-05,
      "loss": 0.09918291866779327,
      "num_input_tokens_seen": 2124194,
      "step": 2064,
      "train_runtime": 4991.5727,
      "train_tokens_per_second": 425.556
    },
    {
      "epoch": 0.8825729244577412,
      "grad_norm": 2.3514387607574463,
      "learning_rate": 1.119144385026738e-05,
      "loss": 0.10322583466768265,
      "num_input_tokens_seen": 2125412,
      "step": 2065,
      "train_runtime": 4993.7244,
      "train_tokens_per_second": 425.617
    },
    {
      "epoch": 0.883000320547067,
      "grad_norm": 2.3750336170196533,
      "learning_rate": 1.118716577540107e-05,
      "loss": 0.089006207883358,
      "num_input_tokens_seen": 2126608,
      "step": 2066,
      "train_runtime": 4995.8808,
      "train_tokens_per_second": 425.672
    },
    {
      "epoch": 0.8834277166363927,
      "grad_norm": 2.6898694038391113,
      "learning_rate": 1.118288770053476e-05,
      "loss": 0.14123940467834473,
      "num_input_tokens_seen": 2127588,
      "step": 2067,
      "train_runtime": 4997.9769,
      "train_tokens_per_second": 425.69
    },
    {
      "epoch": 0.8838551127257186,
      "grad_norm": 2.272538900375366,
      "learning_rate": 1.117860962566845e-05,
      "loss": 0.08137433975934982,
      "num_input_tokens_seen": 2128614,
      "step": 2068,
      "train_runtime": 5000.0668,
      "train_tokens_per_second": 425.717
    },
    {
      "epoch": 0.8842825088150443,
      "grad_norm": 3.188025712966919,
      "learning_rate": 1.117433155080214e-05,
      "loss": 0.1871384084224701,
      "num_input_tokens_seen": 2129810,
      "step": 2069,
      "train_runtime": 5002.1446,
      "train_tokens_per_second": 425.779
    },
    {
      "epoch": 0.8847099049043701,
      "grad_norm": 4.658123970031738,
      "learning_rate": 1.1170053475935829e-05,
      "loss": 0.25402939319610596,
      "num_input_tokens_seen": 2130448,
      "step": 2070,
      "train_runtime": 5004.1452,
      "train_tokens_per_second": 425.737
    },
    {
      "epoch": 0.8851373009936959,
      "grad_norm": 2.4507479667663574,
      "learning_rate": 1.116577540106952e-05,
      "loss": 0.15162120759487152,
      "num_input_tokens_seen": 2131852,
      "step": 2071,
      "train_runtime": 5012.9418,
      "train_tokens_per_second": 425.27
    },
    {
      "epoch": 0.8855646970830217,
      "grad_norm": 3.2056353092193604,
      "learning_rate": 1.116149732620321e-05,
      "loss": 0.14779843389987946,
      "num_input_tokens_seen": 2132686,
      "step": 2072,
      "train_runtime": 5015.0195,
      "train_tokens_per_second": 425.26
    },
    {
      "epoch": 0.8859920931723475,
      "grad_norm": 2.6292271614074707,
      "learning_rate": 1.1157219251336899e-05,
      "loss": 0.2042260468006134,
      "num_input_tokens_seen": 2133802,
      "step": 2073,
      "train_runtime": 5017.252,
      "train_tokens_per_second": 425.293
    },
    {
      "epoch": 0.8864194892616732,
      "grad_norm": 2.4773292541503906,
      "learning_rate": 1.1152941176470588e-05,
      "loss": 0.15483739972114563,
      "num_input_tokens_seen": 2134710,
      "step": 2074,
      "train_runtime": 5019.3906,
      "train_tokens_per_second": 425.293
    },
    {
      "epoch": 0.8868468853509991,
      "grad_norm": 3.203991413116455,
      "learning_rate": 1.1148663101604278e-05,
      "loss": 0.1202445775270462,
      "num_input_tokens_seen": 2135374,
      "step": 2075,
      "train_runtime": 5021.3773,
      "train_tokens_per_second": 425.257
    },
    {
      "epoch": 0.8872742814403248,
      "grad_norm": 3.6493308544158936,
      "learning_rate": 1.1144385026737969e-05,
      "loss": 0.24358196556568146,
      "num_input_tokens_seen": 2136620,
      "step": 2076,
      "train_runtime": 5023.5545,
      "train_tokens_per_second": 425.32
    },
    {
      "epoch": 0.8877016775296506,
      "grad_norm": 3.6222965717315674,
      "learning_rate": 1.1140106951871659e-05,
      "loss": 0.15311744809150696,
      "num_input_tokens_seen": 2137722,
      "step": 2077,
      "train_runtime": 5025.655,
      "train_tokens_per_second": 425.362
    },
    {
      "epoch": 0.8881290736189764,
      "grad_norm": 3.0898115634918213,
      "learning_rate": 1.1135828877005348e-05,
      "loss": 0.1271706521511078,
      "num_input_tokens_seen": 2138862,
      "step": 2078,
      "train_runtime": 5027.7725,
      "train_tokens_per_second": 425.409
    },
    {
      "epoch": 0.8885564697083022,
      "grad_norm": 3.015130043029785,
      "learning_rate": 1.1131550802139037e-05,
      "loss": 0.12016475200653076,
      "num_input_tokens_seen": 2139738,
      "step": 2079,
      "train_runtime": 5029.8225,
      "train_tokens_per_second": 425.41
    },
    {
      "epoch": 0.888983865797628,
      "grad_norm": 4.087200164794922,
      "learning_rate": 1.1127272727272729e-05,
      "loss": 0.15748974680900574,
      "num_input_tokens_seen": 2140384,
      "step": 2080,
      "train_runtime": 5031.8443,
      "train_tokens_per_second": 425.368
    },
    {
      "epoch": 0.8894112618869537,
      "grad_norm": 2.887399911880493,
      "learning_rate": 1.1122994652406418e-05,
      "loss": 0.13596299290657043,
      "num_input_tokens_seen": 2141186,
      "step": 2081,
      "train_runtime": 5033.8677,
      "train_tokens_per_second": 425.356
    },
    {
      "epoch": 0.8898386579762795,
      "grad_norm": 2.561500072479248,
      "learning_rate": 1.1118716577540107e-05,
      "loss": 0.14200156927108765,
      "num_input_tokens_seen": 2142272,
      "step": 2082,
      "train_runtime": 5035.9674,
      "train_tokens_per_second": 425.394
    },
    {
      "epoch": 0.8902660540656053,
      "grad_norm": 3.6333630084991455,
      "learning_rate": 1.1114438502673797e-05,
      "loss": 0.13719424605369568,
      "num_input_tokens_seen": 2143428,
      "step": 2083,
      "train_runtime": 5038.1033,
      "train_tokens_per_second": 425.443
    },
    {
      "epoch": 0.8906934501549311,
      "grad_norm": 2.1444313526153564,
      "learning_rate": 1.1110160427807488e-05,
      "loss": 0.09731752425432205,
      "num_input_tokens_seen": 2144534,
      "step": 2084,
      "train_runtime": 5040.1816,
      "train_tokens_per_second": 425.487
    },
    {
      "epoch": 0.8911208462442569,
      "grad_norm": 2.602187395095825,
      "learning_rate": 1.1105882352941177e-05,
      "loss": 0.1659667044878006,
      "num_input_tokens_seen": 2146040,
      "step": 2085,
      "train_runtime": 5042.4325,
      "train_tokens_per_second": 425.596
    },
    {
      "epoch": 0.8915482423335827,
      "grad_norm": 2.782085657119751,
      "learning_rate": 1.1101604278074867e-05,
      "loss": 0.13991759717464447,
      "num_input_tokens_seen": 2147140,
      "step": 2086,
      "train_runtime": 5044.5511,
      "train_tokens_per_second": 425.635
    },
    {
      "epoch": 0.8919756384229084,
      "grad_norm": 2.631455183029175,
      "learning_rate": 1.1097326203208556e-05,
      "loss": 0.11562193930149078,
      "num_input_tokens_seen": 2148050,
      "step": 2087,
      "train_runtime": 5046.6211,
      "train_tokens_per_second": 425.641
    },
    {
      "epoch": 0.8924030345122342,
      "grad_norm": 6.182000160217285,
      "learning_rate": 1.1093048128342247e-05,
      "loss": 0.09962183982133865,
      "num_input_tokens_seen": 2148750,
      "step": 2088,
      "train_runtime": 5048.6466,
      "train_tokens_per_second": 425.609
    },
    {
      "epoch": 0.89283043060156,
      "grad_norm": 3.3724265098571777,
      "learning_rate": 1.1088770053475937e-05,
      "loss": 0.16166161000728607,
      "num_input_tokens_seen": 2149570,
      "step": 2089,
      "train_runtime": 5050.7075,
      "train_tokens_per_second": 425.598
    },
    {
      "epoch": 0.8932578266908858,
      "grad_norm": 2.070803165435791,
      "learning_rate": 1.1084491978609626e-05,
      "loss": 0.09638729691505432,
      "num_input_tokens_seen": 2150544,
      "step": 2090,
      "train_runtime": 5052.818,
      "train_tokens_per_second": 425.613
    },
    {
      "epoch": 0.8936852227802116,
      "grad_norm": 2.8039610385894775,
      "learning_rate": 1.1080213903743316e-05,
      "loss": 0.13526132702827454,
      "num_input_tokens_seen": 2151614,
      "step": 2091,
      "train_runtime": 5054.9068,
      "train_tokens_per_second": 425.649
    },
    {
      "epoch": 0.8941126188695373,
      "grad_norm": 2.854904890060425,
      "learning_rate": 1.1075935828877007e-05,
      "loss": 0.13203227519989014,
      "num_input_tokens_seen": 2152512,
      "step": 2092,
      "train_runtime": 5056.9533,
      "train_tokens_per_second": 425.654
    },
    {
      "epoch": 0.8945400149588631,
      "grad_norm": 2.8414394855499268,
      "learning_rate": 1.1071657754010696e-05,
      "loss": 0.1113419458270073,
      "num_input_tokens_seen": 2153204,
      "step": 2093,
      "train_runtime": 5058.9412,
      "train_tokens_per_second": 425.623
    },
    {
      "epoch": 0.8949674110481889,
      "grad_norm": 2.2244601249694824,
      "learning_rate": 1.1067379679144386e-05,
      "loss": 0.06751777976751328,
      "num_input_tokens_seen": 2154586,
      "step": 2094,
      "train_runtime": 5061.1303,
      "train_tokens_per_second": 425.712
    },
    {
      "epoch": 0.8953948071375147,
      "grad_norm": 2.958847761154175,
      "learning_rate": 1.1063101604278075e-05,
      "loss": 0.12220863252878189,
      "num_input_tokens_seen": 2155250,
      "step": 2095,
      "train_runtime": 5063.1218,
      "train_tokens_per_second": 425.676
    },
    {
      "epoch": 0.8958222032268405,
      "grad_norm": 1.9352847337722778,
      "learning_rate": 1.1058823529411766e-05,
      "loss": 0.07753197103738785,
      "num_input_tokens_seen": 2156172,
      "step": 2096,
      "train_runtime": 5065.1721,
      "train_tokens_per_second": 425.686
    },
    {
      "epoch": 0.8962495993161662,
      "grad_norm": 2.2991068363189697,
      "learning_rate": 1.1054545454545456e-05,
      "loss": 0.10877975821495056,
      "num_input_tokens_seen": 2157060,
      "step": 2097,
      "train_runtime": 5067.2263,
      "train_tokens_per_second": 425.689
    },
    {
      "epoch": 0.8966769954054921,
      "grad_norm": 2.3581900596618652,
      "learning_rate": 1.1050267379679145e-05,
      "loss": 0.1087910458445549,
      "num_input_tokens_seen": 2157940,
      "step": 2098,
      "train_runtime": 5069.3112,
      "train_tokens_per_second": 425.687
    },
    {
      "epoch": 0.8971043914948178,
      "grad_norm": 3.809530258178711,
      "learning_rate": 1.1045989304812835e-05,
      "loss": 0.12143318355083466,
      "num_input_tokens_seen": 2158944,
      "step": 2099,
      "train_runtime": 5071.4229,
      "train_tokens_per_second": 425.708
    },
    {
      "epoch": 0.8975317875841436,
      "grad_norm": 3.5151312351226807,
      "learning_rate": 1.1041711229946526e-05,
      "loss": 0.18043813109397888,
      "num_input_tokens_seen": 2159674,
      "step": 2100,
      "train_runtime": 5073.4066,
      "train_tokens_per_second": 425.685
    },
    {
      "epoch": 0.8979591836734694,
      "grad_norm": 3.954275608062744,
      "learning_rate": 1.1037433155080215e-05,
      "loss": 0.2579469084739685,
      "num_input_tokens_seen": 2160366,
      "step": 2101,
      "train_runtime": 5081.7292,
      "train_tokens_per_second": 425.124
    },
    {
      "epoch": 0.8983865797627951,
      "grad_norm": 2.8379507064819336,
      "learning_rate": 1.1033155080213905e-05,
      "loss": 0.10394646227359772,
      "num_input_tokens_seen": 2161154,
      "step": 2102,
      "train_runtime": 5083.7792,
      "train_tokens_per_second": 425.108
    },
    {
      "epoch": 0.898813975852121,
      "grad_norm": 3.489959478378296,
      "learning_rate": 1.1028877005347594e-05,
      "loss": 0.12316275388002396,
      "num_input_tokens_seen": 2161938,
      "step": 2103,
      "train_runtime": 5085.921,
      "train_tokens_per_second": 425.083
    },
    {
      "epoch": 0.8992413719414467,
      "grad_norm": 3.760416269302368,
      "learning_rate": 1.1024598930481285e-05,
      "loss": 0.19151145219802856,
      "num_input_tokens_seen": 2162960,
      "step": 2104,
      "train_runtime": 5088.0086,
      "train_tokens_per_second": 425.109
    },
    {
      "epoch": 0.8996687680307726,
      "grad_norm": 4.122915267944336,
      "learning_rate": 1.1020320855614975e-05,
      "loss": 0.2019379436969757,
      "num_input_tokens_seen": 2163668,
      "step": 2105,
      "train_runtime": 5090.0663,
      "train_tokens_per_second": 425.077
    },
    {
      "epoch": 0.9000961641200983,
      "grad_norm": 2.8671376705169678,
      "learning_rate": 1.1016042780748664e-05,
      "loss": 0.12641726434230804,
      "num_input_tokens_seen": 2164382,
      "step": 2106,
      "train_runtime": 5092.0951,
      "train_tokens_per_second": 425.047
    },
    {
      "epoch": 0.900523560209424,
      "grad_norm": 2.865556478500366,
      "learning_rate": 1.1011764705882354e-05,
      "loss": 0.12395767122507095,
      "num_input_tokens_seen": 2165384,
      "step": 2107,
      "train_runtime": 5094.1834,
      "train_tokens_per_second": 425.07
    },
    {
      "epoch": 0.9009509562987499,
      "grad_norm": 3.565251588821411,
      "learning_rate": 1.1007486631016045e-05,
      "loss": 0.19681403040885925,
      "num_input_tokens_seen": 2166104,
      "step": 2108,
      "train_runtime": 5096.2358,
      "train_tokens_per_second": 425.04
    },
    {
      "epoch": 0.9013783523880756,
      "grad_norm": 3.232603073120117,
      "learning_rate": 1.1003208556149734e-05,
      "loss": 0.181961789727211,
      "num_input_tokens_seen": 2166962,
      "step": 2109,
      "train_runtime": 5098.3168,
      "train_tokens_per_second": 425.035
    },
    {
      "epoch": 0.9018057484774015,
      "grad_norm": 2.5656421184539795,
      "learning_rate": 1.0998930481283424e-05,
      "loss": 0.056980133056640625,
      "num_input_tokens_seen": 2167504,
      "step": 2110,
      "train_runtime": 5100.3102,
      "train_tokens_per_second": 424.975
    },
    {
      "epoch": 0.9022331445667272,
      "grad_norm": 4.643036365509033,
      "learning_rate": 1.0994652406417112e-05,
      "loss": 0.2640537917613983,
      "num_input_tokens_seen": 2168670,
      "step": 2111,
      "train_runtime": 5102.4295,
      "train_tokens_per_second": 425.027
    },
    {
      "epoch": 0.902660540656053,
      "grad_norm": 2.278615951538086,
      "learning_rate": 1.0990374331550804e-05,
      "loss": 0.08019732683897018,
      "num_input_tokens_seen": 2169614,
      "step": 2112,
      "train_runtime": 5104.5133,
      "train_tokens_per_second": 425.038
    },
    {
      "epoch": 0.9030879367453788,
      "grad_norm": 1.9510892629623413,
      "learning_rate": 1.0986096256684494e-05,
      "loss": 0.10955323278903961,
      "num_input_tokens_seen": 2170860,
      "step": 2113,
      "train_runtime": 5106.6642,
      "train_tokens_per_second": 425.103
    },
    {
      "epoch": 0.9035153328347045,
      "grad_norm": 2.72688364982605,
      "learning_rate": 1.0981818181818182e-05,
      "loss": 0.16085560619831085,
      "num_input_tokens_seen": 2172150,
      "step": 2114,
      "train_runtime": 5108.822,
      "train_tokens_per_second": 425.176
    },
    {
      "epoch": 0.9039427289240304,
      "grad_norm": 2.535994529724121,
      "learning_rate": 1.0977540106951871e-05,
      "loss": 0.15517297387123108,
      "num_input_tokens_seen": 2173248,
      "step": 2115,
      "train_runtime": 5110.9596,
      "train_tokens_per_second": 425.213
    },
    {
      "epoch": 0.9043701250133561,
      "grad_norm": 2.9148645401000977,
      "learning_rate": 1.0973262032085564e-05,
      "loss": 0.12632083892822266,
      "num_input_tokens_seen": 2174380,
      "step": 2116,
      "train_runtime": 5113.0915,
      "train_tokens_per_second": 425.257
    },
    {
      "epoch": 0.9047975211026819,
      "grad_norm": 2.3214263916015625,
      "learning_rate": 1.0968983957219253e-05,
      "loss": 0.10537756234407425,
      "num_input_tokens_seen": 2175414,
      "step": 2117,
      "train_runtime": 5115.1783,
      "train_tokens_per_second": 425.286
    },
    {
      "epoch": 0.9052249171920077,
      "grad_norm": 2.6031312942504883,
      "learning_rate": 1.0964705882352941e-05,
      "loss": 0.1510811448097229,
      "num_input_tokens_seen": 2176874,
      "step": 2118,
      "train_runtime": 5117.3654,
      "train_tokens_per_second": 425.39
    },
    {
      "epoch": 0.9056523132813334,
      "grad_norm": 3.532231330871582,
      "learning_rate": 1.096042780748663e-05,
      "loss": 0.1757303923368454,
      "num_input_tokens_seen": 2177804,
      "step": 2119,
      "train_runtime": 5119.4193,
      "train_tokens_per_second": 425.401
    },
    {
      "epoch": 0.9060797093706593,
      "grad_norm": 2.547057628631592,
      "learning_rate": 1.0956149732620323e-05,
      "loss": 0.13407796621322632,
      "num_input_tokens_seen": 2179206,
      "step": 2120,
      "train_runtime": 5121.6275,
      "train_tokens_per_second": 425.491
    },
    {
      "epoch": 0.906507105459985,
      "grad_norm": 2.5464515686035156,
      "learning_rate": 1.0951871657754011e-05,
      "loss": 0.08509576320648193,
      "num_input_tokens_seen": 2179886,
      "step": 2121,
      "train_runtime": 5123.6067,
      "train_tokens_per_second": 425.459
    },
    {
      "epoch": 0.9069345015493109,
      "grad_norm": 3.354395627975464,
      "learning_rate": 1.09475935828877e-05,
      "loss": 0.13675421476364136,
      "num_input_tokens_seen": 2180698,
      "step": 2122,
      "train_runtime": 5125.6763,
      "train_tokens_per_second": 425.446
    },
    {
      "epoch": 0.9073618976386366,
      "grad_norm": 2.4906198978424072,
      "learning_rate": 1.094331550802139e-05,
      "loss": 0.15708817541599274,
      "num_input_tokens_seen": 2181852,
      "step": 2123,
      "train_runtime": 5127.7616,
      "train_tokens_per_second": 425.498
    },
    {
      "epoch": 0.9077892937279624,
      "grad_norm": 2.5018856525421143,
      "learning_rate": 1.0939037433155081e-05,
      "loss": 0.13699057698249817,
      "num_input_tokens_seen": 2182806,
      "step": 2124,
      "train_runtime": 5129.8563,
      "train_tokens_per_second": 425.51
    },
    {
      "epoch": 0.9082166898172882,
      "grad_norm": 3.0618326663970947,
      "learning_rate": 1.093475935828877e-05,
      "loss": 0.18096017837524414,
      "num_input_tokens_seen": 2183504,
      "step": 2125,
      "train_runtime": 5131.878,
      "train_tokens_per_second": 425.479
    },
    {
      "epoch": 0.9086440859066139,
      "grad_norm": 2.9401183128356934,
      "learning_rate": 1.093048128342246e-05,
      "loss": 0.20621538162231445,
      "num_input_tokens_seen": 2184810,
      "step": 2126,
      "train_runtime": 5134.0563,
      "train_tokens_per_second": 425.552
    },
    {
      "epoch": 0.9090714819959398,
      "grad_norm": 3.2026705741882324,
      "learning_rate": 1.092620320855615e-05,
      "loss": 0.17546185851097107,
      "num_input_tokens_seen": 2185814,
      "step": 2127,
      "train_runtime": 5136.131,
      "train_tokens_per_second": 425.576
    },
    {
      "epoch": 0.9094988780852655,
      "grad_norm": 3.712503433227539,
      "learning_rate": 1.092192513368984e-05,
      "loss": 0.14737451076507568,
      "num_input_tokens_seen": 2186656,
      "step": 2128,
      "train_runtime": 5138.2015,
      "train_tokens_per_second": 425.568
    },
    {
      "epoch": 0.9099262741745913,
      "grad_norm": 1.7033201456069946,
      "learning_rate": 1.091764705882353e-05,
      "loss": 0.05501745641231537,
      "num_input_tokens_seen": 2187694,
      "step": 2129,
      "train_runtime": 5140.3011,
      "train_tokens_per_second": 425.596
    },
    {
      "epoch": 0.9103536702639171,
      "grad_norm": 2.63810658454895,
      "learning_rate": 1.091336898395722e-05,
      "loss": 0.09936663508415222,
      "num_input_tokens_seen": 2188778,
      "step": 2130,
      "train_runtime": 5142.5206,
      "train_tokens_per_second": 425.624
    },
    {
      "epoch": 0.9107810663532429,
      "grad_norm": 2.8543202877044678,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 0.17764246463775635,
      "num_input_tokens_seen": 2189862,
      "step": 2131,
      "train_runtime": 5151.1029,
      "train_tokens_per_second": 425.125
    },
    {
      "epoch": 0.9112084624425687,
      "grad_norm": 3.8218626976013184,
      "learning_rate": 1.09048128342246e-05,
      "loss": 0.15334482491016388,
      "num_input_tokens_seen": 2190950,
      "step": 2132,
      "train_runtime": 5153.1851,
      "train_tokens_per_second": 425.164
    },
    {
      "epoch": 0.9116358585318944,
      "grad_norm": 2.9003210067749023,
      "learning_rate": 1.090053475935829e-05,
      "loss": 0.1193888783454895,
      "num_input_tokens_seen": 2191624,
      "step": 2133,
      "train_runtime": 5155.2046,
      "train_tokens_per_second": 425.128
    },
    {
      "epoch": 0.9120632546212202,
      "grad_norm": 4.106865406036377,
      "learning_rate": 1.0896256684491979e-05,
      "loss": 0.2729776203632355,
      "num_input_tokens_seen": 2192338,
      "step": 2134,
      "train_runtime": 5157.2343,
      "train_tokens_per_second": 425.1
    },
    {
      "epoch": 0.912490650710546,
      "grad_norm": 4.397752285003662,
      "learning_rate": 1.0891978609625668e-05,
      "loss": 0.13377909362316132,
      "num_input_tokens_seen": 2193204,
      "step": 2135,
      "train_runtime": 5159.2765,
      "train_tokens_per_second": 425.099
    },
    {
      "epoch": 0.9129180467998718,
      "grad_norm": 3.121751070022583,
      "learning_rate": 1.088770053475936e-05,
      "loss": 0.1185264140367508,
      "num_input_tokens_seen": 2193862,
      "step": 2136,
      "train_runtime": 5161.3207,
      "train_tokens_per_second": 425.058
    },
    {
      "epoch": 0.9133454428891976,
      "grad_norm": 2.8448967933654785,
      "learning_rate": 1.0883422459893049e-05,
      "loss": 0.20869427919387817,
      "num_input_tokens_seen": 2195292,
      "step": 2137,
      "train_runtime": 5163.4919,
      "train_tokens_per_second": 425.156
    },
    {
      "epoch": 0.9137728389785233,
      "grad_norm": 2.3537094593048096,
      "learning_rate": 1.0879144385026738e-05,
      "loss": 0.07945031672716141,
      "num_input_tokens_seen": 2196192,
      "step": 2138,
      "train_runtime": 5165.5209,
      "train_tokens_per_second": 425.164
    },
    {
      "epoch": 0.9142002350678491,
      "grad_norm": 2.047752618789673,
      "learning_rate": 1.0874866310160428e-05,
      "loss": 0.09320962429046631,
      "num_input_tokens_seen": 2197194,
      "step": 2139,
      "train_runtime": 5167.6052,
      "train_tokens_per_second": 425.186
    },
    {
      "epoch": 0.9146276311571749,
      "grad_norm": 2.3615562915802,
      "learning_rate": 1.0870588235294119e-05,
      "loss": 0.1290973722934723,
      "num_input_tokens_seen": 2198406,
      "step": 2140,
      "train_runtime": 5169.7525,
      "train_tokens_per_second": 425.244
    },
    {
      "epoch": 0.9150550272465007,
      "grad_norm": 3.5751733779907227,
      "learning_rate": 1.0866310160427808e-05,
      "loss": 0.11524638533592224,
      "num_input_tokens_seen": 2199216,
      "step": 2141,
      "train_runtime": 5171.7611,
      "train_tokens_per_second": 425.235
    },
    {
      "epoch": 0.9154824233358265,
      "grad_norm": 3.339257001876831,
      "learning_rate": 1.0862032085561498e-05,
      "loss": 0.1953415423631668,
      "num_input_tokens_seen": 2200286,
      "step": 2142,
      "train_runtime": 5173.8046,
      "train_tokens_per_second": 425.274
    },
    {
      "epoch": 0.9159098194251523,
      "grad_norm": 1.957407832145691,
      "learning_rate": 1.0857754010695187e-05,
      "loss": 0.10389889776706696,
      "num_input_tokens_seen": 2202008,
      "step": 2143,
      "train_runtime": 5176.0957,
      "train_tokens_per_second": 425.419
    },
    {
      "epoch": 0.916337215514478,
      "grad_norm": 2.46386981010437,
      "learning_rate": 1.0853475935828878e-05,
      "loss": 0.09721916913986206,
      "num_input_tokens_seen": 2203588,
      "step": 2144,
      "train_runtime": 5178.4696,
      "train_tokens_per_second": 425.529
    },
    {
      "epoch": 0.9167646116038038,
      "grad_norm": 2.506495714187622,
      "learning_rate": 1.0849197860962568e-05,
      "loss": 0.12483134865760803,
      "num_input_tokens_seen": 2205476,
      "step": 2145,
      "train_runtime": 5180.7638,
      "train_tokens_per_second": 425.705
    },
    {
      "epoch": 0.9171920076931296,
      "grad_norm": 2.2939774990081787,
      "learning_rate": 1.0844919786096257e-05,
      "loss": 0.16094037890434265,
      "num_input_tokens_seen": 2206608,
      "step": 2146,
      "train_runtime": 5182.8999,
      "train_tokens_per_second": 425.748
    },
    {
      "epoch": 0.9176194037824554,
      "grad_norm": 2.6393582820892334,
      "learning_rate": 1.0840641711229947e-05,
      "loss": 0.15815824270248413,
      "num_input_tokens_seen": 2208032,
      "step": 2147,
      "train_runtime": 5185.0775,
      "train_tokens_per_second": 425.844
    },
    {
      "epoch": 0.9180467998717812,
      "grad_norm": 2.7265353202819824,
      "learning_rate": 1.0836363636363638e-05,
      "loss": 0.13773269951343536,
      "num_input_tokens_seen": 2209026,
      "step": 2148,
      "train_runtime": 5187.1668,
      "train_tokens_per_second": 425.864
    },
    {
      "epoch": 0.9184741959611069,
      "grad_norm": 2.727649688720703,
      "learning_rate": 1.0832085561497327e-05,
      "loss": 0.10120885074138641,
      "num_input_tokens_seen": 2209972,
      "step": 2149,
      "train_runtime": 5189.267,
      "train_tokens_per_second": 425.874
    },
    {
      "epoch": 0.9189015920504328,
      "grad_norm": 2.123680830001831,
      "learning_rate": 1.0827807486631017e-05,
      "loss": 0.11646150052547455,
      "num_input_tokens_seen": 2210930,
      "step": 2150,
      "train_runtime": 5191.3743,
      "train_tokens_per_second": 425.885
    },
    {
      "epoch": 0.9193289881397585,
      "grad_norm": 1.778903841972351,
      "learning_rate": 1.0823529411764706e-05,
      "loss": 0.1022472232580185,
      "num_input_tokens_seen": 2212194,
      "step": 2151,
      "train_runtime": 5193.485,
      "train_tokens_per_second": 425.956
    },
    {
      "epoch": 0.9197563842290843,
      "grad_norm": 3.146822452545166,
      "learning_rate": 1.0819251336898397e-05,
      "loss": 0.1293303668498993,
      "num_input_tokens_seen": 2213710,
      "step": 2152,
      "train_runtime": 5195.7125,
      "train_tokens_per_second": 426.065
    },
    {
      "epoch": 0.9201837803184101,
      "grad_norm": 3.0693538188934326,
      "learning_rate": 1.0814973262032087e-05,
      "loss": 0.11907608807086945,
      "num_input_tokens_seen": 2214468,
      "step": 2153,
      "train_runtime": 5197.7152,
      "train_tokens_per_second": 426.046
    },
    {
      "epoch": 0.9206111764077358,
      "grad_norm": 4.317978382110596,
      "learning_rate": 1.0810695187165776e-05,
      "loss": 0.1422056257724762,
      "num_input_tokens_seen": 2215134,
      "step": 2154,
      "train_runtime": 5199.7024,
      "train_tokens_per_second": 426.012
    },
    {
      "epoch": 0.9210385724970617,
      "grad_norm": 3.077620506286621,
      "learning_rate": 1.0806417112299466e-05,
      "loss": 0.13248398900032043,
      "num_input_tokens_seen": 2216198,
      "step": 2155,
      "train_runtime": 5201.7808,
      "train_tokens_per_second": 426.046
    },
    {
      "epoch": 0.9214659685863874,
      "grad_norm": 2.8807480335235596,
      "learning_rate": 1.0802139037433157e-05,
      "loss": 0.13581492006778717,
      "num_input_tokens_seen": 2216962,
      "step": 2156,
      "train_runtime": 5204.0215,
      "train_tokens_per_second": 426.009
    },
    {
      "epoch": 0.9218933646757133,
      "grad_norm": 2.3715343475341797,
      "learning_rate": 1.0797860962566846e-05,
      "loss": 0.14173632860183716,
      "num_input_tokens_seen": 2218354,
      "step": 2157,
      "train_runtime": 5206.2581,
      "train_tokens_per_second": 426.094
    },
    {
      "epoch": 0.922320760765039,
      "grad_norm": 2.6752946376800537,
      "learning_rate": 1.0793582887700536e-05,
      "loss": 0.109413743019104,
      "num_input_tokens_seen": 2219428,
      "step": 2158,
      "train_runtime": 5208.3697,
      "train_tokens_per_second": 426.127
    },
    {
      "epoch": 0.9227481568543647,
      "grad_norm": 2.5701966285705566,
      "learning_rate": 1.0789304812834225e-05,
      "loss": 0.10824789851903915,
      "num_input_tokens_seen": 2220526,
      "step": 2159,
      "train_runtime": 5210.5067,
      "train_tokens_per_second": 426.163
    },
    {
      "epoch": 0.9231755529436906,
      "grad_norm": 1.5209977626800537,
      "learning_rate": 1.0785026737967916e-05,
      "loss": 0.05084332451224327,
      "num_input_tokens_seen": 2222110,
      "step": 2160,
      "train_runtime": 5212.7347,
      "train_tokens_per_second": 426.285
    },
    {
      "epoch": 0.9236029490330163,
      "grad_norm": 2.1654891967773438,
      "learning_rate": 1.0780748663101606e-05,
      "loss": 0.1114838719367981,
      "num_input_tokens_seen": 2223346,
      "step": 2161,
      "train_runtime": 5221.2387,
      "train_tokens_per_second": 425.827
    },
    {
      "epoch": 0.9240303451223422,
      "grad_norm": 2.767596483230591,
      "learning_rate": 1.0776470588235295e-05,
      "loss": 0.16918615996837616,
      "num_input_tokens_seen": 2224226,
      "step": 2162,
      "train_runtime": 5223.3193,
      "train_tokens_per_second": 425.826
    },
    {
      "epoch": 0.9244577412116679,
      "grad_norm": 2.738677978515625,
      "learning_rate": 1.0772192513368985e-05,
      "loss": 0.08162784576416016,
      "num_input_tokens_seen": 2225198,
      "step": 2163,
      "train_runtime": 5225.3974,
      "train_tokens_per_second": 425.843
    },
    {
      "epoch": 0.9248851373009936,
      "grad_norm": 2.7296721935272217,
      "learning_rate": 1.0767914438502676e-05,
      "loss": 0.08721458911895752,
      "num_input_tokens_seen": 2225796,
      "step": 2164,
      "train_runtime": 5227.4585,
      "train_tokens_per_second": 425.789
    },
    {
      "epoch": 0.9253125333903195,
      "grad_norm": 4.145293235778809,
      "learning_rate": 1.0763636363636365e-05,
      "loss": 0.11372827738523483,
      "num_input_tokens_seen": 2226544,
      "step": 2165,
      "train_runtime": 5229.5087,
      "train_tokens_per_second": 425.765
    },
    {
      "epoch": 0.9257399294796452,
      "grad_norm": 3.3616368770599365,
      "learning_rate": 1.0759358288770055e-05,
      "loss": 0.21249844133853912,
      "num_input_tokens_seen": 2227448,
      "step": 2166,
      "train_runtime": 5231.6021,
      "train_tokens_per_second": 425.768
    },
    {
      "epoch": 0.9261673255689711,
      "grad_norm": 2.7214016914367676,
      "learning_rate": 1.0755080213903744e-05,
      "loss": 0.14507797360420227,
      "num_input_tokens_seen": 2229306,
      "step": 2167,
      "train_runtime": 5233.8521,
      "train_tokens_per_second": 425.94
    },
    {
      "epoch": 0.9265947216582968,
      "grad_norm": 2.3175292015075684,
      "learning_rate": 1.0750802139037435e-05,
      "loss": 0.15931209921836853,
      "num_input_tokens_seen": 2230718,
      "step": 2168,
      "train_runtime": 5236.0199,
      "train_tokens_per_second": 426.033
    },
    {
      "epoch": 0.9270221177476226,
      "grad_norm": 3.5114946365356445,
      "learning_rate": 1.0746524064171125e-05,
      "loss": 0.13204370439052582,
      "num_input_tokens_seen": 2231422,
      "step": 2169,
      "train_runtime": 5238.0654,
      "train_tokens_per_second": 426.001
    },
    {
      "epoch": 0.9274495138369484,
      "grad_norm": 2.4863505363464355,
      "learning_rate": 1.0742245989304814e-05,
      "loss": 0.09473159909248352,
      "num_input_tokens_seen": 2232340,
      "step": 2170,
      "train_runtime": 5240.1543,
      "train_tokens_per_second": 426.007
    },
    {
      "epoch": 0.9278769099262741,
      "grad_norm": 9.001498222351074,
      "learning_rate": 1.0737967914438504e-05,
      "loss": 0.22720074653625488,
      "num_input_tokens_seen": 2233880,
      "step": 2171,
      "train_runtime": 5242.3674,
      "train_tokens_per_second": 426.12
    },
    {
      "epoch": 0.9283043060156,
      "grad_norm": 2.9662041664123535,
      "learning_rate": 1.0733689839572195e-05,
      "loss": 0.10152985155582428,
      "num_input_tokens_seen": 2234980,
      "step": 2172,
      "train_runtime": 5244.443,
      "train_tokens_per_second": 426.162
    },
    {
      "epoch": 0.9287317021049257,
      "grad_norm": 4.028475761413574,
      "learning_rate": 1.0729411764705884e-05,
      "loss": 0.21238811314105988,
      "num_input_tokens_seen": 2235902,
      "step": 2173,
      "train_runtime": 5246.5465,
      "train_tokens_per_second": 426.166
    },
    {
      "epoch": 0.9291590981942516,
      "grad_norm": 3.0487616062164307,
      "learning_rate": 1.0725133689839574e-05,
      "loss": 0.11749480664730072,
      "num_input_tokens_seen": 2236840,
      "step": 2174,
      "train_runtime": 5248.6,
      "train_tokens_per_second": 426.178
    },
    {
      "epoch": 0.9295864942835773,
      "grad_norm": 2.740997552871704,
      "learning_rate": 1.0720855614973263e-05,
      "loss": 0.1353542059659958,
      "num_input_tokens_seen": 2238150,
      "step": 2175,
      "train_runtime": 5250.8079,
      "train_tokens_per_second": 426.249
    },
    {
      "epoch": 0.930013890372903,
      "grad_norm": 2.0356626510620117,
      "learning_rate": 1.0716577540106954e-05,
      "loss": 0.11324270814657211,
      "num_input_tokens_seen": 2239322,
      "step": 2176,
      "train_runtime": 5252.979,
      "train_tokens_per_second": 426.296
    },
    {
      "epoch": 0.9304412864622289,
      "grad_norm": 2.3686628341674805,
      "learning_rate": 1.0712299465240644e-05,
      "loss": 0.09979484975337982,
      "num_input_tokens_seen": 2240178,
      "step": 2177,
      "train_runtime": 5255.0701,
      "train_tokens_per_second": 426.289
    },
    {
      "epoch": 0.9308686825515546,
      "grad_norm": 3.0573229789733887,
      "learning_rate": 1.0708021390374333e-05,
      "loss": 0.12711331248283386,
      "num_input_tokens_seen": 2241096,
      "step": 2178,
      "train_runtime": 5257.2102,
      "train_tokens_per_second": 426.29
    },
    {
      "epoch": 0.9312960786408805,
      "grad_norm": 3.210937261581421,
      "learning_rate": 1.0703743315508021e-05,
      "loss": 0.17777645587921143,
      "num_input_tokens_seen": 2241958,
      "step": 2179,
      "train_runtime": 5259.2869,
      "train_tokens_per_second": 426.286
    },
    {
      "epoch": 0.9317234747302062,
      "grad_norm": 2.386139392852783,
      "learning_rate": 1.0699465240641714e-05,
      "loss": 0.09151815623044968,
      "num_input_tokens_seen": 2242900,
      "step": 2180,
      "train_runtime": 5261.3169,
      "train_tokens_per_second": 426.3
    },
    {
      "epoch": 0.932150870819532,
      "grad_norm": 3.4541666507720947,
      "learning_rate": 1.0695187165775403e-05,
      "loss": 0.15929041802883148,
      "num_input_tokens_seen": 2243884,
      "step": 2181,
      "train_runtime": 5263.3934,
      "train_tokens_per_second": 426.319
    },
    {
      "epoch": 0.9325782669088578,
      "grad_norm": 3.3218729496002197,
      "learning_rate": 1.0690909090909091e-05,
      "loss": 0.12920552492141724,
      "num_input_tokens_seen": 2244708,
      "step": 2182,
      "train_runtime": 5265.4525,
      "train_tokens_per_second": 426.309
    },
    {
      "epoch": 0.9330056629981835,
      "grad_norm": 3.177752733230591,
      "learning_rate": 1.068663101604278e-05,
      "loss": 0.13437598943710327,
      "num_input_tokens_seen": 2245648,
      "step": 2183,
      "train_runtime": 5267.5631,
      "train_tokens_per_second": 426.316
    },
    {
      "epoch": 0.9334330590875094,
      "grad_norm": 2.9457602500915527,
      "learning_rate": 1.068235294117647e-05,
      "loss": 0.11178413778543472,
      "num_input_tokens_seen": 2246654,
      "step": 2184,
      "train_runtime": 5269.6641,
      "train_tokens_per_second": 426.337
    },
    {
      "epoch": 0.9338604551768351,
      "grad_norm": 3.3846898078918457,
      "learning_rate": 1.0678074866310163e-05,
      "loss": 0.15206705033779144,
      "num_input_tokens_seen": 2247496,
      "step": 2185,
      "train_runtime": 5271.7147,
      "train_tokens_per_second": 426.331
    },
    {
      "epoch": 0.9342878512661609,
      "grad_norm": 2.4759533405303955,
      "learning_rate": 1.067379679144385e-05,
      "loss": 0.09942358732223511,
      "num_input_tokens_seen": 2248292,
      "step": 2186,
      "train_runtime": 5273.7229,
      "train_tokens_per_second": 426.32
    },
    {
      "epoch": 0.9347152473554867,
      "grad_norm": 2.98865008354187,
      "learning_rate": 1.066951871657754e-05,
      "loss": 0.1888924539089203,
      "num_input_tokens_seen": 2249290,
      "step": 2187,
      "train_runtime": 5275.8049,
      "train_tokens_per_second": 426.341
    },
    {
      "epoch": 0.9351426434448125,
      "grad_norm": 3.50740909576416,
      "learning_rate": 1.066524064171123e-05,
      "loss": 0.15311260521411896,
      "num_input_tokens_seen": 2250184,
      "step": 2188,
      "train_runtime": 5277.9192,
      "train_tokens_per_second": 426.339
    },
    {
      "epoch": 0.9355700395341383,
      "grad_norm": 2.490849733352661,
      "learning_rate": 1.066096256684492e-05,
      "loss": 0.16449710726737976,
      "num_input_tokens_seen": 2251370,
      "step": 2189,
      "train_runtime": 5280.0296,
      "train_tokens_per_second": 426.393
    },
    {
      "epoch": 0.935997435623464,
      "grad_norm": 1.8838605880737305,
      "learning_rate": 1.065668449197861e-05,
      "loss": 0.08945582807064056,
      "num_input_tokens_seen": 2252332,
      "step": 2190,
      "train_runtime": 5282.0869,
      "train_tokens_per_second": 426.409
    },
    {
      "epoch": 0.9364248317127898,
      "grad_norm": 4.332616806030273,
      "learning_rate": 1.06524064171123e-05,
      "loss": 0.12057185173034668,
      "num_input_tokens_seen": 2252888,
      "step": 2191,
      "train_runtime": 5290.5661,
      "train_tokens_per_second": 425.831
    },
    {
      "epoch": 0.9368522278021156,
      "grad_norm": 2.9275875091552734,
      "learning_rate": 1.0648128342245989e-05,
      "loss": 0.09116809815168381,
      "num_input_tokens_seen": 2254060,
      "step": 2192,
      "train_runtime": 5292.7509,
      "train_tokens_per_second": 425.877
    },
    {
      "epoch": 0.9372796238914414,
      "grad_norm": 3.4034175872802734,
      "learning_rate": 1.064385026737968e-05,
      "loss": 0.22489213943481445,
      "num_input_tokens_seen": 2255156,
      "step": 2193,
      "train_runtime": 5294.8636,
      "train_tokens_per_second": 425.914
    },
    {
      "epoch": 0.9377070199807672,
      "grad_norm": 2.499889850616455,
      "learning_rate": 1.063957219251337e-05,
      "loss": 0.10469234734773636,
      "num_input_tokens_seen": 2256090,
      "step": 2194,
      "train_runtime": 5296.9134,
      "train_tokens_per_second": 425.925
    },
    {
      "epoch": 0.938134416070093,
      "grad_norm": 2.683138370513916,
      "learning_rate": 1.0635294117647059e-05,
      "loss": 0.08835846185684204,
      "num_input_tokens_seen": 2256706,
      "step": 2195,
      "train_runtime": 5298.9077,
      "train_tokens_per_second": 425.881
    },
    {
      "epoch": 0.9385618121594187,
      "grad_norm": 3.7897140979766846,
      "learning_rate": 1.0631016042780748e-05,
      "loss": 0.24589768052101135,
      "num_input_tokens_seen": 2257434,
      "step": 2196,
      "train_runtime": 5300.9293,
      "train_tokens_per_second": 425.856
    },
    {
      "epoch": 0.9389892082487445,
      "grad_norm": 3.4630603790283203,
      "learning_rate": 1.062673796791444e-05,
      "loss": 0.196588933467865,
      "num_input_tokens_seen": 2258554,
      "step": 2197,
      "train_runtime": 5303.0446,
      "train_tokens_per_second": 425.898
    },
    {
      "epoch": 0.9394166043380703,
      "grad_norm": 2.983121395111084,
      "learning_rate": 1.0622459893048129e-05,
      "loss": 0.12487363815307617,
      "num_input_tokens_seen": 2259392,
      "step": 2198,
      "train_runtime": 5305.0673,
      "train_tokens_per_second": 425.893
    },
    {
      "epoch": 0.9398440004273961,
      "grad_norm": 2.822087287902832,
      "learning_rate": 1.0618181818181818e-05,
      "loss": 0.16200515627861023,
      "num_input_tokens_seen": 2260566,
      "step": 2199,
      "train_runtime": 5307.2085,
      "train_tokens_per_second": 425.943
    },
    {
      "epoch": 0.9402713965167219,
      "grad_norm": 1.8860913515090942,
      "learning_rate": 1.0613903743315508e-05,
      "loss": 0.09443388879299164,
      "num_input_tokens_seen": 2262194,
      "step": 2200,
      "train_runtime": 5309.4582,
      "train_tokens_per_second": 426.069
    },
    {
      "epoch": 0.9406987926060476,
      "grad_norm": 3.1899352073669434,
      "learning_rate": 1.0609625668449199e-05,
      "loss": 0.1756967306137085,
      "num_input_tokens_seen": 2263188,
      "step": 2201,
      "train_runtime": 5311.5175,
      "train_tokens_per_second": 426.091
    },
    {
      "epoch": 0.9411261886953735,
      "grad_norm": 2.7726073265075684,
      "learning_rate": 1.0605347593582888e-05,
      "loss": 0.0826355516910553,
      "num_input_tokens_seen": 2263944,
      "step": 2202,
      "train_runtime": 5313.5107,
      "train_tokens_per_second": 426.073
    },
    {
      "epoch": 0.9415535847846992,
      "grad_norm": 4.586921691894531,
      "learning_rate": 1.0601069518716578e-05,
      "loss": 0.20692268013954163,
      "num_input_tokens_seen": 2264748,
      "step": 2203,
      "train_runtime": 5315.5709,
      "train_tokens_per_second": 426.059
    },
    {
      "epoch": 0.941980980874025,
      "grad_norm": 2.3063955307006836,
      "learning_rate": 1.0596791443850267e-05,
      "loss": 0.07637247443199158,
      "num_input_tokens_seen": 2265692,
      "step": 2204,
      "train_runtime": 5317.664,
      "train_tokens_per_second": 426.069
    },
    {
      "epoch": 0.9424083769633508,
      "grad_norm": 3.097294569015503,
      "learning_rate": 1.0592513368983958e-05,
      "loss": 0.1618078649044037,
      "num_input_tokens_seen": 2266638,
      "step": 2205,
      "train_runtime": 5319.723,
      "train_tokens_per_second": 426.082
    },
    {
      "epoch": 0.9428357730526765,
      "grad_norm": 2.5151431560516357,
      "learning_rate": 1.0588235294117648e-05,
      "loss": 0.08781378716230392,
      "num_input_tokens_seen": 2267848,
      "step": 2206,
      "train_runtime": 5321.8432,
      "train_tokens_per_second": 426.14
    },
    {
      "epoch": 0.9432631691420024,
      "grad_norm": 3.21832013130188,
      "learning_rate": 1.0583957219251337e-05,
      "loss": 0.1308228224515915,
      "num_input_tokens_seen": 2268764,
      "step": 2207,
      "train_runtime": 5323.9406,
      "train_tokens_per_second": 426.144
    },
    {
      "epoch": 0.9436905652313281,
      "grad_norm": 2.765855550765991,
      "learning_rate": 1.0579679144385027e-05,
      "loss": 0.11980103701353073,
      "num_input_tokens_seen": 2269744,
      "step": 2208,
      "train_runtime": 5326.0657,
      "train_tokens_per_second": 426.158
    },
    {
      "epoch": 0.944117961320654,
      "grad_norm": 2.2999863624572754,
      "learning_rate": 1.0575401069518718e-05,
      "loss": 0.10366262495517731,
      "num_input_tokens_seen": 2270726,
      "step": 2209,
      "train_runtime": 5328.1526,
      "train_tokens_per_second": 426.175
    },
    {
      "epoch": 0.9445453574099797,
      "grad_norm": 2.7298548221588135,
      "learning_rate": 1.0571122994652407e-05,
      "loss": 0.11786146461963654,
      "num_input_tokens_seen": 2271722,
      "step": 2210,
      "train_runtime": 5330.2559,
      "train_tokens_per_second": 426.194
    },
    {
      "epoch": 0.9449727534993054,
      "grad_norm": 2.8209102153778076,
      "learning_rate": 1.0566844919786097e-05,
      "loss": 0.17704381048679352,
      "num_input_tokens_seen": 2272812,
      "step": 2211,
      "train_runtime": 5332.3726,
      "train_tokens_per_second": 426.229
    },
    {
      "epoch": 0.9454001495886313,
      "grad_norm": 2.208386182785034,
      "learning_rate": 1.0562566844919786e-05,
      "loss": 0.16215816140174866,
      "num_input_tokens_seen": 2274198,
      "step": 2212,
      "train_runtime": 5334.5523,
      "train_tokens_per_second": 426.315
    },
    {
      "epoch": 0.945827545677957,
      "grad_norm": 3.8417880535125732,
      "learning_rate": 1.0558288770053477e-05,
      "loss": 0.18061214685440063,
      "num_input_tokens_seen": 2275008,
      "step": 2213,
      "train_runtime": 5336.5768,
      "train_tokens_per_second": 426.305
    },
    {
      "epoch": 0.9462549417672829,
      "grad_norm": 2.4419734477996826,
      "learning_rate": 1.0554010695187167e-05,
      "loss": 0.14664049446582794,
      "num_input_tokens_seen": 2275952,
      "step": 2214,
      "train_runtime": 5338.621,
      "train_tokens_per_second": 426.318
    },
    {
      "epoch": 0.9466823378566086,
      "grad_norm": 1.8086351156234741,
      "learning_rate": 1.0549732620320856e-05,
      "loss": 0.12906458973884583,
      "num_input_tokens_seen": 2277670,
      "step": 2215,
      "train_runtime": 5340.8503,
      "train_tokens_per_second": 426.462
    },
    {
      "epoch": 0.9471097339459343,
      "grad_norm": 2.719177722930908,
      "learning_rate": 1.0545454545454546e-05,
      "loss": 0.11620186269283295,
      "num_input_tokens_seen": 2278466,
      "step": 2216,
      "train_runtime": 5342.8901,
      "train_tokens_per_second": 426.448
    },
    {
      "epoch": 0.9475371300352602,
      "grad_norm": 2.628976345062256,
      "learning_rate": 1.0541176470588237e-05,
      "loss": 0.14144361019134521,
      "num_input_tokens_seen": 2279650,
      "step": 2217,
      "train_runtime": 5345.0009,
      "train_tokens_per_second": 426.501
    },
    {
      "epoch": 0.9479645261245859,
      "grad_norm": 3.304093599319458,
      "learning_rate": 1.0536898395721926e-05,
      "loss": 0.1491798311471939,
      "num_input_tokens_seen": 2280504,
      "step": 2218,
      "train_runtime": 5347.059,
      "train_tokens_per_second": 426.497
    },
    {
      "epoch": 0.9483919222139118,
      "grad_norm": 2.8636415004730225,
      "learning_rate": 1.0532620320855616e-05,
      "loss": 0.13599589467048645,
      "num_input_tokens_seen": 2281424,
      "step": 2219,
      "train_runtime": 5349.1083,
      "train_tokens_per_second": 426.505
    },
    {
      "epoch": 0.9488193183032375,
      "grad_norm": 2.402021646499634,
      "learning_rate": 1.0528342245989305e-05,
      "loss": 0.08079744130373001,
      "num_input_tokens_seen": 2282100,
      "step": 2220,
      "train_runtime": 5351.1099,
      "train_tokens_per_second": 426.472
    },
    {
      "epoch": 0.9492467143925634,
      "grad_norm": 3.220078945159912,
      "learning_rate": 1.0524064171122996e-05,
      "loss": 0.15267720818519592,
      "num_input_tokens_seen": 2283032,
      "step": 2221,
      "train_runtime": 5359.8284,
      "train_tokens_per_second": 425.952
    },
    {
      "epoch": 0.9496741104818891,
      "grad_norm": 2.245920181274414,
      "learning_rate": 1.0519786096256686e-05,
      "loss": 0.08796598762273788,
      "num_input_tokens_seen": 2284202,
      "step": 2222,
      "train_runtime": 5361.9319,
      "train_tokens_per_second": 426.004
    },
    {
      "epoch": 0.9501015065712148,
      "grad_norm": 2.347975015640259,
      "learning_rate": 1.0515508021390375e-05,
      "loss": 0.06638360023498535,
      "num_input_tokens_seen": 2285314,
      "step": 2223,
      "train_runtime": 5364.1799,
      "train_tokens_per_second": 426.032
    },
    {
      "epoch": 0.9505289026605407,
      "grad_norm": 3.0309362411499023,
      "learning_rate": 1.0511229946524065e-05,
      "loss": 0.1558932512998581,
      "num_input_tokens_seen": 2286416,
      "step": 2224,
      "train_runtime": 5366.4275,
      "train_tokens_per_second": 426.059
    },
    {
      "epoch": 0.9509562987498664,
      "grad_norm": 2.6483371257781982,
      "learning_rate": 1.0506951871657756e-05,
      "loss": 0.1320035755634308,
      "num_input_tokens_seen": 2287386,
      "step": 2225,
      "train_runtime": 5368.505,
      "train_tokens_per_second": 426.075
    },
    {
      "epoch": 0.9513836948391923,
      "grad_norm": 2.654061794281006,
      "learning_rate": 1.0502673796791445e-05,
      "loss": 0.13254696130752563,
      "num_input_tokens_seen": 2288430,
      "step": 2226,
      "train_runtime": 5370.6191,
      "train_tokens_per_second": 426.102
    },
    {
      "epoch": 0.951811090928518,
      "grad_norm": 3.8444437980651855,
      "learning_rate": 1.0498395721925135e-05,
      "loss": 0.20998886227607727,
      "num_input_tokens_seen": 2289104,
      "step": 2227,
      "train_runtime": 5372.5913,
      "train_tokens_per_second": 426.071
    },
    {
      "epoch": 0.9522384870178437,
      "grad_norm": 2.013779878616333,
      "learning_rate": 1.0494117647058824e-05,
      "loss": 0.12033921480178833,
      "num_input_tokens_seen": 2290770,
      "step": 2228,
      "train_runtime": 5374.8506,
      "train_tokens_per_second": 426.202
    },
    {
      "epoch": 0.9526658831071696,
      "grad_norm": 2.6551356315612793,
      "learning_rate": 1.0489839572192515e-05,
      "loss": 0.11887505650520325,
      "num_input_tokens_seen": 2292226,
      "step": 2229,
      "train_runtime": 5377.0349,
      "train_tokens_per_second": 426.299
    },
    {
      "epoch": 0.9530932791964953,
      "grad_norm": 2.649827003479004,
      "learning_rate": 1.0485561497326205e-05,
      "loss": 0.12751273810863495,
      "num_input_tokens_seen": 2293084,
      "step": 2230,
      "train_runtime": 5379.0791,
      "train_tokens_per_second": 426.297
    },
    {
      "epoch": 0.9535206752858212,
      "grad_norm": 1.9154528379440308,
      "learning_rate": 1.0481283422459894e-05,
      "loss": 0.10436928272247314,
      "num_input_tokens_seen": 2294576,
      "step": 2231,
      "train_runtime": 5381.2582,
      "train_tokens_per_second": 426.401
    },
    {
      "epoch": 0.9539480713751469,
      "grad_norm": 2.9512860774993896,
      "learning_rate": 1.0477005347593584e-05,
      "loss": 0.1171833872795105,
      "num_input_tokens_seen": 2295456,
      "step": 2232,
      "train_runtime": 5383.3118,
      "train_tokens_per_second": 426.402
    },
    {
      "epoch": 0.9543754674644727,
      "grad_norm": 2.9378502368927,
      "learning_rate": 1.0472727272727275e-05,
      "loss": 0.12822403013706207,
      "num_input_tokens_seen": 2296482,
      "step": 2233,
      "train_runtime": 5385.3564,
      "train_tokens_per_second": 426.431
    },
    {
      "epoch": 0.9548028635537985,
      "grad_norm": 3.5034008026123047,
      "learning_rate": 1.0468449197860964e-05,
      "loss": 0.1279274970293045,
      "num_input_tokens_seen": 2297264,
      "step": 2234,
      "train_runtime": 5387.3694,
      "train_tokens_per_second": 426.417
    },
    {
      "epoch": 0.9552302596431242,
      "grad_norm": 2.428328275680542,
      "learning_rate": 1.0464171122994654e-05,
      "loss": 0.2062709480524063,
      "num_input_tokens_seen": 2298784,
      "step": 2235,
      "train_runtime": 5389.5736,
      "train_tokens_per_second": 426.524
    },
    {
      "epoch": 0.9556576557324501,
      "grad_norm": 2.7964532375335693,
      "learning_rate": 1.0459893048128343e-05,
      "loss": 0.08819354325532913,
      "num_input_tokens_seen": 2299798,
      "step": 2236,
      "train_runtime": 5391.6384,
      "train_tokens_per_second": 426.549
    },
    {
      "epoch": 0.9560850518217758,
      "grad_norm": 1.3275729417800903,
      "learning_rate": 1.0455614973262034e-05,
      "loss": 0.0548393651843071,
      "num_input_tokens_seen": 2301808,
      "step": 2237,
      "train_runtime": 5393.9812,
      "train_tokens_per_second": 426.736
    },
    {
      "epoch": 0.9565124479111016,
      "grad_norm": 2.5672008991241455,
      "learning_rate": 1.0451336898395724e-05,
      "loss": 0.08960367739200592,
      "num_input_tokens_seen": 2302694,
      "step": 2238,
      "train_runtime": 5396.0196,
      "train_tokens_per_second": 426.739
    },
    {
      "epoch": 0.9569398440004274,
      "grad_norm": 2.848595380783081,
      "learning_rate": 1.0447058823529413e-05,
      "loss": 0.179183691740036,
      "num_input_tokens_seen": 2303744,
      "step": 2239,
      "train_runtime": 5398.123,
      "train_tokens_per_second": 426.768
    },
    {
      "epoch": 0.9573672400897532,
      "grad_norm": 3.3253800868988037,
      "learning_rate": 1.04427807486631e-05,
      "loss": 0.16926611959934235,
      "num_input_tokens_seen": 2304580,
      "step": 2240,
      "train_runtime": 5400.1769,
      "train_tokens_per_second": 426.76
    },
    {
      "epoch": 0.957794636179079,
      "grad_norm": 3.1977779865264893,
      "learning_rate": 1.0438502673796794e-05,
      "loss": 0.13291694223880768,
      "num_input_tokens_seen": 2305478,
      "step": 2241,
      "train_runtime": 5402.2191,
      "train_tokens_per_second": 426.765
    },
    {
      "epoch": 0.9582220322684047,
      "grad_norm": 2.222672939300537,
      "learning_rate": 1.0434224598930483e-05,
      "loss": 0.10854735970497131,
      "num_input_tokens_seen": 2306700,
      "step": 2242,
      "train_runtime": 5404.3714,
      "train_tokens_per_second": 426.821
    },
    {
      "epoch": 0.9586494283577305,
      "grad_norm": 1.9616283178329468,
      "learning_rate": 1.0429946524064173e-05,
      "loss": 0.1156691312789917,
      "num_input_tokens_seen": 2307916,
      "step": 2243,
      "train_runtime": 5406.5203,
      "train_tokens_per_second": 426.876
    },
    {
      "epoch": 0.9590768244470563,
      "grad_norm": 2.538754463195801,
      "learning_rate": 1.042566844919786e-05,
      "loss": 0.11380726844072342,
      "num_input_tokens_seen": 2308914,
      "step": 2244,
      "train_runtime": 5408.6003,
      "train_tokens_per_second": 426.897
    },
    {
      "epoch": 0.9595042205363821,
      "grad_norm": 3.4854772090911865,
      "learning_rate": 1.0421390374331553e-05,
      "loss": 0.15792405605316162,
      "num_input_tokens_seen": 2309942,
      "step": 2245,
      "train_runtime": 5410.7427,
      "train_tokens_per_second": 426.918
    },
    {
      "epoch": 0.9599316166257079,
      "grad_norm": 3.3573896884918213,
      "learning_rate": 1.0417112299465243e-05,
      "loss": 0.1561889946460724,
      "num_input_tokens_seen": 2311726,
      "step": 2246,
      "train_runtime": 5413.0602,
      "train_tokens_per_second": 427.065
    },
    {
      "epoch": 0.9603590127150337,
      "grad_norm": 2.3635971546173096,
      "learning_rate": 1.041283422459893e-05,
      "loss": 0.14996930956840515,
      "num_input_tokens_seen": 2313058,
      "step": 2247,
      "train_runtime": 5415.2033,
      "train_tokens_per_second": 427.141
    },
    {
      "epoch": 0.9607864088043594,
      "grad_norm": 2.4530446529388428,
      "learning_rate": 1.040855614973262e-05,
      "loss": 0.09206946194171906,
      "num_input_tokens_seen": 2313840,
      "step": 2248,
      "train_runtime": 5417.2257,
      "train_tokens_per_second": 427.126
    },
    {
      "epoch": 0.9612138048936852,
      "grad_norm": 2.651663303375244,
      "learning_rate": 1.0404278074866313e-05,
      "loss": 0.08442568778991699,
      "num_input_tokens_seen": 2314458,
      "step": 2249,
      "train_runtime": 5419.2172,
      "train_tokens_per_second": 427.083
    },
    {
      "epoch": 0.961641200983011,
      "grad_norm": 3.452334403991699,
      "learning_rate": 1.04e-05,
      "loss": 0.16761603951454163,
      "num_input_tokens_seen": 2315338,
      "step": 2250,
      "train_runtime": 5421.2615,
      "train_tokens_per_second": 427.085
    },
    {
      "epoch": 0.9620685970723368,
      "grad_norm": 3.322423219680786,
      "learning_rate": 1.039572192513369e-05,
      "loss": 0.09940492361783981,
      "num_input_tokens_seen": 2316192,
      "step": 2251,
      "train_runtime": 5430.1713,
      "train_tokens_per_second": 426.541
    },
    {
      "epoch": 0.9624959931616626,
      "grad_norm": 4.1046929359436035,
      "learning_rate": 1.039144385026738e-05,
      "loss": 0.11258722096681595,
      "num_input_tokens_seen": 2316778,
      "step": 2252,
      "train_runtime": 5432.4252,
      "train_tokens_per_second": 426.472
    },
    {
      "epoch": 0.9629233892509883,
      "grad_norm": 2.8597140312194824,
      "learning_rate": 1.0387165775401072e-05,
      "loss": 0.12939955294132233,
      "num_input_tokens_seen": 2317742,
      "step": 2253,
      "train_runtime": 5434.5044,
      "train_tokens_per_second": 426.486
    },
    {
      "epoch": 0.9633507853403142,
      "grad_norm": 1.3937209844589233,
      "learning_rate": 1.038288770053476e-05,
      "loss": 0.10120612382888794,
      "num_input_tokens_seen": 2321922,
      "step": 2254,
      "train_runtime": 5437.4019,
      "train_tokens_per_second": 427.028
    },
    {
      "epoch": 0.9637781814296399,
      "grad_norm": 2.872166156768799,
      "learning_rate": 1.037860962566845e-05,
      "loss": 0.15230093896389008,
      "num_input_tokens_seen": 2323026,
      "step": 2255,
      "train_runtime": 5439.5505,
      "train_tokens_per_second": 427.062
    },
    {
      "epoch": 0.9642055775189657,
      "grad_norm": 2.6004855632781982,
      "learning_rate": 1.0374331550802139e-05,
      "loss": 0.08310409635305405,
      "num_input_tokens_seen": 2323952,
      "step": 2256,
      "train_runtime": 5441.6664,
      "train_tokens_per_second": 427.066
    },
    {
      "epoch": 0.9646329736082915,
      "grad_norm": 2.575425386428833,
      "learning_rate": 1.037005347593583e-05,
      "loss": 0.1307641565799713,
      "num_input_tokens_seen": 2324858,
      "step": 2257,
      "train_runtime": 5443.784,
      "train_tokens_per_second": 427.067
    },
    {
      "epoch": 0.9650603696976172,
      "grad_norm": 4.051435470581055,
      "learning_rate": 1.036577540106952e-05,
      "loss": 0.2844654619693756,
      "num_input_tokens_seen": 2325884,
      "step": 2258,
      "train_runtime": 5445.9245,
      "train_tokens_per_second": 427.087
    },
    {
      "epoch": 0.9654877657869431,
      "grad_norm": 3.0446956157684326,
      "learning_rate": 1.0361497326203209e-05,
      "loss": 0.22520825266838074,
      "num_input_tokens_seen": 2326978,
      "step": 2259,
      "train_runtime": 5448.0278,
      "train_tokens_per_second": 427.123
    },
    {
      "epoch": 0.9659151618762688,
      "grad_norm": 2.1535825729370117,
      "learning_rate": 1.0357219251336898e-05,
      "loss": 0.14550763368606567,
      "num_input_tokens_seen": 2328416,
      "step": 2260,
      "train_runtime": 5450.2343,
      "train_tokens_per_second": 427.214
    },
    {
      "epoch": 0.9663425579655947,
      "grad_norm": 1.7815985679626465,
      "learning_rate": 1.035294117647059e-05,
      "loss": 0.1046629548072815,
      "num_input_tokens_seen": 2329686,
      "step": 2261,
      "train_runtime": 5452.484,
      "train_tokens_per_second": 427.271
    },
    {
      "epoch": 0.9667699540549204,
      "grad_norm": 2.6957361698150635,
      "learning_rate": 1.0348663101604279e-05,
      "loss": 0.15473268926143646,
      "num_input_tokens_seen": 2330696,
      "step": 2262,
      "train_runtime": 5454.5584,
      "train_tokens_per_second": 427.293
    },
    {
      "epoch": 0.9671973501442461,
      "grad_norm": 2.146594285964966,
      "learning_rate": 1.0344385026737968e-05,
      "loss": 0.12918353080749512,
      "num_input_tokens_seen": 2332482,
      "step": 2263,
      "train_runtime": 5456.8453,
      "train_tokens_per_second": 427.441
    },
    {
      "epoch": 0.967624746233572,
      "grad_norm": 3.9054689407348633,
      "learning_rate": 1.0340106951871658e-05,
      "loss": 0.09738944470882416,
      "num_input_tokens_seen": 2333176,
      "step": 2264,
      "train_runtime": 5458.8675,
      "train_tokens_per_second": 427.41
    },
    {
      "epoch": 0.9680521423228977,
      "grad_norm": 2.7858171463012695,
      "learning_rate": 1.0335828877005349e-05,
      "loss": 0.12292246520519257,
      "num_input_tokens_seen": 2333958,
      "step": 2265,
      "train_runtime": 5460.8811,
      "train_tokens_per_second": 427.396
    },
    {
      "epoch": 0.9684795384122236,
      "grad_norm": 2.3060765266418457,
      "learning_rate": 1.0331550802139038e-05,
      "loss": 0.06947749853134155,
      "num_input_tokens_seen": 2334746,
      "step": 2266,
      "train_runtime": 5462.9097,
      "train_tokens_per_second": 427.381
    },
    {
      "epoch": 0.9689069345015493,
      "grad_norm": 3.6497702598571777,
      "learning_rate": 1.0327272727272728e-05,
      "loss": 0.18004286289215088,
      "num_input_tokens_seen": 2335400,
      "step": 2267,
      "train_runtime": 5464.9113,
      "train_tokens_per_second": 427.345
    },
    {
      "epoch": 0.9693343305908751,
      "grad_norm": 2.7571423053741455,
      "learning_rate": 1.0322994652406417e-05,
      "loss": 0.11387020349502563,
      "num_input_tokens_seen": 2336252,
      "step": 2268,
      "train_runtime": 5466.9622,
      "train_tokens_per_second": 427.34
    },
    {
      "epoch": 0.9697617266802009,
      "grad_norm": 2.0217416286468506,
      "learning_rate": 1.0318716577540108e-05,
      "loss": 0.08127065002918243,
      "num_input_tokens_seen": 2337326,
      "step": 2269,
      "train_runtime": 5469.0231,
      "train_tokens_per_second": 427.375
    },
    {
      "epoch": 0.9701891227695266,
      "grad_norm": 1.7019882202148438,
      "learning_rate": 1.0314438502673798e-05,
      "loss": 0.08605479449033737,
      "num_input_tokens_seen": 2338610,
      "step": 2270,
      "train_runtime": 5471.2288,
      "train_tokens_per_second": 427.438
    },
    {
      "epoch": 0.9706165188588525,
      "grad_norm": 2.190519094467163,
      "learning_rate": 1.0310160427807487e-05,
      "loss": 0.11524763703346252,
      "num_input_tokens_seen": 2339596,
      "step": 2271,
      "train_runtime": 5473.3058,
      "train_tokens_per_second": 427.456
    },
    {
      "epoch": 0.9710439149481782,
      "grad_norm": 2.6079463958740234,
      "learning_rate": 1.0305882352941177e-05,
      "loss": 0.10180608928203583,
      "num_input_tokens_seen": 2340366,
      "step": 2272,
      "train_runtime": 5475.3481,
      "train_tokens_per_second": 427.437
    },
    {
      "epoch": 0.9714713110375041,
      "grad_norm": 2.7155380249023438,
      "learning_rate": 1.0301604278074868e-05,
      "loss": 0.14282962679862976,
      "num_input_tokens_seen": 2341696,
      "step": 2273,
      "train_runtime": 5477.5329,
      "train_tokens_per_second": 427.509
    },
    {
      "epoch": 0.9718987071268298,
      "grad_norm": 2.5958855152130127,
      "learning_rate": 1.0297326203208557e-05,
      "loss": 0.0968252569437027,
      "num_input_tokens_seen": 2342896,
      "step": 2274,
      "train_runtime": 5479.6661,
      "train_tokens_per_second": 427.562
    },
    {
      "epoch": 0.9723261032161555,
      "grad_norm": 3.645071268081665,
      "learning_rate": 1.0293048128342247e-05,
      "loss": 0.18518459796905518,
      "num_input_tokens_seen": 2343656,
      "step": 2275,
      "train_runtime": 5481.7199,
      "train_tokens_per_second": 427.54
    },
    {
      "epoch": 0.9727534993054814,
      "grad_norm": 3.1911346912384033,
      "learning_rate": 1.0288770053475936e-05,
      "loss": 0.09797079861164093,
      "num_input_tokens_seen": 2344640,
      "step": 2276,
      "train_runtime": 5483.7652,
      "train_tokens_per_second": 427.56
    },
    {
      "epoch": 0.9731808953948071,
      "grad_norm": 3.199688196182251,
      "learning_rate": 1.0284491978609627e-05,
      "loss": 0.13134604692459106,
      "num_input_tokens_seen": 2345396,
      "step": 2277,
      "train_runtime": 5485.7867,
      "train_tokens_per_second": 427.54
    },
    {
      "epoch": 0.973608291484133,
      "grad_norm": 3.5392403602600098,
      "learning_rate": 1.0280213903743317e-05,
      "loss": 0.10100971907377243,
      "num_input_tokens_seen": 2346082,
      "step": 2278,
      "train_runtime": 5487.795,
      "train_tokens_per_second": 427.509
    },
    {
      "epoch": 0.9740356875734587,
      "grad_norm": 2.7689268589019775,
      "learning_rate": 1.0275935828877006e-05,
      "loss": 0.10392604768276215,
      "num_input_tokens_seen": 2347168,
      "step": 2279,
      "train_runtime": 5489.876,
      "train_tokens_per_second": 427.545
    },
    {
      "epoch": 0.9744630836627844,
      "grad_norm": 0.8841018080711365,
      "learning_rate": 1.0271657754010696e-05,
      "loss": 0.03436458110809326,
      "num_input_tokens_seen": 2349502,
      "step": 2280,
      "train_runtime": 5492.268,
      "train_tokens_per_second": 427.784
    },
    {
      "epoch": 0.9748904797521103,
      "grad_norm": 3.811055898666382,
      "learning_rate": 1.0267379679144387e-05,
      "loss": 0.23737823963165283,
      "num_input_tokens_seen": 2350714,
      "step": 2281,
      "train_runtime": 5500.9949,
      "train_tokens_per_second": 427.325
    },
    {
      "epoch": 0.975317875841436,
      "grad_norm": 3.6392815113067627,
      "learning_rate": 1.0263101604278076e-05,
      "loss": 0.16858138144016266,
      "num_input_tokens_seen": 2351552,
      "step": 2282,
      "train_runtime": 5503.1836,
      "train_tokens_per_second": 427.308
    },
    {
      "epoch": 0.9757452719307619,
      "grad_norm": 3.2982568740844727,
      "learning_rate": 1.0258823529411766e-05,
      "loss": 0.10996103286743164,
      "num_input_tokens_seen": 2352132,
      "step": 2283,
      "train_runtime": 5505.1415,
      "train_tokens_per_second": 427.261
    },
    {
      "epoch": 0.9761726680200876,
      "grad_norm": 2.979600429534912,
      "learning_rate": 1.0254545454545455e-05,
      "loss": 0.1532599776983261,
      "num_input_tokens_seen": 2353134,
      "step": 2284,
      "train_runtime": 5507.2381,
      "train_tokens_per_second": 427.28
    },
    {
      "epoch": 0.9766000641094134,
      "grad_norm": 3.2876968383789062,
      "learning_rate": 1.0250267379679146e-05,
      "loss": 0.15251268446445465,
      "num_input_tokens_seen": 2354438,
      "step": 2285,
      "train_runtime": 5509.4438,
      "train_tokens_per_second": 427.346
    },
    {
      "epoch": 0.9770274601987392,
      "grad_norm": 3.4809560775756836,
      "learning_rate": 1.0245989304812836e-05,
      "loss": 0.1291067898273468,
      "num_input_tokens_seen": 2355542,
      "step": 2286,
      "train_runtime": 5511.5623,
      "train_tokens_per_second": 427.382
    },
    {
      "epoch": 0.9774548562880649,
      "grad_norm": 3.3865761756896973,
      "learning_rate": 1.0241711229946525e-05,
      "loss": 0.1919110119342804,
      "num_input_tokens_seen": 2356584,
      "step": 2287,
      "train_runtime": 5513.6323,
      "train_tokens_per_second": 427.41
    },
    {
      "epoch": 0.9778822523773908,
      "grad_norm": 1.9207488298416138,
      "learning_rate": 1.0237433155080215e-05,
      "loss": 0.08086986839771271,
      "num_input_tokens_seen": 2357818,
      "step": 2288,
      "train_runtime": 5515.8052,
      "train_tokens_per_second": 427.466
    },
    {
      "epoch": 0.9783096484667165,
      "grad_norm": 2.593085765838623,
      "learning_rate": 1.0233155080213906e-05,
      "loss": 0.1042538434267044,
      "num_input_tokens_seen": 2358784,
      "step": 2289,
      "train_runtime": 5517.8907,
      "train_tokens_per_second": 427.479
    },
    {
      "epoch": 0.9787370445560423,
      "grad_norm": 3.9618942737579346,
      "learning_rate": 1.0228877005347595e-05,
      "loss": 0.15336662530899048,
      "num_input_tokens_seen": 2359760,
      "step": 2290,
      "train_runtime": 5519.9742,
      "train_tokens_per_second": 427.495
    },
    {
      "epoch": 0.9791644406453681,
      "grad_norm": 2.0823564529418945,
      "learning_rate": 1.0224598930481285e-05,
      "loss": 0.12162894755601883,
      "num_input_tokens_seen": 2361362,
      "step": 2291,
      "train_runtime": 5522.2566,
      "train_tokens_per_second": 427.608
    },
    {
      "epoch": 0.9795918367346939,
      "grad_norm": 2.362786054611206,
      "learning_rate": 1.0220320855614974e-05,
      "loss": 0.16087386012077332,
      "num_input_tokens_seen": 2362814,
      "step": 2292,
      "train_runtime": 5524.4503,
      "train_tokens_per_second": 427.701
    },
    {
      "epoch": 0.9800192328240197,
      "grad_norm": 2.899754285812378,
      "learning_rate": 1.0216042780748663e-05,
      "loss": 0.12003335356712341,
      "num_input_tokens_seen": 2363820,
      "step": 2293,
      "train_runtime": 5526.5585,
      "train_tokens_per_second": 427.72
    },
    {
      "epoch": 0.9804466289133454,
      "grad_norm": 2.4319729804992676,
      "learning_rate": 1.0211764705882355e-05,
      "loss": 0.11342653632164001,
      "num_input_tokens_seen": 2364828,
      "step": 2294,
      "train_runtime": 5528.6451,
      "train_tokens_per_second": 427.741
    },
    {
      "epoch": 0.9808740250026712,
      "grad_norm": 2.9359421730041504,
      "learning_rate": 1.0207486631016044e-05,
      "loss": 0.1248394101858139,
      "num_input_tokens_seen": 2365708,
      "step": 2295,
      "train_runtime": 5530.6919,
      "train_tokens_per_second": 427.742
    },
    {
      "epoch": 0.981301421091997,
      "grad_norm": 3.2101783752441406,
      "learning_rate": 1.0203208556149734e-05,
      "loss": 0.16110453009605408,
      "num_input_tokens_seen": 2367262,
      "step": 2296,
      "train_runtime": 5532.9717,
      "train_tokens_per_second": 427.846
    },
    {
      "epoch": 0.9817288171813228,
      "grad_norm": 2.193382740020752,
      "learning_rate": 1.0198930481283423e-05,
      "loss": 0.08950398117303848,
      "num_input_tokens_seen": 2368474,
      "step": 2297,
      "train_runtime": 5535.1533,
      "train_tokens_per_second": 427.897
    },
    {
      "epoch": 0.9821562132706486,
      "grad_norm": 4.416612148284912,
      "learning_rate": 1.0194652406417114e-05,
      "loss": 0.1765228509902954,
      "num_input_tokens_seen": 2369186,
      "step": 2298,
      "train_runtime": 5537.1679,
      "train_tokens_per_second": 427.87
    },
    {
      "epoch": 0.9825836093599744,
      "grad_norm": 4.482437610626221,
      "learning_rate": 1.0190374331550804e-05,
      "loss": 0.1840195655822754,
      "num_input_tokens_seen": 2369774,
      "step": 2299,
      "train_runtime": 5539.1493,
      "train_tokens_per_second": 427.823
    },
    {
      "epoch": 0.9830110054493001,
      "grad_norm": 2.900217056274414,
      "learning_rate": 1.0186096256684493e-05,
      "loss": 0.1668633222579956,
      "num_input_tokens_seen": 2370796,
      "step": 2300,
      "train_runtime": 5541.2375,
      "train_tokens_per_second": 427.846
    },
    {
      "epoch": 0.9834384015386259,
      "grad_norm": 2.4736578464508057,
      "learning_rate": 1.0181818181818182e-05,
      "loss": 0.09531894326210022,
      "num_input_tokens_seen": 2371418,
      "step": 2301,
      "train_runtime": 5543.2438,
      "train_tokens_per_second": 427.803
    },
    {
      "epoch": 0.9838657976279517,
      "grad_norm": 2.863168954849243,
      "learning_rate": 1.0177540106951874e-05,
      "loss": 0.14712025225162506,
      "num_input_tokens_seen": 2372554,
      "step": 2302,
      "train_runtime": 5545.4163,
      "train_tokens_per_second": 427.841
    },
    {
      "epoch": 0.9842931937172775,
      "grad_norm": 2.756906747817993,
      "learning_rate": 1.0173262032085563e-05,
      "loss": 0.08874457329511642,
      "num_input_tokens_seen": 2373208,
      "step": 2303,
      "train_runtime": 5547.4066,
      "train_tokens_per_second": 427.805
    },
    {
      "epoch": 0.9847205898066033,
      "grad_norm": 2.8874385356903076,
      "learning_rate": 1.0168983957219252e-05,
      "loss": 0.17877505719661713,
      "num_input_tokens_seen": 2374430,
      "step": 2304,
      "train_runtime": 5549.5587,
      "train_tokens_per_second": 427.859
    },
    {
      "epoch": 0.985147985895929,
      "grad_norm": 2.769883632659912,
      "learning_rate": 1.016470588235294e-05,
      "loss": 0.08705952763557434,
      "num_input_tokens_seen": 2375258,
      "step": 2305,
      "train_runtime": 5551.6048,
      "train_tokens_per_second": 427.851
    },
    {
      "epoch": 0.9855753819852549,
      "grad_norm": 3.491914987564087,
      "learning_rate": 1.0160427807486633e-05,
      "loss": 0.17664125561714172,
      "num_input_tokens_seen": 2376102,
      "step": 2306,
      "train_runtime": 5553.6291,
      "train_tokens_per_second": 427.847
    },
    {
      "epoch": 0.9860027780745806,
      "grad_norm": 3.1204919815063477,
      "learning_rate": 1.0156149732620322e-05,
      "loss": 0.10803135484457016,
      "num_input_tokens_seen": 2376956,
      "step": 2307,
      "train_runtime": 5555.7173,
      "train_tokens_per_second": 427.84
    },
    {
      "epoch": 0.9864301741639064,
      "grad_norm": 1.9673091173171997,
      "learning_rate": 1.015187165775401e-05,
      "loss": 0.13556765019893646,
      "num_input_tokens_seen": 2378598,
      "step": 2308,
      "train_runtime": 5557.9534,
      "train_tokens_per_second": 427.963
    },
    {
      "epoch": 0.9868575702532322,
      "grad_norm": 1.5314304828643799,
      "learning_rate": 1.01475935828877e-05,
      "loss": 0.06075898930430412,
      "num_input_tokens_seen": 2380048,
      "step": 2309,
      "train_runtime": 5560.1551,
      "train_tokens_per_second": 428.054
    },
    {
      "epoch": 0.9872849663425579,
      "grad_norm": 3.44759202003479,
      "learning_rate": 1.0143315508021393e-05,
      "loss": 0.1153748631477356,
      "num_input_tokens_seen": 2380694,
      "step": 2310,
      "train_runtime": 5562.1571,
      "train_tokens_per_second": 428.016
    },
    {
      "epoch": 0.9877123624318838,
      "grad_norm": 2.410386323928833,
      "learning_rate": 1.0139037433155082e-05,
      "loss": 0.12743216753005981,
      "num_input_tokens_seen": 2381864,
      "step": 2311,
      "train_runtime": 5571.2401,
      "train_tokens_per_second": 427.529
    },
    {
      "epoch": 0.9881397585212095,
      "grad_norm": 2.447937488555908,
      "learning_rate": 1.013475935828877e-05,
      "loss": 0.11452856659889221,
      "num_input_tokens_seen": 2382784,
      "step": 2312,
      "train_runtime": 5573.2989,
      "train_tokens_per_second": 427.536
    },
    {
      "epoch": 0.9885671546105353,
      "grad_norm": 3.982205867767334,
      "learning_rate": 1.013048128342246e-05,
      "loss": 0.2600610554218292,
      "num_input_tokens_seen": 2383892,
      "step": 2313,
      "train_runtime": 5575.3681,
      "train_tokens_per_second": 427.576
    },
    {
      "epoch": 0.9889945506998611,
      "grad_norm": 3.205493927001953,
      "learning_rate": 1.0126203208556152e-05,
      "loss": 0.14277592301368713,
      "num_input_tokens_seen": 2384586,
      "step": 2314,
      "train_runtime": 5577.4022,
      "train_tokens_per_second": 427.544
    },
    {
      "epoch": 0.9894219467891868,
      "grad_norm": 3.331968069076538,
      "learning_rate": 1.012192513368984e-05,
      "loss": 0.15675318241119385,
      "num_input_tokens_seen": 2385480,
      "step": 2315,
      "train_runtime": 5579.4646,
      "train_tokens_per_second": 427.546
    },
    {
      "epoch": 0.9898493428785127,
      "grad_norm": 3.521432638168335,
      "learning_rate": 1.011764705882353e-05,
      "loss": 0.17289461195468903,
      "num_input_tokens_seen": 2386470,
      "step": 2316,
      "train_runtime": 5581.54,
      "train_tokens_per_second": 427.565
    },
    {
      "epoch": 0.9902767389678384,
      "grad_norm": 2.799103021621704,
      "learning_rate": 1.0113368983957219e-05,
      "loss": 0.13407383859157562,
      "num_input_tokens_seen": 2387406,
      "step": 2317,
      "train_runtime": 5583.6535,
      "train_tokens_per_second": 427.571
    },
    {
      "epoch": 0.9907041350571643,
      "grad_norm": 2.995525598526001,
      "learning_rate": 1.010909090909091e-05,
      "loss": 0.17694485187530518,
      "num_input_tokens_seen": 2388554,
      "step": 2318,
      "train_runtime": 5585.769,
      "train_tokens_per_second": 427.614
    },
    {
      "epoch": 0.99113153114649,
      "grad_norm": 2.79598069190979,
      "learning_rate": 1.01048128342246e-05,
      "loss": 0.178671196103096,
      "num_input_tokens_seen": 2389436,
      "step": 2319,
      "train_runtime": 5587.8149,
      "train_tokens_per_second": 427.615
    },
    {
      "epoch": 0.9915589272358158,
      "grad_norm": 2.9074149131774902,
      "learning_rate": 1.0100534759358289e-05,
      "loss": 0.12709584832191467,
      "num_input_tokens_seen": 2390262,
      "step": 2320,
      "train_runtime": 5589.8508,
      "train_tokens_per_second": 427.607
    },
    {
      "epoch": 0.9919863233251416,
      "grad_norm": 2.007220983505249,
      "learning_rate": 1.0096256684491978e-05,
      "loss": 0.11949099600315094,
      "num_input_tokens_seen": 2391706,
      "step": 2321,
      "train_runtime": 5592.0403,
      "train_tokens_per_second": 427.698
    },
    {
      "epoch": 0.9924137194144673,
      "grad_norm": 3.6360058784484863,
      "learning_rate": 1.009197860962567e-05,
      "loss": 0.20659798383712769,
      "num_input_tokens_seen": 2392578,
      "step": 2322,
      "train_runtime": 5594.2029,
      "train_tokens_per_second": 427.689
    },
    {
      "epoch": 0.9928411155037932,
      "grad_norm": 2.549527406692505,
      "learning_rate": 1.0087700534759359e-05,
      "loss": 0.12216900289058685,
      "num_input_tokens_seen": 2393696,
      "step": 2323,
      "train_runtime": 5596.3349,
      "train_tokens_per_second": 427.726
    },
    {
      "epoch": 0.9932685115931189,
      "grad_norm": 3.211176633834839,
      "learning_rate": 1.0083422459893048e-05,
      "loss": 0.1540602445602417,
      "num_input_tokens_seen": 2394578,
      "step": 2324,
      "train_runtime": 5598.3838,
      "train_tokens_per_second": 427.727
    },
    {
      "epoch": 0.9936959076824448,
      "grad_norm": 3.5962655544281006,
      "learning_rate": 1.0079144385026738e-05,
      "loss": 0.16790716350078583,
      "num_input_tokens_seen": 2395308,
      "step": 2325,
      "train_runtime": 5600.4532,
      "train_tokens_per_second": 427.699
    },
    {
      "epoch": 0.9941233037717705,
      "grad_norm": 1.883987545967102,
      "learning_rate": 1.0074866310160429e-05,
      "loss": 0.08721403777599335,
      "num_input_tokens_seen": 2396636,
      "step": 2326,
      "train_runtime": 5602.6203,
      "train_tokens_per_second": 427.771
    },
    {
      "epoch": 0.9945506998610962,
      "grad_norm": 3.5228617191314697,
      "learning_rate": 1.0070588235294118e-05,
      "loss": 0.1339649111032486,
      "num_input_tokens_seen": 2397360,
      "step": 2327,
      "train_runtime": 5604.5983,
      "train_tokens_per_second": 427.749
    },
    {
      "epoch": 0.9949780959504221,
      "grad_norm": 2.3463289737701416,
      "learning_rate": 1.0066310160427808e-05,
      "loss": 0.13015425205230713,
      "num_input_tokens_seen": 2398838,
      "step": 2328,
      "train_runtime": 5606.764,
      "train_tokens_per_second": 427.847
    },
    {
      "epoch": 0.9954054920397478,
      "grad_norm": 3.0615673065185547,
      "learning_rate": 1.0062032085561497e-05,
      "loss": 0.1618393212556839,
      "num_input_tokens_seen": 2399932,
      "step": 2329,
      "train_runtime": 5608.9576,
      "train_tokens_per_second": 427.875
    },
    {
      "epoch": 0.9958328881290737,
      "grad_norm": 3.1222710609436035,
      "learning_rate": 1.0057754010695188e-05,
      "loss": 0.12976691126823425,
      "num_input_tokens_seen": 2400660,
      "step": 2330,
      "train_runtime": 5611.0058,
      "train_tokens_per_second": 427.848
    },
    {
      "epoch": 0.9962602842183994,
      "grad_norm": 2.752530813217163,
      "learning_rate": 1.0053475935828878e-05,
      "loss": 0.11259917914867401,
      "num_input_tokens_seen": 2401812,
      "step": 2331,
      "train_runtime": 5613.1447,
      "train_tokens_per_second": 427.891
    },
    {
      "epoch": 0.9966876803077251,
      "grad_norm": 1.9524004459381104,
      "learning_rate": 1.0049197860962567e-05,
      "loss": 0.05901399254798889,
      "num_input_tokens_seen": 2402658,
      "step": 2332,
      "train_runtime": 5615.1875,
      "train_tokens_per_second": 427.886
    },
    {
      "epoch": 0.997115076397051,
      "grad_norm": 2.7556662559509277,
      "learning_rate": 1.0044919786096257e-05,
      "loss": 0.11250441521406174,
      "num_input_tokens_seen": 2403482,
      "step": 2333,
      "train_runtime": 5617.2401,
      "train_tokens_per_second": 427.876
    },
    {
      "epoch": 0.9975424724863767,
      "grad_norm": 2.751969575881958,
      "learning_rate": 1.0040641711229948e-05,
      "loss": 0.07619868963956833,
      "num_input_tokens_seen": 2404478,
      "step": 2334,
      "train_runtime": 5619.2862,
      "train_tokens_per_second": 427.897
    },
    {
      "epoch": 0.9979698685757026,
      "grad_norm": 3.0440139770507812,
      "learning_rate": 1.0036363636363637e-05,
      "loss": 0.16193558275699615,
      "num_input_tokens_seen": 2405476,
      "step": 2335,
      "train_runtime": 5621.3662,
      "train_tokens_per_second": 427.917
    },
    {
      "epoch": 0.9983972646650283,
      "grad_norm": 2.1244683265686035,
      "learning_rate": 1.0032085561497327e-05,
      "loss": 0.07868650555610657,
      "num_input_tokens_seen": 2406682,
      "step": 2336,
      "train_runtime": 5623.5257,
      "train_tokens_per_second": 427.967
    },
    {
      "epoch": 0.998824660754354,
      "grad_norm": 3.3333520889282227,
      "learning_rate": 1.0027807486631016e-05,
      "loss": 0.12665465474128723,
      "num_input_tokens_seen": 2407814,
      "step": 2337,
      "train_runtime": 5625.6818,
      "train_tokens_per_second": 428.004
    },
    {
      "epoch": 0.9992520568436799,
      "grad_norm": 2.7739481925964355,
      "learning_rate": 1.0023529411764707e-05,
      "loss": 0.08598623424768448,
      "num_input_tokens_seen": 2408526,
      "step": 2338,
      "train_runtime": 5627.7474,
      "train_tokens_per_second": 427.973
    },
    {
      "epoch": 0.9996794529330056,
      "grad_norm": 2.04556941986084,
      "learning_rate": 1.0019251336898397e-05,
      "loss": 0.12641248106956482,
      "num_input_tokens_seen": 2409924,
      "step": 2339,
      "train_runtime": 5630.0213,
      "train_tokens_per_second": 428.049
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.795429229736328,
      "learning_rate": 1.0014973262032086e-05,
      "loss": 0.1748589426279068,
      "num_input_tokens_seen": 2410678,
      "step": 2340,
      "train_runtime": 5631.6409,
      "train_tokens_per_second": 428.06
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5088925957679749,
      "eval_runtime": 58.6107,
      "eval_samples_per_second": 17.028,
      "eval_steps_per_second": 8.514,
      "num_input_tokens_seen": 2410678,
      "step": 2340
    },
    {
      "epoch": 1.0004273960893257,
      "grad_norm": 2.1877224445343018,
      "learning_rate": 1.0010695187165776e-05,
      "loss": 0.11428679525852203,
      "num_input_tokens_seen": 2411554,
      "step": 2341,
      "train_runtime": 5698.88,
      "train_tokens_per_second": 423.163
    },
    {
      "epoch": 1.0008547921786515,
      "grad_norm": 2.252779722213745,
      "learning_rate": 1.0006417112299467e-05,
      "loss": 0.08792650699615479,
      "num_input_tokens_seen": 2412244,
      "step": 2342,
      "train_runtime": 5700.8869,
      "train_tokens_per_second": 423.135
    },
    {
      "epoch": 1.0012821882679774,
      "grad_norm": 2.2379162311553955,
      "learning_rate": 1.0002139037433156e-05,
      "loss": 0.07997279614210129,
      "num_input_tokens_seen": 2412912,
      "step": 2343,
      "train_runtime": 5703.0527,
      "train_tokens_per_second": 423.091
    },
    {
      "epoch": 1.0017095843573032,
      "grad_norm": 1.4324240684509277,
      "learning_rate": 9.997860962566846e-06,
      "loss": 0.052942901849746704,
      "num_input_tokens_seen": 2414128,
      "step": 2344,
      "train_runtime": 5705.216,
      "train_tokens_per_second": 423.144
    },
    {
      "epoch": 1.002136980446629,
      "grad_norm": 2.3981478214263916,
      "learning_rate": 9.993582887700537e-06,
      "loss": 0.10901179909706116,
      "num_input_tokens_seen": 2414992,
      "step": 2345,
      "train_runtime": 5707.3644,
      "train_tokens_per_second": 423.136
    },
    {
      "epoch": 1.0025643765359546,
      "grad_norm": 1.996822714805603,
      "learning_rate": 9.989304812834224e-06,
      "loss": 0.07881154119968414,
      "num_input_tokens_seen": 2415892,
      "step": 2346,
      "train_runtime": 5709.4381,
      "train_tokens_per_second": 423.14
    },
    {
      "epoch": 1.0029917726252804,
      "grad_norm": 3.199432373046875,
      "learning_rate": 9.985026737967916e-06,
      "loss": 0.11375774443149567,
      "num_input_tokens_seen": 2416648,
      "step": 2347,
      "train_runtime": 5711.4828,
      "train_tokens_per_second": 423.121
    },
    {
      "epoch": 1.0034191687146063,
      "grad_norm": 2.537538528442383,
      "learning_rate": 9.980748663101605e-06,
      "loss": 0.09518872201442719,
      "num_input_tokens_seen": 2417946,
      "step": 2348,
      "train_runtime": 5713.6377,
      "train_tokens_per_second": 423.189
    },
    {
      "epoch": 1.003846564803932,
      "grad_norm": 2.334536075592041,
      "learning_rate": 9.976470588235294e-06,
      "loss": 0.11627691984176636,
      "num_input_tokens_seen": 2418850,
      "step": 2349,
      "train_runtime": 5715.7312,
      "train_tokens_per_second": 423.192
    },
    {
      "epoch": 1.0042739608932578,
      "grad_norm": 1.5330848693847656,
      "learning_rate": 9.972192513368984e-06,
      "loss": 0.06604886054992676,
      "num_input_tokens_seen": 2420322,
      "step": 2350,
      "train_runtime": 5717.9435,
      "train_tokens_per_second": 423.285
    },
    {
      "epoch": 1.0047013569825836,
      "grad_norm": 2.3374691009521484,
      "learning_rate": 9.967914438502675e-06,
      "loss": 0.09379670768976212,
      "num_input_tokens_seen": 2421218,
      "step": 2351,
      "train_runtime": 5719.9863,
      "train_tokens_per_second": 423.291
    },
    {
      "epoch": 1.0051287530719093,
      "grad_norm": 1.9925848245620728,
      "learning_rate": 9.963636363636364e-06,
      "loss": 0.05323585867881775,
      "num_input_tokens_seen": 2422046,
      "step": 2352,
      "train_runtime": 5722.0163,
      "train_tokens_per_second": 423.285
    },
    {
      "epoch": 1.0055561491612353,
      "grad_norm": 2.109513759613037,
      "learning_rate": 9.959358288770054e-06,
      "loss": 0.08169494569301605,
      "num_input_tokens_seen": 2422850,
      "step": 2353,
      "train_runtime": 5724.0233,
      "train_tokens_per_second": 423.277
    },
    {
      "epoch": 1.005983545250561,
      "grad_norm": 2.082674026489258,
      "learning_rate": 9.955080213903743e-06,
      "loss": 0.09728004783391953,
      "num_input_tokens_seen": 2424202,
      "step": 2354,
      "train_runtime": 5726.2123,
      "train_tokens_per_second": 423.352
    },
    {
      "epoch": 1.0064109413398867,
      "grad_norm": 2.515268325805664,
      "learning_rate": 9.950802139037435e-06,
      "loss": 0.14416220784187317,
      "num_input_tokens_seen": 2425838,
      "step": 2355,
      "train_runtime": 5728.4172,
      "train_tokens_per_second": 423.474
    },
    {
      "epoch": 1.0068383374292125,
      "grad_norm": 1.7195887565612793,
      "learning_rate": 9.946524064171124e-06,
      "loss": 0.06306891143321991,
      "num_input_tokens_seen": 2427190,
      "step": 2356,
      "train_runtime": 5730.5851,
      "train_tokens_per_second": 423.55
    },
    {
      "epoch": 1.0072657335185382,
      "grad_norm": 2.108156681060791,
      "learning_rate": 9.942245989304813e-06,
      "loss": 0.06050962209701538,
      "num_input_tokens_seen": 2428018,
      "step": 2357,
      "train_runtime": 5732.5968,
      "train_tokens_per_second": 423.546
    },
    {
      "epoch": 1.0076931296078642,
      "grad_norm": 2.2040021419525146,
      "learning_rate": 9.937967914438503e-06,
      "loss": 0.07632505148649216,
      "num_input_tokens_seen": 2428934,
      "step": 2358,
      "train_runtime": 5734.6776,
      "train_tokens_per_second": 423.552
    },
    {
      "epoch": 1.00812052569719,
      "grad_norm": 2.3135409355163574,
      "learning_rate": 9.933689839572194e-06,
      "loss": 0.10747627913951874,
      "num_input_tokens_seen": 2430044,
      "step": 2359,
      "train_runtime": 5736.7963,
      "train_tokens_per_second": 423.589
    },
    {
      "epoch": 1.0085479217865156,
      "grad_norm": 1.6670664548873901,
      "learning_rate": 9.929411764705883e-06,
      "loss": 0.061402734369039536,
      "num_input_tokens_seen": 2431436,
      "step": 2360,
      "train_runtime": 5738.9901,
      "train_tokens_per_second": 423.67
    },
    {
      "epoch": 1.0089753178758414,
      "grad_norm": 4.037785053253174,
      "learning_rate": 9.925133689839573e-06,
      "loss": 0.09310992062091827,
      "num_input_tokens_seen": 2432314,
      "step": 2361,
      "train_runtime": 5741.0622,
      "train_tokens_per_second": 423.67
    },
    {
      "epoch": 1.0094027139651671,
      "grad_norm": 2.36489200592041,
      "learning_rate": 9.920855614973262e-06,
      "loss": 0.07615480571985245,
      "num_input_tokens_seen": 2433344,
      "step": 2362,
      "train_runtime": 5743.1475,
      "train_tokens_per_second": 423.695
    },
    {
      "epoch": 1.009830110054493,
      "grad_norm": 2.688366413116455,
      "learning_rate": 9.916577540106953e-06,
      "loss": 0.08781511336565018,
      "num_input_tokens_seen": 2433966,
      "step": 2363,
      "train_runtime": 5745.1266,
      "train_tokens_per_second": 423.658
    },
    {
      "epoch": 1.0102575061438188,
      "grad_norm": 3.5392396450042725,
      "learning_rate": 9.912299465240643e-06,
      "loss": 0.13966228067874908,
      "num_input_tokens_seen": 2434700,
      "step": 2364,
      "train_runtime": 5747.1283,
      "train_tokens_per_second": 423.638
    },
    {
      "epoch": 1.0106849022331446,
      "grad_norm": 2.560177803039551,
      "learning_rate": 9.908021390374332e-06,
      "loss": 0.10608979314565659,
      "num_input_tokens_seen": 2435642,
      "step": 2365,
      "train_runtime": 5749.1867,
      "train_tokens_per_second": 423.65
    },
    {
      "epoch": 1.0111122983224703,
      "grad_norm": 3.171119213104248,
      "learning_rate": 9.903743315508022e-06,
      "loss": 0.1690225601196289,
      "num_input_tokens_seen": 2436440,
      "step": 2366,
      "train_runtime": 5751.2266,
      "train_tokens_per_second": 423.638
    },
    {
      "epoch": 1.011539694411796,
      "grad_norm": 1.8448641300201416,
      "learning_rate": 9.899465240641713e-06,
      "loss": 0.07290095090866089,
      "num_input_tokens_seen": 2438008,
      "step": 2367,
      "train_runtime": 5753.4694,
      "train_tokens_per_second": 423.746
    },
    {
      "epoch": 1.011967090501122,
      "grad_norm": 2.8242595195770264,
      "learning_rate": 9.895187165775402e-06,
      "loss": 0.08168548345565796,
      "num_input_tokens_seen": 2439072,
      "step": 2368,
      "train_runtime": 5755.5855,
      "train_tokens_per_second": 423.775
    },
    {
      "epoch": 1.0123944865904477,
      "grad_norm": 3.596143960952759,
      "learning_rate": 9.890909090909092e-06,
      "loss": 0.09376522898674011,
      "num_input_tokens_seen": 2439756,
      "step": 2369,
      "train_runtime": 5757.6088,
      "train_tokens_per_second": 423.745
    },
    {
      "epoch": 1.0128218826797735,
      "grad_norm": 2.2868266105651855,
      "learning_rate": 9.886631016042781e-06,
      "loss": 0.07253953069448471,
      "num_input_tokens_seen": 2440878,
      "step": 2370,
      "train_runtime": 5759.7145,
      "train_tokens_per_second": 423.785
    },
    {
      "epoch": 1.0132492787690992,
      "grad_norm": 2.50429368019104,
      "learning_rate": 9.882352941176472e-06,
      "loss": 0.09240512549877167,
      "num_input_tokens_seen": 2441626,
      "step": 2371,
      "train_runtime": 5768.1651,
      "train_tokens_per_second": 423.293
    },
    {
      "epoch": 1.013676674858425,
      "grad_norm": 2.1327872276306152,
      "learning_rate": 9.878074866310162e-06,
      "loss": 0.0722540020942688,
      "num_input_tokens_seen": 2442828,
      "step": 2372,
      "train_runtime": 5770.2456,
      "train_tokens_per_second": 423.349
    },
    {
      "epoch": 1.014104070947751,
      "grad_norm": 5.153519630432129,
      "learning_rate": 9.873796791443851e-06,
      "loss": 0.10753324627876282,
      "num_input_tokens_seen": 2443432,
      "step": 2373,
      "train_runtime": 5772.232,
      "train_tokens_per_second": 423.308
    },
    {
      "epoch": 1.0145314670370766,
      "grad_norm": 2.1998326778411865,
      "learning_rate": 9.86951871657754e-06,
      "loss": 0.0521269217133522,
      "num_input_tokens_seen": 2444114,
      "step": 2374,
      "train_runtime": 5774.2551,
      "train_tokens_per_second": 423.278
    },
    {
      "epoch": 1.0149588631264024,
      "grad_norm": 3.409623384475708,
      "learning_rate": 9.865240641711232e-06,
      "loss": 0.09624180942773819,
      "num_input_tokens_seen": 2444886,
      "step": 2375,
      "train_runtime": 5776.3564,
      "train_tokens_per_second": 423.257
    },
    {
      "epoch": 1.0153862592157281,
      "grad_norm": 1.9307880401611328,
      "learning_rate": 9.86096256684492e-06,
      "loss": 0.07020677626132965,
      "num_input_tokens_seen": 2445884,
      "step": 2376,
      "train_runtime": 5778.4406,
      "train_tokens_per_second": 423.278
    },
    {
      "epoch": 1.0158136553050539,
      "grad_norm": 3.074058771133423,
      "learning_rate": 9.85668449197861e-06,
      "loss": 0.08908405900001526,
      "num_input_tokens_seen": 2446610,
      "step": 2377,
      "train_runtime": 5780.4578,
      "train_tokens_per_second": 423.255
    },
    {
      "epoch": 1.0162410513943798,
      "grad_norm": 2.320885181427002,
      "learning_rate": 9.8524064171123e-06,
      "loss": 0.062562495470047,
      "num_input_tokens_seen": 2447354,
      "step": 2378,
      "train_runtime": 5782.5004,
      "train_tokens_per_second": 423.235
    },
    {
      "epoch": 1.0166684474837056,
      "grad_norm": 1.7606289386749268,
      "learning_rate": 9.848128342245991e-06,
      "loss": 0.06884843111038208,
      "num_input_tokens_seen": 2448558,
      "step": 2379,
      "train_runtime": 5784.6251,
      "train_tokens_per_second": 423.287
    },
    {
      "epoch": 1.0170958435730313,
      "grad_norm": 2.665041446685791,
      "learning_rate": 9.843850267379679e-06,
      "loss": 0.08236092329025269,
      "num_input_tokens_seen": 2449682,
      "step": 2380,
      "train_runtime": 5786.7874,
      "train_tokens_per_second": 423.323
    },
    {
      "epoch": 1.017523239662357,
      "grad_norm": 2.5332484245300293,
      "learning_rate": 9.83957219251337e-06,
      "loss": 0.15148237347602844,
      "num_input_tokens_seen": 2450798,
      "step": 2381,
      "train_runtime": 5788.8815,
      "train_tokens_per_second": 423.363
    },
    {
      "epoch": 1.0179506357516828,
      "grad_norm": 2.1304662227630615,
      "learning_rate": 9.83529411764706e-06,
      "loss": 0.058790091425180435,
      "num_input_tokens_seen": 2452052,
      "step": 2382,
      "train_runtime": 5790.9899,
      "train_tokens_per_second": 423.425
    },
    {
      "epoch": 1.0183780318410087,
      "grad_norm": 2.6087403297424316,
      "learning_rate": 9.83101604278075e-06,
      "loss": 0.11724931001663208,
      "num_input_tokens_seen": 2452918,
      "step": 2383,
      "train_runtime": 5793.1437,
      "train_tokens_per_second": 423.417
    },
    {
      "epoch": 1.0188054279303345,
      "grad_norm": 2.31596302986145,
      "learning_rate": 9.826737967914439e-06,
      "loss": 0.0885908454656601,
      "num_input_tokens_seen": 2453968,
      "step": 2384,
      "train_runtime": 5795.2763,
      "train_tokens_per_second": 423.443
    },
    {
      "epoch": 1.0192328240196602,
      "grad_norm": 2.6740293502807617,
      "learning_rate": 9.82245989304813e-06,
      "loss": 0.10855322331190109,
      "num_input_tokens_seen": 2455208,
      "step": 2385,
      "train_runtime": 5797.4427,
      "train_tokens_per_second": 423.498
    },
    {
      "epoch": 1.019660220108986,
      "grad_norm": 2.1240456104278564,
      "learning_rate": 9.81818181818182e-06,
      "loss": 0.10678456723690033,
      "num_input_tokens_seen": 2456240,
      "step": 2386,
      "train_runtime": 5799.5525,
      "train_tokens_per_second": 423.522
    },
    {
      "epoch": 1.0200876161983117,
      "grad_norm": 1.8659801483154297,
      "learning_rate": 9.813903743315509e-06,
      "loss": 0.06166736036539078,
      "num_input_tokens_seen": 2457874,
      "step": 2387,
      "train_runtime": 5801.7903,
      "train_tokens_per_second": 423.641
    },
    {
      "epoch": 1.0205150122876376,
      "grad_norm": 3.5077083110809326,
      "learning_rate": 9.809625668449198e-06,
      "loss": 0.11333203315734863,
      "num_input_tokens_seen": 2458576,
      "step": 2388,
      "train_runtime": 5803.8339,
      "train_tokens_per_second": 423.612
    },
    {
      "epoch": 1.0209424083769634,
      "grad_norm": 2.2801342010498047,
      "learning_rate": 9.80534759358289e-06,
      "loss": 0.09696577489376068,
      "num_input_tokens_seen": 2460134,
      "step": 2389,
      "train_runtime": 5806.0416,
      "train_tokens_per_second": 423.72
    },
    {
      "epoch": 1.021369804466289,
      "grad_norm": 1.9509514570236206,
      "learning_rate": 9.801069518716579e-06,
      "loss": 0.05764611065387726,
      "num_input_tokens_seen": 2461020,
      "step": 2390,
      "train_runtime": 5808.0591,
      "train_tokens_per_second": 423.725
    },
    {
      "epoch": 1.0217972005556148,
      "grad_norm": 2.473060131072998,
      "learning_rate": 9.796791443850268e-06,
      "loss": 0.09615634381771088,
      "num_input_tokens_seen": 2461950,
      "step": 2391,
      "train_runtime": 5810.0817,
      "train_tokens_per_second": 423.738
    },
    {
      "epoch": 1.0222245966449406,
      "grad_norm": 2.5004258155822754,
      "learning_rate": 9.792513368983958e-06,
      "loss": 0.07503095269203186,
      "num_input_tokens_seen": 2462652,
      "step": 2392,
      "train_runtime": 5812.1233,
      "train_tokens_per_second": 423.71
    },
    {
      "epoch": 1.0226519927342665,
      "grad_norm": 2.369931936264038,
      "learning_rate": 9.788235294117649e-06,
      "loss": 0.09896805137395859,
      "num_input_tokens_seen": 2463670,
      "step": 2393,
      "train_runtime": 5814.2217,
      "train_tokens_per_second": 423.732
    },
    {
      "epoch": 1.0230793888235923,
      "grad_norm": 2.575596809387207,
      "learning_rate": 9.783957219251338e-06,
      "loss": 0.06674186140298843,
      "num_input_tokens_seen": 2464664,
      "step": 2394,
      "train_runtime": 5816.3233,
      "train_tokens_per_second": 423.749
    },
    {
      "epoch": 1.023506784912918,
      "grad_norm": 2.3105413913726807,
      "learning_rate": 9.779679144385028e-06,
      "loss": 0.10718658566474915,
      "num_input_tokens_seen": 2465854,
      "step": 2395,
      "train_runtime": 5818.5034,
      "train_tokens_per_second": 423.795
    },
    {
      "epoch": 1.0239341810022438,
      "grad_norm": 3.1299490928649902,
      "learning_rate": 9.775401069518717e-06,
      "loss": 0.13996629416942596,
      "num_input_tokens_seen": 2466900,
      "step": 2396,
      "train_runtime": 5820.6144,
      "train_tokens_per_second": 423.821
    },
    {
      "epoch": 1.0243615770915697,
      "grad_norm": 2.248750686645508,
      "learning_rate": 9.771122994652408e-06,
      "loss": 0.08786091953516006,
      "num_input_tokens_seen": 2467850,
      "step": 2397,
      "train_runtime": 5822.7239,
      "train_tokens_per_second": 423.831
    },
    {
      "epoch": 1.0247889731808955,
      "grad_norm": 1.6931318044662476,
      "learning_rate": 9.766844919786098e-06,
      "loss": 0.05408065766096115,
      "num_input_tokens_seen": 2469824,
      "step": 2398,
      "train_runtime": 5825.0862,
      "train_tokens_per_second": 423.998
    },
    {
      "epoch": 1.0252163692702212,
      "grad_norm": 2.5773775577545166,
      "learning_rate": 9.762566844919787e-06,
      "loss": 0.08918404579162598,
      "num_input_tokens_seen": 2470964,
      "step": 2399,
      "train_runtime": 5827.2095,
      "train_tokens_per_second": 424.039
    },
    {
      "epoch": 1.025643765359547,
      "grad_norm": 2.606586456298828,
      "learning_rate": 9.758288770053477e-06,
      "loss": 0.06853730231523514,
      "num_input_tokens_seen": 2471930,
      "step": 2400,
      "train_runtime": 5829.2953,
      "train_tokens_per_second": 424.053
    },
    {
      "epoch": 1.0260711614488727,
      "grad_norm": 3.5813770294189453,
      "learning_rate": 9.754010695187166e-06,
      "loss": 0.1254851520061493,
      "num_input_tokens_seen": 2472672,
      "step": 2401,
      "train_runtime": 5837.9546,
      "train_tokens_per_second": 423.551
    },
    {
      "epoch": 1.0264985575381986,
      "grad_norm": 2.3840043544769287,
      "learning_rate": 9.749732620320857e-06,
      "loss": 0.08803518116474152,
      "num_input_tokens_seen": 2473766,
      "step": 2402,
      "train_runtime": 5840.1267,
      "train_tokens_per_second": 423.581
    },
    {
      "epoch": 1.0269259536275244,
      "grad_norm": 3.179842472076416,
      "learning_rate": 9.745454545454547e-06,
      "loss": 0.088936448097229,
      "num_input_tokens_seen": 2474594,
      "step": 2403,
      "train_runtime": 5842.3443,
      "train_tokens_per_second": 423.562
    },
    {
      "epoch": 1.02735334971685,
      "grad_norm": 2.0070273876190186,
      "learning_rate": 9.741176470588236e-06,
      "loss": 0.08375103771686554,
      "num_input_tokens_seen": 2475734,
      "step": 2404,
      "train_runtime": 5844.4822,
      "train_tokens_per_second": 423.602
    },
    {
      "epoch": 1.0277807458061758,
      "grad_norm": 2.1431782245635986,
      "learning_rate": 9.736898395721925e-06,
      "loss": 0.08421497046947479,
      "num_input_tokens_seen": 2476656,
      "step": 2405,
      "train_runtime": 5846.5521,
      "train_tokens_per_second": 423.61
    },
    {
      "epoch": 1.0282081418955016,
      "grad_norm": 1.6754003763198853,
      "learning_rate": 9.732620320855617e-06,
      "loss": 0.05906269699335098,
      "num_input_tokens_seen": 2477932,
      "step": 2406,
      "train_runtime": 5848.7585,
      "train_tokens_per_second": 423.668
    },
    {
      "epoch": 1.0286355379848275,
      "grad_norm": 3.263932466506958,
      "learning_rate": 9.728342245989304e-06,
      "loss": 0.10111347585916519,
      "num_input_tokens_seen": 2478918,
      "step": 2407,
      "train_runtime": 5850.8848,
      "train_tokens_per_second": 423.683
    },
    {
      "epoch": 1.0290629340741533,
      "grad_norm": 1.8592978715896606,
      "learning_rate": 9.724064171122995e-06,
      "loss": 0.06808263808488846,
      "num_input_tokens_seen": 2480044,
      "step": 2408,
      "train_runtime": 5853.0546,
      "train_tokens_per_second": 423.718
    },
    {
      "epoch": 1.029490330163479,
      "grad_norm": 2.580446481704712,
      "learning_rate": 9.719786096256685e-06,
      "loss": 0.08375926315784454,
      "num_input_tokens_seen": 2480716,
      "step": 2409,
      "train_runtime": 5855.0791,
      "train_tokens_per_second": 423.686
    },
    {
      "epoch": 1.0299177262528048,
      "grad_norm": 1.9191920757293701,
      "learning_rate": 9.715508021390374e-06,
      "loss": 0.08534429222345352,
      "num_input_tokens_seen": 2481940,
      "step": 2410,
      "train_runtime": 5857.2097,
      "train_tokens_per_second": 423.741
    },
    {
      "epoch": 1.0303451223421305,
      "grad_norm": 2.676165819168091,
      "learning_rate": 9.711229946524064e-06,
      "loss": 0.10068483650684357,
      "num_input_tokens_seen": 2482946,
      "step": 2411,
      "train_runtime": 5859.3242,
      "train_tokens_per_second": 423.76
    },
    {
      "epoch": 1.0307725184314565,
      "grad_norm": 2.6462061405181885,
      "learning_rate": 9.706951871657755e-06,
      "loss": 0.06492633372545242,
      "num_input_tokens_seen": 2483664,
      "step": 2412,
      "train_runtime": 5861.3688,
      "train_tokens_per_second": 423.734
    },
    {
      "epoch": 1.0311999145207822,
      "grad_norm": 2.6537258625030518,
      "learning_rate": 9.702673796791444e-06,
      "loss": 0.12299549579620361,
      "num_input_tokens_seen": 2484784,
      "step": 2413,
      "train_runtime": 5863.5452,
      "train_tokens_per_second": 423.768
    },
    {
      "epoch": 1.031627310610108,
      "grad_norm": 2.679187297821045,
      "learning_rate": 9.698395721925134e-06,
      "loss": 0.10967110842466354,
      "num_input_tokens_seen": 2485826,
      "step": 2414,
      "train_runtime": 5865.6388,
      "train_tokens_per_second": 423.795
    },
    {
      "epoch": 1.0320547066994337,
      "grad_norm": 2.666874408721924,
      "learning_rate": 9.694117647058823e-06,
      "loss": 0.07112434506416321,
      "num_input_tokens_seen": 2486594,
      "step": 2415,
      "train_runtime": 5867.6634,
      "train_tokens_per_second": 423.779
    },
    {
      "epoch": 1.0324821027887594,
      "grad_norm": 2.3183915615081787,
      "learning_rate": 9.689839572192514e-06,
      "loss": 0.10955458134412766,
      "num_input_tokens_seen": 2487680,
      "step": 2416,
      "train_runtime": 5869.7816,
      "train_tokens_per_second": 423.811
    },
    {
      "epoch": 1.0329094988780854,
      "grad_norm": 3.4333391189575195,
      "learning_rate": 9.685561497326204e-06,
      "loss": 0.12376445531845093,
      "num_input_tokens_seen": 2488640,
      "step": 2417,
      "train_runtime": 5871.8615,
      "train_tokens_per_second": 423.825
    },
    {
      "epoch": 1.033336894967411,
      "grad_norm": 2.136274814605713,
      "learning_rate": 9.681283422459893e-06,
      "loss": 0.07355600595474243,
      "num_input_tokens_seen": 2489554,
      "step": 2418,
      "train_runtime": 5873.8866,
      "train_tokens_per_second": 423.834
    },
    {
      "epoch": 1.0337642910567368,
      "grad_norm": 2.222489833831787,
      "learning_rate": 9.677005347593583e-06,
      "loss": 0.056361593306064606,
      "num_input_tokens_seen": 2490240,
      "step": 2419,
      "train_runtime": 5875.8747,
      "train_tokens_per_second": 423.808
    },
    {
      "epoch": 1.0341916871460626,
      "grad_norm": 2.6595888137817383,
      "learning_rate": 9.672727272727274e-06,
      "loss": 0.09777716547250748,
      "num_input_tokens_seen": 2491280,
      "step": 2420,
      "train_runtime": 5877.942,
      "train_tokens_per_second": 423.835
    },
    {
      "epoch": 1.0346190832353883,
      "grad_norm": 1.7243646383285522,
      "learning_rate": 9.668449197860963e-06,
      "loss": 0.06710144877433777,
      "num_input_tokens_seen": 2492480,
      "step": 2421,
      "train_runtime": 5880.0711,
      "train_tokens_per_second": 423.886
    },
    {
      "epoch": 1.0350464793247143,
      "grad_norm": 3.274920701980591,
      "learning_rate": 9.664171122994653e-06,
      "loss": 0.08725401759147644,
      "num_input_tokens_seen": 2493698,
      "step": 2422,
      "train_runtime": 5882.1957,
      "train_tokens_per_second": 423.94
    },
    {
      "epoch": 1.03547387541404,
      "grad_norm": 2.6410112380981445,
      "learning_rate": 9.659893048128342e-06,
      "loss": 0.11823771893978119,
      "num_input_tokens_seen": 2494918,
      "step": 2423,
      "train_runtime": 5884.3442,
      "train_tokens_per_second": 423.993
    },
    {
      "epoch": 1.0359012715033658,
      "grad_norm": 2.5878045558929443,
      "learning_rate": 9.655614973262033e-06,
      "loss": 0.14652089774608612,
      "num_input_tokens_seen": 2496230,
      "step": 2424,
      "train_runtime": 5886.4966,
      "train_tokens_per_second": 424.06
    },
    {
      "epoch": 1.0363286675926915,
      "grad_norm": 3.307190179824829,
      "learning_rate": 9.651336898395723e-06,
      "loss": 0.11303628236055374,
      "num_input_tokens_seen": 2497182,
      "step": 2425,
      "train_runtime": 5888.6198,
      "train_tokens_per_second": 424.069
    },
    {
      "epoch": 1.0367560636820172,
      "grad_norm": 4.737715721130371,
      "learning_rate": 9.647058823529412e-06,
      "loss": 0.09600888192653656,
      "num_input_tokens_seen": 2498966,
      "step": 2426,
      "train_runtime": 5890.9168,
      "train_tokens_per_second": 424.207
    },
    {
      "epoch": 1.0371834597713432,
      "grad_norm": 2.3335535526275635,
      "learning_rate": 9.642780748663102e-06,
      "loss": 0.08316425234079361,
      "num_input_tokens_seen": 2500034,
      "step": 2427,
      "train_runtime": 5893.0329,
      "train_tokens_per_second": 424.236
    },
    {
      "epoch": 1.037610855860669,
      "grad_norm": 2.93424916267395,
      "learning_rate": 9.638502673796793e-06,
      "loss": 0.14708073437213898,
      "num_input_tokens_seen": 2501340,
      "step": 2428,
      "train_runtime": 5895.1912,
      "train_tokens_per_second": 424.302
    },
    {
      "epoch": 1.0380382519499947,
      "grad_norm": 2.623976469039917,
      "learning_rate": 9.634224598930482e-06,
      "loss": 0.11698875576257706,
      "num_input_tokens_seen": 2502776,
      "step": 2429,
      "train_runtime": 5897.4093,
      "train_tokens_per_second": 424.386
    },
    {
      "epoch": 1.0384656480393204,
      "grad_norm": 2.0000336170196533,
      "learning_rate": 9.629946524064172e-06,
      "loss": 0.08833160251379013,
      "num_input_tokens_seen": 2504208,
      "step": 2430,
      "train_runtime": 5899.5838,
      "train_tokens_per_second": 424.472
    },
    {
      "epoch": 1.0388930441286461,
      "grad_norm": 2.978618621826172,
      "learning_rate": 9.625668449197861e-06,
      "loss": 0.16571927070617676,
      "num_input_tokens_seen": 2505128,
      "step": 2431,
      "train_runtime": 5908.3322,
      "train_tokens_per_second": 423.999
    },
    {
      "epoch": 1.039320440217972,
      "grad_norm": 2.23797345161438,
      "learning_rate": 9.621390374331552e-06,
      "loss": 0.07004649937152863,
      "num_input_tokens_seen": 2506210,
      "step": 2432,
      "train_runtime": 5910.5186,
      "train_tokens_per_second": 424.025
    },
    {
      "epoch": 1.0397478363072978,
      "grad_norm": 2.0627944469451904,
      "learning_rate": 9.617112299465242e-06,
      "loss": 0.07753212004899979,
      "num_input_tokens_seen": 2507078,
      "step": 2433,
      "train_runtime": 5912.6883,
      "train_tokens_per_second": 424.017
    },
    {
      "epoch": 1.0401752323966236,
      "grad_norm": 2.0055627822875977,
      "learning_rate": 9.612834224598931e-06,
      "loss": 0.05498732626438141,
      "num_input_tokens_seen": 2508452,
      "step": 2434,
      "train_runtime": 5914.9208,
      "train_tokens_per_second": 424.089
    },
    {
      "epoch": 1.0406026284859493,
      "grad_norm": 2.9423599243164062,
      "learning_rate": 9.60855614973262e-06,
      "loss": 0.11010020971298218,
      "num_input_tokens_seen": 2509270,
      "step": 2435,
      "train_runtime": 5917.0854,
      "train_tokens_per_second": 424.072
    },
    {
      "epoch": 1.041030024575275,
      "grad_norm": 2.5060904026031494,
      "learning_rate": 9.604278074866312e-06,
      "loss": 0.11796065419912338,
      "num_input_tokens_seen": 2510782,
      "step": 2436,
      "train_runtime": 5919.3991,
      "train_tokens_per_second": 424.162
    },
    {
      "epoch": 1.041457420664601,
      "grad_norm": 2.8321478366851807,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.10900522768497467,
      "num_input_tokens_seen": 2511628,
      "step": 2437,
      "train_runtime": 5921.5572,
      "train_tokens_per_second": 424.15
    },
    {
      "epoch": 1.0418848167539267,
      "grad_norm": 3.812952995300293,
      "learning_rate": 9.59572192513369e-06,
      "loss": 0.10735161602497101,
      "num_input_tokens_seen": 2512272,
      "step": 2438,
      "train_runtime": 5923.5719,
      "train_tokens_per_second": 424.114
    },
    {
      "epoch": 1.0423122128432525,
      "grad_norm": 1.6792380809783936,
      "learning_rate": 9.59144385026738e-06,
      "loss": 0.0646650418639183,
      "num_input_tokens_seen": 2513438,
      "step": 2439,
      "train_runtime": 5925.7319,
      "train_tokens_per_second": 424.157
    },
    {
      "epoch": 1.0427396089325782,
      "grad_norm": 1.7132817506790161,
      "learning_rate": 9.587165775401071e-06,
      "loss": 0.05259291082620621,
      "num_input_tokens_seen": 2514356,
      "step": 2440,
      "train_runtime": 5927.7946,
      "train_tokens_per_second": 424.164
    },
    {
      "epoch": 1.043167005021904,
      "grad_norm": 2.3549015522003174,
      "learning_rate": 9.582887700534759e-06,
      "loss": 0.08349432051181793,
      "num_input_tokens_seen": 2515292,
      "step": 2441,
      "train_runtime": 5929.8538,
      "train_tokens_per_second": 424.174
    },
    {
      "epoch": 1.04359440111123,
      "grad_norm": 2.3505473136901855,
      "learning_rate": 9.57860962566845e-06,
      "loss": 0.08466736972332001,
      "num_input_tokens_seen": 2516546,
      "step": 2442,
      "train_runtime": 5932.0128,
      "train_tokens_per_second": 424.231
    },
    {
      "epoch": 1.0440217972005557,
      "grad_norm": 2.7773830890655518,
      "learning_rate": 9.57433155080214e-06,
      "loss": 0.07543611526489258,
      "num_input_tokens_seen": 2517208,
      "step": 2443,
      "train_runtime": 5934.0376,
      "train_tokens_per_second": 424.198
    },
    {
      "epoch": 1.0444491932898814,
      "grad_norm": 4.043386459350586,
      "learning_rate": 9.570053475935829e-06,
      "loss": 0.08522055298089981,
      "num_input_tokens_seen": 2518212,
      "step": 2444,
      "train_runtime": 5936.1282,
      "train_tokens_per_second": 424.218
    },
    {
      "epoch": 1.0448765893792071,
      "grad_norm": 6.0522637367248535,
      "learning_rate": 9.565775401069519e-06,
      "loss": 0.09105490148067474,
      "num_input_tokens_seen": 2519080,
      "step": 2445,
      "train_runtime": 5938.1846,
      "train_tokens_per_second": 424.217
    },
    {
      "epoch": 1.0453039854685329,
      "grad_norm": 2.9145095348358154,
      "learning_rate": 9.56149732620321e-06,
      "loss": 0.10026349127292633,
      "num_input_tokens_seen": 2519976,
      "step": 2446,
      "train_runtime": 5940.248,
      "train_tokens_per_second": 424.221
    },
    {
      "epoch": 1.0457313815578588,
      "grad_norm": 2.9680044651031494,
      "learning_rate": 9.557219251336899e-06,
      "loss": 0.10473820567131042,
      "num_input_tokens_seen": 2520712,
      "step": 2447,
      "train_runtime": 5942.2754,
      "train_tokens_per_second": 424.2
    },
    {
      "epoch": 1.0461587776471846,
      "grad_norm": 2.586383819580078,
      "learning_rate": 9.552941176470589e-06,
      "loss": 0.08061813563108444,
      "num_input_tokens_seen": 2521502,
      "step": 2448,
      "train_runtime": 5944.254,
      "train_tokens_per_second": 424.191
    },
    {
      "epoch": 1.0465861737365103,
      "grad_norm": 1.7993922233581543,
      "learning_rate": 9.548663101604278e-06,
      "loss": 0.05731932073831558,
      "num_input_tokens_seen": 2523934,
      "step": 2449,
      "train_runtime": 5946.6725,
      "train_tokens_per_second": 424.428
    },
    {
      "epoch": 1.047013569825836,
      "grad_norm": 2.498886823654175,
      "learning_rate": 9.544385026737969e-06,
      "loss": 0.09137125313282013,
      "num_input_tokens_seen": 2525038,
      "step": 2450,
      "train_runtime": 5948.7968,
      "train_tokens_per_second": 424.462
    },
    {
      "epoch": 1.0474409659151618,
      "grad_norm": 2.282993793487549,
      "learning_rate": 9.540106951871659e-06,
      "loss": 0.08812248706817627,
      "num_input_tokens_seen": 2526428,
      "step": 2451,
      "train_runtime": 5950.9498,
      "train_tokens_per_second": 424.542
    },
    {
      "epoch": 1.0478683620044877,
      "grad_norm": 2.5882728099823,
      "learning_rate": 9.535828877005348e-06,
      "loss": 0.09415605664253235,
      "num_input_tokens_seen": 2527440,
      "step": 2452,
      "train_runtime": 5953.0363,
      "train_tokens_per_second": 424.563
    },
    {
      "epoch": 1.0482957580938135,
      "grad_norm": 4.275142669677734,
      "learning_rate": 9.531550802139037e-06,
      "loss": 0.0687527284026146,
      "num_input_tokens_seen": 2528204,
      "step": 2453,
      "train_runtime": 5955.2151,
      "train_tokens_per_second": 424.536
    },
    {
      "epoch": 1.0487231541831392,
      "grad_norm": 3.256615400314331,
      "learning_rate": 9.527272727272729e-06,
      "loss": 0.06413855403661728,
      "num_input_tokens_seen": 2528862,
      "step": 2454,
      "train_runtime": 5957.1844,
      "train_tokens_per_second": 424.506
    },
    {
      "epoch": 1.049150550272465,
      "grad_norm": 3.384748935699463,
      "learning_rate": 9.522994652406418e-06,
      "loss": 0.0843663364648819,
      "num_input_tokens_seen": 2529634,
      "step": 2455,
      "train_runtime": 5959.1832,
      "train_tokens_per_second": 424.493
    },
    {
      "epoch": 1.0495779463617907,
      "grad_norm": 1.8081471920013428,
      "learning_rate": 9.518716577540108e-06,
      "loss": 0.07914873957633972,
      "num_input_tokens_seen": 2531450,
      "step": 2456,
      "train_runtime": 5961.4858,
      "train_tokens_per_second": 424.634
    },
    {
      "epoch": 1.0500053424511167,
      "grad_norm": 1.9101433753967285,
      "learning_rate": 9.514438502673797e-06,
      "loss": 0.04463515803217888,
      "num_input_tokens_seen": 2532710,
      "step": 2457,
      "train_runtime": 5963.624,
      "train_tokens_per_second": 424.693
    },
    {
      "epoch": 1.0504327385404424,
      "grad_norm": 3.349461793899536,
      "learning_rate": 9.510160427807488e-06,
      "loss": 0.08655493706464767,
      "num_input_tokens_seen": 2533830,
      "step": 2458,
      "train_runtime": 5965.7393,
      "train_tokens_per_second": 424.73
    },
    {
      "epoch": 1.0508601346297681,
      "grad_norm": 1.8900494575500488,
      "learning_rate": 9.505882352941178e-06,
      "loss": 0.07853373140096664,
      "num_input_tokens_seen": 2534960,
      "step": 2459,
      "train_runtime": 5967.8869,
      "train_tokens_per_second": 424.767
    },
    {
      "epoch": 1.0512875307190939,
      "grad_norm": 1.8290657997131348,
      "learning_rate": 9.501604278074867e-06,
      "loss": 0.049846798181533813,
      "num_input_tokens_seen": 2536098,
      "step": 2460,
      "train_runtime": 5970.0001,
      "train_tokens_per_second": 424.807
    },
    {
      "epoch": 1.0517149268084196,
      "grad_norm": 3.573796033859253,
      "learning_rate": 9.497326203208556e-06,
      "loss": 0.08434971421957016,
      "num_input_tokens_seen": 2536888,
      "step": 2461,
      "train_runtime": 5978.5897,
      "train_tokens_per_second": 424.329
    },
    {
      "epoch": 1.0521423228977456,
      "grad_norm": 2.914522647857666,
      "learning_rate": 9.493048128342248e-06,
      "loss": 0.09317946434020996,
      "num_input_tokens_seen": 2538444,
      "step": 2462,
      "train_runtime": 5980.9276,
      "train_tokens_per_second": 424.423
    },
    {
      "epoch": 1.0525697189870713,
      "grad_norm": 3.257650136947632,
      "learning_rate": 9.488770053475937e-06,
      "loss": 0.09556391835212708,
      "num_input_tokens_seen": 2539318,
      "step": 2463,
      "train_runtime": 5983.0496,
      "train_tokens_per_second": 424.419
    },
    {
      "epoch": 1.052997115076397,
      "grad_norm": 3.7569570541381836,
      "learning_rate": 9.484491978609626e-06,
      "loss": 0.10677316784858704,
      "num_input_tokens_seen": 2540082,
      "step": 2464,
      "train_runtime": 5985.1147,
      "train_tokens_per_second": 424.4
    },
    {
      "epoch": 1.0534245111657228,
      "grad_norm": 2.6633358001708984,
      "learning_rate": 9.480213903743316e-06,
      "loss": 0.10806825757026672,
      "num_input_tokens_seen": 2541396,
      "step": 2465,
      "train_runtime": 5987.4232,
      "train_tokens_per_second": 424.456
    },
    {
      "epoch": 1.0538519072550485,
      "grad_norm": 2.504791498184204,
      "learning_rate": 9.475935828877007e-06,
      "loss": 0.08049499988555908,
      "num_input_tokens_seen": 2542234,
      "step": 2466,
      "train_runtime": 5989.5275,
      "train_tokens_per_second": 424.447
    },
    {
      "epoch": 1.0542793033443745,
      "grad_norm": 1.8010330200195312,
      "learning_rate": 9.471657754010696e-06,
      "loss": 0.07656461000442505,
      "num_input_tokens_seen": 2543276,
      "step": 2467,
      "train_runtime": 5991.6933,
      "train_tokens_per_second": 424.467
    },
    {
      "epoch": 1.0547066994337002,
      "grad_norm": 3.153151750564575,
      "learning_rate": 9.467379679144386e-06,
      "loss": 0.08555097132921219,
      "num_input_tokens_seen": 2544130,
      "step": 2468,
      "train_runtime": 5993.8528,
      "train_tokens_per_second": 424.457
    },
    {
      "epoch": 1.055134095523026,
      "grad_norm": 3.303352117538452,
      "learning_rate": 9.463101604278075e-06,
      "loss": 0.1573292762041092,
      "num_input_tokens_seen": 2545116,
      "step": 2469,
      "train_runtime": 5996.0568,
      "train_tokens_per_second": 424.465
    },
    {
      "epoch": 1.0555614916123517,
      "grad_norm": 3.2898359298706055,
      "learning_rate": 9.458823529411767e-06,
      "loss": 0.084254689514637,
      "num_input_tokens_seen": 2546268,
      "step": 2470,
      "train_runtime": 5998.2705,
      "train_tokens_per_second": 424.5
    },
    {
      "epoch": 1.0559888877016774,
      "grad_norm": 2.7838451862335205,
      "learning_rate": 9.454545454545456e-06,
      "loss": 0.08452479541301727,
      "num_input_tokens_seen": 2547384,
      "step": 2471,
      "train_runtime": 6000.4856,
      "train_tokens_per_second": 424.53
    },
    {
      "epoch": 1.0564162837910034,
      "grad_norm": 2.2581820487976074,
      "learning_rate": 9.450267379679145e-06,
      "loss": 0.07131518423557281,
      "num_input_tokens_seen": 2548482,
      "step": 2472,
      "train_runtime": 6002.6194,
      "train_tokens_per_second": 424.562
    },
    {
      "epoch": 1.0568436798803291,
      "grad_norm": 2.164290428161621,
      "learning_rate": 9.445989304812835e-06,
      "loss": 0.05863840878009796,
      "num_input_tokens_seen": 2549352,
      "step": 2473,
      "train_runtime": 6004.7653,
      "train_tokens_per_second": 424.555
    },
    {
      "epoch": 1.0572710759696549,
      "grad_norm": 1.770673394203186,
      "learning_rate": 9.441711229946526e-06,
      "loss": 0.06265389919281006,
      "num_input_tokens_seen": 2550902,
      "step": 2474,
      "train_runtime": 6007.0724,
      "train_tokens_per_second": 424.65
    },
    {
      "epoch": 1.0576984720589806,
      "grad_norm": 3.089261054992676,
      "learning_rate": 9.437433155080214e-06,
      "loss": 0.13192465901374817,
      "num_input_tokens_seen": 2551878,
      "step": 2475,
      "train_runtime": 6009.1704,
      "train_tokens_per_second": 424.664
    },
    {
      "epoch": 1.0581258681483063,
      "grad_norm": 5.253305912017822,
      "learning_rate": 9.433155080213905e-06,
      "loss": 0.12144365906715393,
      "num_input_tokens_seen": 2553294,
      "step": 2476,
      "train_runtime": 6011.4069,
      "train_tokens_per_second": 424.742
    },
    {
      "epoch": 1.0585532642376323,
      "grad_norm": 2.399216413497925,
      "learning_rate": 9.428877005347594e-06,
      "loss": 0.08546847105026245,
      "num_input_tokens_seen": 2554620,
      "step": 2477,
      "train_runtime": 6013.6494,
      "train_tokens_per_second": 424.804
    },
    {
      "epoch": 1.058980660326958,
      "grad_norm": 2.4415104389190674,
      "learning_rate": 9.424598930481284e-06,
      "loss": 0.08338084816932678,
      "num_input_tokens_seen": 2555690,
      "step": 2478,
      "train_runtime": 6015.8048,
      "train_tokens_per_second": 424.829
    },
    {
      "epoch": 1.0594080564162838,
      "grad_norm": 4.050954818725586,
      "learning_rate": 9.420320855614973e-06,
      "loss": 0.08639764785766602,
      "num_input_tokens_seen": 2556760,
      "step": 2479,
      "train_runtime": 6017.9178,
      "train_tokens_per_second": 424.858
    },
    {
      "epoch": 1.0598354525056095,
      "grad_norm": 3.0661141872406006,
      "learning_rate": 9.416042780748664e-06,
      "loss": 0.13239911198616028,
      "num_input_tokens_seen": 2557776,
      "step": 2480,
      "train_runtime": 6019.974,
      "train_tokens_per_second": 424.882
    },
    {
      "epoch": 1.0602628485949355,
      "grad_norm": 2.315734624862671,
      "learning_rate": 9.411764705882354e-06,
      "loss": 0.0881490483880043,
      "num_input_tokens_seen": 2559014,
      "step": 2481,
      "train_runtime": 6022.2008,
      "train_tokens_per_second": 424.93
    },
    {
      "epoch": 1.0606902446842612,
      "grad_norm": 2.5126805305480957,
      "learning_rate": 9.407486631016043e-06,
      "loss": 0.0491979606449604,
      "num_input_tokens_seen": 2560306,
      "step": 2482,
      "train_runtime": 6024.3702,
      "train_tokens_per_second": 424.991
    },
    {
      "epoch": 1.061117640773587,
      "grad_norm": 3.510737419128418,
      "learning_rate": 9.403208556149733e-06,
      "loss": 0.0917619988322258,
      "num_input_tokens_seen": 2561080,
      "step": 2483,
      "train_runtime": 6026.4542,
      "train_tokens_per_second": 424.973
    },
    {
      "epoch": 1.0615450368629127,
      "grad_norm": 2.9285781383514404,
      "learning_rate": 9.398930481283424e-06,
      "loss": 0.12614470720291138,
      "num_input_tokens_seen": 2561938,
      "step": 2484,
      "train_runtime": 6028.5376,
      "train_tokens_per_second": 424.968
    },
    {
      "epoch": 1.0619724329522384,
      "grad_norm": 2.4121227264404297,
      "learning_rate": 9.394652406417113e-06,
      "loss": 0.12636646628379822,
      "num_input_tokens_seen": 2562944,
      "step": 2485,
      "train_runtime": 6030.5866,
      "train_tokens_per_second": 424.991
    },
    {
      "epoch": 1.0623998290415644,
      "grad_norm": 2.5880560874938965,
      "learning_rate": 9.390374331550803e-06,
      "loss": 0.0730423554778099,
      "num_input_tokens_seen": 2563880,
      "step": 2486,
      "train_runtime": 6032.6585,
      "train_tokens_per_second": 425.0
    },
    {
      "epoch": 1.0628272251308901,
      "grad_norm": 2.2725648880004883,
      "learning_rate": 9.386096256684492e-06,
      "loss": 0.05836467817425728,
      "num_input_tokens_seen": 2564908,
      "step": 2487,
      "train_runtime": 6034.7434,
      "train_tokens_per_second": 425.024
    },
    {
      "epoch": 1.0632546212202159,
      "grad_norm": 2.2882461547851562,
      "learning_rate": 9.381818181818183e-06,
      "loss": 0.1003018170595169,
      "num_input_tokens_seen": 2565964,
      "step": 2488,
      "train_runtime": 6036.8279,
      "train_tokens_per_second": 425.052
    },
    {
      "epoch": 1.0636820173095416,
      "grad_norm": 2.6683990955352783,
      "learning_rate": 9.377540106951873e-06,
      "loss": 0.12827488780021667,
      "num_input_tokens_seen": 2567030,
      "step": 2489,
      "train_runtime": 6038.9136,
      "train_tokens_per_second": 425.081
    },
    {
      "epoch": 1.0641094133988673,
      "grad_norm": 2.0335280895233154,
      "learning_rate": 9.373262032085562e-06,
      "loss": 0.05924580246210098,
      "num_input_tokens_seen": 2567976,
      "step": 2490,
      "train_runtime": 6041.0089,
      "train_tokens_per_second": 425.091
    },
    {
      "epoch": 1.0645368094881933,
      "grad_norm": 2.5157742500305176,
      "learning_rate": 9.368983957219252e-06,
      "loss": 0.0915893092751503,
      "num_input_tokens_seen": 2568936,
      "step": 2491,
      "train_runtime": 6049.6806,
      "train_tokens_per_second": 424.64
    },
    {
      "epoch": 1.064964205577519,
      "grad_norm": 3.596317768096924,
      "learning_rate": 9.364705882352943e-06,
      "loss": 0.1080373078584671,
      "num_input_tokens_seen": 2569600,
      "step": 2492,
      "train_runtime": 6051.6588,
      "train_tokens_per_second": 424.611
    },
    {
      "epoch": 1.0653916016668448,
      "grad_norm": 2.10650897026062,
      "learning_rate": 9.360427807486632e-06,
      "loss": 0.08839220553636551,
      "num_input_tokens_seen": 2570716,
      "step": 2493,
      "train_runtime": 6053.8066,
      "train_tokens_per_second": 424.645
    },
    {
      "epoch": 1.0658189977561705,
      "grad_norm": 3.5884060859680176,
      "learning_rate": 9.356149732620322e-06,
      "loss": 0.14576679468154907,
      "num_input_tokens_seen": 2571516,
      "step": 2494,
      "train_runtime": 6055.9329,
      "train_tokens_per_second": 424.628
    },
    {
      "epoch": 1.0662463938454962,
      "grad_norm": 3.5533761978149414,
      "learning_rate": 9.351871657754011e-06,
      "loss": 0.11463622003793716,
      "num_input_tokens_seen": 2572310,
      "step": 2495,
      "train_runtime": 6057.98,
      "train_tokens_per_second": 424.615
    },
    {
      "epoch": 1.0666737899348222,
      "grad_norm": 16.307538986206055,
      "learning_rate": 9.347593582887702e-06,
      "loss": 0.2264968901872635,
      "num_input_tokens_seen": 2573646,
      "step": 2496,
      "train_runtime": 6060.1514,
      "train_tokens_per_second": 424.683
    },
    {
      "epoch": 1.067101186024148,
      "grad_norm": 2.061206102371216,
      "learning_rate": 9.343315508021392e-06,
      "loss": 0.06639231741428375,
      "num_input_tokens_seen": 2574818,
      "step": 2497,
      "train_runtime": 6062.2967,
      "train_tokens_per_second": 424.726
    },
    {
      "epoch": 1.0675285821134737,
      "grad_norm": 1.770450472831726,
      "learning_rate": 9.339037433155081e-06,
      "loss": 0.07189320772886276,
      "num_input_tokens_seen": 2576078,
      "step": 2498,
      "train_runtime": 6064.4925,
      "train_tokens_per_second": 424.78
    },
    {
      "epoch": 1.0679559782027994,
      "grad_norm": 3.03733229637146,
      "learning_rate": 9.33475935828877e-06,
      "loss": 0.10361220687627792,
      "num_input_tokens_seen": 2577046,
      "step": 2499,
      "train_runtime": 6066.5911,
      "train_tokens_per_second": 424.793
    },
    {
      "epoch": 1.0683833742921252,
      "grad_norm": 2.5225203037261963,
      "learning_rate": 9.330481283422462e-06,
      "loss": 0.08256438374519348,
      "num_input_tokens_seen": 2577862,
      "step": 2500,
      "train_runtime": 6068.6362,
      "train_tokens_per_second": 424.784
    },
    {
      "epoch": 1.0688107703814511,
      "grad_norm": 2.4388086795806885,
      "learning_rate": 9.326203208556151e-06,
      "loss": 0.06906373053789139,
      "num_input_tokens_seen": 2578680,
      "step": 2501,
      "train_runtime": 6070.738,
      "train_tokens_per_second": 424.772
    },
    {
      "epoch": 1.0692381664707769,
      "grad_norm": 4.451180458068848,
      "learning_rate": 9.32192513368984e-06,
      "loss": 0.15288393199443817,
      "num_input_tokens_seen": 2579326,
      "step": 2502,
      "train_runtime": 6072.7696,
      "train_tokens_per_second": 424.736
    },
    {
      "epoch": 1.0696655625601026,
      "grad_norm": 1.10209059715271,
      "learning_rate": 9.31764705882353e-06,
      "loss": 0.03241486847400665,
      "num_input_tokens_seen": 2581586,
      "step": 2503,
      "train_runtime": 6075.1518,
      "train_tokens_per_second": 424.942
    },
    {
      "epoch": 1.0700929586494283,
      "grad_norm": 2.449856996536255,
      "learning_rate": 9.313368983957221e-06,
      "loss": 0.07203330844640732,
      "num_input_tokens_seen": 2582338,
      "step": 2504,
      "train_runtime": 6077.1794,
      "train_tokens_per_second": 424.924
    },
    {
      "epoch": 1.070520354738754,
      "grad_norm": 3.4753663539886475,
      "learning_rate": 9.30909090909091e-06,
      "loss": 0.08703330159187317,
      "num_input_tokens_seen": 2583168,
      "step": 2505,
      "train_runtime": 6079.2198,
      "train_tokens_per_second": 424.918
    },
    {
      "epoch": 1.07094775082808,
      "grad_norm": 1.6387132406234741,
      "learning_rate": 9.3048128342246e-06,
      "loss": 0.04702633246779442,
      "num_input_tokens_seen": 2584428,
      "step": 2506,
      "train_runtime": 6081.3422,
      "train_tokens_per_second": 424.977
    },
    {
      "epoch": 1.0713751469174058,
      "grad_norm": 2.637098789215088,
      "learning_rate": 9.30053475935829e-06,
      "loss": 0.09241136163473129,
      "num_input_tokens_seen": 2585258,
      "step": 2507,
      "train_runtime": 6083.384,
      "train_tokens_per_second": 424.97
    },
    {
      "epoch": 1.0718025430067315,
      "grad_norm": 2.5830302238464355,
      "learning_rate": 9.296256684491979e-06,
      "loss": 0.08075625449419022,
      "num_input_tokens_seen": 2586070,
      "step": 2508,
      "train_runtime": 6085.4607,
      "train_tokens_per_second": 424.959
    },
    {
      "epoch": 1.0722299390960572,
      "grad_norm": 1.774401068687439,
      "learning_rate": 9.291978609625668e-06,
      "loss": 0.06211743503808975,
      "num_input_tokens_seen": 2587484,
      "step": 2509,
      "train_runtime": 6087.6296,
      "train_tokens_per_second": 425.04
    },
    {
      "epoch": 1.072657335185383,
      "grad_norm": 3.4107000827789307,
      "learning_rate": 9.287700534759358e-06,
      "loss": 0.12342812120914459,
      "num_input_tokens_seen": 2588392,
      "step": 2510,
      "train_runtime": 6089.704,
      "train_tokens_per_second": 425.044
    },
    {
      "epoch": 1.073084731274709,
      "grad_norm": 1.8665857315063477,
      "learning_rate": 9.283422459893049e-06,
      "loss": 0.05985205993056297,
      "num_input_tokens_seen": 2589488,
      "step": 2511,
      "train_runtime": 6091.8471,
      "train_tokens_per_second": 425.074
    },
    {
      "epoch": 1.0735121273640347,
      "grad_norm": 2.3740785121917725,
      "learning_rate": 9.279144385026738e-06,
      "loss": 0.08959895372390747,
      "num_input_tokens_seen": 2590612,
      "step": 2512,
      "train_runtime": 6093.9455,
      "train_tokens_per_second": 425.112
    },
    {
      "epoch": 1.0739395234533604,
      "grad_norm": 4.132309913635254,
      "learning_rate": 9.274866310160428e-06,
      "loss": 0.13241590559482574,
      "num_input_tokens_seen": 2591500,
      "step": 2513,
      "train_runtime": 6096.016,
      "train_tokens_per_second": 425.114
    },
    {
      "epoch": 1.0743669195426861,
      "grad_norm": 2.7073099613189697,
      "learning_rate": 9.270588235294117e-06,
      "loss": 0.10110922157764435,
      "num_input_tokens_seen": 2592246,
      "step": 2514,
      "train_runtime": 6098.084,
      "train_tokens_per_second": 425.092
    },
    {
      "epoch": 1.0747943156320119,
      "grad_norm": 1.4636536836624146,
      "learning_rate": 9.266310160427809e-06,
      "loss": 0.05031413957476616,
      "num_input_tokens_seen": 2593324,
      "step": 2515,
      "train_runtime": 6100.3301,
      "train_tokens_per_second": 425.112
    },
    {
      "epoch": 1.0752217117213378,
      "grad_norm": 3.164484977722168,
      "learning_rate": 9.262032085561498e-06,
      "loss": 0.12404800951480865,
      "num_input_tokens_seen": 2594402,
      "step": 2516,
      "train_runtime": 6102.418,
      "train_tokens_per_second": 425.143
    },
    {
      "epoch": 1.0756491078106636,
      "grad_norm": 2.224860429763794,
      "learning_rate": 9.257754010695187e-06,
      "loss": 0.0686841830611229,
      "num_input_tokens_seen": 2595594,
      "step": 2517,
      "train_runtime": 6104.5673,
      "train_tokens_per_second": 425.189
    },
    {
      "epoch": 1.0760765038999893,
      "grad_norm": 1.763987421989441,
      "learning_rate": 9.253475935828877e-06,
      "loss": 0.044779419898986816,
      "num_input_tokens_seen": 2597202,
      "step": 2518,
      "train_runtime": 6106.8098,
      "train_tokens_per_second": 425.296
    },
    {
      "epoch": 1.076503899989315,
      "grad_norm": 3.291356325149536,
      "learning_rate": 9.249197860962568e-06,
      "loss": 0.1659565269947052,
      "num_input_tokens_seen": 2598682,
      "step": 2519,
      "train_runtime": 6109.0019,
      "train_tokens_per_second": 425.386
    },
    {
      "epoch": 1.0769312960786408,
      "grad_norm": 4.187049388885498,
      "learning_rate": 9.244919786096257e-06,
      "loss": 0.1263025999069214,
      "num_input_tokens_seen": 2599290,
      "step": 2520,
      "train_runtime": 6111.0717,
      "train_tokens_per_second": 425.341
    },
    {
      "epoch": 1.0773586921679668,
      "grad_norm": 2.4789187908172607,
      "learning_rate": 9.240641711229947e-06,
      "loss": 0.05840917304158211,
      "num_input_tokens_seen": 2600320,
      "step": 2521,
      "train_runtime": 6119.8304,
      "train_tokens_per_second": 424.901
    },
    {
      "epoch": 1.0777860882572925,
      "grad_norm": 3.5480587482452393,
      "learning_rate": 9.236363636363636e-06,
      "loss": 0.09059934318065643,
      "num_input_tokens_seen": 2600998,
      "step": 2522,
      "train_runtime": 6121.8122,
      "train_tokens_per_second": 424.874
    },
    {
      "epoch": 1.0782134843466182,
      "grad_norm": 2.734523296356201,
      "learning_rate": 9.232085561497327e-06,
      "loss": 0.10375450551509857,
      "num_input_tokens_seen": 2602492,
      "step": 2523,
      "train_runtime": 6123.99,
      "train_tokens_per_second": 424.967
    },
    {
      "epoch": 1.078640880435944,
      "grad_norm": 3.074808359146118,
      "learning_rate": 9.227807486631017e-06,
      "loss": 0.10524008423089981,
      "num_input_tokens_seen": 2603270,
      "step": 2524,
      "train_runtime": 6126.0093,
      "train_tokens_per_second": 424.954
    },
    {
      "epoch": 1.0790682765252697,
      "grad_norm": 2.966073751449585,
      "learning_rate": 9.223529411764706e-06,
      "loss": 0.06798302382230759,
      "num_input_tokens_seen": 2604098,
      "step": 2525,
      "train_runtime": 6128.0312,
      "train_tokens_per_second": 424.949
    },
    {
      "epoch": 1.0794956726145957,
      "grad_norm": 2.8142268657684326,
      "learning_rate": 9.219251336898396e-06,
      "loss": 0.07415980845689774,
      "num_input_tokens_seen": 2604942,
      "step": 2526,
      "train_runtime": 6130.0892,
      "train_tokens_per_second": 424.944
    },
    {
      "epoch": 1.0799230687039214,
      "grad_norm": 2.471247673034668,
      "learning_rate": 9.214973262032087e-06,
      "loss": 0.08812225610017776,
      "num_input_tokens_seen": 2605822,
      "step": 2527,
      "train_runtime": 6132.1456,
      "train_tokens_per_second": 424.945
    },
    {
      "epoch": 1.0803504647932471,
      "grad_norm": 1.8501207828521729,
      "learning_rate": 9.210695187165776e-06,
      "loss": 0.07012195885181427,
      "num_input_tokens_seen": 2607138,
      "step": 2528,
      "train_runtime": 6134.3808,
      "train_tokens_per_second": 425.004
    },
    {
      "epoch": 1.0807778608825729,
      "grad_norm": 2.5294954776763916,
      "learning_rate": 9.206417112299466e-06,
      "loss": 0.08150813728570938,
      "num_input_tokens_seen": 2608474,
      "step": 2529,
      "train_runtime": 6136.7142,
      "train_tokens_per_second": 425.06
    },
    {
      "epoch": 1.0812052569718986,
      "grad_norm": 3.708526134490967,
      "learning_rate": 9.202139037433155e-06,
      "loss": 0.10302053391933441,
      "num_input_tokens_seen": 2609352,
      "step": 2530,
      "train_runtime": 6138.7855,
      "train_tokens_per_second": 425.06
    },
    {
      "epoch": 1.0816326530612246,
      "grad_norm": 2.5818028450012207,
      "learning_rate": 9.197860962566846e-06,
      "loss": 0.08561830222606659,
      "num_input_tokens_seen": 2610372,
      "step": 2531,
      "train_runtime": 6140.8695,
      "train_tokens_per_second": 425.082
    },
    {
      "epoch": 1.0820600491505503,
      "grad_norm": 2.5761454105377197,
      "learning_rate": 9.193582887700536e-06,
      "loss": 0.10527003556489944,
      "num_input_tokens_seen": 2611530,
      "step": 2532,
      "train_runtime": 6143.0239,
      "train_tokens_per_second": 425.121
    },
    {
      "epoch": 1.082487445239876,
      "grad_norm": 2.6809802055358887,
      "learning_rate": 9.189304812834225e-06,
      "loss": 0.09583239257335663,
      "num_input_tokens_seen": 2612686,
      "step": 2533,
      "train_runtime": 6145.1391,
      "train_tokens_per_second": 425.163
    },
    {
      "epoch": 1.0829148413292018,
      "grad_norm": 2.889986515045166,
      "learning_rate": 9.185026737967915e-06,
      "loss": 0.09865818917751312,
      "num_input_tokens_seen": 2613378,
      "step": 2534,
      "train_runtime": 6147.2692,
      "train_tokens_per_second": 425.128
    },
    {
      "epoch": 1.0833422374185275,
      "grad_norm": 2.403228282928467,
      "learning_rate": 9.180748663101606e-06,
      "loss": 0.09622926265001297,
      "num_input_tokens_seen": 2614408,
      "step": 2535,
      "train_runtime": 6149.3603,
      "train_tokens_per_second": 425.151
    },
    {
      "epoch": 1.0837696335078535,
      "grad_norm": 3.137573480606079,
      "learning_rate": 9.176470588235294e-06,
      "loss": 0.09262220561504364,
      "num_input_tokens_seen": 2615476,
      "step": 2536,
      "train_runtime": 6151.476,
      "train_tokens_per_second": 425.179
    },
    {
      "epoch": 1.0841970295971792,
      "grad_norm": 3.5582549571990967,
      "learning_rate": 9.172192513368985e-06,
      "loss": 0.1597282439470291,
      "num_input_tokens_seen": 2616314,
      "step": 2537,
      "train_runtime": 6153.5418,
      "train_tokens_per_second": 425.172
    },
    {
      "epoch": 1.084624425686505,
      "grad_norm": 2.795961380004883,
      "learning_rate": 9.167914438502674e-06,
      "loss": 0.08986320346593857,
      "num_input_tokens_seen": 2617246,
      "step": 2538,
      "train_runtime": 6155.6361,
      "train_tokens_per_second": 425.179
    },
    {
      "epoch": 1.0850518217758307,
      "grad_norm": 2.2350223064422607,
      "learning_rate": 9.163636363636365e-06,
      "loss": 0.09304250031709671,
      "num_input_tokens_seen": 2618258,
      "step": 2539,
      "train_runtime": 6157.7157,
      "train_tokens_per_second": 425.2
    },
    {
      "epoch": 1.0854792178651564,
      "grad_norm": 5.303435802459717,
      "learning_rate": 9.159358288770053e-06,
      "loss": 0.08296370506286621,
      "num_input_tokens_seen": 2619166,
      "step": 2540,
      "train_runtime": 6159.7577,
      "train_tokens_per_second": 425.206
    },
    {
      "epoch": 1.0859066139544824,
      "grad_norm": 2.2618508338928223,
      "learning_rate": 9.155080213903744e-06,
      "loss": 0.07552073150873184,
      "num_input_tokens_seen": 2620126,
      "step": 2541,
      "train_runtime": 6161.8391,
      "train_tokens_per_second": 425.218
    },
    {
      "epoch": 1.0863340100438081,
      "grad_norm": 1.762229561805725,
      "learning_rate": 9.150802139037434e-06,
      "loss": 0.06286679208278656,
      "num_input_tokens_seen": 2621470,
      "step": 2542,
      "train_runtime": 6163.9897,
      "train_tokens_per_second": 425.288
    },
    {
      "epoch": 1.0867614061331339,
      "grad_norm": 1.8647757768630981,
      "learning_rate": 9.146524064171123e-06,
      "loss": 0.07455813139677048,
      "num_input_tokens_seen": 2623956,
      "step": 2543,
      "train_runtime": 6166.4564,
      "train_tokens_per_second": 425.521
    },
    {
      "epoch": 1.0871888022224596,
      "grad_norm": 2.485333204269409,
      "learning_rate": 9.142245989304813e-06,
      "loss": 0.10116797685623169,
      "num_input_tokens_seen": 2624926,
      "step": 2544,
      "train_runtime": 6168.545,
      "train_tokens_per_second": 425.534
    },
    {
      "epoch": 1.0876161983117854,
      "grad_norm": 1.6954323053359985,
      "learning_rate": 9.137967914438504e-06,
      "loss": 0.0423913411796093,
      "num_input_tokens_seen": 2626152,
      "step": 2545,
      "train_runtime": 6170.6919,
      "train_tokens_per_second": 425.585
    },
    {
      "epoch": 1.0880435944011113,
      "grad_norm": 3.0797646045684814,
      "learning_rate": 9.133689839572193e-06,
      "loss": 0.1578473448753357,
      "num_input_tokens_seen": 2627016,
      "step": 2546,
      "train_runtime": 6172.7449,
      "train_tokens_per_second": 425.583
    },
    {
      "epoch": 1.088470990490437,
      "grad_norm": 3.5497829914093018,
      "learning_rate": 9.129411764705883e-06,
      "loss": 0.12611296772956848,
      "num_input_tokens_seen": 2627798,
      "step": 2547,
      "train_runtime": 6174.7335,
      "train_tokens_per_second": 425.573
    },
    {
      "epoch": 1.0888983865797628,
      "grad_norm": 2.7811191082000732,
      "learning_rate": 9.125133689839572e-06,
      "loss": 0.10591692477464676,
      "num_input_tokens_seen": 2629092,
      "step": 2548,
      "train_runtime": 6176.8832,
      "train_tokens_per_second": 425.634
    },
    {
      "epoch": 1.0893257826690885,
      "grad_norm": 1.9196721315383911,
      "learning_rate": 9.120855614973263e-06,
      "loss": 0.07511201500892639,
      "num_input_tokens_seen": 2630236,
      "step": 2549,
      "train_runtime": 6178.9982,
      "train_tokens_per_second": 425.674
    },
    {
      "epoch": 1.0897531787584143,
      "grad_norm": 3.5732882022857666,
      "learning_rate": 9.116577540106953e-06,
      "loss": 0.1789730340242386,
      "num_input_tokens_seen": 2631200,
      "step": 2550,
      "train_runtime": 6181.0568,
      "train_tokens_per_second": 425.688
    },
    {
      "epoch": 1.0901805748477402,
      "grad_norm": 2.2895474433898926,
      "learning_rate": 9.112299465240642e-06,
      "loss": 0.07485906779766083,
      "num_input_tokens_seen": 2632364,
      "step": 2551,
      "train_runtime": 6189.8056,
      "train_tokens_per_second": 425.274
    },
    {
      "epoch": 1.090607970937066,
      "grad_norm": 2.6540567874908447,
      "learning_rate": 9.108021390374332e-06,
      "loss": 0.054442599415779114,
      "num_input_tokens_seen": 2633204,
      "step": 2552,
      "train_runtime": 6191.8229,
      "train_tokens_per_second": 425.271
    },
    {
      "epoch": 1.0910353670263917,
      "grad_norm": 2.39176869392395,
      "learning_rate": 9.103743315508023e-06,
      "loss": 0.078837551176548,
      "num_input_tokens_seen": 2634234,
      "step": 2553,
      "train_runtime": 6194.0669,
      "train_tokens_per_second": 425.283
    },
    {
      "epoch": 1.0914627631157174,
      "grad_norm": 2.3524105548858643,
      "learning_rate": 9.099465240641712e-06,
      "loss": 0.09189550578594208,
      "num_input_tokens_seen": 2635198,
      "step": 2554,
      "train_runtime": 6196.1704,
      "train_tokens_per_second": 425.295
    },
    {
      "epoch": 1.0918901592050432,
      "grad_norm": 2.4514803886413574,
      "learning_rate": 9.095187165775402e-06,
      "loss": 0.08409885317087173,
      "num_input_tokens_seen": 2635956,
      "step": 2555,
      "train_runtime": 6198.1886,
      "train_tokens_per_second": 425.278
    },
    {
      "epoch": 1.0923175552943691,
      "grad_norm": 2.3649215698242188,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.06314463913440704,
      "num_input_tokens_seen": 2636606,
      "step": 2556,
      "train_runtime": 6200.1454,
      "train_tokens_per_second": 425.249
    },
    {
      "epoch": 1.0927449513836949,
      "grad_norm": 2.2242050170898438,
      "learning_rate": 9.086631016042782e-06,
      "loss": 0.07727743685245514,
      "num_input_tokens_seen": 2637850,
      "step": 2557,
      "train_runtime": 6202.2622,
      "train_tokens_per_second": 425.304
    },
    {
      "epoch": 1.0931723474730206,
      "grad_norm": 2.1953177452087402,
      "learning_rate": 9.082352941176472e-06,
      "loss": 0.09806112945079803,
      "num_input_tokens_seen": 2639198,
      "step": 2558,
      "train_runtime": 6204.4201,
      "train_tokens_per_second": 425.374
    },
    {
      "epoch": 1.0935997435623463,
      "grad_norm": 2.17103910446167,
      "learning_rate": 9.078074866310161e-06,
      "loss": 0.06100569665431976,
      "num_input_tokens_seen": 2640076,
      "step": 2559,
      "train_runtime": 6206.5183,
      "train_tokens_per_second": 425.371
    },
    {
      "epoch": 1.094027139651672,
      "grad_norm": 3.6131880283355713,
      "learning_rate": 9.07379679144385e-06,
      "loss": 0.0696096196770668,
      "num_input_tokens_seen": 2640912,
      "step": 2560,
      "train_runtime": 6208.5815,
      "train_tokens_per_second": 425.365
    },
    {
      "epoch": 1.094454535740998,
      "grad_norm": 3.7856390476226807,
      "learning_rate": 9.069518716577542e-06,
      "loss": 0.11277409642934799,
      "num_input_tokens_seen": 2642292,
      "step": 2561,
      "train_runtime": 6210.766,
      "train_tokens_per_second": 425.437
    },
    {
      "epoch": 1.0948819318303238,
      "grad_norm": 2.5144784450531006,
      "learning_rate": 9.065240641711231e-06,
      "loss": 0.08400897681713104,
      "num_input_tokens_seen": 2643486,
      "step": 2562,
      "train_runtime": 6212.8907,
      "train_tokens_per_second": 425.484
    },
    {
      "epoch": 1.0953093279196495,
      "grad_norm": 2.886179208755493,
      "learning_rate": 9.06096256684492e-06,
      "loss": 0.11095649003982544,
      "num_input_tokens_seen": 2644560,
      "step": 2563,
      "train_runtime": 6215.0488,
      "train_tokens_per_second": 425.509
    },
    {
      "epoch": 1.0957367240089753,
      "grad_norm": 2.416893720626831,
      "learning_rate": 9.05668449197861e-06,
      "loss": 0.0961887538433075,
      "num_input_tokens_seen": 2645550,
      "step": 2564,
      "train_runtime": 6217.1088,
      "train_tokens_per_second": 425.527
    },
    {
      "epoch": 1.096164120098301,
      "grad_norm": 2.321659803390503,
      "learning_rate": 9.052406417112301e-06,
      "loss": 0.060192909091711044,
      "num_input_tokens_seen": 2646420,
      "step": 2565,
      "train_runtime": 6219.1648,
      "train_tokens_per_second": 425.527
    },
    {
      "epoch": 1.096591516187627,
      "grad_norm": 1.7480592727661133,
      "learning_rate": 9.04812834224599e-06,
      "loss": 0.07109661400318146,
      "num_input_tokens_seen": 2647648,
      "step": 2566,
      "train_runtime": 6221.2835,
      "train_tokens_per_second": 425.579
    },
    {
      "epoch": 1.0970189122769527,
      "grad_norm": 2.828977108001709,
      "learning_rate": 9.04385026737968e-06,
      "loss": 0.07299567759037018,
      "num_input_tokens_seen": 2648570,
      "step": 2567,
      "train_runtime": 6223.3506,
      "train_tokens_per_second": 425.586
    },
    {
      "epoch": 1.0974463083662784,
      "grad_norm": 3.1151654720306396,
      "learning_rate": 9.03957219251337e-06,
      "loss": 0.10197779536247253,
      "num_input_tokens_seen": 2649326,
      "step": 2568,
      "train_runtime": 6225.3967,
      "train_tokens_per_second": 425.567
    },
    {
      "epoch": 1.0978737044556042,
      "grad_norm": 4.076137065887451,
      "learning_rate": 9.03529411764706e-06,
      "loss": 0.11928398907184601,
      "num_input_tokens_seen": 2650346,
      "step": 2569,
      "train_runtime": 6227.4496,
      "train_tokens_per_second": 425.591
    },
    {
      "epoch": 1.09830110054493,
      "grad_norm": 1.9719167947769165,
      "learning_rate": 9.031016042780748e-06,
      "loss": 0.05208585411310196,
      "num_input_tokens_seen": 2651832,
      "step": 2570,
      "train_runtime": 6229.6632,
      "train_tokens_per_second": 425.678
    },
    {
      "epoch": 1.0987284966342559,
      "grad_norm": 3.1587843894958496,
      "learning_rate": 9.02673796791444e-06,
      "loss": 0.10116727650165558,
      "num_input_tokens_seen": 2652696,
      "step": 2571,
      "train_runtime": 6231.7094,
      "train_tokens_per_second": 425.677
    },
    {
      "epoch": 1.0991558927235816,
      "grad_norm": 1.8884376287460327,
      "learning_rate": 9.022459893048129e-06,
      "loss": 0.08448435366153717,
      "num_input_tokens_seen": 2654062,
      "step": 2572,
      "train_runtime": 6233.8896,
      "train_tokens_per_second": 425.747
    },
    {
      "epoch": 1.0995832888129073,
      "grad_norm": 1.7889384031295776,
      "learning_rate": 9.01818181818182e-06,
      "loss": 0.0389912873506546,
      "num_input_tokens_seen": 2655048,
      "step": 2573,
      "train_runtime": 6235.9564,
      "train_tokens_per_second": 425.764
    },
    {
      "epoch": 1.100010684902233,
      "grad_norm": 2.9972827434539795,
      "learning_rate": 9.013903743315508e-06,
      "loss": 0.07092035561800003,
      "num_input_tokens_seen": 2656032,
      "step": 2574,
      "train_runtime": 6238.0401,
      "train_tokens_per_second": 425.78
    },
    {
      "epoch": 1.1004380809915588,
      "grad_norm": 2.1166203022003174,
      "learning_rate": 9.009625668449199e-06,
      "loss": 0.06895717233419418,
      "num_input_tokens_seen": 2657772,
      "step": 2575,
      "train_runtime": 6240.3216,
      "train_tokens_per_second": 425.903
    },
    {
      "epoch": 1.1008654770808848,
      "grad_norm": 2.402984142303467,
      "learning_rate": 9.005347593582888e-06,
      "loss": 0.07706868648529053,
      "num_input_tokens_seen": 2658732,
      "step": 2576,
      "train_runtime": 6242.3975,
      "train_tokens_per_second": 425.915
    },
    {
      "epoch": 1.1012928731702105,
      "grad_norm": 2.4520437717437744,
      "learning_rate": 9.001069518716578e-06,
      "loss": 0.08827337622642517,
      "num_input_tokens_seen": 2659752,
      "step": 2577,
      "train_runtime": 6244.4653,
      "train_tokens_per_second": 425.938
    },
    {
      "epoch": 1.1017202692595363,
      "grad_norm": 3.3706552982330322,
      "learning_rate": 8.996791443850267e-06,
      "loss": 0.08362919092178345,
      "num_input_tokens_seen": 2660538,
      "step": 2578,
      "train_runtime": 6246.477,
      "train_tokens_per_second": 425.926
    },
    {
      "epoch": 1.102147665348862,
      "grad_norm": 3.8421828746795654,
      "learning_rate": 8.992513368983958e-06,
      "loss": 0.14808999001979828,
      "num_input_tokens_seen": 2661354,
      "step": 2579,
      "train_runtime": 6248.5049,
      "train_tokens_per_second": 425.919
    },
    {
      "epoch": 1.1025750614381877,
      "grad_norm": 2.08762264251709,
      "learning_rate": 8.988235294117648e-06,
      "loss": 0.06539730727672577,
      "num_input_tokens_seen": 2662054,
      "step": 2580,
      "train_runtime": 6250.5004,
      "train_tokens_per_second": 425.895
    },
    {
      "epoch": 1.1030024575275137,
      "grad_norm": 5.158356189727783,
      "learning_rate": 8.983957219251337e-06,
      "loss": 0.11088553816080093,
      "num_input_tokens_seen": 2662830,
      "step": 2581,
      "train_runtime": 6259.1583,
      "train_tokens_per_second": 425.429
    },
    {
      "epoch": 1.1034298536168394,
      "grad_norm": 2.46724534034729,
      "learning_rate": 8.979679144385027e-06,
      "loss": 0.07529918104410172,
      "num_input_tokens_seen": 2663718,
      "step": 2582,
      "train_runtime": 6261.2312,
      "train_tokens_per_second": 425.43
    },
    {
      "epoch": 1.1038572497061652,
      "grad_norm": 4.171766757965088,
      "learning_rate": 8.975401069518718e-06,
      "loss": 0.11726517230272293,
      "num_input_tokens_seen": 2665074,
      "step": 2583,
      "train_runtime": 6263.4133,
      "train_tokens_per_second": 425.499
    },
    {
      "epoch": 1.104284645795491,
      "grad_norm": 5.988823890686035,
      "learning_rate": 8.971122994652407e-06,
      "loss": 0.12138627469539642,
      "num_input_tokens_seen": 2665798,
      "step": 2584,
      "train_runtime": 6265.4908,
      "train_tokens_per_second": 425.473
    },
    {
      "epoch": 1.1047120418848166,
      "grad_norm": 4.671169281005859,
      "learning_rate": 8.966844919786097e-06,
      "loss": 0.10450796037912369,
      "num_input_tokens_seen": 2666380,
      "step": 2585,
      "train_runtime": 6267.514,
      "train_tokens_per_second": 425.429
    },
    {
      "epoch": 1.1051394379741426,
      "grad_norm": 3.142690896987915,
      "learning_rate": 8.962566844919786e-06,
      "loss": 0.09894594550132751,
      "num_input_tokens_seen": 2667178,
      "step": 2586,
      "train_runtime": 6269.5349,
      "train_tokens_per_second": 425.419
    },
    {
      "epoch": 1.1055668340634683,
      "grad_norm": 2.8497908115386963,
      "learning_rate": 8.958288770053477e-06,
      "loss": 0.1103810966014862,
      "num_input_tokens_seen": 2668086,
      "step": 2587,
      "train_runtime": 6271.6029,
      "train_tokens_per_second": 425.423
    },
    {
      "epoch": 1.105994230152794,
      "grad_norm": 2.5532639026641846,
      "learning_rate": 8.954010695187167e-06,
      "loss": 0.09244312345981598,
      "num_input_tokens_seen": 2668928,
      "step": 2588,
      "train_runtime": 6273.6795,
      "train_tokens_per_second": 425.417
    },
    {
      "epoch": 1.1064216262421198,
      "grad_norm": 2.962883234024048,
      "learning_rate": 8.949732620320856e-06,
      "loss": 0.10032179951667786,
      "num_input_tokens_seen": 2670242,
      "step": 2589,
      "train_runtime": 6275.8322,
      "train_tokens_per_second": 425.48
    },
    {
      "epoch": 1.1068490223314456,
      "grad_norm": 2.0307984352111816,
      "learning_rate": 8.945454545454546e-06,
      "loss": 0.06244519725441933,
      "num_input_tokens_seen": 2671122,
      "step": 2590,
      "train_runtime": 6277.8623,
      "train_tokens_per_second": 425.483
    },
    {
      "epoch": 1.1072764184207715,
      "grad_norm": 2.161024570465088,
      "learning_rate": 8.941176470588237e-06,
      "loss": 0.08688432723283768,
      "num_input_tokens_seen": 2672340,
      "step": 2591,
      "train_runtime": 6280.0155,
      "train_tokens_per_second": 425.531
    },
    {
      "epoch": 1.1077038145100973,
      "grad_norm": 2.793954849243164,
      "learning_rate": 8.936898395721926e-06,
      "loss": 0.07023809105157852,
      "num_input_tokens_seen": 2673234,
      "step": 2592,
      "train_runtime": 6282.1453,
      "train_tokens_per_second": 425.529
    },
    {
      "epoch": 1.108131210599423,
      "grad_norm": 2.716843843460083,
      "learning_rate": 8.932620320855616e-06,
      "loss": 0.10120771825313568,
      "num_input_tokens_seen": 2674702,
      "step": 2593,
      "train_runtime": 6284.4485,
      "train_tokens_per_second": 425.606
    },
    {
      "epoch": 1.1085586066887487,
      "grad_norm": 2.413733720779419,
      "learning_rate": 8.928342245989305e-06,
      "loss": 0.07094432413578033,
      "num_input_tokens_seen": 2675602,
      "step": 2594,
      "train_runtime": 6286.6914,
      "train_tokens_per_second": 425.598
    },
    {
      "epoch": 1.1089860027780745,
      "grad_norm": 2.8126838207244873,
      "learning_rate": 8.924064171122996e-06,
      "loss": 0.08973577618598938,
      "num_input_tokens_seen": 2676792,
      "step": 2595,
      "train_runtime": 6288.8056,
      "train_tokens_per_second": 425.644
    },
    {
      "epoch": 1.1094133988674004,
      "grad_norm": 1.8998647928237915,
      "learning_rate": 8.919786096256686e-06,
      "loss": 0.06264329701662064,
      "num_input_tokens_seen": 2677698,
      "step": 2596,
      "train_runtime": 6290.8786,
      "train_tokens_per_second": 425.648
    },
    {
      "epoch": 1.1098407949567262,
      "grad_norm": 2.1570041179656982,
      "learning_rate": 8.915508021390375e-06,
      "loss": 0.07515907287597656,
      "num_input_tokens_seen": 2678702,
      "step": 2597,
      "train_runtime": 6292.9884,
      "train_tokens_per_second": 425.665
    },
    {
      "epoch": 1.110268191046052,
      "grad_norm": 2.8820595741271973,
      "learning_rate": 8.911229946524065e-06,
      "loss": 0.08067453652620316,
      "num_input_tokens_seen": 2679418,
      "step": 2598,
      "train_runtime": 6295.0155,
      "train_tokens_per_second": 425.641
    },
    {
      "epoch": 1.1106955871353776,
      "grad_norm": 3.3006863594055176,
      "learning_rate": 8.906951871657756e-06,
      "loss": 0.10144171118736267,
      "num_input_tokens_seen": 2681012,
      "step": 2599,
      "train_runtime": 6297.218,
      "train_tokens_per_second": 425.745
    },
    {
      "epoch": 1.1111229832247036,
      "grad_norm": 2.3386714458465576,
      "learning_rate": 8.902673796791445e-06,
      "loss": 0.07165563106536865,
      "num_input_tokens_seen": 2682632,
      "step": 2600,
      "train_runtime": 6299.4293,
      "train_tokens_per_second": 425.853
    },
    {
      "epoch": 1.1115503793140293,
      "grad_norm": 3.2718570232391357,
      "learning_rate": 8.898395721925135e-06,
      "loss": 0.0888989195227623,
      "num_input_tokens_seen": 2683340,
      "step": 2601,
      "train_runtime": 6301.4961,
      "train_tokens_per_second": 425.826
    },
    {
      "epoch": 1.111977775403355,
      "grad_norm": 2.0718796253204346,
      "learning_rate": 8.894117647058824e-06,
      "loss": 0.08921791613101959,
      "num_input_tokens_seen": 2684648,
      "step": 2602,
      "train_runtime": 6303.6537,
      "train_tokens_per_second": 425.888
    },
    {
      "epoch": 1.1124051714926808,
      "grad_norm": 1.9214931726455688,
      "learning_rate": 8.889839572192515e-06,
      "loss": 0.04138598591089249,
      "num_input_tokens_seen": 2685516,
      "step": 2603,
      "train_runtime": 6305.696,
      "train_tokens_per_second": 425.887
    },
    {
      "epoch": 1.1128325675820065,
      "grad_norm": 3.3434629440307617,
      "learning_rate": 8.885561497326203e-06,
      "loss": 0.10868586599826813,
      "num_input_tokens_seen": 2686648,
      "step": 2604,
      "train_runtime": 6307.9632,
      "train_tokens_per_second": 425.914
    },
    {
      "epoch": 1.1132599636713325,
      "grad_norm": 2.715587615966797,
      "learning_rate": 8.881283422459894e-06,
      "loss": 0.08680950105190277,
      "num_input_tokens_seen": 2687476,
      "step": 2605,
      "train_runtime": 6310.0101,
      "train_tokens_per_second": 425.907
    },
    {
      "epoch": 1.1136873597606582,
      "grad_norm": 3.1198415756225586,
      "learning_rate": 8.877005347593584e-06,
      "loss": 0.1386384665966034,
      "num_input_tokens_seen": 2688386,
      "step": 2606,
      "train_runtime": 6312.0721,
      "train_tokens_per_second": 425.912
    },
    {
      "epoch": 1.114114755849984,
      "grad_norm": 3.0754501819610596,
      "learning_rate": 8.872727272727275e-06,
      "loss": 0.07827456295490265,
      "num_input_tokens_seen": 2689104,
      "step": 2607,
      "train_runtime": 6314.0712,
      "train_tokens_per_second": 425.891
    },
    {
      "epoch": 1.1145421519393097,
      "grad_norm": 2.766815662384033,
      "learning_rate": 8.868449197860963e-06,
      "loss": 0.06242305040359497,
      "num_input_tokens_seen": 2689726,
      "step": 2608,
      "train_runtime": 6316.1329,
      "train_tokens_per_second": 425.85
    },
    {
      "epoch": 1.1149695480286355,
      "grad_norm": 2.9260456562042236,
      "learning_rate": 8.864171122994654e-06,
      "loss": 0.06722096353769302,
      "num_input_tokens_seen": 2690360,
      "step": 2609,
      "train_runtime": 6318.0746,
      "train_tokens_per_second": 425.82
    },
    {
      "epoch": 1.1153969441179614,
      "grad_norm": 2.6466591358184814,
      "learning_rate": 8.859893048128343e-06,
      "loss": 0.08745411783456802,
      "num_input_tokens_seen": 2691242,
      "step": 2610,
      "train_runtime": 6320.1518,
      "train_tokens_per_second": 425.819
    },
    {
      "epoch": 1.1158243402072872,
      "grad_norm": 2.320122718811035,
      "learning_rate": 8.855614973262033e-06,
      "loss": 0.04856119304895401,
      "num_input_tokens_seen": 2691864,
      "step": 2611,
      "train_runtime": 6328.8607,
      "train_tokens_per_second": 425.332
    },
    {
      "epoch": 1.116251736296613,
      "grad_norm": 3.1276562213897705,
      "learning_rate": 8.851336898395722e-06,
      "loss": 0.12471344321966171,
      "num_input_tokens_seen": 2692834,
      "step": 2612,
      "train_runtime": 6330.9446,
      "train_tokens_per_second": 425.345
    },
    {
      "epoch": 1.1166791323859386,
      "grad_norm": 3.37678861618042,
      "learning_rate": 8.847058823529413e-06,
      "loss": 0.060979779809713364,
      "num_input_tokens_seen": 2693766,
      "step": 2613,
      "train_runtime": 6333.0852,
      "train_tokens_per_second": 425.348
    },
    {
      "epoch": 1.1171065284752644,
      "grad_norm": 3.7843270301818848,
      "learning_rate": 8.842780748663103e-06,
      "loss": 0.09330800175666809,
      "num_input_tokens_seen": 2694898,
      "step": 2614,
      "train_runtime": 6335.2651,
      "train_tokens_per_second": 425.38
    },
    {
      "epoch": 1.1175339245645903,
      "grad_norm": 4.6900715827941895,
      "learning_rate": 8.838502673796792e-06,
      "loss": 0.13932695984840393,
      "num_input_tokens_seen": 2695682,
      "step": 2615,
      "train_runtime": 6337.2889,
      "train_tokens_per_second": 425.368
    },
    {
      "epoch": 1.117961320653916,
      "grad_norm": 2.6678571701049805,
      "learning_rate": 8.834224598930481e-06,
      "loss": 0.10771764814853668,
      "num_input_tokens_seen": 2696650,
      "step": 2616,
      "train_runtime": 6339.3926,
      "train_tokens_per_second": 425.38
    },
    {
      "epoch": 1.1183887167432418,
      "grad_norm": 2.3856754302978516,
      "learning_rate": 8.829946524064171e-06,
      "loss": 0.06541471183300018,
      "num_input_tokens_seen": 2697664,
      "step": 2617,
      "train_runtime": 6341.512,
      "train_tokens_per_second": 425.398
    },
    {
      "epoch": 1.1188161128325675,
      "grad_norm": 2.3845884799957275,
      "learning_rate": 8.825668449197862e-06,
      "loss": 0.08822523057460785,
      "num_input_tokens_seen": 2698754,
      "step": 2618,
      "train_runtime": 6343.6211,
      "train_tokens_per_second": 425.428
    },
    {
      "epoch": 1.1192435089218933,
      "grad_norm": 3.557136058807373,
      "learning_rate": 8.821390374331552e-06,
      "loss": 0.08408638834953308,
      "num_input_tokens_seen": 2699628,
      "step": 2619,
      "train_runtime": 6345.6715,
      "train_tokens_per_second": 425.428
    },
    {
      "epoch": 1.1196709050112192,
      "grad_norm": 2.9850730895996094,
      "learning_rate": 8.817112299465241e-06,
      "loss": 0.048864059150218964,
      "num_input_tokens_seen": 2700306,
      "step": 2620,
      "train_runtime": 6347.6755,
      "train_tokens_per_second": 425.401
    },
    {
      "epoch": 1.120098301100545,
      "grad_norm": 2.5716195106506348,
      "learning_rate": 8.81283422459893e-06,
      "loss": 0.06911540031433105,
      "num_input_tokens_seen": 2701082,
      "step": 2621,
      "train_runtime": 6349.7376,
      "train_tokens_per_second": 425.385
    },
    {
      "epoch": 1.1205256971898707,
      "grad_norm": 2.721938371658325,
      "learning_rate": 8.808556149732622e-06,
      "loss": 0.08100318908691406,
      "num_input_tokens_seen": 2701912,
      "step": 2622,
      "train_runtime": 6351.8767,
      "train_tokens_per_second": 425.372
    },
    {
      "epoch": 1.1209530932791965,
      "grad_norm": 2.389178991317749,
      "learning_rate": 8.804278074866311e-06,
      "loss": 0.11711670458316803,
      "num_input_tokens_seen": 2703786,
      "step": 2623,
      "train_runtime": 6354.3501,
      "train_tokens_per_second": 425.502
    },
    {
      "epoch": 1.1213804893685222,
      "grad_norm": 2.9732987880706787,
      "learning_rate": 8.8e-06,
      "loss": 0.1260043829679489,
      "num_input_tokens_seen": 2705028,
      "step": 2624,
      "train_runtime": 6356.497,
      "train_tokens_per_second": 425.553
    },
    {
      "epoch": 1.1218078854578482,
      "grad_norm": 2.8061492443084717,
      "learning_rate": 8.79572192513369e-06,
      "loss": 0.0765015110373497,
      "num_input_tokens_seen": 2706344,
      "step": 2625,
      "train_runtime": 6358.6317,
      "train_tokens_per_second": 425.617
    },
    {
      "epoch": 1.122235281547174,
      "grad_norm": 3.7093279361724854,
      "learning_rate": 8.791443850267381e-06,
      "loss": 0.09670396894216537,
      "num_input_tokens_seen": 2707206,
      "step": 2626,
      "train_runtime": 6360.6675,
      "train_tokens_per_second": 425.617
    },
    {
      "epoch": 1.1226626776364996,
      "grad_norm": 2.739389657974243,
      "learning_rate": 8.78716577540107e-06,
      "loss": 0.08371162414550781,
      "num_input_tokens_seen": 2708152,
      "step": 2627,
      "train_runtime": 6362.8322,
      "train_tokens_per_second": 425.621
    },
    {
      "epoch": 1.1230900737258254,
      "grad_norm": 2.908888816833496,
      "learning_rate": 8.78288770053476e-06,
      "loss": 0.11477210372686386,
      "num_input_tokens_seen": 2708942,
      "step": 2628,
      "train_runtime": 6364.8749,
      "train_tokens_per_second": 425.608
    },
    {
      "epoch": 1.123517469815151,
      "grad_norm": 3.44331955909729,
      "learning_rate": 8.77860962566845e-06,
      "loss": 0.1313876211643219,
      "num_input_tokens_seen": 2709744,
      "step": 2629,
      "train_runtime": 6366.9287,
      "train_tokens_per_second": 425.597
    },
    {
      "epoch": 1.123944865904477,
      "grad_norm": 2.745544672012329,
      "learning_rate": 8.77433155080214e-06,
      "loss": 0.04821210727095604,
      "num_input_tokens_seen": 2710754,
      "step": 2630,
      "train_runtime": 6369.0665,
      "train_tokens_per_second": 425.612
    },
    {
      "epoch": 1.1243722619938028,
      "grad_norm": 2.3741164207458496,
      "learning_rate": 8.77005347593583e-06,
      "loss": 0.05976215377449989,
      "num_input_tokens_seen": 2711878,
      "step": 2631,
      "train_runtime": 6371.2411,
      "train_tokens_per_second": 425.644
    },
    {
      "epoch": 1.1247996580831285,
      "grad_norm": 1.262378454208374,
      "learning_rate": 8.76577540106952e-06,
      "loss": 0.037514060735702515,
      "num_input_tokens_seen": 2713688,
      "step": 2632,
      "train_runtime": 6373.538,
      "train_tokens_per_second": 425.774
    },
    {
      "epoch": 1.1252270541724543,
      "grad_norm": 2.302104949951172,
      "learning_rate": 8.761497326203209e-06,
      "loss": 0.058407776057720184,
      "num_input_tokens_seen": 2714632,
      "step": 2633,
      "train_runtime": 6375.6249,
      "train_tokens_per_second": 425.783
    },
    {
      "epoch": 1.12565445026178,
      "grad_norm": 2.742218017578125,
      "learning_rate": 8.7572192513369e-06,
      "loss": 0.08033092319965363,
      "num_input_tokens_seen": 2715508,
      "step": 2634,
      "train_runtime": 6377.6985,
      "train_tokens_per_second": 425.782
    },
    {
      "epoch": 1.126081846351106,
      "grad_norm": 1.60816490650177,
      "learning_rate": 8.752941176470588e-06,
      "loss": 0.06637120246887207,
      "num_input_tokens_seen": 2717434,
      "step": 2635,
      "train_runtime": 6379.9872,
      "train_tokens_per_second": 425.931
    },
    {
      "epoch": 1.1265092424404317,
      "grad_norm": 2.022418975830078,
      "learning_rate": 8.748663101604279e-06,
      "loss": 0.039730578660964966,
      "num_input_tokens_seen": 2718092,
      "step": 2636,
      "train_runtime": 6381.9735,
      "train_tokens_per_second": 425.901
    },
    {
      "epoch": 1.1269366385297575,
      "grad_norm": 2.5975489616394043,
      "learning_rate": 8.744385026737968e-06,
      "loss": 0.11269330978393555,
      "num_input_tokens_seen": 2718962,
      "step": 2637,
      "train_runtime": 6384.0256,
      "train_tokens_per_second": 425.901
    },
    {
      "epoch": 1.1273640346190832,
      "grad_norm": 2.9667372703552246,
      "learning_rate": 8.740106951871658e-06,
      "loss": 0.09536286443471909,
      "num_input_tokens_seen": 2719680,
      "step": 2638,
      "train_runtime": 6386.0096,
      "train_tokens_per_second": 425.881
    },
    {
      "epoch": 1.127791430708409,
      "grad_norm": 2.3580756187438965,
      "learning_rate": 8.735828877005347e-06,
      "loss": 0.08065667003393173,
      "num_input_tokens_seen": 2720836,
      "step": 2639,
      "train_runtime": 6388.1295,
      "train_tokens_per_second": 425.921
    },
    {
      "epoch": 1.1282188267977349,
      "grad_norm": 3.448974609375,
      "learning_rate": 8.731550802139038e-06,
      "loss": 0.07800039649009705,
      "num_input_tokens_seen": 2722444,
      "step": 2640,
      "train_runtime": 6390.368,
      "train_tokens_per_second": 426.023
    },
    {
      "epoch": 1.1286462228870606,
      "grad_norm": 2.6232998371124268,
      "learning_rate": 8.727272727272728e-06,
      "loss": 0.10187067091464996,
      "num_input_tokens_seen": 2723656,
      "step": 2641,
      "train_runtime": 6398.7207,
      "train_tokens_per_second": 425.656
    },
    {
      "epoch": 1.1290736189763864,
      "grad_norm": 2.688281774520874,
      "learning_rate": 8.722994652406417e-06,
      "loss": 0.08508885651826859,
      "num_input_tokens_seen": 2724486,
      "step": 2642,
      "train_runtime": 6400.768,
      "train_tokens_per_second": 425.65
    },
    {
      "epoch": 1.129501015065712,
      "grad_norm": 2.871506690979004,
      "learning_rate": 8.718716577540107e-06,
      "loss": 0.09198738634586334,
      "num_input_tokens_seen": 2725294,
      "step": 2643,
      "train_runtime": 6402.9862,
      "train_tokens_per_second": 425.629
    },
    {
      "epoch": 1.1299284111550378,
      "grad_norm": 1.865782618522644,
      "learning_rate": 8.714438502673798e-06,
      "loss": 0.06879622489213943,
      "num_input_tokens_seen": 2726300,
      "step": 2644,
      "train_runtime": 6405.1109,
      "train_tokens_per_second": 425.644
    },
    {
      "epoch": 1.1303558072443638,
      "grad_norm": 2.213736057281494,
      "learning_rate": 8.710160427807487e-06,
      "loss": 0.04665439948439598,
      "num_input_tokens_seen": 2727268,
      "step": 2645,
      "train_runtime": 6407.1835,
      "train_tokens_per_second": 425.658
    },
    {
      "epoch": 1.1307832033336895,
      "grad_norm": 2.2345521450042725,
      "learning_rate": 8.705882352941177e-06,
      "loss": 0.06750096380710602,
      "num_input_tokens_seen": 2728296,
      "step": 2646,
      "train_runtime": 6409.2622,
      "train_tokens_per_second": 425.68
    },
    {
      "epoch": 1.1312105994230153,
      "grad_norm": 2.2993998527526855,
      "learning_rate": 8.701604278074866e-06,
      "loss": 0.0574868842959404,
      "num_input_tokens_seen": 2729002,
      "step": 2647,
      "train_runtime": 6411.2395,
      "train_tokens_per_second": 425.659
    },
    {
      "epoch": 1.131637995512341,
      "grad_norm": 2.5204601287841797,
      "learning_rate": 8.697326203208557e-06,
      "loss": 0.0829242691397667,
      "num_input_tokens_seen": 2730002,
      "step": 2648,
      "train_runtime": 6413.2911,
      "train_tokens_per_second": 425.679
    },
    {
      "epoch": 1.1320653916016667,
      "grad_norm": 2.1150095462799072,
      "learning_rate": 8.693048128342247e-06,
      "loss": 0.053915947675704956,
      "num_input_tokens_seen": 2730738,
      "step": 2649,
      "train_runtime": 6415.3685,
      "train_tokens_per_second": 425.656
    },
    {
      "epoch": 1.1324927876909927,
      "grad_norm": 2.3570218086242676,
      "learning_rate": 8.688770053475936e-06,
      "loss": 0.0937032625079155,
      "num_input_tokens_seen": 2731792,
      "step": 2650,
      "train_runtime": 6417.4365,
      "train_tokens_per_second": 425.683
    },
    {
      "epoch": 1.1329201837803184,
      "grad_norm": 2.1232759952545166,
      "learning_rate": 8.684491978609626e-06,
      "loss": 0.0683131068944931,
      "num_input_tokens_seen": 2732846,
      "step": 2651,
      "train_runtime": 6419.5167,
      "train_tokens_per_second": 425.709
    },
    {
      "epoch": 1.1333475798696442,
      "grad_norm": 3.5161032676696777,
      "learning_rate": 8.680213903743317e-06,
      "loss": 0.10581836104393005,
      "num_input_tokens_seen": 2733504,
      "step": 2652,
      "train_runtime": 6421.5291,
      "train_tokens_per_second": 425.678
    },
    {
      "epoch": 1.13377497595897,
      "grad_norm": 2.6642963886260986,
      "learning_rate": 8.675935828877006e-06,
      "loss": 0.11502748727798462,
      "num_input_tokens_seen": 2734658,
      "step": 2653,
      "train_runtime": 6423.637,
      "train_tokens_per_second": 425.718
    },
    {
      "epoch": 1.1342023720482959,
      "grad_norm": 3.2872936725616455,
      "learning_rate": 8.671657754010696e-06,
      "loss": 0.07190796732902527,
      "num_input_tokens_seen": 2735978,
      "step": 2654,
      "train_runtime": 6425.7813,
      "train_tokens_per_second": 425.781
    },
    {
      "epoch": 1.1346297681376216,
      "grad_norm": 2.496293544769287,
      "learning_rate": 8.667379679144385e-06,
      "loss": 0.07971389591693878,
      "num_input_tokens_seen": 2736868,
      "step": 2655,
      "train_runtime": 6427.8901,
      "train_tokens_per_second": 425.78
    },
    {
      "epoch": 1.1350571642269474,
      "grad_norm": 1.905638575553894,
      "learning_rate": 8.663101604278076e-06,
      "loss": 0.06217019259929657,
      "num_input_tokens_seen": 2737854,
      "step": 2656,
      "train_runtime": 6429.9512,
      "train_tokens_per_second": 425.797
    },
    {
      "epoch": 1.135484560316273,
      "grad_norm": 2.026752233505249,
      "learning_rate": 8.658823529411766e-06,
      "loss": 0.05629231780767441,
      "num_input_tokens_seen": 2738880,
      "step": 2657,
      "train_runtime": 6432.0445,
      "train_tokens_per_second": 425.818
    },
    {
      "epoch": 1.1359119564055988,
      "grad_norm": 2.248422622680664,
      "learning_rate": 8.654545454545455e-06,
      "loss": 0.07973220199346542,
      "num_input_tokens_seen": 2740298,
      "step": 2658,
      "train_runtime": 6434.1799,
      "train_tokens_per_second": 425.897
    },
    {
      "epoch": 1.1363393524949248,
      "grad_norm": 2.402320146560669,
      "learning_rate": 8.650267379679145e-06,
      "loss": 0.07679715007543564,
      "num_input_tokens_seen": 2741316,
      "step": 2659,
      "train_runtime": 6436.2256,
      "train_tokens_per_second": 425.92
    },
    {
      "epoch": 1.1367667485842505,
      "grad_norm": 1.8567861318588257,
      "learning_rate": 8.645989304812836e-06,
      "loss": 0.06722358614206314,
      "num_input_tokens_seen": 2742708,
      "step": 2660,
      "train_runtime": 6438.3815,
      "train_tokens_per_second": 425.993
    },
    {
      "epoch": 1.1371941446735763,
      "grad_norm": 1.8689320087432861,
      "learning_rate": 8.641711229946525e-06,
      "loss": 0.08618605881929398,
      "num_input_tokens_seen": 2744336,
      "step": 2661,
      "train_runtime": 6440.624,
      "train_tokens_per_second": 426.098
    },
    {
      "epoch": 1.137621540762902,
      "grad_norm": 3.4328114986419678,
      "learning_rate": 8.637433155080215e-06,
      "loss": 0.09289912134408951,
      "num_input_tokens_seen": 2745066,
      "step": 2662,
      "train_runtime": 6442.6775,
      "train_tokens_per_second": 426.075
    },
    {
      "epoch": 1.1380489368522277,
      "grad_norm": 3.1104848384857178,
      "learning_rate": 8.633155080213904e-06,
      "loss": 0.06772873550653458,
      "num_input_tokens_seen": 2745924,
      "step": 2663,
      "train_runtime": 6444.7059,
      "train_tokens_per_second": 426.074
    },
    {
      "epoch": 1.1384763329415537,
      "grad_norm": 2.6316936016082764,
      "learning_rate": 8.628877005347595e-06,
      "loss": 0.08098454028367996,
      "num_input_tokens_seen": 2746630,
      "step": 2664,
      "train_runtime": 6446.725,
      "train_tokens_per_second": 426.05
    },
    {
      "epoch": 1.1389037290308794,
      "grad_norm": 3.3403708934783936,
      "learning_rate": 8.624598930481285e-06,
      "loss": 0.10924291610717773,
      "num_input_tokens_seen": 2747370,
      "step": 2665,
      "train_runtime": 6448.749,
      "train_tokens_per_second": 426.031
    },
    {
      "epoch": 1.1393311251202052,
      "grad_norm": 2.9772796630859375,
      "learning_rate": 8.620320855614974e-06,
      "loss": 0.07570470869541168,
      "num_input_tokens_seen": 2748102,
      "step": 2666,
      "train_runtime": 6450.7652,
      "train_tokens_per_second": 426.012
    },
    {
      "epoch": 1.139758521209531,
      "grad_norm": 2.6464850902557373,
      "learning_rate": 8.616042780748664e-06,
      "loss": 0.09072837978601456,
      "num_input_tokens_seen": 2749400,
      "step": 2667,
      "train_runtime": 6452.8782,
      "train_tokens_per_second": 426.073
    },
    {
      "epoch": 1.1401859172988567,
      "grad_norm": 5.082932949066162,
      "learning_rate": 8.611764705882355e-06,
      "loss": 0.0968312919139862,
      "num_input_tokens_seen": 2750084,
      "step": 2668,
      "train_runtime": 6454.8914,
      "train_tokens_per_second": 426.047
    },
    {
      "epoch": 1.1406133133881826,
      "grad_norm": 3.5218236446380615,
      "learning_rate": 8.607486631016042e-06,
      "loss": 0.15223868191242218,
      "num_input_tokens_seen": 2750886,
      "step": 2669,
      "train_runtime": 6456.8996,
      "train_tokens_per_second": 426.038
    },
    {
      "epoch": 1.1410407094775084,
      "grad_norm": 2.667015790939331,
      "learning_rate": 8.603208556149734e-06,
      "loss": 0.0598151721060276,
      "num_input_tokens_seen": 2751730,
      "step": 2670,
      "train_runtime": 6458.9227,
      "train_tokens_per_second": 426.035
    },
    {
      "epoch": 1.141468105566834,
      "grad_norm": 3.0027811527252197,
      "learning_rate": 8.598930481283423e-06,
      "loss": 0.08696714788675308,
      "num_input_tokens_seen": 2752498,
      "step": 2671,
      "train_runtime": 6467.1559,
      "train_tokens_per_second": 425.612
    },
    {
      "epoch": 1.1418955016561598,
      "grad_norm": 3.6396684646606445,
      "learning_rate": 8.594652406417112e-06,
      "loss": 0.11685916781425476,
      "num_input_tokens_seen": 2753408,
      "step": 2672,
      "train_runtime": 6469.4547,
      "train_tokens_per_second": 425.601
    },
    {
      "epoch": 1.1423228977454856,
      "grad_norm": 2.986107110977173,
      "learning_rate": 8.590374331550802e-06,
      "loss": 0.07806456089019775,
      "num_input_tokens_seen": 2754882,
      "step": 2673,
      "train_runtime": 6471.6667,
      "train_tokens_per_second": 425.684
    },
    {
      "epoch": 1.1427502938348115,
      "grad_norm": 3.2590105533599854,
      "learning_rate": 8.586096256684493e-06,
      "loss": 0.10785083472728729,
      "num_input_tokens_seen": 2755630,
      "step": 2674,
      "train_runtime": 6473.6897,
      "train_tokens_per_second": 425.666
    },
    {
      "epoch": 1.1431776899241373,
      "grad_norm": 1.9124603271484375,
      "learning_rate": 8.581818181818183e-06,
      "loss": 0.05577567219734192,
      "num_input_tokens_seen": 2756786,
      "step": 2675,
      "train_runtime": 6475.801,
      "train_tokens_per_second": 425.706
    },
    {
      "epoch": 1.143605086013463,
      "grad_norm": 3.6917645931243896,
      "learning_rate": 8.577540106951872e-06,
      "loss": 0.1246538758277893,
      "num_input_tokens_seen": 2757538,
      "step": 2676,
      "train_runtime": 6477.7772,
      "train_tokens_per_second": 425.692
    },
    {
      "epoch": 1.1440324821027887,
      "grad_norm": 3.3652679920196533,
      "learning_rate": 8.573262032085561e-06,
      "loss": 0.08964797854423523,
      "num_input_tokens_seen": 2758364,
      "step": 2677,
      "train_runtime": 6479.8357,
      "train_tokens_per_second": 425.684
    },
    {
      "epoch": 1.1444598781921145,
      "grad_norm": 2.346606731414795,
      "learning_rate": 8.568983957219253e-06,
      "loss": 0.09359855949878693,
      "num_input_tokens_seen": 2759700,
      "step": 2678,
      "train_runtime": 6482.0281,
      "train_tokens_per_second": 425.746
    },
    {
      "epoch": 1.1448872742814404,
      "grad_norm": 2.363062620162964,
      "learning_rate": 8.564705882352942e-06,
      "loss": 0.08591769635677338,
      "num_input_tokens_seen": 2760600,
      "step": 2679,
      "train_runtime": 6484.0813,
      "train_tokens_per_second": 425.75
    },
    {
      "epoch": 1.1453146703707662,
      "grad_norm": 2.9432427883148193,
      "learning_rate": 8.560427807486631e-06,
      "loss": 0.06521814316511154,
      "num_input_tokens_seen": 2761468,
      "step": 2680,
      "train_runtime": 6486.1346,
      "train_tokens_per_second": 425.749
    },
    {
      "epoch": 1.145742066460092,
      "grad_norm": 2.3472490310668945,
      "learning_rate": 8.556149732620321e-06,
      "loss": 0.10987380892038345,
      "num_input_tokens_seen": 2762912,
      "step": 2681,
      "train_runtime": 6488.2761,
      "train_tokens_per_second": 425.831
    },
    {
      "epoch": 1.1461694625494177,
      "grad_norm": 2.527367353439331,
      "learning_rate": 8.551871657754012e-06,
      "loss": 0.11642409861087799,
      "num_input_tokens_seen": 2764704,
      "step": 2682,
      "train_runtime": 6490.5527,
      "train_tokens_per_second": 425.958
    },
    {
      "epoch": 1.1465968586387434,
      "grad_norm": 3.351578712463379,
      "learning_rate": 8.547593582887701e-06,
      "loss": 0.06724368780851364,
      "num_input_tokens_seen": 2765410,
      "step": 2683,
      "train_runtime": 6492.5737,
      "train_tokens_per_second": 425.934
    },
    {
      "epoch": 1.1470242547280693,
      "grad_norm": 5.102705001831055,
      "learning_rate": 8.543315508021391e-06,
      "loss": 0.1004955917596817,
      "num_input_tokens_seen": 2766048,
      "step": 2684,
      "train_runtime": 6494.5157,
      "train_tokens_per_second": 425.905
    },
    {
      "epoch": 1.147451650817395,
      "grad_norm": 2.7378604412078857,
      "learning_rate": 8.53903743315508e-06,
      "loss": 0.09332038462162018,
      "num_input_tokens_seen": 2766806,
      "step": 2685,
      "train_runtime": 6496.5741,
      "train_tokens_per_second": 425.887
    },
    {
      "epoch": 1.1478790469067208,
      "grad_norm": 3.1623332500457764,
      "learning_rate": 8.534759358288771e-06,
      "loss": 0.09608341753482819,
      "num_input_tokens_seen": 2767648,
      "step": 2686,
      "train_runtime": 6498.6526,
      "train_tokens_per_second": 425.88
    },
    {
      "epoch": 1.1483064429960466,
      "grad_norm": 2.330491781234741,
      "learning_rate": 8.530481283422461e-06,
      "loss": 0.07984799146652222,
      "num_input_tokens_seen": 2768678,
      "step": 2687,
      "train_runtime": 6500.7332,
      "train_tokens_per_second": 425.902
    },
    {
      "epoch": 1.1487338390853723,
      "grad_norm": 3.4615375995635986,
      "learning_rate": 8.52620320855615e-06,
      "loss": 0.09428500384092331,
      "num_input_tokens_seen": 2769744,
      "step": 2688,
      "train_runtime": 6502.8183,
      "train_tokens_per_second": 425.93
    },
    {
      "epoch": 1.1491612351746983,
      "grad_norm": 3.1374282836914062,
      "learning_rate": 8.52192513368984e-06,
      "loss": 0.08604272454977036,
      "num_input_tokens_seen": 2770574,
      "step": 2689,
      "train_runtime": 6504.8799,
      "train_tokens_per_second": 425.922
    },
    {
      "epoch": 1.149588631264024,
      "grad_norm": 3.0454697608947754,
      "learning_rate": 8.517647058823531e-06,
      "loss": 0.10294269770383835,
      "num_input_tokens_seen": 2771448,
      "step": 2690,
      "train_runtime": 6506.9302,
      "train_tokens_per_second": 425.923
    },
    {
      "epoch": 1.1500160273533497,
      "grad_norm": 2.3497512340545654,
      "learning_rate": 8.51336898395722e-06,
      "loss": 0.09156949073076248,
      "num_input_tokens_seen": 2772622,
      "step": 2691,
      "train_runtime": 6509.0753,
      "train_tokens_per_second": 425.963
    },
    {
      "epoch": 1.1504434234426755,
      "grad_norm": 2.533618688583374,
      "learning_rate": 8.50909090909091e-06,
      "loss": 0.11029275506734848,
      "num_input_tokens_seen": 2774894,
      "step": 2692,
      "train_runtime": 6511.4194,
      "train_tokens_per_second": 426.158
    },
    {
      "epoch": 1.1508708195320012,
      "grad_norm": 2.0940487384796143,
      "learning_rate": 8.5048128342246e-06,
      "loss": 0.07361536473035812,
      "num_input_tokens_seen": 2776040,
      "step": 2693,
      "train_runtime": 6513.5352,
      "train_tokens_per_second": 426.196
    },
    {
      "epoch": 1.1512982156213272,
      "grad_norm": 4.621926784515381,
      "learning_rate": 8.50053475935829e-06,
      "loss": 0.07593721151351929,
      "num_input_tokens_seen": 2777018,
      "step": 2694,
      "train_runtime": 6515.6224,
      "train_tokens_per_second": 426.209
    },
    {
      "epoch": 1.151725611710653,
      "grad_norm": 2.662271022796631,
      "learning_rate": 8.49625668449198e-06,
      "loss": 0.09287705272436142,
      "num_input_tokens_seen": 2778176,
      "step": 2695,
      "train_runtime": 6517.7711,
      "train_tokens_per_second": 426.246
    },
    {
      "epoch": 1.1521530077999786,
      "grad_norm": 2.8439338207244873,
      "learning_rate": 8.49197860962567e-06,
      "loss": 0.09059453010559082,
      "num_input_tokens_seen": 2778874,
      "step": 2696,
      "train_runtime": 6519.7935,
      "train_tokens_per_second": 426.221
    },
    {
      "epoch": 1.1525804038893044,
      "grad_norm": 2.749525785446167,
      "learning_rate": 8.487700534759359e-06,
      "loss": 0.09504384547472,
      "num_input_tokens_seen": 2780042,
      "step": 2697,
      "train_runtime": 6521.9089,
      "train_tokens_per_second": 426.262
    },
    {
      "epoch": 1.1530077999786301,
      "grad_norm": 5.511195182800293,
      "learning_rate": 8.48342245989305e-06,
      "loss": 0.0819120928645134,
      "num_input_tokens_seen": 2780980,
      "step": 2698,
      "train_runtime": 6523.9635,
      "train_tokens_per_second": 426.272
    },
    {
      "epoch": 1.153435196067956,
      "grad_norm": 3.184375286102295,
      "learning_rate": 8.47914438502674e-06,
      "loss": 0.11371450126171112,
      "num_input_tokens_seen": 2782024,
      "step": 2699,
      "train_runtime": 6526.0107,
      "train_tokens_per_second": 426.298
    },
    {
      "epoch": 1.1538625921572818,
      "grad_norm": 2.064286947250366,
      "learning_rate": 8.474866310160429e-06,
      "loss": 0.06332354247570038,
      "num_input_tokens_seen": 2782990,
      "step": 2700,
      "train_runtime": 6528.0638,
      "train_tokens_per_second": 426.312
    },
    {
      "epoch": 1.1542899882466076,
      "grad_norm": 2.1286580562591553,
      "learning_rate": 8.470588235294118e-06,
      "loss": 0.09445683658123016,
      "num_input_tokens_seen": 2784464,
      "step": 2701,
      "train_runtime": 6536.4883,
      "train_tokens_per_second": 425.988
    },
    {
      "epoch": 1.1547173843359333,
      "grad_norm": 2.6281023025512695,
      "learning_rate": 8.46631016042781e-06,
      "loss": 0.06909255683422089,
      "num_input_tokens_seen": 2785448,
      "step": 2702,
      "train_runtime": 6538.6702,
      "train_tokens_per_second": 425.996
    },
    {
      "epoch": 1.155144780425259,
      "grad_norm": 2.761331558227539,
      "learning_rate": 8.462032085561497e-06,
      "loss": 0.09310590475797653,
      "num_input_tokens_seen": 2786576,
      "step": 2703,
      "train_runtime": 6540.8249,
      "train_tokens_per_second": 426.028
    },
    {
      "epoch": 1.155572176514585,
      "grad_norm": 2.4722471237182617,
      "learning_rate": 8.457754010695188e-06,
      "loss": 0.06891348212957382,
      "num_input_tokens_seen": 2787722,
      "step": 2704,
      "train_runtime": 6542.9836,
      "train_tokens_per_second": 426.063
    },
    {
      "epoch": 1.1559995726039107,
      "grad_norm": 2.4333081245422363,
      "learning_rate": 8.453475935828878e-06,
      "loss": 0.043512776494026184,
      "num_input_tokens_seen": 2788436,
      "step": 2705,
      "train_runtime": 6544.995,
      "train_tokens_per_second": 426.041
    },
    {
      "epoch": 1.1564269686932365,
      "grad_norm": 2.408217191696167,
      "learning_rate": 8.449197860962567e-06,
      "loss": 0.06118273735046387,
      "num_input_tokens_seen": 2789302,
      "step": 2706,
      "train_runtime": 6547.1309,
      "train_tokens_per_second": 426.034
    },
    {
      "epoch": 1.1568543647825622,
      "grad_norm": 2.1128132343292236,
      "learning_rate": 8.444919786096257e-06,
      "loss": 0.06444979459047318,
      "num_input_tokens_seen": 2790432,
      "step": 2707,
      "train_runtime": 6549.2783,
      "train_tokens_per_second": 426.067
    },
    {
      "epoch": 1.157281760871888,
      "grad_norm": 2.234581470489502,
      "learning_rate": 8.440641711229948e-06,
      "loss": 0.11141036450862885,
      "num_input_tokens_seen": 2791456,
      "step": 2708,
      "train_runtime": 6551.3717,
      "train_tokens_per_second": 426.087
    },
    {
      "epoch": 1.157709156961214,
      "grad_norm": 2.561875820159912,
      "learning_rate": 8.436363636363637e-06,
      "loss": 0.061126306653022766,
      "num_input_tokens_seen": 2792190,
      "step": 2709,
      "train_runtime": 6553.3761,
      "train_tokens_per_second": 426.069
    },
    {
      "epoch": 1.1581365530505396,
      "grad_norm": 2.9232470989227295,
      "learning_rate": 8.432085561497327e-06,
      "loss": 0.09107153117656708,
      "num_input_tokens_seen": 2793306,
      "step": 2710,
      "train_runtime": 6555.4848,
      "train_tokens_per_second": 426.102
    },
    {
      "epoch": 1.1585639491398654,
      "grad_norm": 2.189206600189209,
      "learning_rate": 8.427807486631016e-06,
      "loss": 0.06996691226959229,
      "num_input_tokens_seen": 2794326,
      "step": 2711,
      "train_runtime": 6557.5283,
      "train_tokens_per_second": 426.125
    },
    {
      "epoch": 1.1589913452291911,
      "grad_norm": 2.8513875007629395,
      "learning_rate": 8.423529411764707e-06,
      "loss": 0.08728228509426117,
      "num_input_tokens_seen": 2795218,
      "step": 2712,
      "train_runtime": 6559.5893,
      "train_tokens_per_second": 426.127
    },
    {
      "epoch": 1.1594187413185169,
      "grad_norm": 2.322314977645874,
      "learning_rate": 8.419251336898397e-06,
      "loss": 0.07986120879650116,
      "num_input_tokens_seen": 2796216,
      "step": 2713,
      "train_runtime": 6561.6983,
      "train_tokens_per_second": 426.142
    },
    {
      "epoch": 1.1598461374078428,
      "grad_norm": 2.479872465133667,
      "learning_rate": 8.414973262032086e-06,
      "loss": 0.06391402333974838,
      "num_input_tokens_seen": 2797098,
      "step": 2714,
      "train_runtime": 6563.7934,
      "train_tokens_per_second": 426.14
    },
    {
      "epoch": 1.1602735334971686,
      "grad_norm": 3.55705189704895,
      "learning_rate": 8.410695187165776e-06,
      "loss": 0.11364055424928665,
      "num_input_tokens_seen": 2797876,
      "step": 2715,
      "train_runtime": 6565.8271,
      "train_tokens_per_second": 426.127
    },
    {
      "epoch": 1.1607009295864943,
      "grad_norm": 2.6096415519714355,
      "learning_rate": 8.406417112299467e-06,
      "loss": 0.13401907682418823,
      "num_input_tokens_seen": 2798910,
      "step": 2716,
      "train_runtime": 6567.9125,
      "train_tokens_per_second": 426.149
    },
    {
      "epoch": 1.16112832567582,
      "grad_norm": 2.2043533325195312,
      "learning_rate": 8.402139037433156e-06,
      "loss": 0.09973882883787155,
      "num_input_tokens_seen": 2800148,
      "step": 2717,
      "train_runtime": 6570.0633,
      "train_tokens_per_second": 426.198
    },
    {
      "epoch": 1.1615557217651458,
      "grad_norm": 2.470081090927124,
      "learning_rate": 8.397860962566846e-06,
      "loss": 0.06988176703453064,
      "num_input_tokens_seen": 2800906,
      "step": 2718,
      "train_runtime": 6572.0754,
      "train_tokens_per_second": 426.183
    },
    {
      "epoch": 1.1619831178544717,
      "grad_norm": 3.067796230316162,
      "learning_rate": 8.393582887700535e-06,
      "loss": 0.0811307430267334,
      "num_input_tokens_seen": 2801718,
      "step": 2719,
      "train_runtime": 6574.0925,
      "train_tokens_per_second": 426.176
    },
    {
      "epoch": 1.1624105139437975,
      "grad_norm": 3.2505218982696533,
      "learning_rate": 8.389304812834226e-06,
      "loss": 0.08843886107206345,
      "num_input_tokens_seen": 2802420,
      "step": 2720,
      "train_runtime": 6576.1063,
      "train_tokens_per_second": 426.152
    },
    {
      "epoch": 1.1628379100331232,
      "grad_norm": 3.2225327491760254,
      "learning_rate": 8.385026737967916e-06,
      "loss": 0.06277766078710556,
      "num_input_tokens_seen": 2803038,
      "step": 2721,
      "train_runtime": 6578.0854,
      "train_tokens_per_second": 426.118
    },
    {
      "epoch": 1.163265306122449,
      "grad_norm": 2.3188552856445312,
      "learning_rate": 8.380748663101605e-06,
      "loss": 0.0782141387462616,
      "num_input_tokens_seen": 2804258,
      "step": 2722,
      "train_runtime": 6580.2041,
      "train_tokens_per_second": 426.166
    },
    {
      "epoch": 1.1636927022117747,
      "grad_norm": 3.044053316116333,
      "learning_rate": 8.376470588235295e-06,
      "loss": 0.07955057919025421,
      "num_input_tokens_seen": 2805078,
      "step": 2723,
      "train_runtime": 6582.2212,
      "train_tokens_per_second": 426.16
    },
    {
      "epoch": 1.1641200983011006,
      "grad_norm": 2.7747609615325928,
      "learning_rate": 8.372192513368986e-06,
      "loss": 0.09291289746761322,
      "num_input_tokens_seen": 2806074,
      "step": 2724,
      "train_runtime": 6584.3475,
      "train_tokens_per_second": 426.173
    },
    {
      "epoch": 1.1645474943904264,
      "grad_norm": 2.7635974884033203,
      "learning_rate": 8.367914438502675e-06,
      "loss": 0.11294674873352051,
      "num_input_tokens_seen": 2807284,
      "step": 2725,
      "train_runtime": 6586.4746,
      "train_tokens_per_second": 426.22
    },
    {
      "epoch": 1.1649748904797521,
      "grad_norm": 2.405041217803955,
      "learning_rate": 8.363636363636365e-06,
      "loss": 0.10040175169706345,
      "num_input_tokens_seen": 2808648,
      "step": 2726,
      "train_runtime": 6588.6336,
      "train_tokens_per_second": 426.287
    },
    {
      "epoch": 1.1654022865690778,
      "grad_norm": 2.0028679370880127,
      "learning_rate": 8.359358288770054e-06,
      "loss": 0.06233786419034004,
      "num_input_tokens_seen": 2809474,
      "step": 2727,
      "train_runtime": 6590.7363,
      "train_tokens_per_second": 426.276
    },
    {
      "epoch": 1.1658296826584036,
      "grad_norm": 2.261458396911621,
      "learning_rate": 8.355080213903743e-06,
      "loss": 0.07772679626941681,
      "num_input_tokens_seen": 2810370,
      "step": 2728,
      "train_runtime": 6592.7501,
      "train_tokens_per_second": 426.282
    },
    {
      "epoch": 1.1662570787477295,
      "grad_norm": 2.2100696563720703,
      "learning_rate": 8.350802139037435e-06,
      "loss": 0.058071255683898926,
      "num_input_tokens_seen": 2810972,
      "step": 2729,
      "train_runtime": 6594.7267,
      "train_tokens_per_second": 426.245
    },
    {
      "epoch": 1.1666844748370553,
      "grad_norm": 3.5097203254699707,
      "learning_rate": 8.346524064171122e-06,
      "loss": 0.1302880048751831,
      "num_input_tokens_seen": 2811980,
      "step": 2730,
      "train_runtime": 6596.8005,
      "train_tokens_per_second": 426.264
    },
    {
      "epoch": 1.167111870926381,
      "grad_norm": 2.097170352935791,
      "learning_rate": 8.342245989304813e-06,
      "loss": 0.07355550676584244,
      "num_input_tokens_seen": 2813264,
      "step": 2731,
      "train_runtime": 6605.1943,
      "train_tokens_per_second": 425.917
    },
    {
      "epoch": 1.1675392670157068,
      "grad_norm": 2.2746753692626953,
      "learning_rate": 8.337967914438503e-06,
      "loss": 0.12391360849142075,
      "num_input_tokens_seen": 2814760,
      "step": 2732,
      "train_runtime": 6607.3828,
      "train_tokens_per_second": 426.002
    },
    {
      "epoch": 1.1679666631050325,
      "grad_norm": 2.543454885482788,
      "learning_rate": 8.333689839572194e-06,
      "loss": 0.11102259904146194,
      "num_input_tokens_seen": 2815868,
      "step": 2733,
      "train_runtime": 6609.5243,
      "train_tokens_per_second": 426.032
    },
    {
      "epoch": 1.1683940591943585,
      "grad_norm": 1.7932102680206299,
      "learning_rate": 8.329411764705882e-06,
      "loss": 0.05068099498748779,
      "num_input_tokens_seen": 2816890,
      "step": 2734,
      "train_runtime": 6611.6147,
      "train_tokens_per_second": 426.052
    },
    {
      "epoch": 1.1688214552836842,
      "grad_norm": 2.764402151107788,
      "learning_rate": 8.325133689839573e-06,
      "loss": 0.10678049921989441,
      "num_input_tokens_seen": 2818174,
      "step": 2735,
      "train_runtime": 6613.8046,
      "train_tokens_per_second": 426.105
    },
    {
      "epoch": 1.16924885137301,
      "grad_norm": 2.8847622871398926,
      "learning_rate": 8.320855614973262e-06,
      "loss": 0.08096059411764145,
      "num_input_tokens_seen": 2819028,
      "step": 2736,
      "train_runtime": 6615.8217,
      "train_tokens_per_second": 426.104
    },
    {
      "epoch": 1.1696762474623357,
      "grad_norm": 3.2103450298309326,
      "learning_rate": 8.316577540106952e-06,
      "loss": 0.07069586962461472,
      "num_input_tokens_seen": 2819830,
      "step": 2737,
      "train_runtime": 6617.8693,
      "train_tokens_per_second": 426.093
    },
    {
      "epoch": 1.1701036435516614,
      "grad_norm": 3.4605236053466797,
      "learning_rate": 8.312299465240641e-06,
      "loss": 0.061676450073719025,
      "num_input_tokens_seen": 2820718,
      "step": 2738,
      "train_runtime": 6619.8791,
      "train_tokens_per_second": 426.098
    },
    {
      "epoch": 1.1705310396409874,
      "grad_norm": 2.972053289413452,
      "learning_rate": 8.308021390374332e-06,
      "loss": 0.07029437273740768,
      "num_input_tokens_seen": 2821584,
      "step": 2739,
      "train_runtime": 6621.9286,
      "train_tokens_per_second": 426.097
    },
    {
      "epoch": 1.170958435730313,
      "grad_norm": 1.8746734857559204,
      "learning_rate": 8.303743315508022e-06,
      "loss": 0.08134539425373077,
      "num_input_tokens_seen": 2822630,
      "step": 2740,
      "train_runtime": 6624.0504,
      "train_tokens_per_second": 426.118
    },
    {
      "epoch": 1.1713858318196388,
      "grad_norm": 2.3372225761413574,
      "learning_rate": 8.299465240641711e-06,
      "loss": 0.06283244490623474,
      "num_input_tokens_seen": 2823352,
      "step": 2741,
      "train_runtime": 6626.0721,
      "train_tokens_per_second": 426.097
    },
    {
      "epoch": 1.1718132279089646,
      "grad_norm": 2.5646891593933105,
      "learning_rate": 8.2951871657754e-06,
      "loss": 0.09890198707580566,
      "num_input_tokens_seen": 2824122,
      "step": 2742,
      "train_runtime": 6628.1503,
      "train_tokens_per_second": 426.08
    },
    {
      "epoch": 1.1722406239982903,
      "grad_norm": 2.6180989742279053,
      "learning_rate": 8.290909090909092e-06,
      "loss": 0.08496618270874023,
      "num_input_tokens_seen": 2825188,
      "step": 2743,
      "train_runtime": 6630.2503,
      "train_tokens_per_second": 426.106
    },
    {
      "epoch": 1.1726680200876163,
      "grad_norm": 2.235213279724121,
      "learning_rate": 8.286631016042781e-06,
      "loss": 0.07360003888607025,
      "num_input_tokens_seen": 2826222,
      "step": 2744,
      "train_runtime": 6632.357,
      "train_tokens_per_second": 426.126
    },
    {
      "epoch": 1.173095416176942,
      "grad_norm": 3.605175018310547,
      "learning_rate": 8.28235294117647e-06,
      "loss": 0.19582922756671906,
      "num_input_tokens_seen": 2827374,
      "step": 2745,
      "train_runtime": 6634.5316,
      "train_tokens_per_second": 426.16
    },
    {
      "epoch": 1.1735228122662678,
      "grad_norm": 1.8830177783966064,
      "learning_rate": 8.27807486631016e-06,
      "loss": 0.06456751376390457,
      "num_input_tokens_seen": 2828634,
      "step": 2746,
      "train_runtime": 6636.6692,
      "train_tokens_per_second": 426.213
    },
    {
      "epoch": 1.1739502083555935,
      "grad_norm": 2.407860040664673,
      "learning_rate": 8.273796791443851e-06,
      "loss": 0.07727548480033875,
      "num_input_tokens_seen": 2829646,
      "step": 2747,
      "train_runtime": 6638.745,
      "train_tokens_per_second": 426.232
    },
    {
      "epoch": 1.1743776044449192,
      "grad_norm": 2.4663100242614746,
      "learning_rate": 8.26951871657754e-06,
      "loss": 0.06858322024345398,
      "num_input_tokens_seen": 2830524,
      "step": 2748,
      "train_runtime": 6640.7949,
      "train_tokens_per_second": 426.233
    },
    {
      "epoch": 1.1748050005342452,
      "grad_norm": 2.6654489040374756,
      "learning_rate": 8.26524064171123e-06,
      "loss": 0.0849633663892746,
      "num_input_tokens_seen": 2831626,
      "step": 2749,
      "train_runtime": 6642.9837,
      "train_tokens_per_second": 426.258
    },
    {
      "epoch": 1.175232396623571,
      "grad_norm": 2.5251452922821045,
      "learning_rate": 8.26096256684492e-06,
      "loss": 0.07822896540164948,
      "num_input_tokens_seen": 2832630,
      "step": 2750,
      "train_runtime": 6645.1044,
      "train_tokens_per_second": 426.273
    },
    {
      "epoch": 1.1756597927128967,
      "grad_norm": 2.760608673095703,
      "learning_rate": 8.256684491978611e-06,
      "loss": 0.07572463154792786,
      "num_input_tokens_seen": 2833528,
      "step": 2751,
      "train_runtime": 6647.2435,
      "train_tokens_per_second": 426.271
    },
    {
      "epoch": 1.1760871888022224,
      "grad_norm": 4.295748710632324,
      "learning_rate": 8.2524064171123e-06,
      "loss": 0.19411985576152802,
      "num_input_tokens_seen": 2834838,
      "step": 2752,
      "train_runtime": 6649.4147,
      "train_tokens_per_second": 426.329
    },
    {
      "epoch": 1.1765145848915481,
      "grad_norm": 4.636139392852783,
      "learning_rate": 8.24812834224599e-06,
      "loss": 0.10699424892663956,
      "num_input_tokens_seen": 2835364,
      "step": 2753,
      "train_runtime": 6651.3976,
      "train_tokens_per_second": 426.281
    },
    {
      "epoch": 1.176941980980874,
      "grad_norm": 4.5621256828308105,
      "learning_rate": 8.24385026737968e-06,
      "loss": 0.07216803729534149,
      "num_input_tokens_seen": 2836190,
      "step": 2754,
      "train_runtime": 6653.4387,
      "train_tokens_per_second": 426.274
    },
    {
      "epoch": 1.1773693770701998,
      "grad_norm": 3.5223195552825928,
      "learning_rate": 8.23957219251337e-06,
      "loss": 0.11959582567214966,
      "num_input_tokens_seen": 2837530,
      "step": 2755,
      "train_runtime": 6655.6158,
      "train_tokens_per_second": 426.336
    },
    {
      "epoch": 1.1777967731595256,
      "grad_norm": 1.3998544216156006,
      "learning_rate": 8.23529411764706e-06,
      "loss": 0.0321936309337616,
      "num_input_tokens_seen": 2838744,
      "step": 2756,
      "train_runtime": 6657.7584,
      "train_tokens_per_second": 426.381
    },
    {
      "epoch": 1.1782241692488513,
      "grad_norm": 2.2441916465759277,
      "learning_rate": 8.23101604278075e-06,
      "loss": 0.07243187725543976,
      "num_input_tokens_seen": 2840084,
      "step": 2757,
      "train_runtime": 6659.9321,
      "train_tokens_per_second": 426.443
    },
    {
      "epoch": 1.178651565338177,
      "grad_norm": 2.9381752014160156,
      "learning_rate": 8.226737967914439e-06,
      "loss": 0.0765104591846466,
      "num_input_tokens_seen": 2840836,
      "step": 2758,
      "train_runtime": 6661.9433,
      "train_tokens_per_second": 426.428
    },
    {
      "epoch": 1.179078961427503,
      "grad_norm": 3.729572296142578,
      "learning_rate": 8.22245989304813e-06,
      "loss": 0.11705591529607773,
      "num_input_tokens_seen": 2841568,
      "step": 2759,
      "train_runtime": 6663.9319,
      "train_tokens_per_second": 426.41
    },
    {
      "epoch": 1.1795063575168288,
      "grad_norm": 2.775517702102661,
      "learning_rate": 8.21818181818182e-06,
      "loss": 0.10254961997270584,
      "num_input_tokens_seen": 2842698,
      "step": 2760,
      "train_runtime": 6666.0617,
      "train_tokens_per_second": 426.443
    },
    {
      "epoch": 1.1799337536061545,
      "grad_norm": 2.1451797485351562,
      "learning_rate": 8.213903743315509e-06,
      "loss": 0.048349082469940186,
      "num_input_tokens_seen": 2843516,
      "step": 2761,
      "train_runtime": 6674.7708,
      "train_tokens_per_second": 426.01
    },
    {
      "epoch": 1.1803611496954802,
      "grad_norm": 2.3256468772888184,
      "learning_rate": 8.209625668449198e-06,
      "loss": 0.06722051650285721,
      "num_input_tokens_seen": 2844970,
      "step": 2762,
      "train_runtime": 6677.0156,
      "train_tokens_per_second": 426.084
    },
    {
      "epoch": 1.180788545784806,
      "grad_norm": 2.4873950481414795,
      "learning_rate": 8.20534759358289e-06,
      "loss": 0.09393316507339478,
      "num_input_tokens_seen": 2846506,
      "step": 2763,
      "train_runtime": 6679.5456,
      "train_tokens_per_second": 426.153
    },
    {
      "epoch": 1.181215941874132,
      "grad_norm": 1.5685741901397705,
      "learning_rate": 8.201069518716577e-06,
      "loss": 0.04840708151459694,
      "num_input_tokens_seen": 2847884,
      "step": 2764,
      "train_runtime": 6681.6985,
      "train_tokens_per_second": 426.222
    },
    {
      "epoch": 1.1816433379634577,
      "grad_norm": 1.8305567502975464,
      "learning_rate": 8.196791443850268e-06,
      "loss": 0.046777788549661636,
      "num_input_tokens_seen": 2848890,
      "step": 2765,
      "train_runtime": 6683.8516,
      "train_tokens_per_second": 426.235
    },
    {
      "epoch": 1.1820707340527834,
      "grad_norm": 4.451944828033447,
      "learning_rate": 8.192513368983958e-06,
      "loss": 0.07781849801540375,
      "num_input_tokens_seen": 2850104,
      "step": 2766,
      "train_runtime": 6685.9593,
      "train_tokens_per_second": 426.282
    },
    {
      "epoch": 1.1824981301421091,
      "grad_norm": 3.108126401901245,
      "learning_rate": 8.188235294117649e-06,
      "loss": 0.09696676582098007,
      "num_input_tokens_seen": 2850980,
      "step": 2767,
      "train_runtime": 6688.04,
      "train_tokens_per_second": 426.28
    },
    {
      "epoch": 1.1829255262314349,
      "grad_norm": 3.0373806953430176,
      "learning_rate": 8.183957219251337e-06,
      "loss": 0.10651018470525742,
      "num_input_tokens_seen": 2851792,
      "step": 2768,
      "train_runtime": 6690.1046,
      "train_tokens_per_second": 426.27
    },
    {
      "epoch": 1.1833529223207608,
      "grad_norm": 3.0271189212799072,
      "learning_rate": 8.179679144385028e-06,
      "loss": 0.09459219872951508,
      "num_input_tokens_seen": 2852826,
      "step": 2769,
      "train_runtime": 6692.1904,
      "train_tokens_per_second": 426.292
    },
    {
      "epoch": 1.1837803184100866,
      "grad_norm": 2.09867262840271,
      "learning_rate": 8.175401069518717e-06,
      "loss": 0.0506097674369812,
      "num_input_tokens_seen": 2853612,
      "step": 2770,
      "train_runtime": 6694.2298,
      "train_tokens_per_second": 426.279
    },
    {
      "epoch": 1.1842077144994123,
      "grad_norm": 2.456977128982544,
      "learning_rate": 8.171122994652407e-06,
      "loss": 0.07622379064559937,
      "num_input_tokens_seen": 2855204,
      "step": 2771,
      "train_runtime": 6696.4379,
      "train_tokens_per_second": 426.377
    },
    {
      "epoch": 1.184635110588738,
      "grad_norm": 3.1064202785491943,
      "learning_rate": 8.166844919786096e-06,
      "loss": 0.0980147272348404,
      "num_input_tokens_seen": 2855986,
      "step": 2772,
      "train_runtime": 6698.5034,
      "train_tokens_per_second": 426.362
    },
    {
      "epoch": 1.1850625066780638,
      "grad_norm": 2.19614315032959,
      "learning_rate": 8.162566844919787e-06,
      "loss": 0.08231724798679352,
      "num_input_tokens_seen": 2857284,
      "step": 2773,
      "train_runtime": 6700.6889,
      "train_tokens_per_second": 426.416
    },
    {
      "epoch": 1.1854899027673897,
      "grad_norm": 2.7485785484313965,
      "learning_rate": 8.158288770053477e-06,
      "loss": 0.1170513853430748,
      "num_input_tokens_seen": 2858210,
      "step": 2774,
      "train_runtime": 6702.7348,
      "train_tokens_per_second": 426.424
    },
    {
      "epoch": 1.1859172988567155,
      "grad_norm": 2.576573371887207,
      "learning_rate": 8.154010695187166e-06,
      "loss": 0.06007217988371849,
      "num_input_tokens_seen": 2859000,
      "step": 2775,
      "train_runtime": 6704.7635,
      "train_tokens_per_second": 426.413
    },
    {
      "epoch": 1.1863446949460412,
      "grad_norm": 1.3909211158752441,
      "learning_rate": 8.149732620320855e-06,
      "loss": 0.060951970517635345,
      "num_input_tokens_seen": 2860828,
      "step": 2776,
      "train_runtime": 6707.0578,
      "train_tokens_per_second": 426.54
    },
    {
      "epoch": 1.186772091035367,
      "grad_norm": 1.9755957126617432,
      "learning_rate": 8.145454545454547e-06,
      "loss": 0.0752592608332634,
      "num_input_tokens_seen": 2862266,
      "step": 2777,
      "train_runtime": 6709.1942,
      "train_tokens_per_second": 426.618
    },
    {
      "epoch": 1.1871994871246927,
      "grad_norm": 2.978173017501831,
      "learning_rate": 8.141176470588236e-06,
      "loss": 0.08630026876926422,
      "num_input_tokens_seen": 2863318,
      "step": 2778,
      "train_runtime": 6711.2836,
      "train_tokens_per_second": 426.642
    },
    {
      "epoch": 1.1876268832140187,
      "grad_norm": 2.4868850708007812,
      "learning_rate": 8.136898395721926e-06,
      "loss": 0.07803753018379211,
      "num_input_tokens_seen": 2864534,
      "step": 2779,
      "train_runtime": 6713.3894,
      "train_tokens_per_second": 426.69
    },
    {
      "epoch": 1.1880542793033444,
      "grad_norm": 3.0123798847198486,
      "learning_rate": 8.132620320855615e-06,
      "loss": 0.09397991001605988,
      "num_input_tokens_seen": 2865628,
      "step": 2780,
      "train_runtime": 6715.5591,
      "train_tokens_per_second": 426.715
    },
    {
      "epoch": 1.1884816753926701,
      "grad_norm": 3.6068315505981445,
      "learning_rate": 8.128342245989306e-06,
      "loss": 0.08335354924201965,
      "num_input_tokens_seen": 2866434,
      "step": 2781,
      "train_runtime": 6717.62,
      "train_tokens_per_second": 426.704
    },
    {
      "epoch": 1.1889090714819959,
      "grad_norm": 2.3114826679229736,
      "learning_rate": 8.124064171122996e-06,
      "loss": 0.08412733674049377,
      "num_input_tokens_seen": 2867234,
      "step": 2782,
      "train_runtime": 6719.6404,
      "train_tokens_per_second": 426.695
    },
    {
      "epoch": 1.1893364675713216,
      "grad_norm": 1.8738144636154175,
      "learning_rate": 8.119786096256685e-06,
      "loss": 0.05668673291802406,
      "num_input_tokens_seen": 2868502,
      "step": 2783,
      "train_runtime": 6721.806,
      "train_tokens_per_second": 426.746
    },
    {
      "epoch": 1.1897638636606476,
      "grad_norm": 2.184523344039917,
      "learning_rate": 8.115508021390374e-06,
      "loss": 0.05665953829884529,
      "num_input_tokens_seen": 2869556,
      "step": 2784,
      "train_runtime": 6723.9139,
      "train_tokens_per_second": 426.769
    },
    {
      "epoch": 1.1901912597499733,
      "grad_norm": 11.423437118530273,
      "learning_rate": 8.111229946524066e-06,
      "loss": 0.15224233269691467,
      "num_input_tokens_seen": 2870234,
      "step": 2785,
      "train_runtime": 6725.9318,
      "train_tokens_per_second": 426.741
    },
    {
      "epoch": 1.190618655839299,
      "grad_norm": 4.016899585723877,
      "learning_rate": 8.106951871657755e-06,
      "loss": 0.12967631220817566,
      "num_input_tokens_seen": 2871520,
      "step": 2786,
      "train_runtime": 6728.0859,
      "train_tokens_per_second": 426.796
    },
    {
      "epoch": 1.1910460519286248,
      "grad_norm": 4.365553855895996,
      "learning_rate": 8.102673796791444e-06,
      "loss": 0.07771287113428116,
      "num_input_tokens_seen": 2872640,
      "step": 2787,
      "train_runtime": 6730.183,
      "train_tokens_per_second": 426.829
    },
    {
      "epoch": 1.1914734480179505,
      "grad_norm": 3.417193651199341,
      "learning_rate": 8.098395721925134e-06,
      "loss": 0.1463349163532257,
      "num_input_tokens_seen": 2873480,
      "step": 2788,
      "train_runtime": 6732.2374,
      "train_tokens_per_second": 426.824
    },
    {
      "epoch": 1.1919008441072765,
      "grad_norm": 2.8998706340789795,
      "learning_rate": 8.094117647058825e-06,
      "loss": 0.07339631766080856,
      "num_input_tokens_seen": 2874280,
      "step": 2789,
      "train_runtime": 6734.331,
      "train_tokens_per_second": 426.81
    },
    {
      "epoch": 1.1923282401966022,
      "grad_norm": 2.5318920612335205,
      "learning_rate": 8.089839572192515e-06,
      "loss": 0.08839103579521179,
      "num_input_tokens_seen": 2875268,
      "step": 2790,
      "train_runtime": 6736.4202,
      "train_tokens_per_second": 426.824
    },
    {
      "epoch": 1.192755636285928,
      "grad_norm": 2.553860902786255,
      "learning_rate": 8.085561497326204e-06,
      "loss": 0.058411236852407455,
      "num_input_tokens_seen": 2876370,
      "step": 2791,
      "train_runtime": 6745.2,
      "train_tokens_per_second": 426.432
    },
    {
      "epoch": 1.1931830323752537,
      "grad_norm": 3.8284435272216797,
      "learning_rate": 8.081283422459893e-06,
      "loss": 0.07421354204416275,
      "num_input_tokens_seen": 2877096,
      "step": 2792,
      "train_runtime": 6747.3774,
      "train_tokens_per_second": 426.402
    },
    {
      "epoch": 1.1936104284645794,
      "grad_norm": 2.812725305557251,
      "learning_rate": 8.077005347593585e-06,
      "loss": 0.0479869470000267,
      "num_input_tokens_seen": 2877996,
      "step": 2793,
      "train_runtime": 6749.4223,
      "train_tokens_per_second": 426.406
    },
    {
      "epoch": 1.1940378245539054,
      "grad_norm": 2.277679443359375,
      "learning_rate": 8.072727272727274e-06,
      "loss": 0.11525529623031616,
      "num_input_tokens_seen": 2880212,
      "step": 2794,
      "train_runtime": 6751.7729,
      "train_tokens_per_second": 426.586
    },
    {
      "epoch": 1.1944652206432311,
      "grad_norm": 2.6919076442718506,
      "learning_rate": 8.068449197860963e-06,
      "loss": 0.08185943961143494,
      "num_input_tokens_seen": 2881484,
      "step": 2795,
      "train_runtime": 6753.9589,
      "train_tokens_per_second": 426.636
    },
    {
      "epoch": 1.1948926167325569,
      "grad_norm": 2.5328187942504883,
      "learning_rate": 8.064171122994653e-06,
      "loss": 0.060951486229896545,
      "num_input_tokens_seen": 2882278,
      "step": 2796,
      "train_runtime": 6755.9624,
      "train_tokens_per_second": 426.627
    },
    {
      "epoch": 1.1953200128218826,
      "grad_norm": 2.5691261291503906,
      "learning_rate": 8.059893048128344e-06,
      "loss": 0.06906014680862427,
      "num_input_tokens_seen": 2883062,
      "step": 2797,
      "train_runtime": 6757.9392,
      "train_tokens_per_second": 426.619
    },
    {
      "epoch": 1.1957474089112083,
      "grad_norm": 2.6108224391937256,
      "learning_rate": 8.055614973262032e-06,
      "loss": 0.09330528974533081,
      "num_input_tokens_seen": 2883984,
      "step": 2798,
      "train_runtime": 6759.9875,
      "train_tokens_per_second": 426.626
    },
    {
      "epoch": 1.1961748050005343,
      "grad_norm": 1.9042527675628662,
      "learning_rate": 8.051336898395723e-06,
      "loss": 0.08516990393400192,
      "num_input_tokens_seen": 2885184,
      "step": 2799,
      "train_runtime": 6762.1257,
      "train_tokens_per_second": 426.668
    },
    {
      "epoch": 1.19660220108986,
      "grad_norm": 2.8938372135162354,
      "learning_rate": 8.047058823529412e-06,
      "loss": 0.08211921155452728,
      "num_input_tokens_seen": 2885868,
      "step": 2800,
      "train_runtime": 6764.1456,
      "train_tokens_per_second": 426.642
    },
    {
      "epoch": 1.1970295971791858,
      "grad_norm": 3.3418660163879395,
      "learning_rate": 8.042780748663103e-06,
      "loss": 0.1417672038078308,
      "num_input_tokens_seen": 2886680,
      "step": 2801,
      "train_runtime": 6766.1981,
      "train_tokens_per_second": 426.632
    },
    {
      "epoch": 1.1974569932685115,
      "grad_norm": 2.593707799911499,
      "learning_rate": 8.038502673796791e-06,
      "loss": 0.09683042019605637,
      "num_input_tokens_seen": 2887896,
      "step": 2802,
      "train_runtime": 6768.3142,
      "train_tokens_per_second": 426.679
    },
    {
      "epoch": 1.1978843893578373,
      "grad_norm": 3.104003667831421,
      "learning_rate": 8.034224598930482e-06,
      "loss": 0.07965659350156784,
      "num_input_tokens_seen": 2888794,
      "step": 2803,
      "train_runtime": 6770.3633,
      "train_tokens_per_second": 426.682
    },
    {
      "epoch": 1.1983117854471632,
      "grad_norm": 2.3621914386749268,
      "learning_rate": 8.029946524064172e-06,
      "loss": 0.08068880438804626,
      "num_input_tokens_seen": 2889590,
      "step": 2804,
      "train_runtime": 6772.3857,
      "train_tokens_per_second": 426.672
    },
    {
      "epoch": 1.198739181536489,
      "grad_norm": 2.863574504852295,
      "learning_rate": 8.025668449197861e-06,
      "loss": 0.08370157331228256,
      "num_input_tokens_seen": 2890482,
      "step": 2805,
      "train_runtime": 6774.5894,
      "train_tokens_per_second": 426.665
    },
    {
      "epoch": 1.1991665776258147,
      "grad_norm": 2.817122459411621,
      "learning_rate": 8.02139037433155e-06,
      "loss": 0.0956934466958046,
      "num_input_tokens_seen": 2891630,
      "step": 2806,
      "train_runtime": 6776.7469,
      "train_tokens_per_second": 426.699
    },
    {
      "epoch": 1.1995939737151404,
      "grad_norm": 3.1060221195220947,
      "learning_rate": 8.017112299465242e-06,
      "loss": 0.08639883249998093,
      "num_input_tokens_seen": 2892254,
      "step": 2807,
      "train_runtime": 6778.7606,
      "train_tokens_per_second": 426.664
    },
    {
      "epoch": 1.2000213698044662,
      "grad_norm": 2.6774814128875732,
      "learning_rate": 8.012834224598931e-06,
      "loss": 0.08912783861160278,
      "num_input_tokens_seen": 2893330,
      "step": 2808,
      "train_runtime": 6780.8837,
      "train_tokens_per_second": 426.689
    },
    {
      "epoch": 1.2000213698044662,
      "eval_loss": 0.5259655117988586,
      "eval_runtime": 57.9652,
      "eval_samples_per_second": 17.217,
      "eval_steps_per_second": 8.609,
      "num_input_tokens_seen": 2893330,
      "step": 2808
    },
    {
      "epoch": 1.2004487658937921,
      "grad_norm": 2.932251214981079,
      "learning_rate": 8.00855614973262e-06,
      "loss": 0.09680595993995667,
      "num_input_tokens_seen": 2894236,
      "step": 2809,
      "train_runtime": 6840.9917,
      "train_tokens_per_second": 423.073
    },
    {
      "epoch": 1.2008761619831179,
      "grad_norm": 2.613478422164917,
      "learning_rate": 8.00427807486631e-06,
      "loss": 0.08638087660074234,
      "num_input_tokens_seen": 2894932,
      "step": 2810,
      "train_runtime": 6843.0378,
      "train_tokens_per_second": 423.048
    },
    {
      "epoch": 1.2013035580724436,
      "grad_norm": 2.7075607776641846,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.07337787002325058,
      "num_input_tokens_seen": 2895814,
      "step": 2811,
      "train_runtime": 6845.0763,
      "train_tokens_per_second": 423.051
    },
    {
      "epoch": 1.2017309541617693,
      "grad_norm": 2.2449817657470703,
      "learning_rate": 7.99572192513369e-06,
      "loss": 0.04989093169569969,
      "num_input_tokens_seen": 2896538,
      "step": 2812,
      "train_runtime": 6847.1325,
      "train_tokens_per_second": 423.029
    },
    {
      "epoch": 1.2021583502510953,
      "grad_norm": 2.069969654083252,
      "learning_rate": 7.99144385026738e-06,
      "loss": 0.0585918165743351,
      "num_input_tokens_seen": 2897914,
      "step": 2813,
      "train_runtime": 6849.3239,
      "train_tokens_per_second": 423.095
    },
    {
      "epoch": 1.202585746340421,
      "grad_norm": 1.98743736743927,
      "learning_rate": 7.98716577540107e-06,
      "loss": 0.07523835450410843,
      "num_input_tokens_seen": 2899346,
      "step": 2814,
      "train_runtime": 6851.576,
      "train_tokens_per_second": 423.165
    },
    {
      "epoch": 1.2030131424297468,
      "grad_norm": 1.943057656288147,
      "learning_rate": 7.98288770053476e-06,
      "loss": 0.08707548677921295,
      "num_input_tokens_seen": 2900616,
      "step": 2815,
      "train_runtime": 6853.7493,
      "train_tokens_per_second": 423.216
    },
    {
      "epoch": 1.2034405385190725,
      "grad_norm": 2.1865179538726807,
      "learning_rate": 7.97860962566845e-06,
      "loss": 0.06938644498586655,
      "num_input_tokens_seen": 2901618,
      "step": 2816,
      "train_runtime": 6855.8715,
      "train_tokens_per_second": 423.231
    },
    {
      "epoch": 1.2038679346083982,
      "grad_norm": 2.6555259227752686,
      "learning_rate": 7.97433155080214e-06,
      "loss": 0.12925024330615997,
      "num_input_tokens_seen": 2902946,
      "step": 2817,
      "train_runtime": 6858.0646,
      "train_tokens_per_second": 423.289
    },
    {
      "epoch": 1.2042953306977242,
      "grad_norm": 3.136338472366333,
      "learning_rate": 7.970053475935829e-06,
      "loss": 0.11719761788845062,
      "num_input_tokens_seen": 2903788,
      "step": 2818,
      "train_runtime": 6860.1263,
      "train_tokens_per_second": 423.285
    },
    {
      "epoch": 1.20472272678705,
      "grad_norm": 3.271803379058838,
      "learning_rate": 7.96577540106952e-06,
      "loss": 0.12141872942447662,
      "num_input_tokens_seen": 2904648,
      "step": 2819,
      "train_runtime": 6862.2059,
      "train_tokens_per_second": 423.282
    },
    {
      "epoch": 1.2051501228763757,
      "grad_norm": 2.9585413932800293,
      "learning_rate": 7.96149732620321e-06,
      "loss": 0.08449549973011017,
      "num_input_tokens_seen": 2905460,
      "step": 2820,
      "train_runtime": 6864.2939,
      "train_tokens_per_second": 423.272
    },
    {
      "epoch": 1.2055775189657014,
      "grad_norm": 1.9990448951721191,
      "learning_rate": 7.9572192513369e-06,
      "loss": 0.03481731936335564,
      "num_input_tokens_seen": 2906738,
      "step": 2821,
      "train_runtime": 6873.0478,
      "train_tokens_per_second": 422.918
    },
    {
      "epoch": 1.2060049150550272,
      "grad_norm": 4.325561046600342,
      "learning_rate": 7.952941176470589e-06,
      "loss": 0.14037759602069855,
      "num_input_tokens_seen": 2907406,
      "step": 2822,
      "train_runtime": 6875.1591,
      "train_tokens_per_second": 422.886
    },
    {
      "epoch": 1.2064323111443531,
      "grad_norm": 3.248270034790039,
      "learning_rate": 7.94866310160428e-06,
      "loss": 0.1077517569065094,
      "num_input_tokens_seen": 2908332,
      "step": 2823,
      "train_runtime": 6877.2179,
      "train_tokens_per_second": 422.894
    },
    {
      "epoch": 1.2068597072336789,
      "grad_norm": 2.6697635650634766,
      "learning_rate": 7.94438502673797e-06,
      "loss": 0.070683054625988,
      "num_input_tokens_seen": 2909240,
      "step": 2824,
      "train_runtime": 6879.3102,
      "train_tokens_per_second": 422.897
    },
    {
      "epoch": 1.2072871033230046,
      "grad_norm": 6.151632785797119,
      "learning_rate": 7.940106951871659e-06,
      "loss": 0.10860487073659897,
      "num_input_tokens_seen": 2910630,
      "step": 2825,
      "train_runtime": 6881.5246,
      "train_tokens_per_second": 422.963
    },
    {
      "epoch": 1.2077144994123303,
      "grad_norm": 3.145603656768799,
      "learning_rate": 7.935828877005348e-06,
      "loss": 0.10445728152990341,
      "num_input_tokens_seen": 2911612,
      "step": 2826,
      "train_runtime": 6883.5792,
      "train_tokens_per_second": 422.979
    },
    {
      "epoch": 1.208141895501656,
      "grad_norm": 2.151937961578369,
      "learning_rate": 7.93155080213904e-06,
      "loss": 0.07243939489126205,
      "num_input_tokens_seen": 2912474,
      "step": 2827,
      "train_runtime": 6885.6675,
      "train_tokens_per_second": 422.976
    },
    {
      "epoch": 1.208569291590982,
      "grad_norm": 3.458822727203369,
      "learning_rate": 7.927272727272729e-06,
      "loss": 0.08526347577571869,
      "num_input_tokens_seen": 2913060,
      "step": 2828,
      "train_runtime": 6887.6348,
      "train_tokens_per_second": 422.941
    },
    {
      "epoch": 1.2089966876803078,
      "grad_norm": 1.8537195920944214,
      "learning_rate": 7.922994652406418e-06,
      "loss": 0.06046935170888901,
      "num_input_tokens_seen": 2914260,
      "step": 2829,
      "train_runtime": 6889.8436,
      "train_tokens_per_second": 422.979
    },
    {
      "epoch": 1.2094240837696335,
      "grad_norm": 2.420975923538208,
      "learning_rate": 7.918716577540108e-06,
      "loss": 0.06824178993701935,
      "num_input_tokens_seen": 2915114,
      "step": 2830,
      "train_runtime": 6891.9053,
      "train_tokens_per_second": 422.977
    },
    {
      "epoch": 1.2098514798589592,
      "grad_norm": 3.2608516216278076,
      "learning_rate": 7.914438502673799e-06,
      "loss": 0.11864794790744781,
      "num_input_tokens_seen": 2916100,
      "step": 2831,
      "train_runtime": 6893.9993,
      "train_tokens_per_second": 422.991
    },
    {
      "epoch": 1.210278875948285,
      "grad_norm": 2.302849054336548,
      "learning_rate": 7.910160427807486e-06,
      "loss": 0.05560072138905525,
      "num_input_tokens_seen": 2916866,
      "step": 2832,
      "train_runtime": 6896.0238,
      "train_tokens_per_second": 422.978
    },
    {
      "epoch": 1.210706272037611,
      "grad_norm": 2.91860294342041,
      "learning_rate": 7.905882352941176e-06,
      "loss": 0.09280073642730713,
      "num_input_tokens_seen": 2917742,
      "step": 2833,
      "train_runtime": 6898.0845,
      "train_tokens_per_second": 422.979
    },
    {
      "epoch": 1.2111336681269367,
      "grad_norm": 2.227351427078247,
      "learning_rate": 7.901604278074867e-06,
      "loss": 0.06476104259490967,
      "num_input_tokens_seen": 2918604,
      "step": 2834,
      "train_runtime": 6900.1122,
      "train_tokens_per_second": 422.979
    },
    {
      "epoch": 1.2115610642162624,
      "grad_norm": 2.196326732635498,
      "learning_rate": 7.897326203208556e-06,
      "loss": 0.06640642881393433,
      "num_input_tokens_seen": 2920110,
      "step": 2835,
      "train_runtime": 6902.3675,
      "train_tokens_per_second": 423.059
    },
    {
      "epoch": 1.2119884603055882,
      "grad_norm": 2.8386425971984863,
      "learning_rate": 7.893048128342246e-06,
      "loss": 0.07271026074886322,
      "num_input_tokens_seen": 2921072,
      "step": 2836,
      "train_runtime": 6904.445,
      "train_tokens_per_second": 423.071
    },
    {
      "epoch": 1.212415856394914,
      "grad_norm": 1.9981520175933838,
      "learning_rate": 7.888770053475935e-06,
      "loss": 0.06900268793106079,
      "num_input_tokens_seen": 2921936,
      "step": 2837,
      "train_runtime": 6906.5864,
      "train_tokens_per_second": 423.065
    },
    {
      "epoch": 1.2128432524842399,
      "grad_norm": 3.183472156524658,
      "learning_rate": 7.884491978609627e-06,
      "loss": 0.11748578399419785,
      "num_input_tokens_seen": 2922878,
      "step": 2838,
      "train_runtime": 6908.652,
      "train_tokens_per_second": 423.075
    },
    {
      "epoch": 1.2132706485735656,
      "grad_norm": 2.30401349067688,
      "learning_rate": 7.880213903743316e-06,
      "loss": 0.06798192858695984,
      "num_input_tokens_seen": 2923898,
      "step": 2839,
      "train_runtime": 6910.7447,
      "train_tokens_per_second": 423.094
    },
    {
      "epoch": 1.2136980446628913,
      "grad_norm": 2.8173458576202393,
      "learning_rate": 7.875935828877005e-06,
      "loss": 0.09715848416090012,
      "num_input_tokens_seen": 2924956,
      "step": 2840,
      "train_runtime": 6912.8286,
      "train_tokens_per_second": 423.12
    },
    {
      "epoch": 1.214125440752217,
      "grad_norm": 3.090378522872925,
      "learning_rate": 7.871657754010695e-06,
      "loss": 0.0925881564617157,
      "num_input_tokens_seen": 2925980,
      "step": 2841,
      "train_runtime": 6914.9945,
      "train_tokens_per_second": 423.136
    },
    {
      "epoch": 1.2145528368415428,
      "grad_norm": 1.9239747524261475,
      "learning_rate": 7.867379679144386e-06,
      "loss": 0.05939649045467377,
      "num_input_tokens_seen": 2926850,
      "step": 2842,
      "train_runtime": 6917.0602,
      "train_tokens_per_second": 423.135
    },
    {
      "epoch": 1.2149802329308688,
      "grad_norm": 2.0961647033691406,
      "learning_rate": 7.863101604278075e-06,
      "loss": 0.06744793057441711,
      "num_input_tokens_seen": 2927960,
      "step": 2843,
      "train_runtime": 6919.1564,
      "train_tokens_per_second": 423.167
    },
    {
      "epoch": 1.2154076290201945,
      "grad_norm": 1.948562502861023,
      "learning_rate": 7.858823529411765e-06,
      "loss": 0.06887713819742203,
      "num_input_tokens_seen": 2928870,
      "step": 2844,
      "train_runtime": 6921.2287,
      "train_tokens_per_second": 423.172
    },
    {
      "epoch": 1.2158350251095202,
      "grad_norm": 3.3425185680389404,
      "learning_rate": 7.854545454545454e-06,
      "loss": 0.09248197078704834,
      "num_input_tokens_seen": 2929656,
      "step": 2845,
      "train_runtime": 6923.2579,
      "train_tokens_per_second": 423.161
    },
    {
      "epoch": 1.216262421198846,
      "grad_norm": 2.802161455154419,
      "learning_rate": 7.850267379679145e-06,
      "loss": 0.08666430413722992,
      "num_input_tokens_seen": 2930468,
      "step": 2846,
      "train_runtime": 6925.34,
      "train_tokens_per_second": 423.151
    },
    {
      "epoch": 1.216689817288172,
      "grad_norm": 2.7891452312469482,
      "learning_rate": 7.845989304812835e-06,
      "loss": 0.09340118616819382,
      "num_input_tokens_seen": 2931556,
      "step": 2847,
      "train_runtime": 6927.428,
      "train_tokens_per_second": 423.181
    },
    {
      "epoch": 1.2171172133774977,
      "grad_norm": 3.0174384117126465,
      "learning_rate": 7.841711229946524e-06,
      "loss": 0.112044557929039,
      "num_input_tokens_seen": 2933032,
      "step": 2848,
      "train_runtime": 6929.655,
      "train_tokens_per_second": 423.258
    },
    {
      "epoch": 1.2175446094668234,
      "grad_norm": 1.6146427392959595,
      "learning_rate": 7.837433155080214e-06,
      "loss": 0.054093532264232635,
      "num_input_tokens_seen": 2934706,
      "step": 2849,
      "train_runtime": 6931.905,
      "train_tokens_per_second": 423.362
    },
    {
      "epoch": 1.2179720055561492,
      "grad_norm": 3.355799674987793,
      "learning_rate": 7.833155080213905e-06,
      "loss": 0.08320631086826324,
      "num_input_tokens_seen": 2935460,
      "step": 2850,
      "train_runtime": 6934.0113,
      "train_tokens_per_second": 423.342
    },
    {
      "epoch": 1.2183994016454749,
      "grad_norm": 3.227617025375366,
      "learning_rate": 7.828877005347594e-06,
      "loss": 0.09887047111988068,
      "num_input_tokens_seen": 2936298,
      "step": 2851,
      "train_runtime": 6942.6121,
      "train_tokens_per_second": 422.939
    },
    {
      "epoch": 1.2188267977348008,
      "grad_norm": 2.6629343032836914,
      "learning_rate": 7.824598930481284e-06,
      "loss": 0.09485273063182831,
      "num_input_tokens_seen": 2937282,
      "step": 2852,
      "train_runtime": 6944.7066,
      "train_tokens_per_second": 422.953
    },
    {
      "epoch": 1.2192541938241266,
      "grad_norm": 2.874451160430908,
      "learning_rate": 7.820320855614973e-06,
      "loss": 0.12541738152503967,
      "num_input_tokens_seen": 2938228,
      "step": 2853,
      "train_runtime": 6946.9853,
      "train_tokens_per_second": 422.95
    },
    {
      "epoch": 1.2196815899134523,
      "grad_norm": 2.3367838859558105,
      "learning_rate": 7.816042780748664e-06,
      "loss": 0.08747762441635132,
      "num_input_tokens_seen": 2939198,
      "step": 2854,
      "train_runtime": 6949.081,
      "train_tokens_per_second": 422.962
    },
    {
      "epoch": 1.220108986002778,
      "grad_norm": 2.5057122707366943,
      "learning_rate": 7.811764705882354e-06,
      "loss": 0.1256774663925171,
      "num_input_tokens_seen": 2940528,
      "step": 2855,
      "train_runtime": 6951.2715,
      "train_tokens_per_second": 423.02
    },
    {
      "epoch": 1.2205363820921038,
      "grad_norm": 1.9073731899261475,
      "learning_rate": 7.807486631016043e-06,
      "loss": 0.07260522991418839,
      "num_input_tokens_seen": 2942088,
      "step": 2856,
      "train_runtime": 6953.5375,
      "train_tokens_per_second": 423.107
    },
    {
      "epoch": 1.2209637781814298,
      "grad_norm": 2.3790485858917236,
      "learning_rate": 7.803208556149733e-06,
      "loss": 0.07577663660049438,
      "num_input_tokens_seen": 2943210,
      "step": 2857,
      "train_runtime": 6955.7918,
      "train_tokens_per_second": 423.131
    },
    {
      "epoch": 1.2213911742707555,
      "grad_norm": 2.46437406539917,
      "learning_rate": 7.798930481283424e-06,
      "loss": 0.0929187759757042,
      "num_input_tokens_seen": 2944140,
      "step": 2858,
      "train_runtime": 6957.9205,
      "train_tokens_per_second": 423.135
    },
    {
      "epoch": 1.2218185703600812,
      "grad_norm": 2.299267530441284,
      "learning_rate": 7.794652406417113e-06,
      "loss": 0.08529412746429443,
      "num_input_tokens_seen": 2945410,
      "step": 2859,
      "train_runtime": 6960.1128,
      "train_tokens_per_second": 423.184
    },
    {
      "epoch": 1.222245966449407,
      "grad_norm": 2.9004669189453125,
      "learning_rate": 7.790374331550803e-06,
      "loss": 0.0760367214679718,
      "num_input_tokens_seen": 2946102,
      "step": 2860,
      "train_runtime": 6962.1379,
      "train_tokens_per_second": 423.161
    },
    {
      "epoch": 1.2226733625387327,
      "grad_norm": 3.453859806060791,
      "learning_rate": 7.786096256684492e-06,
      "loss": 0.12871377170085907,
      "num_input_tokens_seen": 2946862,
      "step": 2861,
      "train_runtime": 6964.1342,
      "train_tokens_per_second": 423.148
    },
    {
      "epoch": 1.2231007586280587,
      "grad_norm": 2.486342430114746,
      "learning_rate": 7.781818181818183e-06,
      "loss": 0.10306844115257263,
      "num_input_tokens_seen": 2947706,
      "step": 2862,
      "train_runtime": 6966.2394,
      "train_tokens_per_second": 423.142
    },
    {
      "epoch": 1.2235281547173844,
      "grad_norm": 2.272264242172241,
      "learning_rate": 7.777540106951871e-06,
      "loss": 0.06444300711154938,
      "num_input_tokens_seen": 2948354,
      "step": 2863,
      "train_runtime": 6968.2818,
      "train_tokens_per_second": 423.111
    },
    {
      "epoch": 1.2239555508067101,
      "grad_norm": 2.3583831787109375,
      "learning_rate": 7.773262032085562e-06,
      "loss": 0.10618904232978821,
      "num_input_tokens_seen": 2949502,
      "step": 2864,
      "train_runtime": 6970.4429,
      "train_tokens_per_second": 423.144
    },
    {
      "epoch": 1.2243829468960359,
      "grad_norm": 2.199553966522217,
      "learning_rate": 7.768983957219252e-06,
      "loss": 0.06423413008451462,
      "num_input_tokens_seen": 2950512,
      "step": 2865,
      "train_runtime": 6972.5889,
      "train_tokens_per_second": 423.159
    },
    {
      "epoch": 1.2248103429853616,
      "grad_norm": 2.9690518379211426,
      "learning_rate": 7.764705882352941e-06,
      "loss": 0.0922478660941124,
      "num_input_tokens_seen": 2951136,
      "step": 2866,
      "train_runtime": 6974.6498,
      "train_tokens_per_second": 423.123
    },
    {
      "epoch": 1.2252377390746876,
      "grad_norm": 3.334764242172241,
      "learning_rate": 7.76042780748663e-06,
      "loss": 0.09897273033857346,
      "num_input_tokens_seen": 2952080,
      "step": 2867,
      "train_runtime": 6976.7405,
      "train_tokens_per_second": 423.132
    },
    {
      "epoch": 1.2256651351640133,
      "grad_norm": 3.082503080368042,
      "learning_rate": 7.756149732620322e-06,
      "loss": 0.10666782408952713,
      "num_input_tokens_seen": 2952916,
      "step": 2868,
      "train_runtime": 6978.811,
      "train_tokens_per_second": 423.126
    },
    {
      "epoch": 1.226092531253339,
      "grad_norm": 2.020904302597046,
      "learning_rate": 7.751871657754011e-06,
      "loss": 0.06418243050575256,
      "num_input_tokens_seen": 2953968,
      "step": 2869,
      "train_runtime": 6980.9706,
      "train_tokens_per_second": 423.146
    },
    {
      "epoch": 1.2265199273426648,
      "grad_norm": 1.7023288011550903,
      "learning_rate": 7.7475935828877e-06,
      "loss": 0.056084923446178436,
      "num_input_tokens_seen": 2955176,
      "step": 2870,
      "train_runtime": 6983.1604,
      "train_tokens_per_second": 423.186
    },
    {
      "epoch": 1.2269473234319905,
      "grad_norm": 2.5139248371124268,
      "learning_rate": 7.74331550802139e-06,
      "loss": 0.06824924796819687,
      "num_input_tokens_seen": 2956064,
      "step": 2871,
      "train_runtime": 6985.2952,
      "train_tokens_per_second": 423.184
    },
    {
      "epoch": 1.2273747195213165,
      "grad_norm": 2.359225034713745,
      "learning_rate": 7.739037433155081e-06,
      "loss": 0.11948101222515106,
      "num_input_tokens_seen": 2957632,
      "step": 2872,
      "train_runtime": 6987.5877,
      "train_tokens_per_second": 423.269
    },
    {
      "epoch": 1.2278021156106422,
      "grad_norm": 2.837101697921753,
      "learning_rate": 7.73475935828877e-06,
      "loss": 0.08400681614875793,
      "num_input_tokens_seen": 2958758,
      "step": 2873,
      "train_runtime": 6989.7942,
      "train_tokens_per_second": 423.297
    },
    {
      "epoch": 1.228229511699968,
      "grad_norm": 2.9417858123779297,
      "learning_rate": 7.73048128342246e-06,
      "loss": 0.08431512117385864,
      "num_input_tokens_seen": 2959658,
      "step": 2874,
      "train_runtime": 6992.1147,
      "train_tokens_per_second": 423.285
    },
    {
      "epoch": 1.2286569077892937,
      "grad_norm": 3.3938148021698,
      "learning_rate": 7.72620320855615e-06,
      "loss": 0.08296189457178116,
      "num_input_tokens_seen": 2960430,
      "step": 2875,
      "train_runtime": 6994.1423,
      "train_tokens_per_second": 423.273
    },
    {
      "epoch": 1.2290843038786194,
      "grad_norm": 2.680143356323242,
      "learning_rate": 7.72192513368984e-06,
      "loss": 0.08281657844781876,
      "num_input_tokens_seen": 2961544,
      "step": 2876,
      "train_runtime": 6996.2808,
      "train_tokens_per_second": 423.303
    },
    {
      "epoch": 1.2295116999679454,
      "grad_norm": 2.1526296138763428,
      "learning_rate": 7.71764705882353e-06,
      "loss": 0.0546220988035202,
      "num_input_tokens_seen": 2962492,
      "step": 2877,
      "train_runtime": 6998.3972,
      "train_tokens_per_second": 423.31
    },
    {
      "epoch": 1.2299390960572711,
      "grad_norm": 2.072197198867798,
      "learning_rate": 7.71336898395722e-06,
      "loss": 0.06942172348499298,
      "num_input_tokens_seen": 2964176,
      "step": 2878,
      "train_runtime": 7000.6483,
      "train_tokens_per_second": 423.415
    },
    {
      "epoch": 1.2303664921465969,
      "grad_norm": 4.803370475769043,
      "learning_rate": 7.709090909090909e-06,
      "loss": 0.13560105860233307,
      "num_input_tokens_seen": 2965500,
      "step": 2879,
      "train_runtime": 7002.8382,
      "train_tokens_per_second": 423.471
    },
    {
      "epoch": 1.2307938882359226,
      "grad_norm": 2.1142685413360596,
      "learning_rate": 7.7048128342246e-06,
      "loss": 0.07338854670524597,
      "num_input_tokens_seen": 2966462,
      "step": 2880,
      "train_runtime": 7004.9031,
      "train_tokens_per_second": 423.484
    },
    {
      "epoch": 1.2312212843252484,
      "grad_norm": 3.6128134727478027,
      "learning_rate": 7.70053475935829e-06,
      "loss": 0.13694776594638824,
      "num_input_tokens_seen": 2967170,
      "step": 2881,
      "train_runtime": 7013.7915,
      "train_tokens_per_second": 423.048
    },
    {
      "epoch": 1.2316486804145743,
      "grad_norm": 2.647808790206909,
      "learning_rate": 7.696256684491979e-06,
      "loss": 0.09182015806436539,
      "num_input_tokens_seen": 2968046,
      "step": 2882,
      "train_runtime": 7015.8816,
      "train_tokens_per_second": 423.047
    },
    {
      "epoch": 1.2320760765039,
      "grad_norm": 4.8416595458984375,
      "learning_rate": 7.691978609625669e-06,
      "loss": 0.10868735611438751,
      "num_input_tokens_seen": 2969132,
      "step": 2883,
      "train_runtime": 7017.9996,
      "train_tokens_per_second": 423.074
    },
    {
      "epoch": 1.2325034725932258,
      "grad_norm": 1.730046272277832,
      "learning_rate": 7.68770053475936e-06,
      "loss": 0.03846367076039314,
      "num_input_tokens_seen": 2969934,
      "step": 2884,
      "train_runtime": 7020.031,
      "train_tokens_per_second": 423.066
    },
    {
      "epoch": 1.2329308686825515,
      "grad_norm": 2.80712890625,
      "learning_rate": 7.683422459893049e-06,
      "loss": 0.10911154747009277,
      "num_input_tokens_seen": 2970916,
      "step": 2885,
      "train_runtime": 7022.1315,
      "train_tokens_per_second": 423.079
    },
    {
      "epoch": 1.2333582647718773,
      "grad_norm": 2.480966806411743,
      "learning_rate": 7.679144385026739e-06,
      "loss": 0.04873647540807724,
      "num_input_tokens_seen": 2971388,
      "step": 2886,
      "train_runtime": 7024.0936,
      "train_tokens_per_second": 423.028
    },
    {
      "epoch": 1.2337856608612032,
      "grad_norm": 2.168053388595581,
      "learning_rate": 7.674866310160428e-06,
      "loss": 0.06658559292554855,
      "num_input_tokens_seen": 2972326,
      "step": 2887,
      "train_runtime": 7026.2106,
      "train_tokens_per_second": 423.034
    },
    {
      "epoch": 1.234213056950529,
      "grad_norm": 2.040863037109375,
      "learning_rate": 7.670588235294119e-06,
      "loss": 0.06779739260673523,
      "num_input_tokens_seen": 2973542,
      "step": 2888,
      "train_runtime": 7028.3737,
      "train_tokens_per_second": 423.077
    },
    {
      "epoch": 1.2346404530398547,
      "grad_norm": 3.933476448059082,
      "learning_rate": 7.666310160427809e-06,
      "loss": 0.10543353110551834,
      "num_input_tokens_seen": 2974400,
      "step": 2889,
      "train_runtime": 7030.4384,
      "train_tokens_per_second": 423.075
    },
    {
      "epoch": 1.2350678491291804,
      "grad_norm": 2.114612102508545,
      "learning_rate": 7.662032085561498e-06,
      "loss": 0.08414778858423233,
      "num_input_tokens_seen": 2975368,
      "step": 2890,
      "train_runtime": 7032.5349,
      "train_tokens_per_second": 423.086
    },
    {
      "epoch": 1.2354952452185062,
      "grad_norm": 3.2001802921295166,
      "learning_rate": 7.657754010695187e-06,
      "loss": 0.10490871965885162,
      "num_input_tokens_seen": 2976270,
      "step": 2891,
      "train_runtime": 7034.6215,
      "train_tokens_per_second": 423.089
    },
    {
      "epoch": 1.2359226413078321,
      "grad_norm": 3.1953108310699463,
      "learning_rate": 7.653475935828879e-06,
      "loss": 0.1332489401102066,
      "num_input_tokens_seen": 2977472,
      "step": 2892,
      "train_runtime": 7036.7692,
      "train_tokens_per_second": 423.131
    },
    {
      "epoch": 1.2363500373971579,
      "grad_norm": 2.920745372772217,
      "learning_rate": 7.649197860962568e-06,
      "loss": 0.0834018886089325,
      "num_input_tokens_seen": 2978222,
      "step": 2893,
      "train_runtime": 7038.8256,
      "train_tokens_per_second": 423.113
    },
    {
      "epoch": 1.2367774334864836,
      "grad_norm": 3.159118175506592,
      "learning_rate": 7.644919786096258e-06,
      "loss": 0.1104327142238617,
      "num_input_tokens_seen": 2979364,
      "step": 2894,
      "train_runtime": 7041.0972,
      "train_tokens_per_second": 423.139
    },
    {
      "epoch": 1.2372048295758094,
      "grad_norm": 4.007479190826416,
      "learning_rate": 7.640641711229947e-06,
      "loss": 0.2195119857788086,
      "num_input_tokens_seen": 2981078,
      "step": 2895,
      "train_runtime": 7043.418,
      "train_tokens_per_second": 423.243
    },
    {
      "epoch": 1.237632225665135,
      "grad_norm": 2.6207010746002197,
      "learning_rate": 7.636363636363638e-06,
      "loss": 0.06317804753780365,
      "num_input_tokens_seen": 2982150,
      "step": 2896,
      "train_runtime": 7045.539,
      "train_tokens_per_second": 423.268
    },
    {
      "epoch": 1.238059621754461,
      "grad_norm": 3.7198662757873535,
      "learning_rate": 7.632085561497326e-06,
      "loss": 0.1306927502155304,
      "num_input_tokens_seen": 2983030,
      "step": 2897,
      "train_runtime": 7047.5661,
      "train_tokens_per_second": 423.271
    },
    {
      "epoch": 1.2384870178437868,
      "grad_norm": 2.76037335395813,
      "learning_rate": 7.627807486631017e-06,
      "loss": 0.07866663485765457,
      "num_input_tokens_seen": 2983862,
      "step": 2898,
      "train_runtime": 7049.6245,
      "train_tokens_per_second": 423.265
    },
    {
      "epoch": 1.2389144139331125,
      "grad_norm": 2.6788196563720703,
      "learning_rate": 7.6235294117647064e-06,
      "loss": 0.08830566704273224,
      "num_input_tokens_seen": 2984688,
      "step": 2899,
      "train_runtime": 7051.7151,
      "train_tokens_per_second": 423.257
    },
    {
      "epoch": 1.2393418100224383,
      "grad_norm": 1.965596079826355,
      "learning_rate": 7.619251336898397e-06,
      "loss": 0.061684973537921906,
      "num_input_tokens_seen": 2985636,
      "step": 2900,
      "train_runtime": 7053.8301,
      "train_tokens_per_second": 423.265
    },
    {
      "epoch": 1.239769206111764,
      "grad_norm": 2.6278157234191895,
      "learning_rate": 7.614973262032086e-06,
      "loss": 0.09667761623859406,
      "num_input_tokens_seen": 2986676,
      "step": 2901,
      "train_runtime": 7055.9581,
      "train_tokens_per_second": 423.284
    },
    {
      "epoch": 1.24019660220109,
      "grad_norm": 3.1461708545684814,
      "learning_rate": 7.6106951871657765e-06,
      "loss": 0.06883089989423752,
      "num_input_tokens_seen": 2987500,
      "step": 2902,
      "train_runtime": 7058.024,
      "train_tokens_per_second": 423.277
    },
    {
      "epoch": 1.2406239982904157,
      "grad_norm": 3.065751552581787,
      "learning_rate": 7.606417112299466e-06,
      "loss": 0.09321364760398865,
      "num_input_tokens_seen": 2988562,
      "step": 2903,
      "train_runtime": 7060.0915,
      "train_tokens_per_second": 423.304
    },
    {
      "epoch": 1.2410513943797414,
      "grad_norm": 2.419569253921509,
      "learning_rate": 7.602139037433156e-06,
      "loss": 0.0910138189792633,
      "num_input_tokens_seen": 2989574,
      "step": 2904,
      "train_runtime": 7062.1784,
      "train_tokens_per_second": 423.322
    },
    {
      "epoch": 1.2414787904690672,
      "grad_norm": 2.703366279602051,
      "learning_rate": 7.597860962566846e-06,
      "loss": 0.08345609903335571,
      "num_input_tokens_seen": 2990368,
      "step": 2905,
      "train_runtime": 7064.2697,
      "train_tokens_per_second": 423.309
    },
    {
      "epoch": 1.241906186558393,
      "grad_norm": 2.8962366580963135,
      "learning_rate": 7.593582887700536e-06,
      "loss": 0.07367713749408722,
      "num_input_tokens_seen": 2991064,
      "step": 2906,
      "train_runtime": 7066.3487,
      "train_tokens_per_second": 423.283
    },
    {
      "epoch": 1.2423335826477189,
      "grad_norm": 3.0173261165618896,
      "learning_rate": 7.589304812834225e-06,
      "loss": 0.10552377998828888,
      "num_input_tokens_seen": 2991910,
      "step": 2907,
      "train_runtime": 7068.4292,
      "train_tokens_per_second": 423.278
    },
    {
      "epoch": 1.2427609787370446,
      "grad_norm": 2.7481191158294678,
      "learning_rate": 7.585026737967916e-06,
      "loss": 0.07555650174617767,
      "num_input_tokens_seen": 2992988,
      "step": 2908,
      "train_runtime": 7070.5554,
      "train_tokens_per_second": 423.303
    },
    {
      "epoch": 1.2431883748263703,
      "grad_norm": 3.0668375492095947,
      "learning_rate": 7.580748663101605e-06,
      "loss": 0.09716898202896118,
      "num_input_tokens_seen": 2993678,
      "step": 2909,
      "train_runtime": 7072.6108,
      "train_tokens_per_second": 423.278
    },
    {
      "epoch": 1.243615770915696,
      "grad_norm": 2.2745583057403564,
      "learning_rate": 7.576470588235295e-06,
      "loss": 0.09371167421340942,
      "num_input_tokens_seen": 2994882,
      "step": 2910,
      "train_runtime": 7074.7554,
      "train_tokens_per_second": 423.32
    },
    {
      "epoch": 1.2440431670050218,
      "grad_norm": 4.9229278564453125,
      "learning_rate": 7.572192513368984e-06,
      "loss": 0.08085934817790985,
      "num_input_tokens_seen": 2996036,
      "step": 2911,
      "train_runtime": 7083.6298,
      "train_tokens_per_second": 422.952
    },
    {
      "epoch": 1.2444705630943478,
      "grad_norm": 2.279939651489258,
      "learning_rate": 7.567914438502675e-06,
      "loss": 0.075149305164814,
      "num_input_tokens_seen": 2997060,
      "step": 2912,
      "train_runtime": 7085.8125,
      "train_tokens_per_second": 422.966
    },
    {
      "epoch": 1.2448979591836735,
      "grad_norm": 2.938504934310913,
      "learning_rate": 7.563636363636364e-06,
      "loss": 0.081467404961586,
      "num_input_tokens_seen": 2997792,
      "step": 2913,
      "train_runtime": 7088.2248,
      "train_tokens_per_second": 422.926
    },
    {
      "epoch": 1.2453253552729993,
      "grad_norm": 3.6294662952423096,
      "learning_rate": 7.559358288770055e-06,
      "loss": 0.10107357800006866,
      "num_input_tokens_seen": 2998780,
      "step": 2914,
      "train_runtime": 7090.346,
      "train_tokens_per_second": 422.938
    },
    {
      "epoch": 1.245752751362325,
      "grad_norm": 3.192488431930542,
      "learning_rate": 7.5550802139037435e-06,
      "loss": 0.12392017245292664,
      "num_input_tokens_seen": 2999772,
      "step": 2915,
      "train_runtime": 7092.5126,
      "train_tokens_per_second": 422.949
    },
    {
      "epoch": 1.2461801474516507,
      "grad_norm": 1.8792623281478882,
      "learning_rate": 7.550802139037434e-06,
      "loss": 0.05888255313038826,
      "num_input_tokens_seen": 3000612,
      "step": 2916,
      "train_runtime": 7094.5412,
      "train_tokens_per_second": 422.947
    },
    {
      "epoch": 1.2466075435409767,
      "grad_norm": 1.5301157236099243,
      "learning_rate": 7.546524064171123e-06,
      "loss": 0.0554562583565712,
      "num_input_tokens_seen": 3002066,
      "step": 2917,
      "train_runtime": 7096.7792,
      "train_tokens_per_second": 423.018
    },
    {
      "epoch": 1.2470349396303024,
      "grad_norm": 2.954228162765503,
      "learning_rate": 7.5422459893048135e-06,
      "loss": 0.07843967527151108,
      "num_input_tokens_seen": 3002924,
      "step": 2918,
      "train_runtime": 7098.8704,
      "train_tokens_per_second": 423.014
    },
    {
      "epoch": 1.2474623357196282,
      "grad_norm": 2.620840072631836,
      "learning_rate": 7.537967914438503e-06,
      "loss": 0.0982760339975357,
      "num_input_tokens_seen": 3004388,
      "step": 2919,
      "train_runtime": 7101.1239,
      "train_tokens_per_second": 423.086
    },
    {
      "epoch": 1.247889731808954,
      "grad_norm": 3.532559394836426,
      "learning_rate": 7.533689839572193e-06,
      "loss": 0.0948241576552391,
      "num_input_tokens_seen": 3005252,
      "step": 2920,
      "train_runtime": 7103.2083,
      "train_tokens_per_second": 423.084
    },
    {
      "epoch": 1.2483171278982796,
      "grad_norm": 3.34596586227417,
      "learning_rate": 7.529411764705883e-06,
      "loss": 0.12633314728736877,
      "num_input_tokens_seen": 3006376,
      "step": 2921,
      "train_runtime": 7105.3749,
      "train_tokens_per_second": 423.113
    },
    {
      "epoch": 1.2487445239876056,
      "grad_norm": 2.737436532974243,
      "learning_rate": 7.525133689839573e-06,
      "loss": 0.10718074440956116,
      "num_input_tokens_seen": 3007246,
      "step": 2922,
      "train_runtime": 7107.4403,
      "train_tokens_per_second": 423.112
    },
    {
      "epoch": 1.2491719200769313,
      "grad_norm": 2.073531150817871,
      "learning_rate": 7.5208556149732624e-06,
      "loss": 0.06771333515644073,
      "num_input_tokens_seen": 3008410,
      "step": 2923,
      "train_runtime": 7109.6154,
      "train_tokens_per_second": 423.147
    },
    {
      "epoch": 1.249599316166257,
      "grad_norm": 1.9747004508972168,
      "learning_rate": 7.516577540106953e-06,
      "loss": 0.04575204849243164,
      "num_input_tokens_seen": 3009346,
      "step": 2924,
      "train_runtime": 7111.6846,
      "train_tokens_per_second": 423.155
    },
    {
      "epoch": 1.2500267122555828,
      "grad_norm": 2.353694200515747,
      "learning_rate": 7.512299465240642e-06,
      "loss": 0.05825520679354668,
      "num_input_tokens_seen": 3010114,
      "step": 2925,
      "train_runtime": 7113.7438,
      "train_tokens_per_second": 423.141
    },
    {
      "epoch": 1.2504541083449086,
      "grad_norm": 4.099815845489502,
      "learning_rate": 7.5080213903743325e-06,
      "loss": 0.14841587841510773,
      "num_input_tokens_seen": 3011484,
      "step": 2926,
      "train_runtime": 7115.9249,
      "train_tokens_per_second": 423.203
    },
    {
      "epoch": 1.2508815044342345,
      "grad_norm": 2.272240400314331,
      "learning_rate": 7.503743315508022e-06,
      "loss": 0.08497324585914612,
      "num_input_tokens_seen": 3012460,
      "step": 2927,
      "train_runtime": 7118.0136,
      "train_tokens_per_second": 423.216
    },
    {
      "epoch": 1.2513089005235603,
      "grad_norm": 2.8117141723632812,
      "learning_rate": 7.499465240641712e-06,
      "loss": 0.12227321416139603,
      "num_input_tokens_seen": 3014190,
      "step": 2928,
      "train_runtime": 7120.292,
      "train_tokens_per_second": 423.324
    },
    {
      "epoch": 1.251736296612886,
      "grad_norm": 2.3796160221099854,
      "learning_rate": 7.495187165775402e-06,
      "loss": 0.04914858564734459,
      "num_input_tokens_seen": 3014772,
      "step": 2929,
      "train_runtime": 7122.287,
      "train_tokens_per_second": 423.287
    },
    {
      "epoch": 1.2521636927022117,
      "grad_norm": 1.7777422666549683,
      "learning_rate": 7.490909090909092e-06,
      "loss": 0.06682610511779785,
      "num_input_tokens_seen": 3016042,
      "step": 2930,
      "train_runtime": 7124.4403,
      "train_tokens_per_second": 423.337
    },
    {
      "epoch": 1.2525910887915375,
      "grad_norm": 3.665255069732666,
      "learning_rate": 7.486631016042781e-06,
      "loss": 0.06342610716819763,
      "num_input_tokens_seen": 3016976,
      "step": 2931,
      "train_runtime": 7126.5124,
      "train_tokens_per_second": 423.345
    },
    {
      "epoch": 1.2530184848808634,
      "grad_norm": 2.206355333328247,
      "learning_rate": 7.482352941176472e-06,
      "loss": 0.06764993816614151,
      "num_input_tokens_seen": 3017772,
      "step": 2932,
      "train_runtime": 7128.6174,
      "train_tokens_per_second": 423.332
    },
    {
      "epoch": 1.2534458809701892,
      "grad_norm": 3.052387237548828,
      "learning_rate": 7.478074866310161e-06,
      "loss": 0.07701212912797928,
      "num_input_tokens_seen": 3018610,
      "step": 2933,
      "train_runtime": 7130.6591,
      "train_tokens_per_second": 423.328
    },
    {
      "epoch": 1.253873277059515,
      "grad_norm": 2.3182268142700195,
      "learning_rate": 7.4737967914438514e-06,
      "loss": 0.1297970414161682,
      "num_input_tokens_seen": 3020178,
      "step": 2934,
      "train_runtime": 7132.8806,
      "train_tokens_per_second": 423.416
    },
    {
      "epoch": 1.2543006731488406,
      "grad_norm": 3.0166118144989014,
      "learning_rate": 7.469518716577541e-06,
      "loss": 0.08625024557113647,
      "num_input_tokens_seen": 3021060,
      "step": 2935,
      "train_runtime": 7134.9574,
      "train_tokens_per_second": 423.417
    },
    {
      "epoch": 1.2547280692381664,
      "grad_norm": 2.9842517375946045,
      "learning_rate": 7.465240641711231e-06,
      "loss": 0.11167801171541214,
      "num_input_tokens_seen": 3022448,
      "step": 2936,
      "train_runtime": 7137.1427,
      "train_tokens_per_second": 423.482
    },
    {
      "epoch": 1.2551554653274923,
      "grad_norm": 2.7223544120788574,
      "learning_rate": 7.460962566844921e-06,
      "loss": 0.04725995659828186,
      "num_input_tokens_seen": 3022992,
      "step": 2937,
      "train_runtime": 7139.0569,
      "train_tokens_per_second": 423.444
    },
    {
      "epoch": 1.255582861416818,
      "grad_norm": 3.089219808578491,
      "learning_rate": 7.456684491978611e-06,
      "loss": 0.10635600984096527,
      "num_input_tokens_seen": 3023878,
      "step": 2938,
      "train_runtime": 7141.1219,
      "train_tokens_per_second": 423.446
    },
    {
      "epoch": 1.2560102575061438,
      "grad_norm": 2.7573907375335693,
      "learning_rate": 7.4524064171123e-06,
      "loss": 0.13875748217105865,
      "num_input_tokens_seen": 3025130,
      "step": 2939,
      "train_runtime": 7143.2857,
      "train_tokens_per_second": 423.493
    },
    {
      "epoch": 1.2564376535954696,
      "grad_norm": 2.022953748703003,
      "learning_rate": 7.448128342245991e-06,
      "loss": 0.08140638470649719,
      "num_input_tokens_seen": 3026354,
      "step": 2940,
      "train_runtime": 7145.4656,
      "train_tokens_per_second": 423.535
    },
    {
      "epoch": 1.2568650496847953,
      "grad_norm": 3.272393226623535,
      "learning_rate": 7.44385026737968e-06,
      "loss": 0.13083159923553467,
      "num_input_tokens_seen": 3027380,
      "step": 2941,
      "train_runtime": 7154.2628,
      "train_tokens_per_second": 423.158
    },
    {
      "epoch": 1.2572924457741212,
      "grad_norm": 1.982572317123413,
      "learning_rate": 7.439572192513369e-06,
      "loss": 0.0995146706700325,
      "num_input_tokens_seen": 3029802,
      "step": 2942,
      "train_runtime": 7156.742,
      "train_tokens_per_second": 423.349
    },
    {
      "epoch": 1.257719841863447,
      "grad_norm": 3.10163950920105,
      "learning_rate": 7.43529411764706e-06,
      "loss": 0.09574081748723984,
      "num_input_tokens_seen": 3030454,
      "step": 2943,
      "train_runtime": 7158.7744,
      "train_tokens_per_second": 423.32
    },
    {
      "epoch": 1.2581472379527727,
      "grad_norm": 3.0127134323120117,
      "learning_rate": 7.4310160427807484e-06,
      "loss": 0.0781400054693222,
      "num_input_tokens_seen": 3031326,
      "step": 2944,
      "train_runtime": 7160.8464,
      "train_tokens_per_second": 423.32
    },
    {
      "epoch": 1.2585746340420985,
      "grad_norm": 1.9683183431625366,
      "learning_rate": 7.426737967914439e-06,
      "loss": 0.07526128739118576,
      "num_input_tokens_seen": 3032784,
      "step": 2945,
      "train_runtime": 7163.0844,
      "train_tokens_per_second": 423.391
    },
    {
      "epoch": 1.2590020301314242,
      "grad_norm": 2.7265121936798096,
      "learning_rate": 7.422459893048128e-06,
      "loss": 0.08381232619285583,
      "num_input_tokens_seen": 3033762,
      "step": 2946,
      "train_runtime": 7165.1435,
      "train_tokens_per_second": 423.406
    },
    {
      "epoch": 1.2594294262207502,
      "grad_norm": 2.7708404064178467,
      "learning_rate": 7.4181818181818185e-06,
      "loss": 0.10159067064523697,
      "num_input_tokens_seen": 3034694,
      "step": 2947,
      "train_runtime": 7167.2784,
      "train_tokens_per_second": 423.41
    },
    {
      "epoch": 1.259856822310076,
      "grad_norm": 2.539111852645874,
      "learning_rate": 7.413903743315508e-06,
      "loss": 0.10891447961330414,
      "num_input_tokens_seen": 3035730,
      "step": 2948,
      "train_runtime": 7169.4713,
      "train_tokens_per_second": 423.425
    },
    {
      "epoch": 1.2602842183994016,
      "grad_norm": 2.6784067153930664,
      "learning_rate": 7.409625668449198e-06,
      "loss": 0.10003319382667542,
      "num_input_tokens_seen": 3036538,
      "step": 2949,
      "train_runtime": 7171.6233,
      "train_tokens_per_second": 423.41
    },
    {
      "epoch": 1.2607116144887274,
      "grad_norm": 2.4050145149230957,
      "learning_rate": 7.405347593582888e-06,
      "loss": 0.08705523610115051,
      "num_input_tokens_seen": 3037786,
      "step": 2950,
      "train_runtime": 7173.8423,
      "train_tokens_per_second": 423.453
    },
    {
      "epoch": 1.261139010578053,
      "grad_norm": 3.7022290229797363,
      "learning_rate": 7.401069518716578e-06,
      "loss": 0.09694761782884598,
      "num_input_tokens_seen": 3038500,
      "step": 2951,
      "train_runtime": 7175.9263,
      "train_tokens_per_second": 423.43
    },
    {
      "epoch": 1.261566406667379,
      "grad_norm": 2.862175226211548,
      "learning_rate": 7.396791443850267e-06,
      "loss": 0.10729025304317474,
      "num_input_tokens_seen": 3039416,
      "step": 2952,
      "train_runtime": 7177.9955,
      "train_tokens_per_second": 423.435
    },
    {
      "epoch": 1.2619938027567048,
      "grad_norm": 2.244023561477661,
      "learning_rate": 7.392513368983958e-06,
      "loss": 0.05060585215687752,
      "num_input_tokens_seen": 3040176,
      "step": 2953,
      "train_runtime": 7180.1855,
      "train_tokens_per_second": 423.412
    },
    {
      "epoch": 1.2624211988460305,
      "grad_norm": 2.366894006729126,
      "learning_rate": 7.388235294117647e-06,
      "loss": 0.05431223660707474,
      "num_input_tokens_seen": 3041154,
      "step": 2954,
      "train_runtime": 7182.2722,
      "train_tokens_per_second": 423.425
    },
    {
      "epoch": 1.2628485949353563,
      "grad_norm": 3.590388536453247,
      "learning_rate": 7.383957219251337e-06,
      "loss": 0.10859466344118118,
      "num_input_tokens_seen": 3041974,
      "step": 2955,
      "train_runtime": 7184.3743,
      "train_tokens_per_second": 423.415
    },
    {
      "epoch": 1.263275991024682,
      "grad_norm": 2.237165927886963,
      "learning_rate": 7.379679144385027e-06,
      "loss": 0.07725930958986282,
      "num_input_tokens_seen": 3043170,
      "step": 2956,
      "train_runtime": 7186.5322,
      "train_tokens_per_second": 423.455
    },
    {
      "epoch": 1.263703387114008,
      "grad_norm": 2.8432765007019043,
      "learning_rate": 7.375401069518717e-06,
      "loss": 0.09574152529239655,
      "num_input_tokens_seen": 3044402,
      "step": 2957,
      "train_runtime": 7188.6796,
      "train_tokens_per_second": 423.499
    },
    {
      "epoch": 1.2641307832033337,
      "grad_norm": 2.175550699234009,
      "learning_rate": 7.371122994652407e-06,
      "loss": 0.06777040660381317,
      "num_input_tokens_seen": 3045510,
      "step": 2958,
      "train_runtime": 7190.8261,
      "train_tokens_per_second": 423.527
    },
    {
      "epoch": 1.2645581792926595,
      "grad_norm": 2.986528158187866,
      "learning_rate": 7.366844919786097e-06,
      "loss": 0.15197508037090302,
      "num_input_tokens_seen": 3046528,
      "step": 2959,
      "train_runtime": 7192.908,
      "train_tokens_per_second": 423.546
    },
    {
      "epoch": 1.2649855753819852,
      "grad_norm": 2.2462847232818604,
      "learning_rate": 7.362566844919786e-06,
      "loss": 0.07876642793416977,
      "num_input_tokens_seen": 3048014,
      "step": 2960,
      "train_runtime": 7195.1337,
      "train_tokens_per_second": 423.622
    },
    {
      "epoch": 1.265412971471311,
      "grad_norm": 2.8731985092163086,
      "learning_rate": 7.358288770053477e-06,
      "loss": 0.071068674325943,
      "num_input_tokens_seen": 3049246,
      "step": 2961,
      "train_runtime": 7197.2982,
      "train_tokens_per_second": 423.665
    },
    {
      "epoch": 1.265840367560637,
      "grad_norm": 3.0250935554504395,
      "learning_rate": 7.354010695187166e-06,
      "loss": 0.10351283103227615,
      "num_input_tokens_seen": 3050088,
      "step": 2962,
      "train_runtime": 7199.4573,
      "train_tokens_per_second": 423.655
    },
    {
      "epoch": 1.2662677636499626,
      "grad_norm": 2.763643264770508,
      "learning_rate": 7.349732620320856e-06,
      "loss": 0.07316923886537552,
      "num_input_tokens_seen": 3050988,
      "step": 2963,
      "train_runtime": 7201.552,
      "train_tokens_per_second": 423.657
    },
    {
      "epoch": 1.2666951597392884,
      "grad_norm": 2.818345785140991,
      "learning_rate": 7.345454545454546e-06,
      "loss": 0.08374558389186859,
      "num_input_tokens_seen": 3052072,
      "step": 2964,
      "train_runtime": 7203.67,
      "train_tokens_per_second": 423.683
    },
    {
      "epoch": 1.267122555828614,
      "grad_norm": 2.974126100540161,
      "learning_rate": 7.341176470588236e-06,
      "loss": 0.1368192881345749,
      "num_input_tokens_seen": 3052984,
      "step": 2965,
      "train_runtime": 7205.7308,
      "train_tokens_per_second": 423.688
    },
    {
      "epoch": 1.2675499519179398,
      "grad_norm": 2.4413678646087646,
      "learning_rate": 7.3368983957219256e-06,
      "loss": 0.062972292304039,
      "num_input_tokens_seen": 3053904,
      "step": 2966,
      "train_runtime": 7207.7925,
      "train_tokens_per_second": 423.695
    },
    {
      "epoch": 1.2679773480072658,
      "grad_norm": 2.8165433406829834,
      "learning_rate": 7.332620320855616e-06,
      "loss": 0.09501127898693085,
      "num_input_tokens_seen": 3054740,
      "step": 2967,
      "train_runtime": 7209.814,
      "train_tokens_per_second": 423.692
    },
    {
      "epoch": 1.2684047440965915,
      "grad_norm": 2.9385902881622314,
      "learning_rate": 7.328342245989305e-06,
      "loss": 0.10646688938140869,
      "num_input_tokens_seen": 3056286,
      "step": 2968,
      "train_runtime": 7212.0179,
      "train_tokens_per_second": 423.777
    },
    {
      "epoch": 1.2688321401859173,
      "grad_norm": 3.7117159366607666,
      "learning_rate": 7.324064171122996e-06,
      "loss": 0.09627954661846161,
      "num_input_tokens_seen": 3057050,
      "step": 2969,
      "train_runtime": 7214.0406,
      "train_tokens_per_second": 423.764
    },
    {
      "epoch": 1.269259536275243,
      "grad_norm": 1.66433846950531,
      "learning_rate": 7.319786096256685e-06,
      "loss": 0.06451142579317093,
      "num_input_tokens_seen": 3058172,
      "step": 2970,
      "train_runtime": 7216.1612,
      "train_tokens_per_second": 423.795
    },
    {
      "epoch": 1.2696869323645688,
      "grad_norm": 2.7369394302368164,
      "learning_rate": 7.315508021390375e-06,
      "loss": 0.08050864934921265,
      "num_input_tokens_seen": 3059114,
      "step": 2971,
      "train_runtime": 7225.0413,
      "train_tokens_per_second": 423.404
    },
    {
      "epoch": 1.2701143284538947,
      "grad_norm": 2.4905636310577393,
      "learning_rate": 7.311229946524065e-06,
      "loss": 0.06311958283185959,
      "num_input_tokens_seen": 3060174,
      "step": 2972,
      "train_runtime": 7227.11,
      "train_tokens_per_second": 423.43
    },
    {
      "epoch": 1.2705417245432205,
      "grad_norm": 2.404844284057617,
      "learning_rate": 7.306951871657755e-06,
      "loss": 0.05939580500125885,
      "num_input_tokens_seen": 3061080,
      "step": 2973,
      "train_runtime": 7229.3493,
      "train_tokens_per_second": 423.424
    },
    {
      "epoch": 1.2709691206325462,
      "grad_norm": 2.6710987091064453,
      "learning_rate": 7.302673796791444e-06,
      "loss": 0.13894209265708923,
      "num_input_tokens_seen": 3062558,
      "step": 2974,
      "train_runtime": 7231.5213,
      "train_tokens_per_second": 423.501
    },
    {
      "epoch": 1.271396516721872,
      "grad_norm": 2.849973201751709,
      "learning_rate": 7.298395721925135e-06,
      "loss": 0.11251353472471237,
      "num_input_tokens_seen": 3063870,
      "step": 2975,
      "train_runtime": 7233.708,
      "train_tokens_per_second": 423.555
    },
    {
      "epoch": 1.2718239128111977,
      "grad_norm": 2.4804179668426514,
      "learning_rate": 7.294117647058823e-06,
      "loss": 0.10212527215480804,
      "num_input_tokens_seen": 3064920,
      "step": 2976,
      "train_runtime": 7235.8292,
      "train_tokens_per_second": 423.576
    },
    {
      "epoch": 1.2722513089005236,
      "grad_norm": 3.3878254890441895,
      "learning_rate": 7.2898395721925145e-06,
      "loss": 0.20068228244781494,
      "num_input_tokens_seen": 3065972,
      "step": 2977,
      "train_runtime": 7237.9569,
      "train_tokens_per_second": 423.596
    },
    {
      "epoch": 1.2726787049898494,
      "grad_norm": 3.1920878887176514,
      "learning_rate": 7.285561497326203e-06,
      "loss": 0.11708036065101624,
      "num_input_tokens_seen": 3066740,
      "step": 2978,
      "train_runtime": 7239.973,
      "train_tokens_per_second": 423.584
    },
    {
      "epoch": 1.273106101079175,
      "grad_norm": 2.858785390853882,
      "learning_rate": 7.2812834224598934e-06,
      "loss": 0.11348709464073181,
      "num_input_tokens_seen": 3067800,
      "step": 2979,
      "train_runtime": 7242.1248,
      "train_tokens_per_second": 423.605
    },
    {
      "epoch": 1.2735334971685008,
      "grad_norm": 3.5289299488067627,
      "learning_rate": 7.277005347593583e-06,
      "loss": 0.11311467736959457,
      "num_input_tokens_seen": 3068474,
      "step": 2980,
      "train_runtime": 7244.1318,
      "train_tokens_per_second": 423.581
    },
    {
      "epoch": 1.2739608932578266,
      "grad_norm": 2.6099350452423096,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.08651623129844666,
      "num_input_tokens_seen": 3069356,
      "step": 2981,
      "train_runtime": 7246.255,
      "train_tokens_per_second": 423.578
    },
    {
      "epoch": 1.2743882893471525,
      "grad_norm": 2.4583418369293213,
      "learning_rate": 7.268449197860963e-06,
      "loss": 0.09893359988927841,
      "num_input_tokens_seen": 3070462,
      "step": 2982,
      "train_runtime": 7248.367,
      "train_tokens_per_second": 423.607
    },
    {
      "epoch": 1.2748156854364783,
      "grad_norm": 2.880175828933716,
      "learning_rate": 7.264171122994653e-06,
      "loss": 0.1534269005060196,
      "num_input_tokens_seen": 3071384,
      "step": 2983,
      "train_runtime": 7250.3815,
      "train_tokens_per_second": 423.617
    },
    {
      "epoch": 1.275243081525804,
      "grad_norm": 3.4426698684692383,
      "learning_rate": 7.259893048128342e-06,
      "loss": 0.0703113004565239,
      "num_input_tokens_seen": 3072138,
      "step": 2984,
      "train_runtime": 7252.3964,
      "train_tokens_per_second": 423.603
    },
    {
      "epoch": 1.2756704776151297,
      "grad_norm": 3.566863775253296,
      "learning_rate": 7.255614973262033e-06,
      "loss": 0.120570108294487,
      "num_input_tokens_seen": 3073342,
      "step": 2985,
      "train_runtime": 7254.5062,
      "train_tokens_per_second": 423.646
    },
    {
      "epoch": 1.2760978737044555,
      "grad_norm": 2.0629470348358154,
      "learning_rate": 7.251336898395722e-06,
      "loss": 0.050238534808158875,
      "num_input_tokens_seen": 3074396,
      "step": 2986,
      "train_runtime": 7256.5934,
      "train_tokens_per_second": 423.669
    },
    {
      "epoch": 1.2765252697937814,
      "grad_norm": 3.396526575088501,
      "learning_rate": 7.247058823529412e-06,
      "loss": 0.09168101102113724,
      "num_input_tokens_seen": 3075092,
      "step": 2987,
      "train_runtime": 7258.6006,
      "train_tokens_per_second": 423.648
    },
    {
      "epoch": 1.2769526658831072,
      "grad_norm": 2.337019443511963,
      "learning_rate": 7.242780748663102e-06,
      "loss": 0.07134734094142914,
      "num_input_tokens_seen": 3075776,
      "step": 2988,
      "train_runtime": 7260.6084,
      "train_tokens_per_second": 423.625
    },
    {
      "epoch": 1.277380061972433,
      "grad_norm": 2.160491704940796,
      "learning_rate": 7.238502673796792e-06,
      "loss": 0.10964909195899963,
      "num_input_tokens_seen": 3077336,
      "step": 2989,
      "train_runtime": 7262.8133,
      "train_tokens_per_second": 423.711
    },
    {
      "epoch": 1.2778074580617587,
      "grad_norm": 2.9258317947387695,
      "learning_rate": 7.2342245989304816e-06,
      "loss": 0.11585892736911774,
      "num_input_tokens_seen": 3078078,
      "step": 2990,
      "train_runtime": 7264.8889,
      "train_tokens_per_second": 423.692
    },
    {
      "epoch": 1.2782348541510844,
      "grad_norm": 2.223729133605957,
      "learning_rate": 7.229946524064172e-06,
      "loss": 0.054368749260902405,
      "num_input_tokens_seen": 3078888,
      "step": 2991,
      "train_runtime": 7266.8982,
      "train_tokens_per_second": 423.687
    },
    {
      "epoch": 1.2786622502404104,
      "grad_norm": 2.646707534790039,
      "learning_rate": 7.225668449197861e-06,
      "loss": 0.06866642832756042,
      "num_input_tokens_seen": 3079550,
      "step": 2992,
      "train_runtime": 7268.9418,
      "train_tokens_per_second": 423.659
    },
    {
      "epoch": 1.279089646329736,
      "grad_norm": 1.551003098487854,
      "learning_rate": 7.221390374331552e-06,
      "loss": 0.06737552583217621,
      "num_input_tokens_seen": 3081252,
      "step": 2993,
      "train_runtime": 7271.1833,
      "train_tokens_per_second": 423.762
    },
    {
      "epoch": 1.2795170424190618,
      "grad_norm": 4.425068378448486,
      "learning_rate": 7.217112299465241e-06,
      "loss": 0.18755845725536346,
      "num_input_tokens_seen": 3082642,
      "step": 2994,
      "train_runtime": 7273.4332,
      "train_tokens_per_second": 423.822
    },
    {
      "epoch": 1.2799444385083876,
      "grad_norm": 2.225693941116333,
      "learning_rate": 7.212834224598931e-06,
      "loss": 0.09273100644350052,
      "num_input_tokens_seen": 3083644,
      "step": 2995,
      "train_runtime": 7275.5686,
      "train_tokens_per_second": 423.835
    },
    {
      "epoch": 1.2803718345977133,
      "grad_norm": 2.2131972312927246,
      "learning_rate": 7.208556149732621e-06,
      "loss": 0.09236171841621399,
      "num_input_tokens_seen": 3085096,
      "step": 2996,
      "train_runtime": 7277.8116,
      "train_tokens_per_second": 423.904
    },
    {
      "epoch": 1.2807992306870393,
      "grad_norm": 2.118729591369629,
      "learning_rate": 7.204278074866311e-06,
      "loss": 0.0640028715133667,
      "num_input_tokens_seen": 3086270,
      "step": 2997,
      "train_runtime": 7279.9187,
      "train_tokens_per_second": 423.943
    },
    {
      "epoch": 1.281226626776365,
      "grad_norm": 2.5736396312713623,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.07748084515333176,
      "num_input_tokens_seen": 3087164,
      "step": 2998,
      "train_runtime": 7281.9759,
      "train_tokens_per_second": 423.946
    },
    {
      "epoch": 1.2816540228656907,
      "grad_norm": 2.87827205657959,
      "learning_rate": 7.195721925133691e-06,
      "loss": 0.08067092299461365,
      "num_input_tokens_seen": 3087900,
      "step": 2999,
      "train_runtime": 7283.9746,
      "train_tokens_per_second": 423.931
    },
    {
      "epoch": 1.2820814189550165,
      "grad_norm": 2.8396527767181396,
      "learning_rate": 7.19144385026738e-06,
      "loss": 0.08095599710941315,
      "num_input_tokens_seen": 3088772,
      "step": 3000,
      "train_runtime": 7286.0723,
      "train_tokens_per_second": 423.928
    },
    {
      "epoch": 1.2825088150443422,
      "grad_norm": 2.690825939178467,
      "learning_rate": 7.1871657754010706e-06,
      "loss": 0.1062483862042427,
      "num_input_tokens_seen": 3089816,
      "step": 3001,
      "train_runtime": 7294.792,
      "train_tokens_per_second": 423.565
    },
    {
      "epoch": 1.2829362111336682,
      "grad_norm": 2.9456281661987305,
      "learning_rate": 7.18288770053476e-06,
      "loss": 0.11716188490390778,
      "num_input_tokens_seen": 3090760,
      "step": 3002,
      "train_runtime": 7296.8825,
      "train_tokens_per_second": 423.573
    },
    {
      "epoch": 1.283363607222994,
      "grad_norm": 2.6339001655578613,
      "learning_rate": 7.17860962566845e-06,
      "loss": 0.08602765947580338,
      "num_input_tokens_seen": 3091570,
      "step": 3003,
      "train_runtime": 7299.1123,
      "train_tokens_per_second": 423.554
    },
    {
      "epoch": 1.2837910033123197,
      "grad_norm": 3.105644941329956,
      "learning_rate": 7.17433155080214e-06,
      "loss": 0.08951435983181,
      "num_input_tokens_seen": 3092560,
      "step": 3004,
      "train_runtime": 7301.1614,
      "train_tokens_per_second": 423.571
    },
    {
      "epoch": 1.2842183994016454,
      "grad_norm": 1.7576477527618408,
      "learning_rate": 7.17005347593583e-06,
      "loss": 0.052186835557222366,
      "num_input_tokens_seen": 3094294,
      "step": 3005,
      "train_runtime": 7303.4429,
      "train_tokens_per_second": 423.676
    },
    {
      "epoch": 1.2846457954909711,
      "grad_norm": 2.681201696395874,
      "learning_rate": 7.1657754010695195e-06,
      "loss": 0.059617698192596436,
      "num_input_tokens_seen": 3095138,
      "step": 3006,
      "train_runtime": 7305.5214,
      "train_tokens_per_second": 423.671
    },
    {
      "epoch": 1.285073191580297,
      "grad_norm": 4.363328456878662,
      "learning_rate": 7.16149732620321e-06,
      "loss": 0.14670848846435547,
      "num_input_tokens_seen": 3095786,
      "step": 3007,
      "train_runtime": 7307.5178,
      "train_tokens_per_second": 423.644
    },
    {
      "epoch": 1.2855005876696228,
      "grad_norm": 2.1391069889068604,
      "learning_rate": 7.157219251336898e-06,
      "loss": 0.10724630951881409,
      "num_input_tokens_seen": 3097300,
      "step": 3008,
      "train_runtime": 7309.7184,
      "train_tokens_per_second": 423.724
    },
    {
      "epoch": 1.2859279837589486,
      "grad_norm": 3.1217257976531982,
      "learning_rate": 7.1529411764705895e-06,
      "loss": 0.09327880293130875,
      "num_input_tokens_seen": 3098856,
      "step": 3009,
      "train_runtime": 7311.9635,
      "train_tokens_per_second": 423.806
    },
    {
      "epoch": 1.2863553798482743,
      "grad_norm": 2.1741437911987305,
      "learning_rate": 7.148663101604278e-06,
      "loss": 0.05772768333554268,
      "num_input_tokens_seen": 3099682,
      "step": 3010,
      "train_runtime": 7314.0006,
      "train_tokens_per_second": 423.801
    },
    {
      "epoch": 1.2867827759376,
      "grad_norm": 2.7755205631256104,
      "learning_rate": 7.144385026737969e-06,
      "loss": 0.11665832996368408,
      "num_input_tokens_seen": 3100658,
      "step": 3011,
      "train_runtime": 7316.0977,
      "train_tokens_per_second": 423.813
    },
    {
      "epoch": 1.287210172026926,
      "grad_norm": 2.30713152885437,
      "learning_rate": 7.140106951871658e-06,
      "loss": 0.07408469170331955,
      "num_input_tokens_seen": 3101772,
      "step": 3012,
      "train_runtime": 7318.1514,
      "train_tokens_per_second": 423.846
    },
    {
      "epoch": 1.2876375681162517,
      "grad_norm": 2.4536685943603516,
      "learning_rate": 7.135828877005348e-06,
      "loss": 0.08536134660243988,
      "num_input_tokens_seen": 3102906,
      "step": 3013,
      "train_runtime": 7320.2379,
      "train_tokens_per_second": 423.88
    },
    {
      "epoch": 1.2880649642055775,
      "grad_norm": 3.521406412124634,
      "learning_rate": 7.131550802139038e-06,
      "loss": 0.10339319705963135,
      "num_input_tokens_seen": 3103976,
      "step": 3014,
      "train_runtime": 7322.3341,
      "train_tokens_per_second": 423.905
    },
    {
      "epoch": 1.2884923602949034,
      "grad_norm": 2.5345664024353027,
      "learning_rate": 7.127272727272728e-06,
      "loss": 0.10013153403997421,
      "num_input_tokens_seen": 3104932,
      "step": 3015,
      "train_runtime": 7324.401,
      "train_tokens_per_second": 423.916
    },
    {
      "epoch": 1.288919756384229,
      "grad_norm": 2.035928249359131,
      "learning_rate": 7.122994652406417e-06,
      "loss": 0.08597824722528458,
      "num_input_tokens_seen": 3105924,
      "step": 3016,
      "train_runtime": 7326.4888,
      "train_tokens_per_second": 423.931
    },
    {
      "epoch": 1.289347152473555,
      "grad_norm": 3.1361629962921143,
      "learning_rate": 7.118716577540108e-06,
      "loss": 0.11318005621433258,
      "num_input_tokens_seen": 3106784,
      "step": 3017,
      "train_runtime": 7328.5533,
      "train_tokens_per_second": 423.929
    },
    {
      "epoch": 1.2897745485628807,
      "grad_norm": 2.8527190685272217,
      "learning_rate": 7.114438502673797e-06,
      "loss": 0.09566424041986465,
      "num_input_tokens_seen": 3107672,
      "step": 3018,
      "train_runtime": 7330.6137,
      "train_tokens_per_second": 423.931
    },
    {
      "epoch": 1.2902019446522064,
      "grad_norm": 2.5158822536468506,
      "learning_rate": 7.110160427807487e-06,
      "loss": 0.07096898555755615,
      "num_input_tokens_seen": 3108620,
      "step": 3019,
      "train_runtime": 7332.6776,
      "train_tokens_per_second": 423.941
    },
    {
      "epoch": 1.2906293407415323,
      "grad_norm": 2.812882661819458,
      "learning_rate": 7.105882352941177e-06,
      "loss": 0.10749098658561707,
      "num_input_tokens_seen": 3109912,
      "step": 3020,
      "train_runtime": 7334.821,
      "train_tokens_per_second": 423.993
    },
    {
      "epoch": 1.2910567368308579,
      "grad_norm": 2.430882692337036,
      "learning_rate": 7.101604278074867e-06,
      "loss": 0.07211394608020782,
      "num_input_tokens_seen": 3110904,
      "step": 3021,
      "train_runtime": 7336.8693,
      "train_tokens_per_second": 424.01
    },
    {
      "epoch": 1.2914841329201838,
      "grad_norm": 2.4072933197021484,
      "learning_rate": 7.0973262032085565e-06,
      "loss": 0.08512874692678452,
      "num_input_tokens_seen": 3111826,
      "step": 3022,
      "train_runtime": 7338.9381,
      "train_tokens_per_second": 424.016
    },
    {
      "epoch": 1.2919115290095096,
      "grad_norm": 2.674919366836548,
      "learning_rate": 7.093048128342247e-06,
      "loss": 0.06882139295339584,
      "num_input_tokens_seen": 3112546,
      "step": 3023,
      "train_runtime": 7340.9654,
      "train_tokens_per_second": 423.997
    },
    {
      "epoch": 1.2923389250988353,
      "grad_norm": 2.1940455436706543,
      "learning_rate": 7.088770053475936e-06,
      "loss": 0.053770191967487335,
      "num_input_tokens_seen": 3113880,
      "step": 3024,
      "train_runtime": 7343.1515,
      "train_tokens_per_second": 424.052
    },
    {
      "epoch": 1.2927663211881613,
      "grad_norm": 2.0841023921966553,
      "learning_rate": 7.0844919786096266e-06,
      "loss": 0.06542058289051056,
      "num_input_tokens_seen": 3115018,
      "step": 3025,
      "train_runtime": 7345.3787,
      "train_tokens_per_second": 424.079
    },
    {
      "epoch": 1.2931937172774868,
      "grad_norm": 1.4411394596099854,
      "learning_rate": 7.080213903743316e-06,
      "loss": 0.044625215232372284,
      "num_input_tokens_seen": 3116772,
      "step": 3026,
      "train_runtime": 7347.7206,
      "train_tokens_per_second": 424.182
    },
    {
      "epoch": 1.2936211133668127,
      "grad_norm": 2.1820149421691895,
      "learning_rate": 7.075935828877006e-06,
      "loss": 0.06833451241254807,
      "num_input_tokens_seen": 3117590,
      "step": 3027,
      "train_runtime": 7349.7332,
      "train_tokens_per_second": 424.177
    },
    {
      "epoch": 1.2940485094561385,
      "grad_norm": 2.389446258544922,
      "learning_rate": 7.071657754010696e-06,
      "loss": 0.10026880353689194,
      "num_input_tokens_seen": 3118734,
      "step": 3028,
      "train_runtime": 7351.8677,
      "train_tokens_per_second": 424.21
    },
    {
      "epoch": 1.2944759055454642,
      "grad_norm": 2.3412301540374756,
      "learning_rate": 7.067379679144386e-06,
      "loss": 0.07555332779884338,
      "num_input_tokens_seen": 3119610,
      "step": 3029,
      "train_runtime": 7353.9259,
      "train_tokens_per_second": 424.21
    },
    {
      "epoch": 1.2949033016347902,
      "grad_norm": 2.6127512454986572,
      "learning_rate": 7.0631016042780755e-06,
      "loss": 0.0979684367775917,
      "num_input_tokens_seen": 3120986,
      "step": 3030,
      "train_runtime": 7356.1024,
      "train_tokens_per_second": 424.272
    },
    {
      "epoch": 1.2953306977241157,
      "grad_norm": 1.8076595067977905,
      "learning_rate": 7.058823529411766e-06,
      "loss": 0.05226100981235504,
      "num_input_tokens_seen": 3122304,
      "step": 3031,
      "train_runtime": 7364.9665,
      "train_tokens_per_second": 423.94
    },
    {
      "epoch": 1.2957580938134416,
      "grad_norm": 2.1378977298736572,
      "learning_rate": 7.054545454545455e-06,
      "loss": 0.0676194578409195,
      "num_input_tokens_seen": 3123400,
      "step": 3032,
      "train_runtime": 7367.1523,
      "train_tokens_per_second": 423.963
    },
    {
      "epoch": 1.2961854899027674,
      "grad_norm": 2.258054256439209,
      "learning_rate": 7.0502673796791455e-06,
      "loss": 0.07752581685781479,
      "num_input_tokens_seen": 3124474,
      "step": 3033,
      "train_runtime": 7369.3583,
      "train_tokens_per_second": 423.982
    },
    {
      "epoch": 1.2966128859920931,
      "grad_norm": 2.367128849029541,
      "learning_rate": 7.045989304812835e-06,
      "loss": 0.07573604583740234,
      "num_input_tokens_seen": 3125546,
      "step": 3034,
      "train_runtime": 7371.4422,
      "train_tokens_per_second": 424.007
    },
    {
      "epoch": 1.297040282081419,
      "grad_norm": 2.9613142013549805,
      "learning_rate": 7.041711229946525e-06,
      "loss": 0.09954195469617844,
      "num_input_tokens_seen": 3126568,
      "step": 3035,
      "train_runtime": 7373.524,
      "train_tokens_per_second": 424.026
    },
    {
      "epoch": 1.2974676781707448,
      "grad_norm": 3.532851219177246,
      "learning_rate": 7.037433155080215e-06,
      "loss": 0.09093168377876282,
      "num_input_tokens_seen": 3127474,
      "step": 3036,
      "train_runtime": 7375.6051,
      "train_tokens_per_second": 424.029
    },
    {
      "epoch": 1.2978950742600706,
      "grad_norm": 3.328404426574707,
      "learning_rate": 7.033155080213905e-06,
      "loss": 0.1490725725889206,
      "num_input_tokens_seen": 3128286,
      "step": 3037,
      "train_runtime": 7377.6712,
      "train_tokens_per_second": 424.021
    },
    {
      "epoch": 1.2983224703493963,
      "grad_norm": 1.9930859804153442,
      "learning_rate": 7.0288770053475944e-06,
      "loss": 0.071568563580513,
      "num_input_tokens_seen": 3129562,
      "step": 3038,
      "train_runtime": 7379.8758,
      "train_tokens_per_second": 424.067
    },
    {
      "epoch": 1.298749866438722,
      "grad_norm": 3.499483108520508,
      "learning_rate": 7.024598930481285e-06,
      "loss": 0.08723960816860199,
      "num_input_tokens_seen": 3130352,
      "step": 3039,
      "train_runtime": 7381.9143,
      "train_tokens_per_second": 424.057
    },
    {
      "epoch": 1.299177262528048,
      "grad_norm": 2.7622833251953125,
      "learning_rate": 7.020320855614974e-06,
      "loss": 0.09654998034238815,
      "num_input_tokens_seen": 3131122,
      "step": 3040,
      "train_runtime": 7383.9449,
      "train_tokens_per_second": 424.045
    },
    {
      "epoch": 1.2996046586173737,
      "grad_norm": 2.9033401012420654,
      "learning_rate": 7.0160427807486645e-06,
      "loss": 0.09674963355064392,
      "num_input_tokens_seen": 3132330,
      "step": 3041,
      "train_runtime": 7386.1516,
      "train_tokens_per_second": 424.081
    },
    {
      "epoch": 1.3000320547066995,
      "grad_norm": 2.604142189025879,
      "learning_rate": 7.011764705882353e-06,
      "loss": 0.09368482977151871,
      "num_input_tokens_seen": 3133322,
      "step": 3042,
      "train_runtime": 7388.2946,
      "train_tokens_per_second": 424.093
    },
    {
      "epoch": 1.3004594507960252,
      "grad_norm": 2.385263681411743,
      "learning_rate": 7.007486631016044e-06,
      "loss": 0.10400746017694473,
      "num_input_tokens_seen": 3134570,
      "step": 3043,
      "train_runtime": 7390.4408,
      "train_tokens_per_second": 424.138
    },
    {
      "epoch": 1.300886846885351,
      "grad_norm": 2.639981508255005,
      "learning_rate": 7.003208556149733e-06,
      "loss": 0.13504458963871002,
      "num_input_tokens_seen": 3135918,
      "step": 3044,
      "train_runtime": 7392.6324,
      "train_tokens_per_second": 424.195
    },
    {
      "epoch": 1.301314242974677,
      "grad_norm": 2.1733832359313965,
      "learning_rate": 6.998930481283424e-06,
      "loss": 0.06582202762365341,
      "num_input_tokens_seen": 3137060,
      "step": 3045,
      "train_runtime": 7394.7576,
      "train_tokens_per_second": 424.228
    },
    {
      "epoch": 1.3017416390640026,
      "grad_norm": 3.159576416015625,
      "learning_rate": 6.9946524064171125e-06,
      "loss": 0.07929767668247223,
      "num_input_tokens_seen": 3137722,
      "step": 3046,
      "train_runtime": 7396.8436,
      "train_tokens_per_second": 424.197
    },
    {
      "epoch": 1.3021690351533284,
      "grad_norm": 1.9610964059829712,
      "learning_rate": 6.990374331550803e-06,
      "loss": 0.0629386231303215,
      "num_input_tokens_seen": 3138922,
      "step": 3047,
      "train_runtime": 7398.9644,
      "train_tokens_per_second": 424.238
    },
    {
      "epoch": 1.3025964312426541,
      "grad_norm": 2.345449209213257,
      "learning_rate": 6.986096256684492e-06,
      "loss": 0.07290294021368027,
      "num_input_tokens_seen": 3139788,
      "step": 3048,
      "train_runtime": 7401.0418,
      "train_tokens_per_second": 424.236
    },
    {
      "epoch": 1.3030238273319799,
      "grad_norm": 2.531792640686035,
      "learning_rate": 6.981818181818183e-06,
      "loss": 0.05829806998372078,
      "num_input_tokens_seen": 3140678,
      "step": 3049,
      "train_runtime": 7403.1163,
      "train_tokens_per_second": 424.237
    },
    {
      "epoch": 1.3034512234213058,
      "grad_norm": 1.87013840675354,
      "learning_rate": 6.977540106951872e-06,
      "loss": 0.0795431062579155,
      "num_input_tokens_seen": 3142044,
      "step": 3050,
      "train_runtime": 7405.2994,
      "train_tokens_per_second": 424.297
    },
    {
      "epoch": 1.3038786195106316,
      "grad_norm": 2.370725393295288,
      "learning_rate": 6.9732620320855615e-06,
      "loss": 0.0632067620754242,
      "num_input_tokens_seen": 3143798,
      "step": 3051,
      "train_runtime": 7407.5542,
      "train_tokens_per_second": 424.404
    },
    {
      "epoch": 1.3043060155999573,
      "grad_norm": 2.9113504886627197,
      "learning_rate": 6.968983957219252e-06,
      "loss": 0.042246293276548386,
      "num_input_tokens_seen": 3144388,
      "step": 3052,
      "train_runtime": 7409.5774,
      "train_tokens_per_second": 424.368
    },
    {
      "epoch": 1.304733411689283,
      "grad_norm": 4.423725605010986,
      "learning_rate": 6.964705882352941e-06,
      "loss": 0.06097486615180969,
      "num_input_tokens_seen": 3145058,
      "step": 3053,
      "train_runtime": 7411.5962,
      "train_tokens_per_second": 424.343
    },
    {
      "epoch": 1.3051608077786088,
      "grad_norm": 2.5406594276428223,
      "learning_rate": 6.9604278074866315e-06,
      "loss": 0.07264570146799088,
      "num_input_tokens_seen": 3146128,
      "step": 3054,
      "train_runtime": 7413.6958,
      "train_tokens_per_second": 424.367
    },
    {
      "epoch": 1.3055882038679347,
      "grad_norm": 2.289741039276123,
      "learning_rate": 6.956149732620321e-06,
      "loss": 0.05240107327699661,
      "num_input_tokens_seen": 3146998,
      "step": 3055,
      "train_runtime": 7415.719,
      "train_tokens_per_second": 424.369
    },
    {
      "epoch": 1.3060155999572605,
      "grad_norm": 1.955268383026123,
      "learning_rate": 6.951871657754011e-06,
      "loss": 0.06604617834091187,
      "num_input_tokens_seen": 3148164,
      "step": 3056,
      "train_runtime": 7417.8684,
      "train_tokens_per_second": 424.403
    },
    {
      "epoch": 1.3064429960465862,
      "grad_norm": 2.243042469024658,
      "learning_rate": 6.947593582887701e-06,
      "loss": 0.08481141924858093,
      "num_input_tokens_seen": 3149206,
      "step": 3057,
      "train_runtime": 7419.9563,
      "train_tokens_per_second": 424.424
    },
    {
      "epoch": 1.306870392135912,
      "grad_norm": 2.78591251373291,
      "learning_rate": 6.943315508021391e-06,
      "loss": 0.10786169767379761,
      "num_input_tokens_seen": 3150110,
      "step": 3058,
      "train_runtime": 7422.011,
      "train_tokens_per_second": 424.428
    },
    {
      "epoch": 1.3072977882252377,
      "grad_norm": 2.2791383266448975,
      "learning_rate": 6.9390374331550804e-06,
      "loss": 0.08200730383396149,
      "num_input_tokens_seen": 3151468,
      "step": 3059,
      "train_runtime": 7424.1611,
      "train_tokens_per_second": 424.488
    },
    {
      "epoch": 1.3077251843145636,
      "grad_norm": 2.2670657634735107,
      "learning_rate": 6.934759358288771e-06,
      "loss": 0.07363367825746536,
      "num_input_tokens_seen": 3152410,
      "step": 3060,
      "train_runtime": 7426.2034,
      "train_tokens_per_second": 424.498
    },
    {
      "epoch": 1.3081525804038894,
      "grad_norm": 2.2953970432281494,
      "learning_rate": 6.93048128342246e-06,
      "loss": 0.04841242730617523,
      "num_input_tokens_seen": 3153580,
      "step": 3061,
      "train_runtime": 7435.0197,
      "train_tokens_per_second": 424.152
    },
    {
      "epoch": 1.3085799764932151,
      "grad_norm": 2.86474347114563,
      "learning_rate": 6.9262032085561505e-06,
      "loss": 0.049897633492946625,
      "num_input_tokens_seen": 3154504,
      "step": 3062,
      "train_runtime": 7437.2679,
      "train_tokens_per_second": 424.148
    },
    {
      "epoch": 1.3090073725825409,
      "grad_norm": 2.233349084854126,
      "learning_rate": 6.92192513368984e-06,
      "loss": 0.11106890439987183,
      "num_input_tokens_seen": 3156300,
      "step": 3063,
      "train_runtime": 7439.5169,
      "train_tokens_per_second": 424.261
    },
    {
      "epoch": 1.3094347686718666,
      "grad_norm": 3.18144154548645,
      "learning_rate": 6.91764705882353e-06,
      "loss": 0.0804712176322937,
      "num_input_tokens_seen": 3157306,
      "step": 3064,
      "train_runtime": 7441.6268,
      "train_tokens_per_second": 424.276
    },
    {
      "epoch": 1.3098621647611925,
      "grad_norm": 2.4771389961242676,
      "learning_rate": 6.91336898395722e-06,
      "loss": 0.05247760936617851,
      "num_input_tokens_seen": 3157980,
      "step": 3065,
      "train_runtime": 7443.6084,
      "train_tokens_per_second": 424.254
    },
    {
      "epoch": 1.3102895608505183,
      "grad_norm": 2.279043197631836,
      "learning_rate": 6.90909090909091e-06,
      "loss": 0.04659748449921608,
      "num_input_tokens_seen": 3158662,
      "step": 3066,
      "train_runtime": 7445.6018,
      "train_tokens_per_second": 424.232
    },
    {
      "epoch": 1.310716956939844,
      "grad_norm": 2.101201057434082,
      "learning_rate": 6.904812834224599e-06,
      "loss": 0.06005648523569107,
      "num_input_tokens_seen": 3159780,
      "step": 3067,
      "train_runtime": 7447.7474,
      "train_tokens_per_second": 424.26
    },
    {
      "epoch": 1.3111443530291698,
      "grad_norm": 3.5098633766174316,
      "learning_rate": 6.90053475935829e-06,
      "loss": 0.12493811547756195,
      "num_input_tokens_seen": 3160762,
      "step": 3068,
      "train_runtime": 7449.8496,
      "train_tokens_per_second": 424.272
    },
    {
      "epoch": 1.3115717491184955,
      "grad_norm": 2.609753370285034,
      "learning_rate": 6.896256684491979e-06,
      "loss": 0.10243531316518784,
      "num_input_tokens_seen": 3161672,
      "step": 3069,
      "train_runtime": 7451.8958,
      "train_tokens_per_second": 424.278
    },
    {
      "epoch": 1.3119991452078215,
      "grad_norm": 2.270583152770996,
      "learning_rate": 6.891978609625669e-06,
      "loss": 0.11233691871166229,
      "num_input_tokens_seen": 3163168,
      "step": 3070,
      "train_runtime": 7454.0997,
      "train_tokens_per_second": 424.353
    },
    {
      "epoch": 1.3124265412971472,
      "grad_norm": 2.7746379375457764,
      "learning_rate": 6.887700534759358e-06,
      "loss": 0.11023058742284775,
      "num_input_tokens_seen": 3164030,
      "step": 3071,
      "train_runtime": 7456.1817,
      "train_tokens_per_second": 424.35
    },
    {
      "epoch": 1.312853937386473,
      "grad_norm": 2.94696044921875,
      "learning_rate": 6.883422459893049e-06,
      "loss": 0.1009884849190712,
      "num_input_tokens_seen": 3164998,
      "step": 3072,
      "train_runtime": 7458.2548,
      "train_tokens_per_second": 424.362
    },
    {
      "epoch": 1.3132813334757987,
      "grad_norm": 4.1987810134887695,
      "learning_rate": 6.879144385026738e-06,
      "loss": 0.08677253127098083,
      "num_input_tokens_seen": 3165648,
      "step": 3073,
      "train_runtime": 7460.3047,
      "train_tokens_per_second": 424.332
    },
    {
      "epoch": 1.3137087295651244,
      "grad_norm": 2.8691184520721436,
      "learning_rate": 6.874866310160429e-06,
      "loss": 0.10081765055656433,
      "num_input_tokens_seen": 3166990,
      "step": 3074,
      "train_runtime": 7462.46,
      "train_tokens_per_second": 424.39
    },
    {
      "epoch": 1.3141361256544504,
      "grad_norm": 2.11684513092041,
      "learning_rate": 6.8705882352941175e-06,
      "loss": 0.07259511947631836,
      "num_input_tokens_seen": 3168236,
      "step": 3075,
      "train_runtime": 7464.6439,
      "train_tokens_per_second": 424.432
    },
    {
      "epoch": 1.314563521743776,
      "grad_norm": 2.465240955352783,
      "learning_rate": 6.866310160427808e-06,
      "loss": 0.06386428326368332,
      "num_input_tokens_seen": 3168898,
      "step": 3076,
      "train_runtime": 7466.714,
      "train_tokens_per_second": 424.403
    },
    {
      "epoch": 1.3149909178331018,
      "grad_norm": 2.7865781784057617,
      "learning_rate": 6.862032085561497e-06,
      "loss": 0.05655631050467491,
      "num_input_tokens_seen": 3169736,
      "step": 3077,
      "train_runtime": 7468.7832,
      "train_tokens_per_second": 424.398
    },
    {
      "epoch": 1.3154183139224276,
      "grad_norm": 3.729416608810425,
      "learning_rate": 6.8577540106951875e-06,
      "loss": 0.09011790156364441,
      "num_input_tokens_seen": 3170618,
      "step": 3078,
      "train_runtime": 7470.8087,
      "train_tokens_per_second": 424.401
    },
    {
      "epoch": 1.3158457100117533,
      "grad_norm": 3.140462636947632,
      "learning_rate": 6.853475935828877e-06,
      "loss": 0.1120164692401886,
      "num_input_tokens_seen": 3171422,
      "step": 3079,
      "train_runtime": 7472.884,
      "train_tokens_per_second": 424.391
    },
    {
      "epoch": 1.3162731061010793,
      "grad_norm": 2.1219489574432373,
      "learning_rate": 6.849197860962567e-06,
      "loss": 0.06099581718444824,
      "num_input_tokens_seen": 3172372,
      "step": 3080,
      "train_runtime": 7474.9819,
      "train_tokens_per_second": 424.399
    },
    {
      "epoch": 1.316700502190405,
      "grad_norm": 2.388225555419922,
      "learning_rate": 6.844919786096257e-06,
      "loss": 0.07526946067810059,
      "num_input_tokens_seen": 3173228,
      "step": 3081,
      "train_runtime": 7477.0284,
      "train_tokens_per_second": 424.397
    },
    {
      "epoch": 1.3171278982797308,
      "grad_norm": 2.454058885574341,
      "learning_rate": 6.840641711229947e-06,
      "loss": 0.0884820967912674,
      "num_input_tokens_seen": 3174322,
      "step": 3082,
      "train_runtime": 7479.1303,
      "train_tokens_per_second": 424.424
    },
    {
      "epoch": 1.3175552943690565,
      "grad_norm": 3.4711782932281494,
      "learning_rate": 6.8363636363636364e-06,
      "loss": 0.06224628910422325,
      "num_input_tokens_seen": 3175220,
      "step": 3083,
      "train_runtime": 7481.2066,
      "train_tokens_per_second": 424.426
    },
    {
      "epoch": 1.3179826904583822,
      "grad_norm": 4.044619083404541,
      "learning_rate": 6.832085561497327e-06,
      "loss": 0.13691066205501556,
      "num_input_tokens_seen": 3176010,
      "step": 3084,
      "train_runtime": 7483.3093,
      "train_tokens_per_second": 424.412
    },
    {
      "epoch": 1.3184100865477082,
      "grad_norm": 2.34553861618042,
      "learning_rate": 6.827807486631016e-06,
      "loss": 0.07405084371566772,
      "num_input_tokens_seen": 3176834,
      "step": 3085,
      "train_runtime": 7485.3985,
      "train_tokens_per_second": 424.404
    },
    {
      "epoch": 1.318837482637034,
      "grad_norm": 2.4816157817840576,
      "learning_rate": 6.8235294117647065e-06,
      "loss": 0.052815310657024384,
      "num_input_tokens_seen": 3177836,
      "step": 3086,
      "train_runtime": 7487.4831,
      "train_tokens_per_second": 424.42
    },
    {
      "epoch": 1.3192648787263597,
      "grad_norm": 2.9102537631988525,
      "learning_rate": 6.819251336898396e-06,
      "loss": 0.10762530565261841,
      "num_input_tokens_seen": 3178868,
      "step": 3087,
      "train_runtime": 7489.5586,
      "train_tokens_per_second": 424.44
    },
    {
      "epoch": 1.3196922748156854,
      "grad_norm": 3.6969783306121826,
      "learning_rate": 6.814973262032086e-06,
      "loss": 0.13340641558170319,
      "num_input_tokens_seen": 3179948,
      "step": 3088,
      "train_runtime": 7491.6729,
      "train_tokens_per_second": 424.464
    },
    {
      "epoch": 1.3201196709050111,
      "grad_norm": 2.1342952251434326,
      "learning_rate": 6.810695187165776e-06,
      "loss": 0.062183134257793427,
      "num_input_tokens_seen": 3180812,
      "step": 3089,
      "train_runtime": 7493.688,
      "train_tokens_per_second": 424.465
    },
    {
      "epoch": 1.320547066994337,
      "grad_norm": 6.565427303314209,
      "learning_rate": 6.806417112299466e-06,
      "loss": 0.12299738824367523,
      "num_input_tokens_seen": 3181614,
      "step": 3090,
      "train_runtime": 7495.7337,
      "train_tokens_per_second": 424.457
    },
    {
      "epoch": 1.3209744630836628,
      "grad_norm": 1.5610580444335938,
      "learning_rate": 6.802139037433155e-06,
      "loss": 0.047917917370796204,
      "num_input_tokens_seen": 3183208,
      "step": 3091,
      "train_runtime": 7504.6671,
      "train_tokens_per_second": 424.164
    },
    {
      "epoch": 1.3214018591729886,
      "grad_norm": 2.8131675720214844,
      "learning_rate": 6.797860962566846e-06,
      "loss": 0.07366964966058731,
      "num_input_tokens_seen": 3184086,
      "step": 3092,
      "train_runtime": 7506.9263,
      "train_tokens_per_second": 424.153
    },
    {
      "epoch": 1.3218292552623143,
      "grad_norm": 2.299919843673706,
      "learning_rate": 6.793582887700535e-06,
      "loss": 0.07928738743066788,
      "num_input_tokens_seen": 3184846,
      "step": 3093,
      "train_runtime": 7508.9698,
      "train_tokens_per_second": 424.139
    },
    {
      "epoch": 1.32225665135164,
      "grad_norm": 2.5078322887420654,
      "learning_rate": 6.7893048128342254e-06,
      "loss": 0.08887529373168945,
      "num_input_tokens_seen": 3186240,
      "step": 3094,
      "train_runtime": 7511.2524,
      "train_tokens_per_second": 424.196
    },
    {
      "epoch": 1.322684047440966,
      "grad_norm": 1.9447228908538818,
      "learning_rate": 6.785026737967915e-06,
      "loss": 0.08179621398448944,
      "num_input_tokens_seen": 3187796,
      "step": 3095,
      "train_runtime": 7513.474,
      "train_tokens_per_second": 424.277
    },
    {
      "epoch": 1.3231114435302918,
      "grad_norm": 3.1969754695892334,
      "learning_rate": 6.780748663101605e-06,
      "loss": 0.08298946917057037,
      "num_input_tokens_seen": 3189332,
      "step": 3096,
      "train_runtime": 7515.7048,
      "train_tokens_per_second": 424.356
    },
    {
      "epoch": 1.3235388396196175,
      "grad_norm": 2.5398192405700684,
      "learning_rate": 6.776470588235295e-06,
      "loss": 0.09162995219230652,
      "num_input_tokens_seen": 3190378,
      "step": 3097,
      "train_runtime": 7517.8576,
      "train_tokens_per_second": 424.373
    },
    {
      "epoch": 1.3239662357089432,
      "grad_norm": 2.6871464252471924,
      "learning_rate": 6.772192513368985e-06,
      "loss": 0.05782339349389076,
      "num_input_tokens_seen": 3191108,
      "step": 3098,
      "train_runtime": 7519.887,
      "train_tokens_per_second": 424.356
    },
    {
      "epoch": 1.324393631798269,
      "grad_norm": 2.4937820434570312,
      "learning_rate": 6.767914438502674e-06,
      "loss": 0.09907757490873337,
      "num_input_tokens_seen": 3192432,
      "step": 3099,
      "train_runtime": 7522.0524,
      "train_tokens_per_second": 424.41
    },
    {
      "epoch": 1.324821027887595,
      "grad_norm": 3.5931575298309326,
      "learning_rate": 6.763636363636365e-06,
      "loss": 0.1524689793586731,
      "num_input_tokens_seen": 3193192,
      "step": 3100,
      "train_runtime": 7524.106,
      "train_tokens_per_second": 424.395
    },
    {
      "epoch": 1.3252484239769207,
      "grad_norm": 3.1266043186187744,
      "learning_rate": 6.759358288770054e-06,
      "loss": 0.06912095844745636,
      "num_input_tokens_seen": 3193864,
      "step": 3101,
      "train_runtime": 7526.1934,
      "train_tokens_per_second": 424.366
    },
    {
      "epoch": 1.3256758200662464,
      "grad_norm": 1.976914405822754,
      "learning_rate": 6.755080213903744e-06,
      "loss": 0.07313459366559982,
      "num_input_tokens_seen": 3195204,
      "step": 3102,
      "train_runtime": 7528.4074,
      "train_tokens_per_second": 424.42
    },
    {
      "epoch": 1.3261032161555721,
      "grad_norm": 3.273916482925415,
      "learning_rate": 6.750802139037434e-06,
      "loss": 0.09610220789909363,
      "num_input_tokens_seen": 3196008,
      "step": 3103,
      "train_runtime": 7530.4705,
      "train_tokens_per_second": 424.41
    },
    {
      "epoch": 1.3265306122448979,
      "grad_norm": 2.6544153690338135,
      "learning_rate": 6.746524064171124e-06,
      "loss": 0.06252476572990417,
      "num_input_tokens_seen": 3197224,
      "step": 3104,
      "train_runtime": 7532.6532,
      "train_tokens_per_second": 424.449
    },
    {
      "epoch": 1.3269580083342238,
      "grad_norm": 2.23612642288208,
      "learning_rate": 6.742245989304813e-06,
      "loss": 0.050831139087677,
      "num_input_tokens_seen": 3198530,
      "step": 3105,
      "train_runtime": 7534.8636,
      "train_tokens_per_second": 424.497
    },
    {
      "epoch": 1.3273854044235496,
      "grad_norm": 2.608036994934082,
      "learning_rate": 6.737967914438504e-06,
      "loss": 0.07467533648014069,
      "num_input_tokens_seen": 3199426,
      "step": 3106,
      "train_runtime": 7536.9543,
      "train_tokens_per_second": 424.499
    },
    {
      "epoch": 1.3278128005128753,
      "grad_norm": 2.664032459259033,
      "learning_rate": 6.7336898395721925e-06,
      "loss": 0.08931900560855865,
      "num_input_tokens_seen": 3200508,
      "step": 3107,
      "train_runtime": 7539.0648,
      "train_tokens_per_second": 424.523
    },
    {
      "epoch": 1.328240196602201,
      "grad_norm": 2.920459747314453,
      "learning_rate": 6.729411764705884e-06,
      "loss": 0.12720629572868347,
      "num_input_tokens_seen": 3201420,
      "step": 3108,
      "train_runtime": 7541.1838,
      "train_tokens_per_second": 424.525
    },
    {
      "epoch": 1.3286675926915268,
      "grad_norm": 2.4971139430999756,
      "learning_rate": 6.725133689839572e-06,
      "loss": 0.06938755512237549,
      "num_input_tokens_seen": 3202332,
      "step": 3109,
      "train_runtime": 7543.326,
      "train_tokens_per_second": 424.525
    },
    {
      "epoch": 1.3290949887808527,
      "grad_norm": 2.570857048034668,
      "learning_rate": 6.7208556149732625e-06,
      "loss": 0.09026563167572021,
      "num_input_tokens_seen": 3203396,
      "step": 3110,
      "train_runtime": 7545.4445,
      "train_tokens_per_second": 424.547
    },
    {
      "epoch": 1.3295223848701785,
      "grad_norm": 3.0758118629455566,
      "learning_rate": 6.716577540106952e-06,
      "loss": 0.10461202263832092,
      "num_input_tokens_seen": 3204304,
      "step": 3111,
      "train_runtime": 7547.632,
      "train_tokens_per_second": 424.544
    },
    {
      "epoch": 1.3299497809595042,
      "grad_norm": 4.081035137176514,
      "learning_rate": 6.712299465240642e-06,
      "loss": 0.12959963083267212,
      "num_input_tokens_seen": 3205074,
      "step": 3112,
      "train_runtime": 7549.6602,
      "train_tokens_per_second": 424.532
    },
    {
      "epoch": 1.33037717704883,
      "grad_norm": 3.0498290061950684,
      "learning_rate": 6.708021390374332e-06,
      "loss": 0.0922926515340805,
      "num_input_tokens_seen": 3206216,
      "step": 3113,
      "train_runtime": 7551.8489,
      "train_tokens_per_second": 424.56
    },
    {
      "epoch": 1.3308045731381557,
      "grad_norm": 3.1709518432617188,
      "learning_rate": 6.703743315508022e-06,
      "loss": 0.11085475236177444,
      "num_input_tokens_seen": 3207130,
      "step": 3114,
      "train_runtime": 7553.9126,
      "train_tokens_per_second": 424.565
    },
    {
      "epoch": 1.3312319692274817,
      "grad_norm": 3.0828921794891357,
      "learning_rate": 6.699465240641711e-06,
      "loss": 0.08031412959098816,
      "num_input_tokens_seen": 3207922,
      "step": 3115,
      "train_runtime": 7555.9127,
      "train_tokens_per_second": 424.558
    },
    {
      "epoch": 1.3316593653168074,
      "grad_norm": 6.81247091293335,
      "learning_rate": 6.695187165775402e-06,
      "loss": 0.07905543595552444,
      "num_input_tokens_seen": 3209014,
      "step": 3116,
      "train_runtime": 7558.0246,
      "train_tokens_per_second": 424.584
    },
    {
      "epoch": 1.3320867614061331,
      "grad_norm": 4.130695343017578,
      "learning_rate": 6.690909090909091e-06,
      "loss": 0.1136293113231659,
      "num_input_tokens_seen": 3210022,
      "step": 3117,
      "train_runtime": 7560.1119,
      "train_tokens_per_second": 424.6
    },
    {
      "epoch": 1.3325141574954589,
      "grad_norm": 2.7064762115478516,
      "learning_rate": 6.6866310160427814e-06,
      "loss": 0.0745459794998169,
      "num_input_tokens_seen": 3210888,
      "step": 3118,
      "train_runtime": 7562.2223,
      "train_tokens_per_second": 424.596
    },
    {
      "epoch": 1.3329415535847846,
      "grad_norm": 2.6442806720733643,
      "learning_rate": 6.682352941176471e-06,
      "loss": 0.08416484296321869,
      "num_input_tokens_seen": 3211746,
      "step": 3119,
      "train_runtime": 7564.2535,
      "train_tokens_per_second": 424.595
    },
    {
      "epoch": 1.3333689496741106,
      "grad_norm": 1.8397011756896973,
      "learning_rate": 6.678074866310161e-06,
      "loss": 0.06384040415287018,
      "num_input_tokens_seen": 3213106,
      "step": 3120,
      "train_runtime": 7566.5208,
      "train_tokens_per_second": 424.648
    },
    {
      "epoch": 1.3337963457634363,
      "grad_norm": 2.006704092025757,
      "learning_rate": 6.673796791443851e-06,
      "loss": 0.062374405562877655,
      "num_input_tokens_seen": 3214000,
      "step": 3121,
      "train_runtime": 7575.7411,
      "train_tokens_per_second": 424.249
    },
    {
      "epoch": 1.334223741852762,
      "grad_norm": 2.6489882469177246,
      "learning_rate": 6.669518716577541e-06,
      "loss": 0.10289392620325089,
      "num_input_tokens_seen": 3215234,
      "step": 3122,
      "train_runtime": 7577.9407,
      "train_tokens_per_second": 424.289
    },
    {
      "epoch": 1.3346511379420878,
      "grad_norm": 3.345651626586914,
      "learning_rate": 6.66524064171123e-06,
      "loss": 0.102023646235466,
      "num_input_tokens_seen": 3215954,
      "step": 3123,
      "train_runtime": 7579.9753,
      "train_tokens_per_second": 424.27
    },
    {
      "epoch": 1.3350785340314135,
      "grad_norm": 2.2507457733154297,
      "learning_rate": 6.660962566844921e-06,
      "loss": 0.1047191321849823,
      "num_input_tokens_seen": 3217424,
      "step": 3124,
      "train_runtime": 7582.1733,
      "train_tokens_per_second": 424.341
    },
    {
      "epoch": 1.3355059301207395,
      "grad_norm": 2.5344626903533936,
      "learning_rate": 6.65668449197861e-06,
      "loss": 0.07828375697135925,
      "num_input_tokens_seen": 3218498,
      "step": 3125,
      "train_runtime": 7584.2663,
      "train_tokens_per_second": 424.365
    },
    {
      "epoch": 1.3359333262100652,
      "grad_norm": 2.8161840438842773,
      "learning_rate": 6.6524064171123e-06,
      "loss": 0.08589472621679306,
      "num_input_tokens_seen": 3219368,
      "step": 3126,
      "train_runtime": 7586.3759,
      "train_tokens_per_second": 424.362
    },
    {
      "epoch": 1.336360722299391,
      "grad_norm": 4.029890537261963,
      "learning_rate": 6.64812834224599e-06,
      "loss": 0.21060766279697418,
      "num_input_tokens_seen": 3220682,
      "step": 3127,
      "train_runtime": 7588.5272,
      "train_tokens_per_second": 424.415
    },
    {
      "epoch": 1.3367881183887167,
      "grad_norm": 3.2842509746551514,
      "learning_rate": 6.64385026737968e-06,
      "loss": 0.11375979334115982,
      "num_input_tokens_seen": 3221522,
      "step": 3128,
      "train_runtime": 7590.651,
      "train_tokens_per_second": 424.407
    },
    {
      "epoch": 1.3372155144780424,
      "grad_norm": 2.2260725498199463,
      "learning_rate": 6.6395721925133696e-06,
      "loss": 0.04220890253782272,
      "num_input_tokens_seen": 3222400,
      "step": 3129,
      "train_runtime": 7592.7594,
      "train_tokens_per_second": 424.404
    },
    {
      "epoch": 1.3376429105673684,
      "grad_norm": 2.689847230911255,
      "learning_rate": 6.63529411764706e-06,
      "loss": 0.07509066164493561,
      "num_input_tokens_seen": 3223306,
      "step": 3130,
      "train_runtime": 7594.8283,
      "train_tokens_per_second": 424.408
    },
    {
      "epoch": 1.3380703066566941,
      "grad_norm": 3.381842613220215,
      "learning_rate": 6.631016042780749e-06,
      "loss": 0.11579402536153793,
      "num_input_tokens_seen": 3224116,
      "step": 3131,
      "train_runtime": 7596.8708,
      "train_tokens_per_second": 424.401
    },
    {
      "epoch": 1.3384977027460199,
      "grad_norm": 2.8190507888793945,
      "learning_rate": 6.62673796791444e-06,
      "loss": 0.07199320942163467,
      "num_input_tokens_seen": 3224880,
      "step": 3132,
      "train_runtime": 7598.8906,
      "train_tokens_per_second": 424.388
    },
    {
      "epoch": 1.3389250988353456,
      "grad_norm": 2.4699301719665527,
      "learning_rate": 6.622459893048129e-06,
      "loss": 0.0774102583527565,
      "num_input_tokens_seen": 3226060,
      "step": 3133,
      "train_runtime": 7601.0476,
      "train_tokens_per_second": 424.423
    },
    {
      "epoch": 1.3393524949246713,
      "grad_norm": 1.870215654373169,
      "learning_rate": 6.618181818181819e-06,
      "loss": 0.06685984134674072,
      "num_input_tokens_seen": 3227376,
      "step": 3134,
      "train_runtime": 7603.3535,
      "train_tokens_per_second": 424.467
    },
    {
      "epoch": 1.3397798910139973,
      "grad_norm": 2.2541487216949463,
      "learning_rate": 6.613903743315509e-06,
      "loss": 0.05152186378836632,
      "num_input_tokens_seen": 3228192,
      "step": 3135,
      "train_runtime": 7605.4052,
      "train_tokens_per_second": 424.46
    },
    {
      "epoch": 1.340207287103323,
      "grad_norm": 3.4293160438537598,
      "learning_rate": 6.609625668449199e-06,
      "loss": 0.09395182132720947,
      "num_input_tokens_seen": 3229324,
      "step": 3136,
      "train_runtime": 7607.5531,
      "train_tokens_per_second": 424.489
    },
    {
      "epoch": 1.3406346831926488,
      "grad_norm": 2.1196675300598145,
      "learning_rate": 6.6053475935828885e-06,
      "loss": 0.049595557153224945,
      "num_input_tokens_seen": 3230240,
      "step": 3137,
      "train_runtime": 7609.6069,
      "train_tokens_per_second": 424.495
    },
    {
      "epoch": 1.3410620792819745,
      "grad_norm": 2.0202667713165283,
      "learning_rate": 6.601069518716579e-06,
      "loss": 0.07142960280179977,
      "num_input_tokens_seen": 3231114,
      "step": 3138,
      "train_runtime": 7611.6673,
      "train_tokens_per_second": 424.495
    },
    {
      "epoch": 1.3414894753713003,
      "grad_norm": 1.9164057970046997,
      "learning_rate": 6.596791443850267e-06,
      "loss": 0.06770211458206177,
      "num_input_tokens_seen": 3232778,
      "step": 3139,
      "train_runtime": 7613.9101,
      "train_tokens_per_second": 424.588
    },
    {
      "epoch": 1.3419168714606262,
      "grad_norm": 2.2521464824676514,
      "learning_rate": 6.5925133689839586e-06,
      "loss": 0.0636894628405571,
      "num_input_tokens_seen": 3233844,
      "step": 3140,
      "train_runtime": 7615.9939,
      "train_tokens_per_second": 424.612
    },
    {
      "epoch": 1.342344267549952,
      "grad_norm": 2.4593846797943115,
      "learning_rate": 6.588235294117647e-06,
      "loss": 0.07364197075366974,
      "num_input_tokens_seen": 3234692,
      "step": 3141,
      "train_runtime": 7618.0169,
      "train_tokens_per_second": 424.611
    },
    {
      "epoch": 1.3427716636392777,
      "grad_norm": 3.288740396499634,
      "learning_rate": 6.583957219251338e-06,
      "loss": 0.09747792035341263,
      "num_input_tokens_seen": 3235608,
      "step": 3142,
      "train_runtime": 7620.1,
      "train_tokens_per_second": 424.615
    },
    {
      "epoch": 1.3431990597286034,
      "grad_norm": 2.2910284996032715,
      "learning_rate": 6.579679144385027e-06,
      "loss": 0.07498076558113098,
      "num_input_tokens_seen": 3236522,
      "step": 3143,
      "train_runtime": 7622.1882,
      "train_tokens_per_second": 424.618
    },
    {
      "epoch": 1.3436264558179292,
      "grad_norm": 2.2119545936584473,
      "learning_rate": 6.575401069518717e-06,
      "loss": 0.06378357112407684,
      "num_input_tokens_seen": 3237140,
      "step": 3144,
      "train_runtime": 7624.18,
      "train_tokens_per_second": 424.589
    },
    {
      "epoch": 1.3440538519072551,
      "grad_norm": 3.214130163192749,
      "learning_rate": 6.571122994652407e-06,
      "loss": 0.08002831786870956,
      "num_input_tokens_seen": 3237814,
      "step": 3145,
      "train_runtime": 7626.2812,
      "train_tokens_per_second": 424.56
    },
    {
      "epoch": 1.3444812479965809,
      "grad_norm": 3.826883316040039,
      "learning_rate": 6.566844919786097e-06,
      "loss": 0.12311968207359314,
      "num_input_tokens_seen": 3238762,
      "step": 3146,
      "train_runtime": 7628.3517,
      "train_tokens_per_second": 424.569
    },
    {
      "epoch": 1.3449086440859066,
      "grad_norm": 4.192801475524902,
      "learning_rate": 6.562566844919786e-06,
      "loss": 0.10599485039710999,
      "num_input_tokens_seen": 3239436,
      "step": 3147,
      "train_runtime": 7630.3393,
      "train_tokens_per_second": 424.547
    },
    {
      "epoch": 1.3453360401752323,
      "grad_norm": 2.884986639022827,
      "learning_rate": 6.558288770053477e-06,
      "loss": 0.08564812690019608,
      "num_input_tokens_seen": 3240318,
      "step": 3148,
      "train_runtime": 7632.3879,
      "train_tokens_per_second": 424.548
    },
    {
      "epoch": 1.345763436264558,
      "grad_norm": 1.9095141887664795,
      "learning_rate": 6.554010695187166e-06,
      "loss": 0.05019308254122734,
      "num_input_tokens_seen": 3241032,
      "step": 3149,
      "train_runtime": 7634.3811,
      "train_tokens_per_second": 424.531
    },
    {
      "epoch": 1.346190832353884,
      "grad_norm": 2.0331380367279053,
      "learning_rate": 6.549732620320856e-06,
      "loss": 0.05688313767313957,
      "num_input_tokens_seen": 3242110,
      "step": 3150,
      "train_runtime": 7636.4823,
      "train_tokens_per_second": 424.555
    },
    {
      "epoch": 1.3466182284432098,
      "grad_norm": 3.029564380645752,
      "learning_rate": 6.545454545454546e-06,
      "loss": 0.10305293649435043,
      "num_input_tokens_seen": 3243518,
      "step": 3151,
      "train_runtime": 7645.2682,
      "train_tokens_per_second": 424.252
    },
    {
      "epoch": 1.3470456245325355,
      "grad_norm": 2.2749345302581787,
      "learning_rate": 6.541176470588236e-06,
      "loss": 0.09431841969490051,
      "num_input_tokens_seen": 3245382,
      "step": 3152,
      "train_runtime": 7647.559,
      "train_tokens_per_second": 424.368
    },
    {
      "epoch": 1.3474730206218613,
      "grad_norm": 2.6047770977020264,
      "learning_rate": 6.536898395721926e-06,
      "loss": 0.09861025214195251,
      "num_input_tokens_seen": 3246666,
      "step": 3153,
      "train_runtime": 7649.7573,
      "train_tokens_per_second": 424.414
    },
    {
      "epoch": 1.347900416711187,
      "grad_norm": 2.601649284362793,
      "learning_rate": 6.532620320855616e-06,
      "loss": 0.07444708049297333,
      "num_input_tokens_seen": 3247776,
      "step": 3154,
      "train_runtime": 7651.9825,
      "train_tokens_per_second": 424.436
    },
    {
      "epoch": 1.348327812800513,
      "grad_norm": 2.5074522495269775,
      "learning_rate": 6.528342245989305e-06,
      "loss": 0.08883404731750488,
      "num_input_tokens_seen": 3248536,
      "step": 3155,
      "train_runtime": 7653.9973,
      "train_tokens_per_second": 424.423
    },
    {
      "epoch": 1.3487552088898387,
      "grad_norm": 1.6762393712997437,
      "learning_rate": 6.524064171122996e-06,
      "loss": 0.04432428628206253,
      "num_input_tokens_seen": 3249608,
      "step": 3156,
      "train_runtime": 7656.1165,
      "train_tokens_per_second": 424.446
    },
    {
      "epoch": 1.3491826049791644,
      "grad_norm": 3.06726336479187,
      "learning_rate": 6.519786096256685e-06,
      "loss": 0.11071242392063141,
      "num_input_tokens_seen": 3250512,
      "step": 3157,
      "train_runtime": 7658.1751,
      "train_tokens_per_second": 424.45
    },
    {
      "epoch": 1.3496100010684902,
      "grad_norm": 2.862823963165283,
      "learning_rate": 6.515508021390375e-06,
      "loss": 0.1063152328133583,
      "num_input_tokens_seen": 3251470,
      "step": 3158,
      "train_runtime": 7660.2353,
      "train_tokens_per_second": 424.461
    },
    {
      "epoch": 1.350037397157816,
      "grad_norm": 3.865882396697998,
      "learning_rate": 6.511229946524065e-06,
      "loss": 0.11060947179794312,
      "num_input_tokens_seen": 3252774,
      "step": 3159,
      "train_runtime": 7662.3633,
      "train_tokens_per_second": 424.513
    },
    {
      "epoch": 1.3504647932471419,
      "grad_norm": 3.1284728050231934,
      "learning_rate": 6.506951871657754e-06,
      "loss": 0.11533777415752411,
      "num_input_tokens_seen": 3253910,
      "step": 3160,
      "train_runtime": 7664.4706,
      "train_tokens_per_second": 424.545
    },
    {
      "epoch": 1.3508921893364676,
      "grad_norm": 2.9496543407440186,
      "learning_rate": 6.5026737967914445e-06,
      "loss": 0.06632790714502335,
      "num_input_tokens_seen": 3254644,
      "step": 3161,
      "train_runtime": 7666.5196,
      "train_tokens_per_second": 424.527
    },
    {
      "epoch": 1.3513195854257933,
      "grad_norm": 2.239867925643921,
      "learning_rate": 6.498395721925134e-06,
      "loss": 0.0695294663310051,
      "num_input_tokens_seen": 3255744,
      "step": 3162,
      "train_runtime": 7668.6331,
      "train_tokens_per_second": 424.553
    },
    {
      "epoch": 1.351746981515119,
      "grad_norm": 3.2851932048797607,
      "learning_rate": 6.494117647058824e-06,
      "loss": 0.09921103715896606,
      "num_input_tokens_seen": 3257164,
      "step": 3163,
      "train_runtime": 7670.8119,
      "train_tokens_per_second": 424.618
    },
    {
      "epoch": 1.3521743776044448,
      "grad_norm": 3.820545196533203,
      "learning_rate": 6.489839572192514e-06,
      "loss": 0.1404322385787964,
      "num_input_tokens_seen": 3258186,
      "step": 3164,
      "train_runtime": 7672.8864,
      "train_tokens_per_second": 424.636
    },
    {
      "epoch": 1.3526017736937708,
      "grad_norm": 2.379157781600952,
      "learning_rate": 6.485561497326204e-06,
      "loss": 0.06789929419755936,
      "num_input_tokens_seen": 3259484,
      "step": 3165,
      "train_runtime": 7675.0344,
      "train_tokens_per_second": 424.687
    },
    {
      "epoch": 1.3530291697830965,
      "grad_norm": 2.4860281944274902,
      "learning_rate": 6.4812834224598935e-06,
      "loss": 0.07726884633302689,
      "num_input_tokens_seen": 3260464,
      "step": 3166,
      "train_runtime": 7677.1108,
      "train_tokens_per_second": 424.699
    },
    {
      "epoch": 1.3534565658724222,
      "grad_norm": 3.3024070262908936,
      "learning_rate": 6.477005347593584e-06,
      "loss": 0.1286133974790573,
      "num_input_tokens_seen": 3261152,
      "step": 3167,
      "train_runtime": 7679.153,
      "train_tokens_per_second": 424.676
    },
    {
      "epoch": 1.353883961961748,
      "grad_norm": 2.6740643978118896,
      "learning_rate": 6.472727272727272e-06,
      "loss": 0.08236388862133026,
      "num_input_tokens_seen": 3262730,
      "step": 3168,
      "train_runtime": 7681.3795,
      "train_tokens_per_second": 424.758
    },
    {
      "epoch": 1.3543113580510737,
      "grad_norm": 2.360847234725952,
      "learning_rate": 6.4684491978609635e-06,
      "loss": 0.10281451791524887,
      "num_input_tokens_seen": 3263640,
      "step": 3169,
      "train_runtime": 7683.3981,
      "train_tokens_per_second": 424.765
    },
    {
      "epoch": 1.3547387541403997,
      "grad_norm": 2.7765560150146484,
      "learning_rate": 6.464171122994652e-06,
      "loss": 0.10607416927814484,
      "num_input_tokens_seen": 3264980,
      "step": 3170,
      "train_runtime": 7685.6401,
      "train_tokens_per_second": 424.816
    },
    {
      "epoch": 1.3551661502297254,
      "grad_norm": 2.682161569595337,
      "learning_rate": 6.459893048128343e-06,
      "loss": 0.09889735281467438,
      "num_input_tokens_seen": 3265910,
      "step": 3171,
      "train_runtime": 7687.6968,
      "train_tokens_per_second": 424.823
    },
    {
      "epoch": 1.3555935463190512,
      "grad_norm": 1.7430683374404907,
      "learning_rate": 6.455614973262032e-06,
      "loss": 0.06315062195062637,
      "num_input_tokens_seen": 3266904,
      "step": 3172,
      "train_runtime": 7689.7756,
      "train_tokens_per_second": 424.837
    },
    {
      "epoch": 1.356020942408377,
      "grad_norm": 2.373918294906616,
      "learning_rate": 6.451336898395722e-06,
      "loss": 0.0726034939289093,
      "num_input_tokens_seen": 3267664,
      "step": 3173,
      "train_runtime": 7691.7623,
      "train_tokens_per_second": 424.826
    },
    {
      "epoch": 1.3564483384977026,
      "grad_norm": 3.0013420581817627,
      "learning_rate": 6.4470588235294116e-06,
      "loss": 0.11568459868431091,
      "num_input_tokens_seen": 3268618,
      "step": 3174,
      "train_runtime": 7693.8173,
      "train_tokens_per_second": 424.837
    },
    {
      "epoch": 1.3568757345870286,
      "grad_norm": 2.996323823928833,
      "learning_rate": 6.442780748663102e-06,
      "loss": 0.1038421168923378,
      "num_input_tokens_seen": 3269598,
      "step": 3175,
      "train_runtime": 7695.9036,
      "train_tokens_per_second": 424.849
    },
    {
      "epoch": 1.3573031306763543,
      "grad_norm": 2.336603879928589,
      "learning_rate": 6.438502673796791e-06,
      "loss": 0.07186711579561234,
      "num_input_tokens_seen": 3270622,
      "step": 3176,
      "train_runtime": 7698.0293,
      "train_tokens_per_second": 424.865
    },
    {
      "epoch": 1.35773052676568,
      "grad_norm": 1.8723411560058594,
      "learning_rate": 6.434224598930482e-06,
      "loss": 0.06434863060712814,
      "num_input_tokens_seen": 3272038,
      "step": 3177,
      "train_runtime": 7700.2273,
      "train_tokens_per_second": 424.927
    },
    {
      "epoch": 1.3581579228550058,
      "grad_norm": 2.6882195472717285,
      "learning_rate": 6.429946524064171e-06,
      "loss": 0.1429688185453415,
      "num_input_tokens_seen": 3273066,
      "step": 3178,
      "train_runtime": 7702.3331,
      "train_tokens_per_second": 424.945
    },
    {
      "epoch": 1.3585853189443315,
      "grad_norm": 2.995041847229004,
      "learning_rate": 6.425668449197861e-06,
      "loss": 0.09537284076213837,
      "num_input_tokens_seen": 3273844,
      "step": 3179,
      "train_runtime": 7704.3742,
      "train_tokens_per_second": 424.933
    },
    {
      "epoch": 1.3590127150336575,
      "grad_norm": 2.7519783973693848,
      "learning_rate": 6.421390374331551e-06,
      "loss": 0.09719207882881165,
      "num_input_tokens_seen": 3274778,
      "step": 3180,
      "train_runtime": 7706.4238,
      "train_tokens_per_second": 424.941
    },
    {
      "epoch": 1.3594401111229832,
      "grad_norm": 2.245267868041992,
      "learning_rate": 6.417112299465241e-06,
      "loss": 0.08162558078765869,
      "num_input_tokens_seen": 3275972,
      "step": 3181,
      "train_runtime": 7714.8184,
      "train_tokens_per_second": 424.634
    },
    {
      "epoch": 1.359867507212309,
      "grad_norm": 3.112441062927246,
      "learning_rate": 6.4128342245989305e-06,
      "loss": 0.10832598060369492,
      "num_input_tokens_seen": 3276938,
      "step": 3182,
      "train_runtime": 7716.9322,
      "train_tokens_per_second": 424.643
    },
    {
      "epoch": 1.3602949033016347,
      "grad_norm": 1.7854994535446167,
      "learning_rate": 6.408556149732621e-06,
      "loss": 0.03599024564027786,
      "num_input_tokens_seen": 3278072,
      "step": 3183,
      "train_runtime": 7719.1001,
      "train_tokens_per_second": 424.67
    },
    {
      "epoch": 1.3607222993909605,
      "grad_norm": 3.4036924839019775,
      "learning_rate": 6.40427807486631e-06,
      "loss": 0.06701241433620453,
      "num_input_tokens_seen": 3278774,
      "step": 3184,
      "train_runtime": 7721.1234,
      "train_tokens_per_second": 424.65
    },
    {
      "epoch": 1.3611496954802864,
      "grad_norm": 2.578629970550537,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.10280337929725647,
      "num_input_tokens_seen": 3280296,
      "step": 3185,
      "train_runtime": 7723.3386,
      "train_tokens_per_second": 424.725
    },
    {
      "epoch": 1.3615770915696122,
      "grad_norm": 3.8966360092163086,
      "learning_rate": 6.39572192513369e-06,
      "loss": 0.14626756310462952,
      "num_input_tokens_seen": 3281082,
      "step": 3186,
      "train_runtime": 7725.3846,
      "train_tokens_per_second": 424.714
    },
    {
      "epoch": 1.362004487658938,
      "grad_norm": 4.28747034072876,
      "learning_rate": 6.39144385026738e-06,
      "loss": 0.12739111483097076,
      "num_input_tokens_seen": 3282362,
      "step": 3187,
      "train_runtime": 7727.556,
      "train_tokens_per_second": 424.761
    },
    {
      "epoch": 1.3624318837482636,
      "grad_norm": 3.4851841926574707,
      "learning_rate": 6.38716577540107e-06,
      "loss": 0.07526595890522003,
      "num_input_tokens_seen": 3283254,
      "step": 3188,
      "train_runtime": 7729.6648,
      "train_tokens_per_second": 424.76
    },
    {
      "epoch": 1.3628592798375894,
      "grad_norm": 3.4750003814697266,
      "learning_rate": 6.38288770053476e-06,
      "loss": 0.09182319045066833,
      "num_input_tokens_seen": 3284418,
      "step": 3189,
      "train_runtime": 7731.8394,
      "train_tokens_per_second": 424.791
    },
    {
      "epoch": 1.3632866759269153,
      "grad_norm": 3.5623397827148438,
      "learning_rate": 6.3786096256684495e-06,
      "loss": 0.1354844570159912,
      "num_input_tokens_seen": 3285346,
      "step": 3190,
      "train_runtime": 7733.8771,
      "train_tokens_per_second": 424.799
    },
    {
      "epoch": 1.363714072016241,
      "grad_norm": 3.087920904159546,
      "learning_rate": 6.37433155080214e-06,
      "loss": 0.10777562856674194,
      "num_input_tokens_seen": 3286140,
      "step": 3191,
      "train_runtime": 7735.8946,
      "train_tokens_per_second": 424.791
    },
    {
      "epoch": 1.3641414681055668,
      "grad_norm": 2.4586117267608643,
      "learning_rate": 6.370053475935829e-06,
      "loss": 0.11487673223018646,
      "num_input_tokens_seen": 3287342,
      "step": 3192,
      "train_runtime": 7738.0136,
      "train_tokens_per_second": 424.83
    },
    {
      "epoch": 1.3645688641948925,
      "grad_norm": 2.5379438400268555,
      "learning_rate": 6.3657754010695195e-06,
      "loss": 0.08158126473426819,
      "num_input_tokens_seen": 3288068,
      "step": 3193,
      "train_runtime": 7740.0271,
      "train_tokens_per_second": 424.813
    },
    {
      "epoch": 1.3649962602842183,
      "grad_norm": 2.3742177486419678,
      "learning_rate": 6.361497326203209e-06,
      "loss": 0.09661485999822617,
      "num_input_tokens_seen": 3289172,
      "step": 3194,
      "train_runtime": 7742.141,
      "train_tokens_per_second": 424.84
    },
    {
      "epoch": 1.3654236563735442,
      "grad_norm": 2.754796266555786,
      "learning_rate": 6.357219251336899e-06,
      "loss": 0.11201435327529907,
      "num_input_tokens_seen": 3290482,
      "step": 3195,
      "train_runtime": 7744.3333,
      "train_tokens_per_second": 424.889
    },
    {
      "epoch": 1.36585105246287,
      "grad_norm": 3.072242498397827,
      "learning_rate": 6.352941176470589e-06,
      "loss": 0.09425794333219528,
      "num_input_tokens_seen": 3291380,
      "step": 3196,
      "train_runtime": 7746.3824,
      "train_tokens_per_second": 424.893
    },
    {
      "epoch": 1.3662784485521957,
      "grad_norm": 2.5668187141418457,
      "learning_rate": 6.348663101604279e-06,
      "loss": 0.09268469363451004,
      "num_input_tokens_seen": 3292146,
      "step": 3197,
      "train_runtime": 7748.3807,
      "train_tokens_per_second": 424.882
    },
    {
      "epoch": 1.3667058446415215,
      "grad_norm": 3.569673538208008,
      "learning_rate": 6.3443850267379684e-06,
      "loss": 0.11805763840675354,
      "num_input_tokens_seen": 3292906,
      "step": 3198,
      "train_runtime": 7750.4117,
      "train_tokens_per_second": 424.869
    },
    {
      "epoch": 1.3671332407308472,
      "grad_norm": 2.6122937202453613,
      "learning_rate": 6.340106951871659e-06,
      "loss": 0.08259596675634384,
      "num_input_tokens_seen": 3293944,
      "step": 3199,
      "train_runtime": 7752.5446,
      "train_tokens_per_second": 424.886
    },
    {
      "epoch": 1.3675606368201731,
      "grad_norm": 2.524951934814453,
      "learning_rate": 6.335828877005348e-06,
      "loss": 0.09323089569807053,
      "num_input_tokens_seen": 3294696,
      "step": 3200,
      "train_runtime": 7754.5906,
      "train_tokens_per_second": 424.87
    },
    {
      "epoch": 1.3679880329094989,
      "grad_norm": 2.4553985595703125,
      "learning_rate": 6.3315508021390385e-06,
      "loss": 0.11925337463617325,
      "num_input_tokens_seen": 3296142,
      "step": 3201,
      "train_runtime": 7756.7619,
      "train_tokens_per_second": 424.938
    },
    {
      "epoch": 1.3684154289988246,
      "grad_norm": 3.1260263919830322,
      "learning_rate": 6.327272727272727e-06,
      "loss": 0.09206196665763855,
      "num_input_tokens_seen": 3297164,
      "step": 3202,
      "train_runtime": 7758.8571,
      "train_tokens_per_second": 424.955
    },
    {
      "epoch": 1.3688428250881504,
      "grad_norm": 2.3616137504577637,
      "learning_rate": 6.322994652406418e-06,
      "loss": 0.08467519283294678,
      "num_input_tokens_seen": 3298542,
      "step": 3203,
      "train_runtime": 7761.0347,
      "train_tokens_per_second": 425.013
    },
    {
      "epoch": 1.369270221177476,
      "grad_norm": 2.5820698738098145,
      "learning_rate": 6.318716577540107e-06,
      "loss": 0.0881558209657669,
      "num_input_tokens_seen": 3299500,
      "step": 3204,
      "train_runtime": 7763.1263,
      "train_tokens_per_second": 425.022
    },
    {
      "epoch": 1.369697617266802,
      "grad_norm": 3.3584587574005127,
      "learning_rate": 6.314438502673798e-06,
      "loss": 0.13167576491832733,
      "num_input_tokens_seen": 3300406,
      "step": 3205,
      "train_runtime": 7765.1786,
      "train_tokens_per_second": 425.026
    },
    {
      "epoch": 1.3701250133561278,
      "grad_norm": 3.0288429260253906,
      "learning_rate": 6.3101604278074865e-06,
      "loss": 0.08793671429157257,
      "num_input_tokens_seen": 3301086,
      "step": 3206,
      "train_runtime": 7767.1598,
      "train_tokens_per_second": 425.006
    },
    {
      "epoch": 1.3705524094454535,
      "grad_norm": 1.759108543395996,
      "learning_rate": 6.305882352941177e-06,
      "loss": 0.06125544011592865,
      "num_input_tokens_seen": 3302922,
      "step": 3207,
      "train_runtime": 7769.4243,
      "train_tokens_per_second": 425.118
    },
    {
      "epoch": 1.3709798055347795,
      "grad_norm": 3.0307743549346924,
      "learning_rate": 6.301604278074866e-06,
      "loss": 0.12302170693874359,
      "num_input_tokens_seen": 3304154,
      "step": 3208,
      "train_runtime": 7771.5456,
      "train_tokens_per_second": 425.16
    },
    {
      "epoch": 1.371407201624105,
      "grad_norm": 2.2818562984466553,
      "learning_rate": 6.2973262032085566e-06,
      "loss": 0.056075967848300934,
      "num_input_tokens_seen": 3305124,
      "step": 3209,
      "train_runtime": 7773.6444,
      "train_tokens_per_second": 425.17
    },
    {
      "epoch": 1.371834597713431,
      "grad_norm": 2.10774827003479,
      "learning_rate": 6.293048128342246e-06,
      "loss": 0.06893950700759888,
      "num_input_tokens_seen": 3306108,
      "step": 3210,
      "train_runtime": 7775.7067,
      "train_tokens_per_second": 425.184
    },
    {
      "epoch": 1.3722619938027567,
      "grad_norm": 2.7141215801239014,
      "learning_rate": 6.288770053475936e-06,
      "loss": 0.10845555365085602,
      "num_input_tokens_seen": 3307284,
      "step": 3211,
      "train_runtime": 7784.0884,
      "train_tokens_per_second": 424.877
    },
    {
      "epoch": 1.3726893898920824,
      "grad_norm": 2.9843757152557373,
      "learning_rate": 6.284491978609626e-06,
      "loss": 0.1299024522304535,
      "num_input_tokens_seen": 3308394,
      "step": 3212,
      "train_runtime": 7786.2149,
      "train_tokens_per_second": 424.904
    },
    {
      "epoch": 1.3731167859814084,
      "grad_norm": 2.4188318252563477,
      "learning_rate": 6.280213903743316e-06,
      "loss": 0.07884763181209564,
      "num_input_tokens_seen": 3309298,
      "step": 3213,
      "train_runtime": 7788.5997,
      "train_tokens_per_second": 424.89
    },
    {
      "epoch": 1.373544182070734,
      "grad_norm": 2.6896750926971436,
      "learning_rate": 6.2759358288770055e-06,
      "loss": 0.10796432197093964,
      "num_input_tokens_seen": 3310758,
      "step": 3214,
      "train_runtime": 7790.7413,
      "train_tokens_per_second": 424.961
    },
    {
      "epoch": 1.3739715781600599,
      "grad_norm": 2.2479846477508545,
      "learning_rate": 6.271657754010696e-06,
      "loss": 0.07536740601062775,
      "num_input_tokens_seen": 3312158,
      "step": 3215,
      "train_runtime": 7792.9437,
      "train_tokens_per_second": 425.02
    },
    {
      "epoch": 1.3743989742493856,
      "grad_norm": 2.9850735664367676,
      "learning_rate": 6.267379679144385e-06,
      "loss": 0.09404246509075165,
      "num_input_tokens_seen": 3313224,
      "step": 3216,
      "train_runtime": 7795.0619,
      "train_tokens_per_second": 425.041
    },
    {
      "epoch": 1.3748263703387114,
      "grad_norm": 2.559086322784424,
      "learning_rate": 6.2631016042780755e-06,
      "loss": 0.1048872321844101,
      "num_input_tokens_seen": 3314132,
      "step": 3217,
      "train_runtime": 7797.1175,
      "train_tokens_per_second": 425.046
    },
    {
      "epoch": 1.3752537664280373,
      "grad_norm": 2.6748650074005127,
      "learning_rate": 6.258823529411765e-06,
      "loss": 0.11120984703302383,
      "num_input_tokens_seen": 3315556,
      "step": 3218,
      "train_runtime": 7799.2893,
      "train_tokens_per_second": 425.11
    },
    {
      "epoch": 1.3756811625173628,
      "grad_norm": 4.306665420532227,
      "learning_rate": 6.254545454545455e-06,
      "loss": 0.10187392681837082,
      "num_input_tokens_seen": 3316754,
      "step": 3219,
      "train_runtime": 7801.399,
      "train_tokens_per_second": 425.149
    },
    {
      "epoch": 1.3761085586066888,
      "grad_norm": 1.6322040557861328,
      "learning_rate": 6.250267379679145e-06,
      "loss": 0.07576320320367813,
      "num_input_tokens_seen": 3319018,
      "step": 3220,
      "train_runtime": 7803.795,
      "train_tokens_per_second": 425.308
    },
    {
      "epoch": 1.3765359546960145,
      "grad_norm": 3.8274483680725098,
      "learning_rate": 6.245989304812835e-06,
      "loss": 0.09648226946592331,
      "num_input_tokens_seen": 3319694,
      "step": 3221,
      "train_runtime": 7805.8143,
      "train_tokens_per_second": 425.285
    },
    {
      "epoch": 1.3769633507853403,
      "grad_norm": 2.8847217559814453,
      "learning_rate": 6.2417112299465244e-06,
      "loss": 0.07623672485351562,
      "num_input_tokens_seen": 3320444,
      "step": 3222,
      "train_runtime": 7807.83,
      "train_tokens_per_second": 425.271
    },
    {
      "epoch": 1.3773907468746662,
      "grad_norm": 2.2220189571380615,
      "learning_rate": 6.237433155080215e-06,
      "loss": 0.0994056910276413,
      "num_input_tokens_seen": 3322766,
      "step": 3223,
      "train_runtime": 7810.2513,
      "train_tokens_per_second": 425.437
    },
    {
      "epoch": 1.3778181429639917,
      "grad_norm": 2.475769519805908,
      "learning_rate": 6.233155080213904e-06,
      "loss": 0.07649150490760803,
      "num_input_tokens_seen": 3323868,
      "step": 3224,
      "train_runtime": 7812.4277,
      "train_tokens_per_second": 425.459
    },
    {
      "epoch": 1.3782455390533177,
      "grad_norm": 2.271890163421631,
      "learning_rate": 6.2288770053475945e-06,
      "loss": 0.10229285806417465,
      "num_input_tokens_seen": 3325204,
      "step": 3225,
      "train_runtime": 7814.5906,
      "train_tokens_per_second": 425.512
    },
    {
      "epoch": 1.3786729351426434,
      "grad_norm": 3.6042072772979736,
      "learning_rate": 6.224598930481284e-06,
      "loss": 0.17576083540916443,
      "num_input_tokens_seen": 3326244,
      "step": 3226,
      "train_runtime": 7816.7153,
      "train_tokens_per_second": 425.53
    },
    {
      "epoch": 1.3791003312319692,
      "grad_norm": 4.114015579223633,
      "learning_rate": 6.220320855614974e-06,
      "loss": 0.13761650025844574,
      "num_input_tokens_seen": 3327552,
      "step": 3227,
      "train_runtime": 7818.9416,
      "train_tokens_per_second": 425.576
    },
    {
      "epoch": 1.3795277273212951,
      "grad_norm": 3.441895008087158,
      "learning_rate": 6.216042780748664e-06,
      "loss": 0.09298841655254364,
      "num_input_tokens_seen": 3328282,
      "step": 3228,
      "train_runtime": 7820.9631,
      "train_tokens_per_second": 425.559
    },
    {
      "epoch": 1.3799551234106207,
      "grad_norm": 2.1415321826934814,
      "learning_rate": 6.211764705882354e-06,
      "loss": 0.05958282947540283,
      "num_input_tokens_seen": 3329422,
      "step": 3229,
      "train_runtime": 7823.0754,
      "train_tokens_per_second": 425.59
    },
    {
      "epoch": 1.3803825194999466,
      "grad_norm": 2.828928232192993,
      "learning_rate": 6.207486631016043e-06,
      "loss": 0.10203618556261063,
      "num_input_tokens_seen": 3330476,
      "step": 3230,
      "train_runtime": 7825.1545,
      "train_tokens_per_second": 425.612
    },
    {
      "epoch": 1.3808099155892724,
      "grad_norm": 4.01011323928833,
      "learning_rate": 6.203208556149734e-06,
      "loss": 0.1174144297838211,
      "num_input_tokens_seen": 3331124,
      "step": 3231,
      "train_runtime": 7827.1648,
      "train_tokens_per_second": 425.585
    },
    {
      "epoch": 1.381237311678598,
      "grad_norm": 2.849505662918091,
      "learning_rate": 6.198930481283423e-06,
      "loss": 0.08645319193601608,
      "num_input_tokens_seen": 3331968,
      "step": 3232,
      "train_runtime": 7829.2195,
      "train_tokens_per_second": 425.581
    },
    {
      "epoch": 1.381664707767924,
      "grad_norm": 2.091911554336548,
      "learning_rate": 6.1946524064171134e-06,
      "loss": 0.061108991503715515,
      "num_input_tokens_seen": 3333166,
      "step": 3233,
      "train_runtime": 7831.3448,
      "train_tokens_per_second": 425.619
    },
    {
      "epoch": 1.3820921038572498,
      "grad_norm": 3.3379859924316406,
      "learning_rate": 6.190374331550803e-06,
      "loss": 0.11296980082988739,
      "num_input_tokens_seen": 3333980,
      "step": 3234,
      "train_runtime": 7833.3626,
      "train_tokens_per_second": 425.613
    },
    {
      "epoch": 1.3825194999465755,
      "grad_norm": 2.765662908554077,
      "learning_rate": 6.186096256684493e-06,
      "loss": 0.07650604844093323,
      "num_input_tokens_seen": 3335564,
      "step": 3235,
      "train_runtime": 7835.5906,
      "train_tokens_per_second": 425.694
    },
    {
      "epoch": 1.3829468960359013,
      "grad_norm": 2.891916275024414,
      "learning_rate": 6.181818181818182e-06,
      "loss": 0.06541066616773605,
      "num_input_tokens_seen": 3336810,
      "step": 3236,
      "train_runtime": 7837.764,
      "train_tokens_per_second": 425.735
    },
    {
      "epoch": 1.383374292125227,
      "grad_norm": 2.093151807785034,
      "learning_rate": 6.177540106951873e-06,
      "loss": 0.07020801305770874,
      "num_input_tokens_seen": 3337786,
      "step": 3237,
      "train_runtime": 7839.8166,
      "train_tokens_per_second": 425.748
    },
    {
      "epoch": 1.383801688214553,
      "grad_norm": 2.0807228088378906,
      "learning_rate": 6.1732620320855615e-06,
      "loss": 0.09419591724872589,
      "num_input_tokens_seen": 3338878,
      "step": 3238,
      "train_runtime": 7841.9428,
      "train_tokens_per_second": 425.772
    },
    {
      "epoch": 1.3842290843038787,
      "grad_norm": 3.4361512660980225,
      "learning_rate": 6.168983957219253e-06,
      "loss": 0.11589787900447845,
      "num_input_tokens_seen": 3339822,
      "step": 3239,
      "train_runtime": 7844.0139,
      "train_tokens_per_second": 425.78
    },
    {
      "epoch": 1.3846564803932044,
      "grad_norm": 2.43046236038208,
      "learning_rate": 6.164705882352941e-06,
      "loss": 0.06651141494512558,
      "num_input_tokens_seen": 3340934,
      "step": 3240,
      "train_runtime": 7846.1423,
      "train_tokens_per_second": 425.806
    },
    {
      "epoch": 1.3850838764825302,
      "grad_norm": 3.2076175212860107,
      "learning_rate": 6.1604278074866315e-06,
      "loss": 0.1519242823123932,
      "num_input_tokens_seen": 3342036,
      "step": 3241,
      "train_runtime": 7854.9541,
      "train_tokens_per_second": 425.469
    },
    {
      "epoch": 1.385511272571856,
      "grad_norm": 2.9235315322875977,
      "learning_rate": 6.156149732620321e-06,
      "loss": 0.08675135672092438,
      "num_input_tokens_seen": 3342836,
      "step": 3242,
      "train_runtime": 7857.0448,
      "train_tokens_per_second": 425.457
    },
    {
      "epoch": 1.3859386686611819,
      "grad_norm": 3.4454190731048584,
      "learning_rate": 6.151871657754011e-06,
      "loss": 0.0739680826663971,
      "num_input_tokens_seen": 3343546,
      "step": 3243,
      "train_runtime": 7859.2758,
      "train_tokens_per_second": 425.427
    },
    {
      "epoch": 1.3863660647505076,
      "grad_norm": 3.0161163806915283,
      "learning_rate": 6.147593582887701e-06,
      "loss": 0.055188506841659546,
      "num_input_tokens_seen": 3344216,
      "step": 3244,
      "train_runtime": 7861.324,
      "train_tokens_per_second": 425.401
    },
    {
      "epoch": 1.3867934608398333,
      "grad_norm": 3.4926955699920654,
      "learning_rate": 6.143315508021391e-06,
      "loss": 0.11531645059585571,
      "num_input_tokens_seen": 3344990,
      "step": 3245,
      "train_runtime": 7863.3465,
      "train_tokens_per_second": 425.39
    },
    {
      "epoch": 1.387220856929159,
      "grad_norm": 2.739166021347046,
      "learning_rate": 6.1390374331550805e-06,
      "loss": 0.07723821699619293,
      "num_input_tokens_seen": 3345798,
      "step": 3246,
      "train_runtime": 7865.3728,
      "train_tokens_per_second": 425.383
    },
    {
      "epoch": 1.3876482530184848,
      "grad_norm": 2.535468339920044,
      "learning_rate": 6.134759358288771e-06,
      "loss": 0.09975472092628479,
      "num_input_tokens_seen": 3346956,
      "step": 3247,
      "train_runtime": 7867.5149,
      "train_tokens_per_second": 425.415
    },
    {
      "epoch": 1.3880756491078108,
      "grad_norm": 2.196161985397339,
      "learning_rate": 6.13048128342246e-06,
      "loss": 0.08604616671800613,
      "num_input_tokens_seen": 3348918,
      "step": 3248,
      "train_runtime": 7869.8792,
      "train_tokens_per_second": 425.536
    },
    {
      "epoch": 1.3885030451971365,
      "grad_norm": 2.617840051651001,
      "learning_rate": 6.1262032085561505e-06,
      "loss": 0.09151151776313782,
      "num_input_tokens_seen": 3349608,
      "step": 3249,
      "train_runtime": 7871.9382,
      "train_tokens_per_second": 425.512
    },
    {
      "epoch": 1.3889304412864623,
      "grad_norm": 2.439768075942993,
      "learning_rate": 6.12192513368984e-06,
      "loss": 0.098683200776577,
      "num_input_tokens_seen": 3351060,
      "step": 3250,
      "train_runtime": 7874.2035,
      "train_tokens_per_second": 425.574
    },
    {
      "epoch": 1.389357837375788,
      "grad_norm": 1.9714925289154053,
      "learning_rate": 6.11764705882353e-06,
      "loss": 0.06090874969959259,
      "num_input_tokens_seen": 3351784,
      "step": 3251,
      "train_runtime": 7876.2825,
      "train_tokens_per_second": 425.554
    },
    {
      "epoch": 1.3897852334651137,
      "grad_norm": 3.021029233932495,
      "learning_rate": 6.11336898395722e-06,
      "loss": 0.1005382388830185,
      "num_input_tokens_seen": 3352838,
      "step": 3252,
      "train_runtime": 7878.4455,
      "train_tokens_per_second": 425.571
    },
    {
      "epoch": 1.3902126295544397,
      "grad_norm": 2.7178242206573486,
      "learning_rate": 6.10909090909091e-06,
      "loss": 0.13861191272735596,
      "num_input_tokens_seen": 3354390,
      "step": 3253,
      "train_runtime": 7880.6542,
      "train_tokens_per_second": 425.649
    },
    {
      "epoch": 1.3906400256437654,
      "grad_norm": 4.7103447914123535,
      "learning_rate": 6.104812834224599e-06,
      "loss": 0.10404442250728607,
      "num_input_tokens_seen": 3355422,
      "step": 3254,
      "train_runtime": 7882.7686,
      "train_tokens_per_second": 425.665
    },
    {
      "epoch": 1.3910674217330912,
      "grad_norm": 2.4716503620147705,
      "learning_rate": 6.10053475935829e-06,
      "loss": 0.06483546644449234,
      "num_input_tokens_seen": 3356166,
      "step": 3255,
      "train_runtime": 7884.8226,
      "train_tokens_per_second": 425.649
    },
    {
      "epoch": 1.391494817822417,
      "grad_norm": 2.449089288711548,
      "learning_rate": 6.096256684491979e-06,
      "loss": 0.0674515962600708,
      "num_input_tokens_seen": 3357178,
      "step": 3256,
      "train_runtime": 7886.964,
      "train_tokens_per_second": 425.662
    },
    {
      "epoch": 1.3919222139117426,
      "grad_norm": 2.07804799079895,
      "learning_rate": 6.0919786096256694e-06,
      "loss": 0.07267894595861435,
      "num_input_tokens_seen": 3358174,
      "step": 3257,
      "train_runtime": 7889.0536,
      "train_tokens_per_second": 425.675
    },
    {
      "epoch": 1.3923496100010686,
      "grad_norm": 2.101487159729004,
      "learning_rate": 6.087700534759359e-06,
      "loss": 0.0976356789469719,
      "num_input_tokens_seen": 3359174,
      "step": 3258,
      "train_runtime": 7891.2149,
      "train_tokens_per_second": 425.685
    },
    {
      "epoch": 1.3927770060903943,
      "grad_norm": 2.7273178100585938,
      "learning_rate": 6.083422459893049e-06,
      "loss": 0.11177102476358414,
      "num_input_tokens_seen": 3360114,
      "step": 3259,
      "train_runtime": 7893.2804,
      "train_tokens_per_second": 425.693
    },
    {
      "epoch": 1.39320440217972,
      "grad_norm": 1.8124136924743652,
      "learning_rate": 6.079144385026739e-06,
      "loss": 0.06550105661153793,
      "num_input_tokens_seen": 3361138,
      "step": 3260,
      "train_runtime": 7895.365,
      "train_tokens_per_second": 425.71
    },
    {
      "epoch": 1.3936317982690458,
      "grad_norm": 2.7312512397766113,
      "learning_rate": 6.074866310160429e-06,
      "loss": 0.06788568198680878,
      "num_input_tokens_seen": 3361994,
      "step": 3261,
      "train_runtime": 7897.3823,
      "train_tokens_per_second": 425.71
    },
    {
      "epoch": 1.3940591943583716,
      "grad_norm": 2.6394524574279785,
      "learning_rate": 6.070588235294118e-06,
      "loss": 0.09610903263092041,
      "num_input_tokens_seen": 3362964,
      "step": 3262,
      "train_runtime": 7899.4652,
      "train_tokens_per_second": 425.72
    },
    {
      "epoch": 1.3944865904476975,
      "grad_norm": 2.760646104812622,
      "learning_rate": 6.066310160427809e-06,
      "loss": 0.06169695779681206,
      "num_input_tokens_seen": 3363918,
      "step": 3263,
      "train_runtime": 7901.6277,
      "train_tokens_per_second": 425.725
    },
    {
      "epoch": 1.3949139865370233,
      "grad_norm": 2.8675825595855713,
      "learning_rate": 6.062032085561498e-06,
      "loss": 0.08873794972896576,
      "num_input_tokens_seen": 3364968,
      "step": 3264,
      "train_runtime": 7903.7118,
      "train_tokens_per_second": 425.745
    },
    {
      "epoch": 1.395341382626349,
      "grad_norm": 2.1839499473571777,
      "learning_rate": 6.057754010695188e-06,
      "loss": 0.07397759705781937,
      "num_input_tokens_seen": 3365880,
      "step": 3265,
      "train_runtime": 7905.7644,
      "train_tokens_per_second": 425.75
    },
    {
      "epoch": 1.3957687787156747,
      "grad_norm": 2.5855510234832764,
      "learning_rate": 6.053475935828878e-06,
      "loss": 0.05796331912279129,
      "num_input_tokens_seen": 3366844,
      "step": 3266,
      "train_runtime": 7907.9275,
      "train_tokens_per_second": 425.756
    },
    {
      "epoch": 1.3961961748050005,
      "grad_norm": 2.515497922897339,
      "learning_rate": 6.049197860962568e-06,
      "loss": 0.09895144402980804,
      "num_input_tokens_seen": 3368118,
      "step": 3267,
      "train_runtime": 7910.1973,
      "train_tokens_per_second": 425.794
    },
    {
      "epoch": 1.3966235708943264,
      "grad_norm": 3.038151502609253,
      "learning_rate": 6.044919786096258e-06,
      "loss": 0.08927208185195923,
      "num_input_tokens_seen": 3368890,
      "step": 3268,
      "train_runtime": 7912.2153,
      "train_tokens_per_second": 425.783
    },
    {
      "epoch": 1.3970509669836522,
      "grad_norm": 1.5889421701431274,
      "learning_rate": 6.040641711229946e-06,
      "loss": 0.04282796382904053,
      "num_input_tokens_seen": 3370000,
      "step": 3269,
      "train_runtime": 7914.3488,
      "train_tokens_per_second": 425.809
    },
    {
      "epoch": 1.397478363072978,
      "grad_norm": 3.962462902069092,
      "learning_rate": 6.0363636363636365e-06,
      "loss": 0.11668793857097626,
      "num_input_tokens_seen": 3370638,
      "step": 3270,
      "train_runtime": 7916.3317,
      "train_tokens_per_second": 425.783
    },
    {
      "epoch": 1.3979057591623036,
      "grad_norm": 2.4602210521698,
      "learning_rate": 6.032085561497326e-06,
      "loss": 0.12678514420986176,
      "num_input_tokens_seen": 3372222,
      "step": 3271,
      "train_runtime": 7925.236,
      "train_tokens_per_second": 425.504
    },
    {
      "epoch": 1.3983331552516294,
      "grad_norm": 3.434453248977661,
      "learning_rate": 6.027807486631016e-06,
      "loss": 0.1236409917473793,
      "num_input_tokens_seen": 3373236,
      "step": 3272,
      "train_runtime": 7927.4754,
      "train_tokens_per_second": 425.512
    },
    {
      "epoch": 1.3987605513409553,
      "grad_norm": 2.778341293334961,
      "learning_rate": 6.023529411764706e-06,
      "loss": 0.159600168466568,
      "num_input_tokens_seen": 3374566,
      "step": 3273,
      "train_runtime": 7929.6698,
      "train_tokens_per_second": 425.562
    },
    {
      "epoch": 1.399187947430281,
      "grad_norm": 3.214672803878784,
      "learning_rate": 6.019251336898396e-06,
      "loss": 0.07167714834213257,
      "num_input_tokens_seen": 3375176,
      "step": 3274,
      "train_runtime": 7931.7042,
      "train_tokens_per_second": 425.53
    },
    {
      "epoch": 1.3996153435196068,
      "grad_norm": 3.492658853530884,
      "learning_rate": 6.014973262032085e-06,
      "loss": 0.07649270445108414,
      "num_input_tokens_seen": 3375926,
      "step": 3275,
      "train_runtime": 7933.7272,
      "train_tokens_per_second": 425.516
    },
    {
      "epoch": 1.4000427396089326,
      "grad_norm": 2.721825361251831,
      "learning_rate": 6.010695187165776e-06,
      "loss": 0.05536902695894241,
      "num_input_tokens_seen": 3376746,
      "step": 3276,
      "train_runtime": 7935.7973,
      "train_tokens_per_second": 425.508
    },
    {
      "epoch": 1.4000427396089326,
      "eval_loss": 0.5228173136711121,
      "eval_runtime": 58.4858,
      "eval_samples_per_second": 17.064,
      "eval_steps_per_second": 8.532,
      "num_input_tokens_seen": 3376746,
      "step": 3276
    },
    {
      "epoch": 1.4004701356982583,
      "grad_norm": 2.9269371032714844,
      "learning_rate": 6.006417112299465e-06,
      "loss": 0.11548300087451935,
      "num_input_tokens_seen": 3377638,
      "step": 3277,
      "train_runtime": 7996.4036,
      "train_tokens_per_second": 422.395
    },
    {
      "epoch": 1.4008975317875842,
      "grad_norm": 4.1711883544921875,
      "learning_rate": 6.0021390374331554e-06,
      "loss": 0.10242607444524765,
      "num_input_tokens_seen": 3378530,
      "step": 3278,
      "train_runtime": 7998.452,
      "train_tokens_per_second": 422.398
    },
    {
      "epoch": 1.40132492787691,
      "grad_norm": 2.544394016265869,
      "learning_rate": 5.997860962566845e-06,
      "loss": 0.08933587372303009,
      "num_input_tokens_seen": 3379626,
      "step": 3279,
      "train_runtime": 8000.5639,
      "train_tokens_per_second": 422.423
    },
    {
      "epoch": 1.4017523239662357,
      "grad_norm": 1.9375145435333252,
      "learning_rate": 5.993582887700535e-06,
      "loss": 0.055566221475601196,
      "num_input_tokens_seen": 3380814,
      "step": 3280,
      "train_runtime": 8002.713,
      "train_tokens_per_second": 422.458
    },
    {
      "epoch": 1.4021797200555615,
      "grad_norm": 2.98313307762146,
      "learning_rate": 5.989304812834225e-06,
      "loss": 0.053592294454574585,
      "num_input_tokens_seen": 3381844,
      "step": 3281,
      "train_runtime": 8004.8024,
      "train_tokens_per_second": 422.477
    },
    {
      "epoch": 1.4026071161448872,
      "grad_norm": 2.6182615756988525,
      "learning_rate": 5.985026737967915e-06,
      "loss": 0.07741963863372803,
      "num_input_tokens_seen": 3382824,
      "step": 3282,
      "train_runtime": 8007.0197,
      "train_tokens_per_second": 422.482
    },
    {
      "epoch": 1.4030345122342132,
      "grad_norm": 3.2741074562072754,
      "learning_rate": 5.980748663101604e-06,
      "loss": 0.09766849130392075,
      "num_input_tokens_seen": 3383700,
      "step": 3283,
      "train_runtime": 8009.1251,
      "train_tokens_per_second": 422.481
    },
    {
      "epoch": 1.403461908323539,
      "grad_norm": 2.3035178184509277,
      "learning_rate": 5.976470588235295e-06,
      "loss": 0.09213754534721375,
      "num_input_tokens_seen": 3384786,
      "step": 3284,
      "train_runtime": 8011.2104,
      "train_tokens_per_second": 422.506
    },
    {
      "epoch": 1.4038893044128646,
      "grad_norm": 2.2309117317199707,
      "learning_rate": 5.972192513368984e-06,
      "loss": 0.07901042699813843,
      "num_input_tokens_seen": 3385842,
      "step": 3285,
      "train_runtime": 8013.3411,
      "train_tokens_per_second": 422.526
    },
    {
      "epoch": 1.4043167005021904,
      "grad_norm": 2.4701178073883057,
      "learning_rate": 5.967914438502674e-06,
      "loss": 0.083577960729599,
      "num_input_tokens_seen": 3386646,
      "step": 3286,
      "train_runtime": 8015.3636,
      "train_tokens_per_second": 422.519
    },
    {
      "epoch": 1.4047440965915161,
      "grad_norm": 3.515695810317993,
      "learning_rate": 5.963636363636364e-06,
      "loss": 0.06064355745911598,
      "num_input_tokens_seen": 3387208,
      "step": 3287,
      "train_runtime": 8017.3234,
      "train_tokens_per_second": 422.486
    },
    {
      "epoch": 1.405171492680842,
      "grad_norm": 1.9999638795852661,
      "learning_rate": 5.959358288770054e-06,
      "loss": 0.05568959563970566,
      "num_input_tokens_seen": 3388584,
      "step": 3288,
      "train_runtime": 8019.5323,
      "train_tokens_per_second": 422.541
    },
    {
      "epoch": 1.4055988887701678,
      "grad_norm": 3.764004707336426,
      "learning_rate": 5.9550802139037436e-06,
      "loss": 0.08620937168598175,
      "num_input_tokens_seen": 3389254,
      "step": 3289,
      "train_runtime": 8021.5274,
      "train_tokens_per_second": 422.52
    },
    {
      "epoch": 1.4060262848594935,
      "grad_norm": 3.606846809387207,
      "learning_rate": 5.950802139037434e-06,
      "loss": 0.10050512850284576,
      "num_input_tokens_seen": 3389826,
      "step": 3290,
      "train_runtime": 8023.5298,
      "train_tokens_per_second": 422.486
    },
    {
      "epoch": 1.4064536809488193,
      "grad_norm": 1.91462242603302,
      "learning_rate": 5.946524064171123e-06,
      "loss": 0.07654330134391785,
      "num_input_tokens_seen": 3392164,
      "step": 3291,
      "train_runtime": 8025.9654,
      "train_tokens_per_second": 422.649
    },
    {
      "epoch": 1.406881077038145,
      "grad_norm": 2.7579917907714844,
      "learning_rate": 5.942245989304814e-06,
      "loss": 0.05657302588224411,
      "num_input_tokens_seen": 3392902,
      "step": 3292,
      "train_runtime": 8028.0984,
      "train_tokens_per_second": 422.628
    },
    {
      "epoch": 1.407308473127471,
      "grad_norm": 2.145709753036499,
      "learning_rate": 5.937967914438503e-06,
      "loss": 0.0642455592751503,
      "num_input_tokens_seen": 3393572,
      "step": 3293,
      "train_runtime": 8030.1262,
      "train_tokens_per_second": 422.605
    },
    {
      "epoch": 1.4077358692167967,
      "grad_norm": 2.3519108295440674,
      "learning_rate": 5.933689839572193e-06,
      "loss": 0.08043119311332703,
      "num_input_tokens_seen": 3394936,
      "step": 3294,
      "train_runtime": 8032.3303,
      "train_tokens_per_second": 422.659
    },
    {
      "epoch": 1.4081632653061225,
      "grad_norm": 2.605780601501465,
      "learning_rate": 5.929411764705883e-06,
      "loss": 0.08094091713428497,
      "num_input_tokens_seen": 3395768,
      "step": 3295,
      "train_runtime": 8034.3804,
      "train_tokens_per_second": 422.655
    },
    {
      "epoch": 1.4085906613954482,
      "grad_norm": 2.5442302227020264,
      "learning_rate": 5.925133689839573e-06,
      "loss": 0.10728247463703156,
      "num_input_tokens_seen": 3397042,
      "step": 3296,
      "train_runtime": 8036.5455,
      "train_tokens_per_second": 422.699
    },
    {
      "epoch": 1.409018057484774,
      "grad_norm": 3.6697146892547607,
      "learning_rate": 5.9208556149732625e-06,
      "loss": 0.0980713739991188,
      "num_input_tokens_seen": 3397924,
      "step": 3297,
      "train_runtime": 8038.6002,
      "train_tokens_per_second": 422.701
    },
    {
      "epoch": 1.4094454535741,
      "grad_norm": 3.2714905738830566,
      "learning_rate": 5.916577540106953e-06,
      "loss": 0.08470352739095688,
      "num_input_tokens_seen": 3398894,
      "step": 3298,
      "train_runtime": 8040.6876,
      "train_tokens_per_second": 422.712
    },
    {
      "epoch": 1.4098728496634256,
      "grad_norm": 2.5557289123535156,
      "learning_rate": 5.912299465240641e-06,
      "loss": 0.06353791803121567,
      "num_input_tokens_seen": 3399736,
      "step": 3299,
      "train_runtime": 8042.7452,
      "train_tokens_per_second": 422.708
    },
    {
      "epoch": 1.4103002457527514,
      "grad_norm": 3.2656362056732178,
      "learning_rate": 5.9080213903743326e-06,
      "loss": 0.1370503306388855,
      "num_input_tokens_seen": 3400642,
      "step": 3300,
      "train_runtime": 8044.8074,
      "train_tokens_per_second": 422.713
    },
    {
      "epoch": 1.410727641842077,
      "grad_norm": 1.4520354270935059,
      "learning_rate": 5.903743315508021e-06,
      "loss": 0.03744306415319443,
      "num_input_tokens_seen": 3401634,
      "step": 3301,
      "train_runtime": 8053.4928,
      "train_tokens_per_second": 422.38
    },
    {
      "epoch": 1.4111550379314028,
      "grad_norm": 2.6194164752960205,
      "learning_rate": 5.899465240641712e-06,
      "loss": 0.06881201267242432,
      "num_input_tokens_seen": 3402664,
      "step": 3302,
      "train_runtime": 8055.5522,
      "train_tokens_per_second": 422.4
    },
    {
      "epoch": 1.4115824340207288,
      "grad_norm": 2.7365314960479736,
      "learning_rate": 5.895187165775401e-06,
      "loss": 0.11222471296787262,
      "num_input_tokens_seen": 3404336,
      "step": 3303,
      "train_runtime": 8058.0261,
      "train_tokens_per_second": 422.478
    },
    {
      "epoch": 1.4120098301100545,
      "grad_norm": 2.707691192626953,
      "learning_rate": 5.890909090909091e-06,
      "loss": 0.10020206868648529,
      "num_input_tokens_seen": 3405616,
      "step": 3304,
      "train_runtime": 8060.2164,
      "train_tokens_per_second": 422.522
    },
    {
      "epoch": 1.4124372261993803,
      "grad_norm": 2.190894842147827,
      "learning_rate": 5.886631016042781e-06,
      "loss": 0.08316968381404877,
      "num_input_tokens_seen": 3407272,
      "step": 3305,
      "train_runtime": 8062.4888,
      "train_tokens_per_second": 422.608
    },
    {
      "epoch": 1.412864622288706,
      "grad_norm": 2.6829841136932373,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.08554250001907349,
      "num_input_tokens_seen": 3407982,
      "step": 3306,
      "train_runtime": 8064.5385,
      "train_tokens_per_second": 422.589
    },
    {
      "epoch": 1.4132920183780318,
      "grad_norm": 4.562150001525879,
      "learning_rate": 5.87807486631016e-06,
      "loss": 0.0754498690366745,
      "num_input_tokens_seen": 3409010,
      "step": 3307,
      "train_runtime": 8066.6763,
      "train_tokens_per_second": 422.604
    },
    {
      "epoch": 1.4137194144673577,
      "grad_norm": 2.3213205337524414,
      "learning_rate": 5.873796791443851e-06,
      "loss": 0.0716032087802887,
      "num_input_tokens_seen": 3410246,
      "step": 3308,
      "train_runtime": 8068.8521,
      "train_tokens_per_second": 422.643
    },
    {
      "epoch": 1.4141468105566835,
      "grad_norm": 2.7951669692993164,
      "learning_rate": 5.86951871657754e-06,
      "loss": 0.0717962235212326,
      "num_input_tokens_seen": 3410970,
      "step": 3309,
      "train_runtime": 8070.8783,
      "train_tokens_per_second": 422.627
    },
    {
      "epoch": 1.4145742066460092,
      "grad_norm": 2.3975892066955566,
      "learning_rate": 5.86524064171123e-06,
      "loss": 0.10723154246807098,
      "num_input_tokens_seen": 3411910,
      "step": 3310,
      "train_runtime": 8072.9585,
      "train_tokens_per_second": 422.634
    },
    {
      "epoch": 1.415001602735335,
      "grad_norm": 2.0501773357391357,
      "learning_rate": 5.86096256684492e-06,
      "loss": 0.06581204384565353,
      "num_input_tokens_seen": 3412864,
      "step": 3311,
      "train_runtime": 8075.0418,
      "train_tokens_per_second": 422.644
    },
    {
      "epoch": 1.4154289988246607,
      "grad_norm": 3.418687343597412,
      "learning_rate": 5.85668449197861e-06,
      "loss": 0.09550642967224121,
      "num_input_tokens_seen": 3413604,
      "step": 3312,
      "train_runtime": 8077.0561,
      "train_tokens_per_second": 422.63
    },
    {
      "epoch": 1.4158563949139866,
      "grad_norm": 2.386291265487671,
      "learning_rate": 5.8524064171123e-06,
      "loss": 0.076219841837883,
      "num_input_tokens_seen": 3414552,
      "step": 3313,
      "train_runtime": 8079.124,
      "train_tokens_per_second": 422.639
    },
    {
      "epoch": 1.4162837910033124,
      "grad_norm": 2.824305534362793,
      "learning_rate": 5.84812834224599e-06,
      "loss": 0.09948301315307617,
      "num_input_tokens_seen": 3415394,
      "step": 3314,
      "train_runtime": 8081.2181,
      "train_tokens_per_second": 422.634
    },
    {
      "epoch": 1.416711187092638,
      "grad_norm": 3.9405288696289062,
      "learning_rate": 5.843850267379679e-06,
      "loss": 0.07686260342597961,
      "num_input_tokens_seen": 3416112,
      "step": 3315,
      "train_runtime": 8083.3316,
      "train_tokens_per_second": 422.612
    },
    {
      "epoch": 1.4171385831819638,
      "grad_norm": 2.819261312484741,
      "learning_rate": 5.83957219251337e-06,
      "loss": 0.11186288297176361,
      "num_input_tokens_seen": 3417374,
      "step": 3316,
      "train_runtime": 8085.5186,
      "train_tokens_per_second": 422.654
    },
    {
      "epoch": 1.4175659792712896,
      "grad_norm": 2.714190721511841,
      "learning_rate": 5.835294117647059e-06,
      "loss": 0.06278377026319504,
      "num_input_tokens_seen": 3418006,
      "step": 3317,
      "train_runtime": 8087.548,
      "train_tokens_per_second": 422.626
    },
    {
      "epoch": 1.4179933753606155,
      "grad_norm": 2.5089035034179688,
      "learning_rate": 5.831016042780749e-06,
      "loss": 0.0700489729642868,
      "num_input_tokens_seen": 3418738,
      "step": 3318,
      "train_runtime": 8089.6899,
      "train_tokens_per_second": 422.604
    },
    {
      "epoch": 1.4184207714499413,
      "grad_norm": 2.058255434036255,
      "learning_rate": 5.826737967914439e-06,
      "loss": 0.07212214171886444,
      "num_input_tokens_seen": 3420098,
      "step": 3319,
      "train_runtime": 8091.915,
      "train_tokens_per_second": 422.656
    },
    {
      "epoch": 1.418848167539267,
      "grad_norm": 2.40096378326416,
      "learning_rate": 5.822459893048129e-06,
      "loss": 0.07313577085733414,
      "num_input_tokens_seen": 3420958,
      "step": 3320,
      "train_runtime": 8094.0063,
      "train_tokens_per_second": 422.653
    },
    {
      "epoch": 1.4192755636285928,
      "grad_norm": 2.945573568344116,
      "learning_rate": 5.8181818181818185e-06,
      "loss": 0.10425429046154022,
      "num_input_tokens_seen": 3421848,
      "step": 3321,
      "train_runtime": 8096.0691,
      "train_tokens_per_second": 422.655
    },
    {
      "epoch": 1.4197029597179185,
      "grad_norm": 4.813085079193115,
      "learning_rate": 5.813903743315509e-06,
      "loss": 0.10666569322347641,
      "num_input_tokens_seen": 3422528,
      "step": 3322,
      "train_runtime": 8098.147,
      "train_tokens_per_second": 422.631
    },
    {
      "epoch": 1.4201303558072444,
      "grad_norm": 2.222306251525879,
      "learning_rate": 5.809625668449198e-06,
      "loss": 0.07510511577129364,
      "num_input_tokens_seen": 3423812,
      "step": 3323,
      "train_runtime": 8100.2616,
      "train_tokens_per_second": 422.679
    },
    {
      "epoch": 1.4205577518965702,
      "grad_norm": 2.2628870010375977,
      "learning_rate": 5.8053475935828886e-06,
      "loss": 0.06365840882062912,
      "num_input_tokens_seen": 3424664,
      "step": 3324,
      "train_runtime": 8102.2914,
      "train_tokens_per_second": 422.678
    },
    {
      "epoch": 1.420985147985896,
      "grad_norm": 3.2907354831695557,
      "learning_rate": 5.801069518716578e-06,
      "loss": 0.09870798885822296,
      "num_input_tokens_seen": 3425506,
      "step": 3325,
      "train_runtime": 8104.3527,
      "train_tokens_per_second": 422.675
    },
    {
      "epoch": 1.4214125440752217,
      "grad_norm": 3.198361396789551,
      "learning_rate": 5.796791443850268e-06,
      "loss": 0.09451224654912949,
      "num_input_tokens_seen": 3426584,
      "step": 3326,
      "train_runtime": 8106.4991,
      "train_tokens_per_second": 422.696
    },
    {
      "epoch": 1.4218399401645474,
      "grad_norm": 3.202866554260254,
      "learning_rate": 5.792513368983958e-06,
      "loss": 0.08676502853631973,
      "num_input_tokens_seen": 3427396,
      "step": 3327,
      "train_runtime": 8108.5508,
      "train_tokens_per_second": 422.689
    },
    {
      "epoch": 1.4222673362538734,
      "grad_norm": 5.930995464324951,
      "learning_rate": 5.788235294117648e-06,
      "loss": 0.09072207659482956,
      "num_input_tokens_seen": 3428824,
      "step": 3328,
      "train_runtime": 8110.7922,
      "train_tokens_per_second": 422.748
    },
    {
      "epoch": 1.422694732343199,
      "grad_norm": 3.1066980361938477,
      "learning_rate": 5.7839572192513375e-06,
      "loss": 0.10657098889350891,
      "num_input_tokens_seen": 3429514,
      "step": 3329,
      "train_runtime": 8112.7824,
      "train_tokens_per_second": 422.73
    },
    {
      "epoch": 1.4231221284325248,
      "grad_norm": 1.743547797203064,
      "learning_rate": 5.779679144385028e-06,
      "loss": 0.05457301437854767,
      "num_input_tokens_seen": 3430646,
      "step": 3330,
      "train_runtime": 8115.0892,
      "train_tokens_per_second": 422.749
    },
    {
      "epoch": 1.4235495245218506,
      "grad_norm": 2.8788514137268066,
      "learning_rate": 5.775401069518717e-06,
      "loss": 0.11445778608322144,
      "num_input_tokens_seen": 3432116,
      "step": 3331,
      "train_runtime": 8124.0387,
      "train_tokens_per_second": 422.464
    },
    {
      "epoch": 1.4239769206111763,
      "grad_norm": 3.868516206741333,
      "learning_rate": 5.7711229946524075e-06,
      "loss": 0.09987609088420868,
      "num_input_tokens_seen": 3432712,
      "step": 3332,
      "train_runtime": 8126.0717,
      "train_tokens_per_second": 422.432
    },
    {
      "epoch": 1.4244043167005023,
      "grad_norm": 1.805632472038269,
      "learning_rate": 5.766844919786096e-06,
      "loss": 0.07219941914081573,
      "num_input_tokens_seen": 3434014,
      "step": 3333,
      "train_runtime": 8128.2483,
      "train_tokens_per_second": 422.479
    },
    {
      "epoch": 1.424831712789828,
      "grad_norm": 2.495518684387207,
      "learning_rate": 5.762566844919787e-06,
      "loss": 0.08428099751472473,
      "num_input_tokens_seen": 3434822,
      "step": 3334,
      "train_runtime": 8130.3134,
      "train_tokens_per_second": 422.471
    },
    {
      "epoch": 1.4252591088791537,
      "grad_norm": 2.4579107761383057,
      "learning_rate": 5.758288770053476e-06,
      "loss": 0.09721948206424713,
      "num_input_tokens_seen": 3436342,
      "step": 3335,
      "train_runtime": 8132.637,
      "train_tokens_per_second": 422.537
    },
    {
      "epoch": 1.4256865049684795,
      "grad_norm": 4.116396903991699,
      "learning_rate": 5.754010695187167e-06,
      "loss": 0.12898507714271545,
      "num_input_tokens_seen": 3437160,
      "step": 3336,
      "train_runtime": 8134.7023,
      "train_tokens_per_second": 422.531
    },
    {
      "epoch": 1.4261139010578052,
      "grad_norm": 2.652043581008911,
      "learning_rate": 5.749732620320856e-06,
      "loss": 0.06861100345849991,
      "num_input_tokens_seen": 3437986,
      "step": 3337,
      "train_runtime": 8136.7504,
      "train_tokens_per_second": 422.526
    },
    {
      "epoch": 1.4265412971471312,
      "grad_norm": 3.158684492111206,
      "learning_rate": 5.745454545454546e-06,
      "loss": 0.07517271488904953,
      "num_input_tokens_seen": 3438746,
      "step": 3338,
      "train_runtime": 8138.7835,
      "train_tokens_per_second": 422.514
    },
    {
      "epoch": 1.426968693236457,
      "grad_norm": 2.2263336181640625,
      "learning_rate": 5.741176470588235e-06,
      "loss": 0.062450092285871506,
      "num_input_tokens_seen": 3439710,
      "step": 3339,
      "train_runtime": 8140.8374,
      "train_tokens_per_second": 422.525
    },
    {
      "epoch": 1.4273960893257827,
      "grad_norm": 3.6909103393554688,
      "learning_rate": 5.736898395721926e-06,
      "loss": 0.11271899938583374,
      "num_input_tokens_seen": 3440330,
      "step": 3340,
      "train_runtime": 8142.8141,
      "train_tokens_per_second": 422.499
    },
    {
      "epoch": 1.4278234854151084,
      "grad_norm": 1.6725077629089355,
      "learning_rate": 5.732620320855615e-06,
      "loss": 0.0480821318924427,
      "num_input_tokens_seen": 3441264,
      "step": 3341,
      "train_runtime": 8144.9915,
      "train_tokens_per_second": 422.501
    },
    {
      "epoch": 1.4282508815044341,
      "grad_norm": 3.02956485748291,
      "learning_rate": 5.728342245989305e-06,
      "loss": 0.10927972197532654,
      "num_input_tokens_seen": 3442282,
      "step": 3342,
      "train_runtime": 8147.1691,
      "train_tokens_per_second": 422.513
    },
    {
      "epoch": 1.42867827759376,
      "grad_norm": 2.569598913192749,
      "learning_rate": 5.724064171122995e-06,
      "loss": 0.04497809708118439,
      "num_input_tokens_seen": 3443004,
      "step": 3343,
      "train_runtime": 8149.2002,
      "train_tokens_per_second": 422.496
    },
    {
      "epoch": 1.4291056736830858,
      "grad_norm": 1.9206790924072266,
      "learning_rate": 5.719786096256685e-06,
      "loss": 0.07463937252759933,
      "num_input_tokens_seen": 3444348,
      "step": 3344,
      "train_runtime": 8151.4019,
      "train_tokens_per_second": 422.547
    },
    {
      "epoch": 1.4295330697724116,
      "grad_norm": 3.100486993789673,
      "learning_rate": 5.7155080213903745e-06,
      "loss": 0.07265537232160568,
      "num_input_tokens_seen": 3445336,
      "step": 3345,
      "train_runtime": 8153.5082,
      "train_tokens_per_second": 422.559
    },
    {
      "epoch": 1.4299604658617373,
      "grad_norm": 2.538771867752075,
      "learning_rate": 5.711229946524065e-06,
      "loss": 0.101160928606987,
      "num_input_tokens_seen": 3446364,
      "step": 3346,
      "train_runtime": 8155.7078,
      "train_tokens_per_second": 422.571
    },
    {
      "epoch": 1.430387861951063,
      "grad_norm": 2.903012752532959,
      "learning_rate": 5.706951871657754e-06,
      "loss": 0.09906446933746338,
      "num_input_tokens_seen": 3447158,
      "step": 3347,
      "train_runtime": 8157.7787,
      "train_tokens_per_second": 422.561
    },
    {
      "epoch": 1.430815258040389,
      "grad_norm": 3.4784984588623047,
      "learning_rate": 5.702673796791445e-06,
      "loss": 0.12654723227024078,
      "num_input_tokens_seen": 3448074,
      "step": 3348,
      "train_runtime": 8159.9385,
      "train_tokens_per_second": 422.561
    },
    {
      "epoch": 1.4312426541297147,
      "grad_norm": 2.8888609409332275,
      "learning_rate": 5.698395721925134e-06,
      "loss": 0.08391845226287842,
      "num_input_tokens_seen": 3449190,
      "step": 3349,
      "train_runtime": 8162.0231,
      "train_tokens_per_second": 422.59
    },
    {
      "epoch": 1.4316700502190405,
      "grad_norm": 2.6996712684631348,
      "learning_rate": 5.694117647058824e-06,
      "loss": 0.10848686099052429,
      "num_input_tokens_seen": 3450114,
      "step": 3350,
      "train_runtime": 8164.1669,
      "train_tokens_per_second": 422.592
    },
    {
      "epoch": 1.4320974463083662,
      "grad_norm": 2.638010025024414,
      "learning_rate": 5.689839572192514e-06,
      "loss": 0.06812462210655212,
      "num_input_tokens_seen": 3450882,
      "step": 3351,
      "train_runtime": 8166.2183,
      "train_tokens_per_second": 422.58
    },
    {
      "epoch": 1.432524842397692,
      "grad_norm": 1.608130693435669,
      "learning_rate": 5.685561497326204e-06,
      "loss": 0.06419845670461655,
      "num_input_tokens_seen": 3452208,
      "step": 3352,
      "train_runtime": 8168.4642,
      "train_tokens_per_second": 422.626
    },
    {
      "epoch": 1.432952238487018,
      "grad_norm": 2.9062299728393555,
      "learning_rate": 5.6812834224598935e-06,
      "loss": 0.07693544030189514,
      "num_input_tokens_seen": 3453014,
      "step": 3353,
      "train_runtime": 8170.5475,
      "train_tokens_per_second": 422.617
    },
    {
      "epoch": 1.4333796345763437,
      "grad_norm": 1.127708911895752,
      "learning_rate": 5.677005347593584e-06,
      "loss": 0.03639684244990349,
      "num_input_tokens_seen": 3454954,
      "step": 3354,
      "train_runtime": 8172.8804,
      "train_tokens_per_second": 422.734
    },
    {
      "epoch": 1.4338070306656694,
      "grad_norm": 2.1426429748535156,
      "learning_rate": 5.672727272727273e-06,
      "loss": 0.06263867020606995,
      "num_input_tokens_seen": 3456072,
      "step": 3355,
      "train_runtime": 8174.9586,
      "train_tokens_per_second": 422.763
    },
    {
      "epoch": 1.4342344267549951,
      "grad_norm": 2.771253824234009,
      "learning_rate": 5.6684491978609635e-06,
      "loss": 0.08619582653045654,
      "num_input_tokens_seen": 3457078,
      "step": 3356,
      "train_runtime": 8177.0134,
      "train_tokens_per_second": 422.78
    },
    {
      "epoch": 1.4346618228443209,
      "grad_norm": 2.44169545173645,
      "learning_rate": 5.664171122994653e-06,
      "loss": 0.06517580896615982,
      "num_input_tokens_seen": 3457792,
      "step": 3357,
      "train_runtime": 8179.0374,
      "train_tokens_per_second": 422.763
    },
    {
      "epoch": 1.4350892189336468,
      "grad_norm": 3.2420051097869873,
      "learning_rate": 5.659893048128343e-06,
      "loss": 0.09686869382858276,
      "num_input_tokens_seen": 3458602,
      "step": 3358,
      "train_runtime": 8181.0818,
      "train_tokens_per_second": 422.756
    },
    {
      "epoch": 1.4355166150229726,
      "grad_norm": 3.132455825805664,
      "learning_rate": 5.655614973262033e-06,
      "loss": 0.08758358657360077,
      "num_input_tokens_seen": 3459278,
      "step": 3359,
      "train_runtime": 8183.0658,
      "train_tokens_per_second": 422.736
    },
    {
      "epoch": 1.4359440111122983,
      "grad_norm": 2.7000677585601807,
      "learning_rate": 5.651336898395723e-06,
      "loss": 0.07843629270792007,
      "num_input_tokens_seen": 3460328,
      "step": 3360,
      "train_runtime": 8185.1773,
      "train_tokens_per_second": 422.755
    },
    {
      "epoch": 1.436371407201624,
      "grad_norm": 2.8171896934509277,
      "learning_rate": 5.6470588235294125e-06,
      "loss": 0.0721723884344101,
      "num_input_tokens_seen": 3461146,
      "step": 3361,
      "train_runtime": 8194.0761,
      "train_tokens_per_second": 422.396
    },
    {
      "epoch": 1.4367988032909498,
      "grad_norm": 2.8880884647369385,
      "learning_rate": 5.642780748663103e-06,
      "loss": 0.05921054631471634,
      "num_input_tokens_seen": 3461900,
      "step": 3362,
      "train_runtime": 8196.1972,
      "train_tokens_per_second": 422.379
    },
    {
      "epoch": 1.4372261993802757,
      "grad_norm": 3.226677179336548,
      "learning_rate": 5.638502673796792e-06,
      "loss": 0.11903731524944305,
      "num_input_tokens_seen": 3463580,
      "step": 3363,
      "train_runtime": 8198.5634,
      "train_tokens_per_second": 422.462
    },
    {
      "epoch": 1.4376535954696015,
      "grad_norm": 1.6061899662017822,
      "learning_rate": 5.6342245989304825e-06,
      "loss": 0.04425713047385216,
      "num_input_tokens_seen": 3464586,
      "step": 3364,
      "train_runtime": 8200.8167,
      "train_tokens_per_second": 422.468
    },
    {
      "epoch": 1.4380809915589272,
      "grad_norm": 2.3322904109954834,
      "learning_rate": 5.629946524064172e-06,
      "loss": 0.06715641915798187,
      "num_input_tokens_seen": 3465548,
      "step": 3365,
      "train_runtime": 8203.1201,
      "train_tokens_per_second": 422.467
    },
    {
      "epoch": 1.438508387648253,
      "grad_norm": 2.8087809085845947,
      "learning_rate": 5.625668449197862e-06,
      "loss": 0.06320177018642426,
      "num_input_tokens_seen": 3466286,
      "step": 3366,
      "train_runtime": 8205.1548,
      "train_tokens_per_second": 422.452
    },
    {
      "epoch": 1.4389357837375787,
      "grad_norm": 3.234147787094116,
      "learning_rate": 5.621390374331551e-06,
      "loss": 0.07897357642650604,
      "num_input_tokens_seen": 3467024,
      "step": 3367,
      "train_runtime": 8207.2212,
      "train_tokens_per_second": 422.436
    },
    {
      "epoch": 1.4393631798269046,
      "grad_norm": 2.737189769744873,
      "learning_rate": 5.617112299465242e-06,
      "loss": 0.08145792037248611,
      "num_input_tokens_seen": 3467804,
      "step": 3368,
      "train_runtime": 8209.2396,
      "train_tokens_per_second": 422.427
    },
    {
      "epoch": 1.4397905759162304,
      "grad_norm": 2.7741196155548096,
      "learning_rate": 5.6128342245989306e-06,
      "loss": 0.1259549856185913,
      "num_input_tokens_seen": 3469114,
      "step": 3369,
      "train_runtime": 8211.3905,
      "train_tokens_per_second": 422.476
    },
    {
      "epoch": 1.4402179720055561,
      "grad_norm": 2.974273204803467,
      "learning_rate": 5.608556149732622e-06,
      "loss": 0.07337890565395355,
      "num_input_tokens_seen": 3469832,
      "step": 3370,
      "train_runtime": 8213.473,
      "train_tokens_per_second": 422.456
    },
    {
      "epoch": 1.4406453680948819,
      "grad_norm": 3.4330101013183594,
      "learning_rate": 5.60427807486631e-06,
      "loss": 0.09360972046852112,
      "num_input_tokens_seen": 3470740,
      "step": 3371,
      "train_runtime": 8215.6387,
      "train_tokens_per_second": 422.455
    },
    {
      "epoch": 1.4410727641842076,
      "grad_norm": 2.310922861099243,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.0742092877626419,
      "num_input_tokens_seen": 3471684,
      "step": 3372,
      "train_runtime": 8217.7899,
      "train_tokens_per_second": 422.46
    },
    {
      "epoch": 1.4415001602735336,
      "grad_norm": 3.569305419921875,
      "learning_rate": 5.59572192513369e-06,
      "loss": 0.10085339844226837,
      "num_input_tokens_seen": 3472384,
      "step": 3373,
      "train_runtime": 8219.8268,
      "train_tokens_per_second": 422.44
    },
    {
      "epoch": 1.4419275563628593,
      "grad_norm": 3.0879979133605957,
      "learning_rate": 5.59144385026738e-06,
      "loss": 0.06029132753610611,
      "num_input_tokens_seen": 3473258,
      "step": 3374,
      "train_runtime": 8221.89,
      "train_tokens_per_second": 422.44
    },
    {
      "epoch": 1.442354952452185,
      "grad_norm": 2.059502124786377,
      "learning_rate": 5.58716577540107e-06,
      "loss": 0.03092336282134056,
      "num_input_tokens_seen": 3474226,
      "step": 3375,
      "train_runtime": 8223.9475,
      "train_tokens_per_second": 422.452
    },
    {
      "epoch": 1.4427823485415108,
      "grad_norm": 2.2955570220947266,
      "learning_rate": 5.58288770053476e-06,
      "loss": 0.048130616545677185,
      "num_input_tokens_seen": 3475244,
      "step": 3376,
      "train_runtime": 8226.0419,
      "train_tokens_per_second": 422.469
    },
    {
      "epoch": 1.4432097446308365,
      "grad_norm": 2.7250657081604004,
      "learning_rate": 5.5786096256684495e-06,
      "loss": 0.08422693610191345,
      "num_input_tokens_seen": 3475994,
      "step": 3377,
      "train_runtime": 8228.0659,
      "train_tokens_per_second": 422.456
    },
    {
      "epoch": 1.4436371407201625,
      "grad_norm": 2.998913288116455,
      "learning_rate": 5.574331550802139e-06,
      "loss": 0.08139088749885559,
      "num_input_tokens_seen": 3477770,
      "step": 3378,
      "train_runtime": 8230.3395,
      "train_tokens_per_second": 422.555
    },
    {
      "epoch": 1.4440645368094882,
      "grad_norm": 2.543578624725342,
      "learning_rate": 5.570053475935829e-06,
      "loss": 0.12946677207946777,
      "num_input_tokens_seen": 3479482,
      "step": 3379,
      "train_runtime": 8232.6176,
      "train_tokens_per_second": 422.646
    },
    {
      "epoch": 1.444491932898814,
      "grad_norm": 2.661632537841797,
      "learning_rate": 5.565775401069519e-06,
      "loss": 0.07993008941411972,
      "num_input_tokens_seen": 3480278,
      "step": 3380,
      "train_runtime": 8234.6443,
      "train_tokens_per_second": 422.639
    },
    {
      "epoch": 1.4449193289881397,
      "grad_norm": 3.101353645324707,
      "learning_rate": 5.561497326203209e-06,
      "loss": 0.15083709359169006,
      "num_input_tokens_seen": 3481240,
      "step": 3381,
      "train_runtime": 8236.7788,
      "train_tokens_per_second": 422.646
    },
    {
      "epoch": 1.4453467250774654,
      "grad_norm": 2.812164783477783,
      "learning_rate": 5.5572192513368984e-06,
      "loss": 0.09883914887905121,
      "num_input_tokens_seen": 3482460,
      "step": 3382,
      "train_runtime": 8238.9317,
      "train_tokens_per_second": 422.683
    },
    {
      "epoch": 1.4457741211667914,
      "grad_norm": 2.372804880142212,
      "learning_rate": 5.552941176470589e-06,
      "loss": 0.08400600403547287,
      "num_input_tokens_seen": 3483718,
      "step": 3383,
      "train_runtime": 8241.051,
      "train_tokens_per_second": 422.727
    },
    {
      "epoch": 1.4462015172561171,
      "grad_norm": 3.749724864959717,
      "learning_rate": 5.548663101604278e-06,
      "loss": 0.1743989884853363,
      "num_input_tokens_seen": 3484536,
      "step": 3384,
      "train_runtime": 8243.0736,
      "train_tokens_per_second": 422.723
    },
    {
      "epoch": 1.4466289133454429,
      "grad_norm": 3.051945209503174,
      "learning_rate": 5.5443850267379685e-06,
      "loss": 0.13166695833206177,
      "num_input_tokens_seen": 3485586,
      "step": 3385,
      "train_runtime": 8245.1255,
      "train_tokens_per_second": 422.745
    },
    {
      "epoch": 1.4470563094347686,
      "grad_norm": 1.5855789184570312,
      "learning_rate": 5.540106951871658e-06,
      "loss": 0.0466139018535614,
      "num_input_tokens_seen": 3486818,
      "step": 3386,
      "train_runtime": 8247.2447,
      "train_tokens_per_second": 422.786
    },
    {
      "epoch": 1.4474837055240943,
      "grad_norm": 2.6617164611816406,
      "learning_rate": 5.535828877005348e-06,
      "loss": 0.10417041927576065,
      "num_input_tokens_seen": 3487750,
      "step": 3387,
      "train_runtime": 8249.3002,
      "train_tokens_per_second": 422.793
    },
    {
      "epoch": 1.4479111016134203,
      "grad_norm": 3.13396954536438,
      "learning_rate": 5.531550802139038e-06,
      "loss": 0.06696020066738129,
      "num_input_tokens_seen": 3488566,
      "step": 3388,
      "train_runtime": 8251.3547,
      "train_tokens_per_second": 422.787
    },
    {
      "epoch": 1.448338497702746,
      "grad_norm": 2.83705735206604,
      "learning_rate": 5.527272727272728e-06,
      "loss": 0.0697670727968216,
      "num_input_tokens_seen": 3489408,
      "step": 3389,
      "train_runtime": 8253.4078,
      "train_tokens_per_second": 422.784
    },
    {
      "epoch": 1.4487658937920718,
      "grad_norm": 3.204883575439453,
      "learning_rate": 5.522994652406417e-06,
      "loss": 0.13399052619934082,
      "num_input_tokens_seen": 3490594,
      "step": 3390,
      "train_runtime": 8255.583,
      "train_tokens_per_second": 422.816
    },
    {
      "epoch": 1.4491932898813975,
      "grad_norm": 2.9109880924224854,
      "learning_rate": 5.518716577540108e-06,
      "loss": 0.09673532843589783,
      "num_input_tokens_seen": 3491528,
      "step": 3391,
      "train_runtime": 8264.3021,
      "train_tokens_per_second": 422.483
    },
    {
      "epoch": 1.4496206859707232,
      "grad_norm": 3.5054914951324463,
      "learning_rate": 5.514438502673797e-06,
      "loss": 0.10953780263662338,
      "num_input_tokens_seen": 3492236,
      "step": 3392,
      "train_runtime": 8266.2936,
      "train_tokens_per_second": 422.467
    },
    {
      "epoch": 1.4500480820600492,
      "grad_norm": 1.947727918624878,
      "learning_rate": 5.510160427807487e-06,
      "loss": 0.09512528032064438,
      "num_input_tokens_seen": 3494042,
      "step": 3393,
      "train_runtime": 8268.6239,
      "train_tokens_per_second": 422.566
    },
    {
      "epoch": 1.450475478149375,
      "grad_norm": 2.8469960689544678,
      "learning_rate": 5.505882352941177e-06,
      "loss": 0.07383649051189423,
      "num_input_tokens_seen": 3494992,
      "step": 3394,
      "train_runtime": 8270.7068,
      "train_tokens_per_second": 422.575
    },
    {
      "epoch": 1.4509028742387007,
      "grad_norm": 3.785738945007324,
      "learning_rate": 5.501604278074867e-06,
      "loss": 0.10403358936309814,
      "num_input_tokens_seen": 3495698,
      "step": 3395,
      "train_runtime": 8272.7259,
      "train_tokens_per_second": 422.557
    },
    {
      "epoch": 1.4513302703280264,
      "grad_norm": 2.586132526397705,
      "learning_rate": 5.497326203208556e-06,
      "loss": 0.07209435105323792,
      "num_input_tokens_seen": 3497202,
      "step": 3396,
      "train_runtime": 8274.9042,
      "train_tokens_per_second": 422.627
    },
    {
      "epoch": 1.4517576664173522,
      "grad_norm": 2.6013777256011963,
      "learning_rate": 5.493048128342247e-06,
      "loss": 0.07098311930894852,
      "num_input_tokens_seen": 3498130,
      "step": 3397,
      "train_runtime": 8276.9821,
      "train_tokens_per_second": 422.634
    },
    {
      "epoch": 1.4521850625066781,
      "grad_norm": 1.5699478387832642,
      "learning_rate": 5.4887700534759355e-06,
      "loss": 0.06787033379077911,
      "num_input_tokens_seen": 3499556,
      "step": 3398,
      "train_runtime": 8279.19,
      "train_tokens_per_second": 422.693
    },
    {
      "epoch": 1.4526124585960039,
      "grad_norm": 3.692495822906494,
      "learning_rate": 5.484491978609627e-06,
      "loss": 0.13066597282886505,
      "num_input_tokens_seen": 3500258,
      "step": 3399,
      "train_runtime": 8281.2447,
      "train_tokens_per_second": 422.673
    },
    {
      "epoch": 1.4530398546853296,
      "grad_norm": 4.033546447753906,
      "learning_rate": 5.480213903743315e-06,
      "loss": 0.09929617494344711,
      "num_input_tokens_seen": 3500900,
      "step": 3400,
      "train_runtime": 8283.2973,
      "train_tokens_per_second": 422.646
    },
    {
      "epoch": 1.4534672507746556,
      "grad_norm": 1.5992916822433472,
      "learning_rate": 5.4759358288770055e-06,
      "loss": 0.05435958504676819,
      "num_input_tokens_seen": 3502164,
      "step": 3401,
      "train_runtime": 8285.4583,
      "train_tokens_per_second": 422.688
    },
    {
      "epoch": 1.453894646863981,
      "grad_norm": 2.5853278636932373,
      "learning_rate": 5.471657754010695e-06,
      "loss": 0.0704997181892395,
      "num_input_tokens_seen": 3503062,
      "step": 3402,
      "train_runtime": 8287.527,
      "train_tokens_per_second": 422.691
    },
    {
      "epoch": 1.454322042953307,
      "grad_norm": 2.21325945854187,
      "learning_rate": 5.467379679144385e-06,
      "loss": 0.06303384155035019,
      "num_input_tokens_seen": 3503800,
      "step": 3403,
      "train_runtime": 8289.6252,
      "train_tokens_per_second": 422.673
    },
    {
      "epoch": 1.4547494390426328,
      "grad_norm": 3.524198532104492,
      "learning_rate": 5.463101604278075e-06,
      "loss": 0.09854979813098907,
      "num_input_tokens_seen": 3504602,
      "step": 3404,
      "train_runtime": 8291.6786,
      "train_tokens_per_second": 422.665
    },
    {
      "epoch": 1.4551768351319585,
      "grad_norm": 3.52382755279541,
      "learning_rate": 5.458823529411765e-06,
      "loss": 0.06434839218854904,
      "num_input_tokens_seen": 3505770,
      "step": 3405,
      "train_runtime": 8293.8228,
      "train_tokens_per_second": 422.697
    },
    {
      "epoch": 1.4556042312212845,
      "grad_norm": 2.3930070400238037,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.0762820616364479,
      "num_input_tokens_seen": 3506734,
      "step": 3406,
      "train_runtime": 8295.901,
      "train_tokens_per_second": 422.707
    },
    {
      "epoch": 1.45603162731061,
      "grad_norm": 2.7942795753479004,
      "learning_rate": 5.450267379679145e-06,
      "loss": 0.09349891543388367,
      "num_input_tokens_seen": 3507646,
      "step": 3407,
      "train_runtime": 8297.9914,
      "train_tokens_per_second": 422.71
    },
    {
      "epoch": 1.456459023399936,
      "grad_norm": 3.1380910873413086,
      "learning_rate": 5.445989304812834e-06,
      "loss": 0.09565715491771698,
      "num_input_tokens_seen": 3509150,
      "step": 3408,
      "train_runtime": 8300.1985,
      "train_tokens_per_second": 422.779
    },
    {
      "epoch": 1.4568864194892617,
      "grad_norm": 2.1942691802978516,
      "learning_rate": 5.4417112299465245e-06,
      "loss": 0.054045066237449646,
      "num_input_tokens_seen": 3509874,
      "step": 3409,
      "train_runtime": 8302.2423,
      "train_tokens_per_second": 422.762
    },
    {
      "epoch": 1.4573138155785874,
      "grad_norm": 3.4737846851348877,
      "learning_rate": 5.437433155080214e-06,
      "loss": 0.1164722889661789,
      "num_input_tokens_seen": 3510676,
      "step": 3410,
      "train_runtime": 8304.3026,
      "train_tokens_per_second": 422.754
    },
    {
      "epoch": 1.4577412116679134,
      "grad_norm": 2.253200054168701,
      "learning_rate": 5.433155080213904e-06,
      "loss": 0.11444836854934692,
      "num_input_tokens_seen": 3511926,
      "step": 3411,
      "train_runtime": 8306.4713,
      "train_tokens_per_second": 422.794
    },
    {
      "epoch": 1.4581686077572389,
      "grad_norm": 2.9953954219818115,
      "learning_rate": 5.428877005347594e-06,
      "loss": 0.13085690140724182,
      "num_input_tokens_seen": 3512958,
      "step": 3412,
      "train_runtime": 8308.5339,
      "train_tokens_per_second": 422.813
    },
    {
      "epoch": 1.4585960038465648,
      "grad_norm": 2.8434152603149414,
      "learning_rate": 5.424598930481284e-06,
      "loss": 0.09413914382457733,
      "num_input_tokens_seen": 3513898,
      "step": 3413,
      "train_runtime": 8310.6039,
      "train_tokens_per_second": 422.821
    },
    {
      "epoch": 1.4590233999358906,
      "grad_norm": 2.961225748062134,
      "learning_rate": 5.420320855614973e-06,
      "loss": 0.11663661152124405,
      "num_input_tokens_seen": 3514840,
      "step": 3414,
      "train_runtime": 8312.7508,
      "train_tokens_per_second": 422.825
    },
    {
      "epoch": 1.4594507960252163,
      "grad_norm": 2.2813780307769775,
      "learning_rate": 5.416042780748664e-06,
      "loss": 0.08790498971939087,
      "num_input_tokens_seen": 3516314,
      "step": 3415,
      "train_runtime": 8315.0166,
      "train_tokens_per_second": 422.887
    },
    {
      "epoch": 1.4598781921145423,
      "grad_norm": 2.6718039512634277,
      "learning_rate": 5.411764705882353e-06,
      "loss": 0.0762622207403183,
      "num_input_tokens_seen": 3517254,
      "step": 3416,
      "train_runtime": 8317.1149,
      "train_tokens_per_second": 422.894
    },
    {
      "epoch": 1.4603055882038678,
      "grad_norm": 3.686269760131836,
      "learning_rate": 5.4074866310160434e-06,
      "loss": 0.08884076774120331,
      "num_input_tokens_seen": 3518342,
      "step": 3417,
      "train_runtime": 8319.2992,
      "train_tokens_per_second": 422.913
    },
    {
      "epoch": 1.4607329842931938,
      "grad_norm": 2.682656764984131,
      "learning_rate": 5.403208556149733e-06,
      "loss": 0.13189348578453064,
      "num_input_tokens_seen": 3519606,
      "step": 3418,
      "train_runtime": 8321.5507,
      "train_tokens_per_second": 422.951
    },
    {
      "epoch": 1.4611603803825195,
      "grad_norm": 7.003383636474609,
      "learning_rate": 5.398930481283423e-06,
      "loss": 0.04883841052651405,
      "num_input_tokens_seen": 3520424,
      "step": 3419,
      "train_runtime": 8323.6568,
      "train_tokens_per_second": 422.942
    },
    {
      "epoch": 1.4615877764718452,
      "grad_norm": 3.0600647926330566,
      "learning_rate": 5.394652406417113e-06,
      "loss": 0.11123605072498322,
      "num_input_tokens_seen": 3521612,
      "step": 3420,
      "train_runtime": 8325.8075,
      "train_tokens_per_second": 422.975
    },
    {
      "epoch": 1.4620151725611712,
      "grad_norm": 2.620290756225586,
      "learning_rate": 5.390374331550803e-06,
      "loss": 0.09981417655944824,
      "num_input_tokens_seen": 3522782,
      "step": 3421,
      "train_runtime": 8334.5692,
      "train_tokens_per_second": 422.671
    },
    {
      "epoch": 1.4624425686504967,
      "grad_norm": 1.8304671049118042,
      "learning_rate": 5.386096256684492e-06,
      "loss": 0.06130238622426987,
      "num_input_tokens_seen": 3524114,
      "step": 3422,
      "train_runtime": 8336.8198,
      "train_tokens_per_second": 422.717
    },
    {
      "epoch": 1.4628699647398227,
      "grad_norm": 3.1707441806793213,
      "learning_rate": 5.381818181818183e-06,
      "loss": 0.1021229550242424,
      "num_input_tokens_seen": 3525058,
      "step": 3423,
      "train_runtime": 8338.9186,
      "train_tokens_per_second": 422.724
    },
    {
      "epoch": 1.4632973608291484,
      "grad_norm": 2.6195712089538574,
      "learning_rate": 5.377540106951872e-06,
      "loss": 0.12342748790979385,
      "num_input_tokens_seen": 3526074,
      "step": 3424,
      "train_runtime": 8341.0818,
      "train_tokens_per_second": 422.736
    },
    {
      "epoch": 1.4637247569184741,
      "grad_norm": 3.001678228378296,
      "learning_rate": 5.373262032085562e-06,
      "loss": 0.08639542758464813,
      "num_input_tokens_seen": 3526878,
      "step": 3425,
      "train_runtime": 8343.1503,
      "train_tokens_per_second": 422.727
    },
    {
      "epoch": 1.4641521530078,
      "grad_norm": 2.942387104034424,
      "learning_rate": 5.368983957219252e-06,
      "loss": 0.11888884007930756,
      "num_input_tokens_seen": 3527834,
      "step": 3426,
      "train_runtime": 8345.329,
      "train_tokens_per_second": 422.732
    },
    {
      "epoch": 1.4645795490971258,
      "grad_norm": 3.7468812465667725,
      "learning_rate": 5.364705882352942e-06,
      "loss": 0.10000333935022354,
      "num_input_tokens_seen": 3528562,
      "step": 3427,
      "train_runtime": 8347.3307,
      "train_tokens_per_second": 422.717
    },
    {
      "epoch": 1.4650069451864516,
      "grad_norm": 2.5948994159698486,
      "learning_rate": 5.3604278074866316e-06,
      "loss": 0.10086463391780853,
      "num_input_tokens_seen": 3529832,
      "step": 3428,
      "train_runtime": 8349.5017,
      "train_tokens_per_second": 422.76
    },
    {
      "epoch": 1.4654343412757773,
      "grad_norm": 2.230076789855957,
      "learning_rate": 5.356149732620322e-06,
      "loss": 0.09517347067594528,
      "num_input_tokens_seen": 3531034,
      "step": 3429,
      "train_runtime": 8351.6207,
      "train_tokens_per_second": 422.796
    },
    {
      "epoch": 1.465861737365103,
      "grad_norm": 2.726621627807617,
      "learning_rate": 5.3518716577540105e-06,
      "loss": 0.10943930596113205,
      "num_input_tokens_seen": 3532090,
      "step": 3430,
      "train_runtime": 8353.7035,
      "train_tokens_per_second": 422.817
    },
    {
      "epoch": 1.466289133454429,
      "grad_norm": 3.3502588272094727,
      "learning_rate": 5.347593582887702e-06,
      "loss": 0.1161169558763504,
      "num_input_tokens_seen": 3533044,
      "step": 3431,
      "train_runtime": 8355.7607,
      "train_tokens_per_second": 422.827
    },
    {
      "epoch": 1.4667165295437548,
      "grad_norm": 2.2125492095947266,
      "learning_rate": 5.34331550802139e-06,
      "loss": 0.060301244258880615,
      "num_input_tokens_seen": 3533864,
      "step": 3432,
      "train_runtime": 8357.8302,
      "train_tokens_per_second": 422.821
    },
    {
      "epoch": 1.4671439256330805,
      "grad_norm": 2.5336380004882812,
      "learning_rate": 5.339037433155081e-06,
      "loss": 0.13681688904762268,
      "num_input_tokens_seen": 3535292,
      "step": 3433,
      "train_runtime": 8360.0613,
      "train_tokens_per_second": 422.879
    },
    {
      "epoch": 1.4675713217224062,
      "grad_norm": 2.2956624031066895,
      "learning_rate": 5.33475935828877e-06,
      "loss": 0.05246267467737198,
      "num_input_tokens_seen": 3536248,
      "step": 3434,
      "train_runtime": 8362.1942,
      "train_tokens_per_second": 422.885
    },
    {
      "epoch": 1.467998717811732,
      "grad_norm": 2.4161217212677,
      "learning_rate": 5.33048128342246e-06,
      "loss": 0.09199666976928711,
      "num_input_tokens_seen": 3537698,
      "step": 3435,
      "train_runtime": 8364.4485,
      "train_tokens_per_second": 422.945
    },
    {
      "epoch": 1.468426113901058,
      "grad_norm": 2.7502171993255615,
      "learning_rate": 5.32620320855615e-06,
      "loss": 0.11626212298870087,
      "num_input_tokens_seen": 3538998,
      "step": 3436,
      "train_runtime": 8366.6264,
      "train_tokens_per_second": 422.99
    },
    {
      "epoch": 1.4688535099903837,
      "grad_norm": 3.2480013370513916,
      "learning_rate": 5.32192513368984e-06,
      "loss": 0.08362399786710739,
      "num_input_tokens_seen": 3539902,
      "step": 3437,
      "train_runtime": 8368.9535,
      "train_tokens_per_second": 422.98
    },
    {
      "epoch": 1.4692809060797094,
      "grad_norm": 2.008153200149536,
      "learning_rate": 5.317647058823529e-06,
      "loss": 0.1897895783185959,
      "num_input_tokens_seen": 3541228,
      "step": 3438,
      "train_runtime": 8371.178,
      "train_tokens_per_second": 423.026
    },
    {
      "epoch": 1.4697083021690351,
      "grad_norm": 2.816376209259033,
      "learning_rate": 5.31336898395722e-06,
      "loss": 0.11472421884536743,
      "num_input_tokens_seen": 3542404,
      "step": 3439,
      "train_runtime": 8373.3475,
      "train_tokens_per_second": 423.057
    },
    {
      "epoch": 1.4701356982583609,
      "grad_norm": 2.9486799240112305,
      "learning_rate": 5.309090909090909e-06,
      "loss": 0.06911687552928925,
      "num_input_tokens_seen": 3543462,
      "step": 3440,
      "train_runtime": 8375.4062,
      "train_tokens_per_second": 423.079
    },
    {
      "epoch": 1.4705630943476868,
      "grad_norm": 2.5832102298736572,
      "learning_rate": 5.3048128342245995e-06,
      "loss": 0.1230064183473587,
      "num_input_tokens_seen": 3544484,
      "step": 3441,
      "train_runtime": 8377.4918,
      "train_tokens_per_second": 423.096
    },
    {
      "epoch": 1.4709904904370126,
      "grad_norm": 2.9772675037384033,
      "learning_rate": 5.300534759358289e-06,
      "loss": 0.16411425173282623,
      "num_input_tokens_seen": 3545460,
      "step": 3442,
      "train_runtime": 8379.6544,
      "train_tokens_per_second": 423.103
    },
    {
      "epoch": 1.4714178865263383,
      "grad_norm": 3.372304677963257,
      "learning_rate": 5.296256684491979e-06,
      "loss": 0.15941262245178223,
      "num_input_tokens_seen": 3546342,
      "step": 3443,
      "train_runtime": 8381.7183,
      "train_tokens_per_second": 423.104
    },
    {
      "epoch": 1.471845282615664,
      "grad_norm": 1.5999884605407715,
      "learning_rate": 5.291978609625669e-06,
      "loss": 0.07250265777111053,
      "num_input_tokens_seen": 3547926,
      "step": 3444,
      "train_runtime": 8383.9563,
      "train_tokens_per_second": 423.18
    },
    {
      "epoch": 1.4722726787049898,
      "grad_norm": 2.5111491680145264,
      "learning_rate": 5.287700534759359e-06,
      "loss": 0.073473259806633,
      "num_input_tokens_seen": 3548690,
      "step": 3445,
      "train_runtime": 8385.9834,
      "train_tokens_per_second": 423.169
    },
    {
      "epoch": 1.4727000747943157,
      "grad_norm": 1.8669384717941284,
      "learning_rate": 5.283422459893048e-06,
      "loss": 0.07209983468055725,
      "num_input_tokens_seen": 3550136,
      "step": 3446,
      "train_runtime": 8388.1591,
      "train_tokens_per_second": 423.232
    },
    {
      "epoch": 1.4731274708836415,
      "grad_norm": 2.1105058193206787,
      "learning_rate": 5.279144385026739e-06,
      "loss": 0.07174131274223328,
      "num_input_tokens_seen": 3551504,
      "step": 3447,
      "train_runtime": 8390.3067,
      "train_tokens_per_second": 423.287
    },
    {
      "epoch": 1.4735548669729672,
      "grad_norm": 2.0642714500427246,
      "learning_rate": 5.274866310160428e-06,
      "loss": 0.06614141166210175,
      "num_input_tokens_seen": 3552516,
      "step": 3448,
      "train_runtime": 8392.4425,
      "train_tokens_per_second": 423.299
    },
    {
      "epoch": 1.473982263062293,
      "grad_norm": 3.1498465538024902,
      "learning_rate": 5.270588235294118e-06,
      "loss": 0.09876853972673416,
      "num_input_tokens_seen": 3553364,
      "step": 3449,
      "train_runtime": 8394.5008,
      "train_tokens_per_second": 423.297
    },
    {
      "epoch": 1.4744096591516187,
      "grad_norm": 1.7724378108978271,
      "learning_rate": 5.266310160427808e-06,
      "loss": 0.03674537315964699,
      "num_input_tokens_seen": 3553950,
      "step": 3450,
      "train_runtime": 8396.5029,
      "train_tokens_per_second": 423.266
    },
    {
      "epoch": 1.4748370552409447,
      "grad_norm": 2.847510576248169,
      "learning_rate": 5.262032085561498e-06,
      "loss": 0.10168350487947464,
      "num_input_tokens_seen": 3555214,
      "step": 3451,
      "train_runtime": 8405.2848,
      "train_tokens_per_second": 422.974
    },
    {
      "epoch": 1.4752644513302704,
      "grad_norm": 2.6584270000457764,
      "learning_rate": 5.257754010695188e-06,
      "loss": 0.16335546970367432,
      "num_input_tokens_seen": 3556016,
      "step": 3452,
      "train_runtime": 8407.3966,
      "train_tokens_per_second": 422.963
    },
    {
      "epoch": 1.4756918474195961,
      "grad_norm": 3.1546804904937744,
      "learning_rate": 5.253475935828878e-06,
      "loss": 0.10889051109552383,
      "num_input_tokens_seen": 3556974,
      "step": 3453,
      "train_runtime": 8409.4877,
      "train_tokens_per_second": 422.972
    },
    {
      "epoch": 1.4761192435089219,
      "grad_norm": 3.6013145446777344,
      "learning_rate": 5.249197860962567e-06,
      "loss": 0.09238863736391068,
      "num_input_tokens_seen": 3557656,
      "step": 3454,
      "train_runtime": 8411.5329,
      "train_tokens_per_second": 422.95
    },
    {
      "epoch": 1.4765466395982476,
      "grad_norm": 2.0311074256896973,
      "learning_rate": 5.244919786096258e-06,
      "loss": 0.05239570140838623,
      "num_input_tokens_seen": 3558732,
      "step": 3455,
      "train_runtime": 8413.6426,
      "train_tokens_per_second": 422.972
    },
    {
      "epoch": 1.4769740356875736,
      "grad_norm": 3.5176239013671875,
      "learning_rate": 5.240641711229947e-06,
      "loss": 0.0817013829946518,
      "num_input_tokens_seen": 3559394,
      "step": 3456,
      "train_runtime": 8415.6328,
      "train_tokens_per_second": 422.95
    },
    {
      "epoch": 1.4774014317768993,
      "grad_norm": 2.136615037918091,
      "learning_rate": 5.236363636363637e-06,
      "loss": 0.08257411420345306,
      "num_input_tokens_seen": 3560308,
      "step": 3457,
      "train_runtime": 8417.7218,
      "train_tokens_per_second": 422.954
    },
    {
      "epoch": 1.477828827866225,
      "grad_norm": 2.2736639976501465,
      "learning_rate": 5.232085561497327e-06,
      "loss": 0.08371192961931229,
      "num_input_tokens_seen": 3561506,
      "step": 3458,
      "train_runtime": 8419.8506,
      "train_tokens_per_second": 422.989
    },
    {
      "epoch": 1.4782562239555508,
      "grad_norm": 1.6897475719451904,
      "learning_rate": 5.227807486631017e-06,
      "loss": 0.04890358820557594,
      "num_input_tokens_seen": 3562716,
      "step": 3459,
      "train_runtime": 8422.0198,
      "train_tokens_per_second": 423.024
    },
    {
      "epoch": 1.4786836200448765,
      "grad_norm": 2.692368745803833,
      "learning_rate": 5.2235294117647065e-06,
      "loss": 0.07973666489124298,
      "num_input_tokens_seen": 3563588,
      "step": 3460,
      "train_runtime": 8424.0966,
      "train_tokens_per_second": 423.023
    },
    {
      "epoch": 1.4791110161342025,
      "grad_norm": 2.952913761138916,
      "learning_rate": 5.219251336898397e-06,
      "loss": 0.11168647557497025,
      "num_input_tokens_seen": 3564550,
      "step": 3461,
      "train_runtime": 8426.1744,
      "train_tokens_per_second": 423.033
    },
    {
      "epoch": 1.4795384122235282,
      "grad_norm": 3.4418411254882812,
      "learning_rate": 5.214973262032086e-06,
      "loss": 0.08286750316619873,
      "num_input_tokens_seen": 3565530,
      "step": 3462,
      "train_runtime": 8428.2967,
      "train_tokens_per_second": 423.043
    },
    {
      "epoch": 1.479965808312854,
      "grad_norm": 3.002765417098999,
      "learning_rate": 5.2106951871657766e-06,
      "loss": 0.10810767114162445,
      "num_input_tokens_seen": 3566486,
      "step": 3463,
      "train_runtime": 8430.4465,
      "train_tokens_per_second": 423.048
    },
    {
      "epoch": 1.4803932044021797,
      "grad_norm": 2.9465012550354004,
      "learning_rate": 5.206417112299465e-06,
      "loss": 0.12109486758708954,
      "num_input_tokens_seen": 3567464,
      "step": 3464,
      "train_runtime": 8432.5307,
      "train_tokens_per_second": 423.06
    },
    {
      "epoch": 1.4808206004915054,
      "grad_norm": 2.2203128337860107,
      "learning_rate": 5.202139037433156e-06,
      "loss": 0.06570500880479813,
      "num_input_tokens_seen": 3568406,
      "step": 3465,
      "train_runtime": 8434.6323,
      "train_tokens_per_second": 423.066
    },
    {
      "epoch": 1.4812479965808314,
      "grad_norm": 2.604599952697754,
      "learning_rate": 5.197860962566845e-06,
      "loss": 0.07831765711307526,
      "num_input_tokens_seen": 3569380,
      "step": 3466,
      "train_runtime": 8436.6886,
      "train_tokens_per_second": 423.078
    },
    {
      "epoch": 1.4816753926701571,
      "grad_norm": 2.78183650970459,
      "learning_rate": 5.193582887700536e-06,
      "loss": 0.10326939821243286,
      "num_input_tokens_seen": 3570578,
      "step": 3467,
      "train_runtime": 8438.8033,
      "train_tokens_per_second": 423.114
    },
    {
      "epoch": 1.4821027887594829,
      "grad_norm": 2.882122278213501,
      "learning_rate": 5.189304812834225e-06,
      "loss": 0.09420980513095856,
      "num_input_tokens_seen": 3571538,
      "step": 3468,
      "train_runtime": 8440.8535,
      "train_tokens_per_second": 423.125
    },
    {
      "epoch": 1.4825301848488086,
      "grad_norm": 2.774277448654175,
      "learning_rate": 5.185026737967915e-06,
      "loss": 0.10477755963802338,
      "num_input_tokens_seen": 3572466,
      "step": 3469,
      "train_runtime": 8442.9331,
      "train_tokens_per_second": 423.131
    },
    {
      "epoch": 1.4829575809381343,
      "grad_norm": 2.6454672813415527,
      "learning_rate": 5.180748663101604e-06,
      "loss": 0.08873839676380157,
      "num_input_tokens_seen": 3573842,
      "step": 3470,
      "train_runtime": 8445.1457,
      "train_tokens_per_second": 423.183
    },
    {
      "epoch": 1.4833849770274603,
      "grad_norm": 3.13795804977417,
      "learning_rate": 5.176470588235295e-06,
      "loss": 0.08118026703596115,
      "num_input_tokens_seen": 3574652,
      "step": 3471,
      "train_runtime": 8447.2139,
      "train_tokens_per_second": 423.175
    },
    {
      "epoch": 1.483812373116786,
      "grad_norm": 1.8243160247802734,
      "learning_rate": 5.172192513368984e-06,
      "loss": 0.05078557878732681,
      "num_input_tokens_seen": 3576118,
      "step": 3472,
      "train_runtime": 8449.408,
      "train_tokens_per_second": 423.239
    },
    {
      "epoch": 1.4842397692061118,
      "grad_norm": 2.634284019470215,
      "learning_rate": 5.167914438502674e-06,
      "loss": 0.0728183388710022,
      "num_input_tokens_seen": 3576850,
      "step": 3473,
      "train_runtime": 8451.461,
      "train_tokens_per_second": 423.223
    },
    {
      "epoch": 1.4846671652954375,
      "grad_norm": 2.5489344596862793,
      "learning_rate": 5.163636363636364e-06,
      "loss": 0.08887887001037598,
      "num_input_tokens_seen": 3577670,
      "step": 3474,
      "train_runtime": 8453.4868,
      "train_tokens_per_second": 423.218
    },
    {
      "epoch": 1.4850945613847633,
      "grad_norm": 3.1579835414886475,
      "learning_rate": 5.159358288770054e-06,
      "loss": 0.07928609102964401,
      "num_input_tokens_seen": 3578458,
      "step": 3475,
      "train_runtime": 8455.5255,
      "train_tokens_per_second": 423.209
    },
    {
      "epoch": 1.4855219574740892,
      "grad_norm": 2.656999111175537,
      "learning_rate": 5.155080213903744e-06,
      "loss": 0.11406660825014114,
      "num_input_tokens_seen": 3579212,
      "step": 3476,
      "train_runtime": 8457.6903,
      "train_tokens_per_second": 423.19
    },
    {
      "epoch": 1.485949353563415,
      "grad_norm": 4.294614315032959,
      "learning_rate": 5.150802139037434e-06,
      "loss": 0.1402263641357422,
      "num_input_tokens_seen": 3580500,
      "step": 3477,
      "train_runtime": 8459.8512,
      "train_tokens_per_second": 423.234
    },
    {
      "epoch": 1.4863767496527407,
      "grad_norm": 3.577075719833374,
      "learning_rate": 5.146524064171123e-06,
      "loss": 0.1078961193561554,
      "num_input_tokens_seen": 3581280,
      "step": 3478,
      "train_runtime": 8461.975,
      "train_tokens_per_second": 423.22
    },
    {
      "epoch": 1.4868041457420664,
      "grad_norm": 2.1636970043182373,
      "learning_rate": 5.142245989304814e-06,
      "loss": 0.09547430276870728,
      "num_input_tokens_seen": 3583034,
      "step": 3479,
      "train_runtime": 8464.2985,
      "train_tokens_per_second": 423.311
    },
    {
      "epoch": 1.4872315418313922,
      "grad_norm": 2.7201411724090576,
      "learning_rate": 5.137967914438503e-06,
      "loss": 0.09909026324748993,
      "num_input_tokens_seen": 3583846,
      "step": 3480,
      "train_runtime": 8466.3273,
      "train_tokens_per_second": 423.306
    },
    {
      "epoch": 1.4876589379207181,
      "grad_norm": 3.4193906784057617,
      "learning_rate": 5.133689839572193e-06,
      "loss": 0.05734444037079811,
      "num_input_tokens_seen": 3584650,
      "step": 3481,
      "train_runtime": 8475.0719,
      "train_tokens_per_second": 422.964
    },
    {
      "epoch": 1.4880863340100439,
      "grad_norm": 2.037862539291382,
      "learning_rate": 5.129411764705883e-06,
      "loss": 0.05727183073759079,
      "num_input_tokens_seen": 3585896,
      "step": 3482,
      "train_runtime": 8477.2564,
      "train_tokens_per_second": 423.002
    },
    {
      "epoch": 1.4885137300993696,
      "grad_norm": 3.5194313526153564,
      "learning_rate": 5.125133689839573e-06,
      "loss": 0.1218360960483551,
      "num_input_tokens_seen": 3586630,
      "step": 3483,
      "train_runtime": 8479.3229,
      "train_tokens_per_second": 422.985
    },
    {
      "epoch": 1.4889411261886953,
      "grad_norm": 2.0693070888519287,
      "learning_rate": 5.1208556149732626e-06,
      "loss": 0.08063124865293503,
      "num_input_tokens_seen": 3587772,
      "step": 3484,
      "train_runtime": 8481.4351,
      "train_tokens_per_second": 423.015
    },
    {
      "epoch": 1.489368522278021,
      "grad_norm": 3.8299968242645264,
      "learning_rate": 5.116577540106953e-06,
      "loss": 0.10842914879322052,
      "num_input_tokens_seen": 3589070,
      "step": 3485,
      "train_runtime": 8483.5976,
      "train_tokens_per_second": 423.06
    },
    {
      "epoch": 1.489795918367347,
      "grad_norm": 2.5851049423217773,
      "learning_rate": 5.112299465240642e-06,
      "loss": 0.0744856745004654,
      "num_input_tokens_seen": 3589990,
      "step": 3486,
      "train_runtime": 8485.6504,
      "train_tokens_per_second": 423.066
    },
    {
      "epoch": 1.4902233144566728,
      "grad_norm": 2.627002239227295,
      "learning_rate": 5.108021390374332e-06,
      "loss": 0.08011586219072342,
      "num_input_tokens_seen": 3590824,
      "step": 3487,
      "train_runtime": 8487.7026,
      "train_tokens_per_second": 423.062
    },
    {
      "epoch": 1.4906507105459985,
      "grad_norm": 2.822197437286377,
      "learning_rate": 5.103743315508022e-06,
      "loss": 0.1270557940006256,
      "num_input_tokens_seen": 3592014,
      "step": 3488,
      "train_runtime": 8489.8198,
      "train_tokens_per_second": 423.097
    },
    {
      "epoch": 1.4910781066353243,
      "grad_norm": 2.8751938343048096,
      "learning_rate": 5.0994652406417115e-06,
      "loss": 0.12426812201738358,
      "num_input_tokens_seen": 3593174,
      "step": 3489,
      "train_runtime": 8491.9395,
      "train_tokens_per_second": 423.128
    },
    {
      "epoch": 1.49150550272465,
      "grad_norm": 2.637859344482422,
      "learning_rate": 5.095187165775402e-06,
      "loss": 0.08521677553653717,
      "num_input_tokens_seen": 3594054,
      "step": 3490,
      "train_runtime": 8493.9965,
      "train_tokens_per_second": 423.129
    },
    {
      "epoch": 1.491932898813976,
      "grad_norm": 2.4682505130767822,
      "learning_rate": 5.090909090909091e-06,
      "loss": 0.06862461566925049,
      "num_input_tokens_seen": 3594820,
      "step": 3491,
      "train_runtime": 8496.0327,
      "train_tokens_per_second": 423.117
    },
    {
      "epoch": 1.4923602949033017,
      "grad_norm": 3.118116617202759,
      "learning_rate": 5.0866310160427815e-06,
      "loss": 0.1124785989522934,
      "num_input_tokens_seen": 3595562,
      "step": 3492,
      "train_runtime": 8498.1151,
      "train_tokens_per_second": 423.101
    },
    {
      "epoch": 1.4927876909926274,
      "grad_norm": 1.9591572284698486,
      "learning_rate": 5.08235294117647e-06,
      "loss": 0.09464672207832336,
      "num_input_tokens_seen": 3597142,
      "step": 3493,
      "train_runtime": 8500.372,
      "train_tokens_per_second": 423.175
    },
    {
      "epoch": 1.4932150870819532,
      "grad_norm": 4.782775402069092,
      "learning_rate": 5.078074866310161e-06,
      "loss": 0.1299022138118744,
      "num_input_tokens_seen": 3598654,
      "step": 3494,
      "train_runtime": 8502.6312,
      "train_tokens_per_second": 423.24
    },
    {
      "epoch": 1.493642483171279,
      "grad_norm": 2.768278121948242,
      "learning_rate": 5.07379679144385e-06,
      "loss": 0.10148009657859802,
      "num_input_tokens_seen": 3599626,
      "step": 3495,
      "train_runtime": 8504.7155,
      "train_tokens_per_second": 423.251
    },
    {
      "epoch": 1.4940698792606049,
      "grad_norm": 2.4285728931427,
      "learning_rate": 5.069518716577541e-06,
      "loss": 0.08303365856409073,
      "num_input_tokens_seen": 3600584,
      "step": 3496,
      "train_runtime": 8506.8138,
      "train_tokens_per_second": 423.259
    },
    {
      "epoch": 1.4944972753499306,
      "grad_norm": 2.906404972076416,
      "learning_rate": 5.06524064171123e-06,
      "loss": 0.07900217175483704,
      "num_input_tokens_seen": 3601320,
      "step": 3497,
      "train_runtime": 8508.811,
      "train_tokens_per_second": 423.246
    },
    {
      "epoch": 1.4949246714392563,
      "grad_norm": 2.8383841514587402,
      "learning_rate": 5.06096256684492e-06,
      "loss": 0.08480820804834366,
      "num_input_tokens_seen": 3602104,
      "step": 3498,
      "train_runtime": 8510.8449,
      "train_tokens_per_second": 423.237
    },
    {
      "epoch": 1.495352067528582,
      "grad_norm": 1.7837657928466797,
      "learning_rate": 5.056684491978609e-06,
      "loss": 0.04530474543571472,
      "num_input_tokens_seen": 3603074,
      "step": 3499,
      "train_runtime": 8512.91,
      "train_tokens_per_second": 423.248
    },
    {
      "epoch": 1.4957794636179078,
      "grad_norm": 1.8503044843673706,
      "learning_rate": 5.0524064171123e-06,
      "loss": 0.045571133494377136,
      "num_input_tokens_seen": 3603946,
      "step": 3500,
      "train_runtime": 8514.9644,
      "train_tokens_per_second": 423.249
    },
    {
      "epoch": 1.4962068597072338,
      "grad_norm": 3.4572830200195312,
      "learning_rate": 5.048128342245989e-06,
      "loss": 0.10345087945461273,
      "num_input_tokens_seen": 3604912,
      "step": 3501,
      "train_runtime": 8517.1063,
      "train_tokens_per_second": 423.255
    },
    {
      "epoch": 1.4966342557965595,
      "grad_norm": 2.5880701541900635,
      "learning_rate": 5.043850267379679e-06,
      "loss": 0.08722706139087677,
      "num_input_tokens_seen": 3605644,
      "step": 3502,
      "train_runtime": 8519.1514,
      "train_tokens_per_second": 423.24
    },
    {
      "epoch": 1.4970616518858852,
      "grad_norm": 2.522632122039795,
      "learning_rate": 5.039572192513369e-06,
      "loss": 0.11368047446012497,
      "num_input_tokens_seen": 3606748,
      "step": 3503,
      "train_runtime": 8521.2704,
      "train_tokens_per_second": 423.264
    },
    {
      "epoch": 1.497489047975211,
      "grad_norm": 5.514909744262695,
      "learning_rate": 5.035294117647059e-06,
      "loss": 0.1331629753112793,
      "num_input_tokens_seen": 3607682,
      "step": 3504,
      "train_runtime": 8523.4109,
      "train_tokens_per_second": 423.267
    },
    {
      "epoch": 1.4979164440645367,
      "grad_norm": 2.5853452682495117,
      "learning_rate": 5.0310160427807485e-06,
      "loss": 0.055814750492572784,
      "num_input_tokens_seen": 3608966,
      "step": 3505,
      "train_runtime": 8525.5599,
      "train_tokens_per_second": 423.311
    },
    {
      "epoch": 1.4983438401538627,
      "grad_norm": 3.2321200370788574,
      "learning_rate": 5.026737967914439e-06,
      "loss": 0.13172556459903717,
      "num_input_tokens_seen": 3609806,
      "step": 3506,
      "train_runtime": 8527.6203,
      "train_tokens_per_second": 423.308
    },
    {
      "epoch": 1.4987712362431884,
      "grad_norm": 3.129863739013672,
      "learning_rate": 5.022459893048128e-06,
      "loss": 0.0940563827753067,
      "num_input_tokens_seen": 3610544,
      "step": 3507,
      "train_runtime": 8529.6387,
      "train_tokens_per_second": 423.294
    },
    {
      "epoch": 1.4991986323325142,
      "grad_norm": 2.606651782989502,
      "learning_rate": 5.0181818181818186e-06,
      "loss": 0.11145033687353134,
      "num_input_tokens_seen": 3611986,
      "step": 3508,
      "train_runtime": 8531.8424,
      "train_tokens_per_second": 423.354
    },
    {
      "epoch": 1.49962602842184,
      "grad_norm": 2.441556692123413,
      "learning_rate": 5.013903743315508e-06,
      "loss": 0.06653917580842972,
      "num_input_tokens_seen": 3612900,
      "step": 3509,
      "train_runtime": 8533.9113,
      "train_tokens_per_second": 423.358
    },
    {
      "epoch": 1.5000534245111656,
      "grad_norm": 5.1997222900390625,
      "learning_rate": 5.009625668449198e-06,
      "loss": 0.10496173053979874,
      "num_input_tokens_seen": 3613656,
      "step": 3510,
      "train_runtime": 8535.97,
      "train_tokens_per_second": 423.345
    },
    {
      "epoch": 1.5004808206004916,
      "grad_norm": 3.1247751712799072,
      "learning_rate": 5.005347593582888e-06,
      "loss": 0.10358445346355438,
      "num_input_tokens_seen": 3614490,
      "step": 3511,
      "train_runtime": 8544.3174,
      "train_tokens_per_second": 423.029
    },
    {
      "epoch": 1.5009082166898173,
      "grad_norm": 2.868659019470215,
      "learning_rate": 5.001069518716578e-06,
      "loss": 0.08242306113243103,
      "num_input_tokens_seen": 3615200,
      "step": 3512,
      "train_runtime": 8546.3694,
      "train_tokens_per_second": 423.01
    },
    {
      "epoch": 1.501335612779143,
      "grad_norm": 3.333298444747925,
      "learning_rate": 4.996791443850268e-06,
      "loss": 0.07817995548248291,
      "num_input_tokens_seen": 3616154,
      "step": 3513,
      "train_runtime": 8548.5881,
      "train_tokens_per_second": 423.012
    },
    {
      "epoch": 1.5017630088684688,
      "grad_norm": 3.1215336322784424,
      "learning_rate": 4.992513368983958e-06,
      "loss": 0.12454163283109665,
      "num_input_tokens_seen": 3617242,
      "step": 3514,
      "train_runtime": 8550.7275,
      "train_tokens_per_second": 423.033
    },
    {
      "epoch": 1.5021904049577945,
      "grad_norm": 2.46378231048584,
      "learning_rate": 4.988235294117647e-06,
      "loss": 0.05583369731903076,
      "num_input_tokens_seen": 3617946,
      "step": 3515,
      "train_runtime": 8552.7518,
      "train_tokens_per_second": 423.015
    },
    {
      "epoch": 1.5026178010471205,
      "grad_norm": 2.6294753551483154,
      "learning_rate": 4.9839572192513375e-06,
      "loss": 0.07819390296936035,
      "num_input_tokens_seen": 3619108,
      "step": 3516,
      "train_runtime": 8554.9581,
      "train_tokens_per_second": 423.042
    },
    {
      "epoch": 1.5030451971364462,
      "grad_norm": 2.5538837909698486,
      "learning_rate": 4.979679144385027e-06,
      "loss": 0.07925939559936523,
      "num_input_tokens_seen": 3619834,
      "step": 3517,
      "train_runtime": 8557.0006,
      "train_tokens_per_second": 423.026
    },
    {
      "epoch": 1.503472593225772,
      "grad_norm": 3.946864366531372,
      "learning_rate": 4.975401069518717e-06,
      "loss": 0.18219003081321716,
      "num_input_tokens_seen": 3620758,
      "step": 3518,
      "train_runtime": 8559.0578,
      "train_tokens_per_second": 423.032
    },
    {
      "epoch": 1.5038999893150977,
      "grad_norm": 2.776991128921509,
      "learning_rate": 4.971122994652407e-06,
      "loss": 0.09152452647686005,
      "num_input_tokens_seen": 3621780,
      "step": 3519,
      "train_runtime": 8561.1408,
      "train_tokens_per_second": 423.049
    },
    {
      "epoch": 1.5043273854044235,
      "grad_norm": 3.3464083671569824,
      "learning_rate": 4.966844919786097e-06,
      "loss": 0.07054004073143005,
      "num_input_tokens_seen": 3622286,
      "step": 3520,
      "train_runtime": 8563.0991,
      "train_tokens_per_second": 423.011
    },
    {
      "epoch": 1.5047547814937494,
      "grad_norm": 2.448549270629883,
      "learning_rate": 4.9625668449197864e-06,
      "loss": 0.07452575117349625,
      "num_input_tokens_seen": 3623258,
      "step": 3521,
      "train_runtime": 8565.1991,
      "train_tokens_per_second": 423.021
    },
    {
      "epoch": 1.5051821775830752,
      "grad_norm": 2.4839916229248047,
      "learning_rate": 4.958288770053477e-06,
      "loss": 0.0875958651304245,
      "num_input_tokens_seen": 3624150,
      "step": 3522,
      "train_runtime": 8567.2523,
      "train_tokens_per_second": 423.024
    },
    {
      "epoch": 1.505609573672401,
      "grad_norm": 2.109748125076294,
      "learning_rate": 4.954010695187166e-06,
      "loss": 0.06728150695562363,
      "num_input_tokens_seen": 3625330,
      "step": 3523,
      "train_runtime": 8569.3318,
      "train_tokens_per_second": 423.059
    },
    {
      "epoch": 1.5060369697617266,
      "grad_norm": 2.2448651790618896,
      "learning_rate": 4.9497326203208565e-06,
      "loss": 0.07205653190612793,
      "num_input_tokens_seen": 3626370,
      "step": 3524,
      "train_runtime": 8571.4497,
      "train_tokens_per_second": 423.075
    },
    {
      "epoch": 1.5064643658510524,
      "grad_norm": 2.5308432579040527,
      "learning_rate": 4.945454545454546e-06,
      "loss": 0.09219775348901749,
      "num_input_tokens_seen": 3627456,
      "step": 3525,
      "train_runtime": 8573.5777,
      "train_tokens_per_second": 423.097
    },
    {
      "epoch": 1.5068917619403783,
      "grad_norm": 3.1706414222717285,
      "learning_rate": 4.941176470588236e-06,
      "loss": 0.061785317957401276,
      "num_input_tokens_seen": 3628396,
      "step": 3526,
      "train_runtime": 8575.6669,
      "train_tokens_per_second": 423.104
    },
    {
      "epoch": 1.507319158029704,
      "grad_norm": 2.0880186557769775,
      "learning_rate": 4.936898395721926e-06,
      "loss": 0.085757315158844,
      "num_input_tokens_seen": 3629646,
      "step": 3527,
      "train_runtime": 8577.8162,
      "train_tokens_per_second": 423.143
    },
    {
      "epoch": 1.5077465541190298,
      "grad_norm": 2.876124620437622,
      "learning_rate": 4.932620320855616e-06,
      "loss": 0.09343293309211731,
      "num_input_tokens_seen": 3630838,
      "step": 3528,
      "train_runtime": 8579.9545,
      "train_tokens_per_second": 423.177
    },
    {
      "epoch": 1.5081739502083555,
      "grad_norm": 3.6000235080718994,
      "learning_rate": 4.928342245989305e-06,
      "loss": 0.08634567260742188,
      "num_input_tokens_seen": 3631472,
      "step": 3529,
      "train_runtime": 8581.9436,
      "train_tokens_per_second": 423.153
    },
    {
      "epoch": 1.5086013462976813,
      "grad_norm": 2.11653733253479,
      "learning_rate": 4.924064171122996e-06,
      "loss": 0.07055184245109558,
      "num_input_tokens_seen": 3632704,
      "step": 3530,
      "train_runtime": 8584.1136,
      "train_tokens_per_second": 423.189
    },
    {
      "epoch": 1.5090287423870072,
      "grad_norm": 2.221297264099121,
      "learning_rate": 4.919786096256685e-06,
      "loss": 0.07774901390075684,
      "num_input_tokens_seen": 3633946,
      "step": 3531,
      "train_runtime": 8586.2955,
      "train_tokens_per_second": 423.226
    },
    {
      "epoch": 1.509456138476333,
      "grad_norm": 2.4940123558044434,
      "learning_rate": 4.915508021390375e-06,
      "loss": 0.09609182924032211,
      "num_input_tokens_seen": 3634958,
      "step": 3532,
      "train_runtime": 8588.4219,
      "train_tokens_per_second": 423.239
    },
    {
      "epoch": 1.5098835345656587,
      "grad_norm": 3.6815807819366455,
      "learning_rate": 4.911229946524065e-06,
      "loss": 0.07095140218734741,
      "num_input_tokens_seen": 3635952,
      "step": 3533,
      "train_runtime": 8590.4839,
      "train_tokens_per_second": 423.253
    },
    {
      "epoch": 1.5103109306549845,
      "grad_norm": 2.821209192276001,
      "learning_rate": 4.906951871657754e-06,
      "loss": 0.06957799196243286,
      "num_input_tokens_seen": 3636766,
      "step": 3534,
      "train_runtime": 8592.5461,
      "train_tokens_per_second": 423.247
    },
    {
      "epoch": 1.5107383267443102,
      "grad_norm": 2.082606077194214,
      "learning_rate": 4.902673796791445e-06,
      "loss": 0.06596414744853973,
      "num_input_tokens_seen": 3637806,
      "step": 3535,
      "train_runtime": 8594.6247,
      "train_tokens_per_second": 423.265
    },
    {
      "epoch": 1.5111657228336361,
      "grad_norm": 2.1967203617095947,
      "learning_rate": 4.898395721925134e-06,
      "loss": 0.06473173201084137,
      "num_input_tokens_seen": 3638872,
      "step": 3536,
      "train_runtime": 8596.7642,
      "train_tokens_per_second": 423.284
    },
    {
      "epoch": 1.5115931189229619,
      "grad_norm": 2.4574005603790283,
      "learning_rate": 4.894117647058824e-06,
      "loss": 0.12428296357393265,
      "num_input_tokens_seen": 3640364,
      "step": 3537,
      "train_runtime": 8598.9708,
      "train_tokens_per_second": 423.349
    },
    {
      "epoch": 1.5120205150122876,
      "grad_norm": 2.1085562705993652,
      "learning_rate": 4.889839572192514e-06,
      "loss": 0.06412091106176376,
      "num_input_tokens_seen": 3641484,
      "step": 3538,
      "train_runtime": 8601.0831,
      "train_tokens_per_second": 423.375
    },
    {
      "epoch": 1.5124479111016136,
      "grad_norm": 2.595482349395752,
      "learning_rate": 4.885561497326204e-06,
      "loss": 0.06665106117725372,
      "num_input_tokens_seen": 3642326,
      "step": 3539,
      "train_runtime": 8603.1389,
      "train_tokens_per_second": 423.372
    },
    {
      "epoch": 1.512875307190939,
      "grad_norm": 2.2363498210906982,
      "learning_rate": 4.8812834224598935e-06,
      "loss": 0.06344731152057648,
      "num_input_tokens_seen": 3643502,
      "step": 3540,
      "train_runtime": 8605.2634,
      "train_tokens_per_second": 423.404
    },
    {
      "epoch": 1.513302703280265,
      "grad_norm": 2.690617799758911,
      "learning_rate": 4.877005347593583e-06,
      "loss": 0.1108677089214325,
      "num_input_tokens_seen": 3644406,
      "step": 3541,
      "train_runtime": 8613.6948,
      "train_tokens_per_second": 423.094
    },
    {
      "epoch": 1.5137300993695908,
      "grad_norm": 3.1423168182373047,
      "learning_rate": 4.872727272727273e-06,
      "loss": 0.09781505167484283,
      "num_input_tokens_seen": 3645192,
      "step": 3542,
      "train_runtime": 8615.7386,
      "train_tokens_per_second": 423.085
    },
    {
      "epoch": 1.5141574954589165,
      "grad_norm": 2.874504327774048,
      "learning_rate": 4.868449197860963e-06,
      "loss": 0.11589546501636505,
      "num_input_tokens_seen": 3646134,
      "step": 3543,
      "train_runtime": 8617.8246,
      "train_tokens_per_second": 423.092
    },
    {
      "epoch": 1.5145848915482425,
      "grad_norm": 2.688213586807251,
      "learning_rate": 4.864171122994652e-06,
      "loss": 0.07865557074546814,
      "num_input_tokens_seen": 3647116,
      "step": 3544,
      "train_runtime": 8619.9233,
      "train_tokens_per_second": 423.103
    },
    {
      "epoch": 1.515012287637568,
      "grad_norm": 2.785665988922119,
      "learning_rate": 4.8598930481283425e-06,
      "loss": 0.13384030759334564,
      "num_input_tokens_seen": 3648542,
      "step": 3545,
      "train_runtime": 8622.1125,
      "train_tokens_per_second": 423.161
    },
    {
      "epoch": 1.515439683726894,
      "grad_norm": 2.1130268573760986,
      "learning_rate": 4.855614973262032e-06,
      "loss": 0.04683662950992584,
      "num_input_tokens_seen": 3649582,
      "step": 3546,
      "train_runtime": 8624.2186,
      "train_tokens_per_second": 423.178
    },
    {
      "epoch": 1.5158670798162197,
      "grad_norm": 3.0648772716522217,
      "learning_rate": 4.851336898395722e-06,
      "loss": 0.1367878019809723,
      "num_input_tokens_seen": 3650584,
      "step": 3547,
      "train_runtime": 8626.3315,
      "train_tokens_per_second": 423.191
    },
    {
      "epoch": 1.5162944759055454,
      "grad_norm": 4.038280487060547,
      "learning_rate": 4.847058823529412e-06,
      "loss": 0.10116203129291534,
      "num_input_tokens_seen": 3651216,
      "step": 3548,
      "train_runtime": 8628.3617,
      "train_tokens_per_second": 423.164
    },
    {
      "epoch": 1.5167218719948714,
      "grad_norm": 4.316404342651367,
      "learning_rate": 4.842780748663102e-06,
      "loss": 0.08583742380142212,
      "num_input_tokens_seen": 3652058,
      "step": 3549,
      "train_runtime": 8630.4695,
      "train_tokens_per_second": 423.159
    },
    {
      "epoch": 1.517149268084197,
      "grad_norm": 2.402935028076172,
      "learning_rate": 4.838502673796791e-06,
      "loss": 0.08893301337957382,
      "num_input_tokens_seen": 3653200,
      "step": 3550,
      "train_runtime": 8632.6056,
      "train_tokens_per_second": 423.186
    },
    {
      "epoch": 1.5175766641735229,
      "grad_norm": 2.695525884628296,
      "learning_rate": 4.834224598930482e-06,
      "loss": 0.08129577338695526,
      "num_input_tokens_seen": 3653862,
      "step": 3551,
      "train_runtime": 8634.6028,
      "train_tokens_per_second": 423.165
    },
    {
      "epoch": 1.5180040602628486,
      "grad_norm": 2.369659662246704,
      "learning_rate": 4.829946524064171e-06,
      "loss": 0.08968549966812134,
      "num_input_tokens_seen": 3654836,
      "step": 3552,
      "train_runtime": 8636.6773,
      "train_tokens_per_second": 423.176
    },
    {
      "epoch": 1.5184314563521744,
      "grad_norm": 4.375683307647705,
      "learning_rate": 4.825668449197861e-06,
      "loss": 0.10574985295534134,
      "num_input_tokens_seen": 3655676,
      "step": 3553,
      "train_runtime": 8638.7415,
      "train_tokens_per_second": 423.172
    },
    {
      "epoch": 1.5188588524415003,
      "grad_norm": 2.773441791534424,
      "learning_rate": 4.821390374331551e-06,
      "loss": 0.09314025938510895,
      "num_input_tokens_seen": 3657138,
      "step": 3554,
      "train_runtime": 8641.0978,
      "train_tokens_per_second": 423.226
    },
    {
      "epoch": 1.5192862485308258,
      "grad_norm": 3.531966209411621,
      "learning_rate": 4.817112299465241e-06,
      "loss": 0.06882999837398529,
      "num_input_tokens_seen": 3658066,
      "step": 3555,
      "train_runtime": 8643.1606,
      "train_tokens_per_second": 423.232
    },
    {
      "epoch": 1.5197136446201518,
      "grad_norm": 2.7457752227783203,
      "learning_rate": 4.812834224598931e-06,
      "loss": 0.11129777878522873,
      "num_input_tokens_seen": 3658884,
      "step": 3556,
      "train_runtime": 8645.1887,
      "train_tokens_per_second": 423.228
    },
    {
      "epoch": 1.5201410407094775,
      "grad_norm": 2.219132900238037,
      "learning_rate": 4.808556149732621e-06,
      "loss": 0.04661089926958084,
      "num_input_tokens_seen": 3659718,
      "step": 3557,
      "train_runtime": 8647.2586,
      "train_tokens_per_second": 423.223
    },
    {
      "epoch": 1.5205684367988033,
      "grad_norm": 2.838327646255493,
      "learning_rate": 4.80427807486631e-06,
      "loss": 0.11034586280584335,
      "num_input_tokens_seen": 3660798,
      "step": 3558,
      "train_runtime": 8649.3516,
      "train_tokens_per_second": 423.245
    },
    {
      "epoch": 1.5209958328881292,
      "grad_norm": 2.868974447250366,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.09107831865549088,
      "num_input_tokens_seen": 3662264,
      "step": 3559,
      "train_runtime": 8651.5353,
      "train_tokens_per_second": 423.308
    },
    {
      "epoch": 1.5214232289774547,
      "grad_norm": 2.219090223312378,
      "learning_rate": 4.79572192513369e-06,
      "loss": 0.05881549045443535,
      "num_input_tokens_seen": 3663284,
      "step": 3560,
      "train_runtime": 8653.6207,
      "train_tokens_per_second": 423.324
    },
    {
      "epoch": 1.5218506250667807,
      "grad_norm": 2.2947075366973877,
      "learning_rate": 4.7914438502673795e-06,
      "loss": 0.09609011560678482,
      "num_input_tokens_seen": 3664528,
      "step": 3561,
      "train_runtime": 8655.7772,
      "train_tokens_per_second": 423.362
    },
    {
      "epoch": 1.5222780211561064,
      "grad_norm": 2.6582374572753906,
      "learning_rate": 4.78716577540107e-06,
      "loss": 0.0881572738289833,
      "num_input_tokens_seen": 3665326,
      "step": 3562,
      "train_runtime": 8657.7944,
      "train_tokens_per_second": 423.356
    },
    {
      "epoch": 1.5227054172454322,
      "grad_norm": 2.1353437900543213,
      "learning_rate": 4.782887700534759e-06,
      "loss": 0.08960959315299988,
      "num_input_tokens_seen": 3666748,
      "step": 3563,
      "train_runtime": 8660.0053,
      "train_tokens_per_second": 423.412
    },
    {
      "epoch": 1.5231328133347581,
      "grad_norm": 3.5151259899139404,
      "learning_rate": 4.7786096256684496e-06,
      "loss": 0.08680078387260437,
      "num_input_tokens_seen": 3667402,
      "step": 3564,
      "train_runtime": 8661.9914,
      "train_tokens_per_second": 423.39
    },
    {
      "epoch": 1.5235602094240837,
      "grad_norm": 3.3296053409576416,
      "learning_rate": 4.774331550802139e-06,
      "loss": 0.09696043282747269,
      "num_input_tokens_seen": 3668078,
      "step": 3565,
      "train_runtime": 8664.0232,
      "train_tokens_per_second": 423.369
    },
    {
      "epoch": 1.5239876055134096,
      "grad_norm": 2.413674831390381,
      "learning_rate": 4.770053475935829e-06,
      "loss": 0.08138687163591385,
      "num_input_tokens_seen": 3669300,
      "step": 3566,
      "train_runtime": 8666.157,
      "train_tokens_per_second": 423.406
    },
    {
      "epoch": 1.5244150016027354,
      "grad_norm": 2.5076205730438232,
      "learning_rate": 4.765775401069519e-06,
      "loss": 0.06423820555210114,
      "num_input_tokens_seen": 3670048,
      "step": 3567,
      "train_runtime": 8668.2095,
      "train_tokens_per_second": 423.392
    },
    {
      "epoch": 1.524842397692061,
      "grad_norm": 2.161332607269287,
      "learning_rate": 4.761497326203209e-06,
      "loss": 0.06576938927173615,
      "num_input_tokens_seen": 3671052,
      "step": 3568,
      "train_runtime": 8670.3015,
      "train_tokens_per_second": 423.405
    },
    {
      "epoch": 1.525269793781387,
      "grad_norm": 2.467561960220337,
      "learning_rate": 4.7572192513368985e-06,
      "loss": 0.057123128324747086,
      "num_input_tokens_seen": 3672232,
      "step": 3569,
      "train_runtime": 8672.4405,
      "train_tokens_per_second": 423.437
    },
    {
      "epoch": 1.5256971898707126,
      "grad_norm": 2.314906358718872,
      "learning_rate": 4.752941176470589e-06,
      "loss": 0.09102455526590347,
      "num_input_tokens_seen": 3673146,
      "step": 3570,
      "train_runtime": 8674.4993,
      "train_tokens_per_second": 423.442
    },
    {
      "epoch": 1.5261245859600385,
      "grad_norm": 3.9919824600219727,
      "learning_rate": 4.748663101604278e-06,
      "loss": 0.13130563497543335,
      "num_input_tokens_seen": 3674052,
      "step": 3571,
      "train_runtime": 8683.0978,
      "train_tokens_per_second": 423.127
    },
    {
      "epoch": 1.5265519820493643,
      "grad_norm": 3.405505657196045,
      "learning_rate": 4.7443850267379685e-06,
      "loss": 0.10028623044490814,
      "num_input_tokens_seen": 3675134,
      "step": 3572,
      "train_runtime": 8685.2214,
      "train_tokens_per_second": 423.148
    },
    {
      "epoch": 1.52697937813869,
      "grad_norm": 2.2997593879699707,
      "learning_rate": 4.740106951871658e-06,
      "loss": 0.0854053944349289,
      "num_input_tokens_seen": 3676550,
      "step": 3573,
      "train_runtime": 8687.4943,
      "train_tokens_per_second": 423.2
    },
    {
      "epoch": 1.527406774228016,
      "grad_norm": 3.0073964595794678,
      "learning_rate": 4.735828877005348e-06,
      "loss": 0.05891802906990051,
      "num_input_tokens_seen": 3677318,
      "step": 3574,
      "train_runtime": 8689.5116,
      "train_tokens_per_second": 423.19
    },
    {
      "epoch": 1.5278341703173415,
      "grad_norm": 3.070615530014038,
      "learning_rate": 4.731550802139038e-06,
      "loss": 0.11514905095100403,
      "num_input_tokens_seen": 3678310,
      "step": 3575,
      "train_runtime": 8691.602,
      "train_tokens_per_second": 423.203
    },
    {
      "epoch": 1.5282615664066674,
      "grad_norm": 2.3874621391296387,
      "learning_rate": 4.727272727272728e-06,
      "loss": 0.08202254772186279,
      "num_input_tokens_seen": 3679302,
      "step": 3576,
      "train_runtime": 8693.7024,
      "train_tokens_per_second": 423.215
    },
    {
      "epoch": 1.5286889624959932,
      "grad_norm": 2.8542189598083496,
      "learning_rate": 4.7229946524064174e-06,
      "loss": 0.2057594358921051,
      "num_input_tokens_seen": 3682632,
      "step": 3577,
      "train_runtime": 8696.3469,
      "train_tokens_per_second": 423.469
    },
    {
      "epoch": 1.529116358585319,
      "grad_norm": 3.4465928077697754,
      "learning_rate": 4.718716577540107e-06,
      "loss": 0.06976252794265747,
      "num_input_tokens_seen": 3683404,
      "step": 3578,
      "train_runtime": 8698.4066,
      "train_tokens_per_second": 423.457
    },
    {
      "epoch": 1.5295437546746449,
      "grad_norm": 3.1718413829803467,
      "learning_rate": 4.714438502673797e-06,
      "loss": 0.08370421081781387,
      "num_input_tokens_seen": 3684402,
      "step": 3579,
      "train_runtime": 8700.481,
      "train_tokens_per_second": 423.471
    },
    {
      "epoch": 1.5299711507639704,
      "grad_norm": 3.515944004058838,
      "learning_rate": 4.710160427807487e-06,
      "loss": 0.1102132722735405,
      "num_input_tokens_seen": 3685524,
      "step": 3580,
      "train_runtime": 8702.6006,
      "train_tokens_per_second": 423.497
    },
    {
      "epoch": 1.5303985468532963,
      "grad_norm": 2.4737038612365723,
      "learning_rate": 4.705882352941177e-06,
      "loss": 0.08640070259571075,
      "num_input_tokens_seen": 3686692,
      "step": 3581,
      "train_runtime": 8704.7422,
      "train_tokens_per_second": 423.527
    },
    {
      "epoch": 1.530825942942622,
      "grad_norm": 3.1986310482025146,
      "learning_rate": 4.701604278074866e-06,
      "loss": 0.13651110231876373,
      "num_input_tokens_seen": 3687594,
      "step": 3582,
      "train_runtime": 8706.7929,
      "train_tokens_per_second": 423.531
    },
    {
      "epoch": 1.5312533390319478,
      "grad_norm": 3.6478352546691895,
      "learning_rate": 4.697326203208557e-06,
      "loss": 0.10898452252149582,
      "num_input_tokens_seen": 3688340,
      "step": 3583,
      "train_runtime": 8708.8134,
      "train_tokens_per_second": 423.518
    },
    {
      "epoch": 1.5316807351212738,
      "grad_norm": 2.9175631999969482,
      "learning_rate": 4.693048128342246e-06,
      "loss": 0.07430899143218994,
      "num_input_tokens_seen": 3689224,
      "step": 3584,
      "train_runtime": 8710.8741,
      "train_tokens_per_second": 423.519
    },
    {
      "epoch": 1.5321081312105993,
      "grad_norm": 2.845033884048462,
      "learning_rate": 4.688770053475936e-06,
      "loss": 0.08749985694885254,
      "num_input_tokens_seen": 3690152,
      "step": 3585,
      "train_runtime": 8712.9521,
      "train_tokens_per_second": 423.525
    },
    {
      "epoch": 1.5325355272999253,
      "grad_norm": 3.592252731323242,
      "learning_rate": 4.684491978609626e-06,
      "loss": 0.13111191987991333,
      "num_input_tokens_seen": 3690984,
      "step": 3586,
      "train_runtime": 8714.9906,
      "train_tokens_per_second": 423.521
    },
    {
      "epoch": 1.532962923389251,
      "grad_norm": 2.714939832687378,
      "learning_rate": 4.680213903743316e-06,
      "loss": 0.10098357498645782,
      "num_input_tokens_seen": 3692010,
      "step": 3587,
      "train_runtime": 8717.1166,
      "train_tokens_per_second": 423.536
    },
    {
      "epoch": 1.5333903194785767,
      "grad_norm": 2.8202216625213623,
      "learning_rate": 4.6759358288770056e-06,
      "loss": 0.08065255731344223,
      "num_input_tokens_seen": 3693006,
      "step": 3588,
      "train_runtime": 8719.1684,
      "train_tokens_per_second": 423.55
    },
    {
      "epoch": 1.5338177155679027,
      "grad_norm": 2.5369462966918945,
      "learning_rate": 4.671657754010696e-06,
      "loss": 0.07120746374130249,
      "num_input_tokens_seen": 3694132,
      "step": 3589,
      "train_runtime": 8721.2802,
      "train_tokens_per_second": 423.577
    },
    {
      "epoch": 1.5342451116572282,
      "grad_norm": 2.47664737701416,
      "learning_rate": 4.667379679144385e-06,
      "loss": 0.0982205867767334,
      "num_input_tokens_seen": 3695768,
      "step": 3590,
      "train_runtime": 8723.6191,
      "train_tokens_per_second": 423.651
    },
    {
      "epoch": 1.5346725077465542,
      "grad_norm": 1.9944649934768677,
      "learning_rate": 4.663101604278076e-06,
      "loss": 0.04601496085524559,
      "num_input_tokens_seen": 3696788,
      "step": 3591,
      "train_runtime": 8725.6879,
      "train_tokens_per_second": 423.667
    },
    {
      "epoch": 1.53509990383588,
      "grad_norm": 2.4604697227478027,
      "learning_rate": 4.658823529411765e-06,
      "loss": 0.06514716148376465,
      "num_input_tokens_seen": 3697840,
      "step": 3592,
      "train_runtime": 8727.8502,
      "train_tokens_per_second": 423.683
    },
    {
      "epoch": 1.5355272999252056,
      "grad_norm": 2.6117424964904785,
      "learning_rate": 4.654545454545455e-06,
      "loss": 0.11933612823486328,
      "num_input_tokens_seen": 3699010,
      "step": 3593,
      "train_runtime": 8730.0155,
      "train_tokens_per_second": 423.712
    },
    {
      "epoch": 1.5359546960145316,
      "grad_norm": 2.248175621032715,
      "learning_rate": 4.650267379679145e-06,
      "loss": 0.08991917222738266,
      "num_input_tokens_seen": 3700278,
      "step": 3594,
      "train_runtime": 8732.1409,
      "train_tokens_per_second": 423.754
    },
    {
      "epoch": 1.5363820921038571,
      "grad_norm": 2.2596352100372314,
      "learning_rate": 4.645989304812834e-06,
      "loss": 0.09629835933446884,
      "num_input_tokens_seen": 3701786,
      "step": 3595,
      "train_runtime": 8734.3549,
      "train_tokens_per_second": 423.819
    },
    {
      "epoch": 1.536809488193183,
      "grad_norm": 3.0452094078063965,
      "learning_rate": 4.6417112299465245e-06,
      "loss": 0.09905613213777542,
      "num_input_tokens_seen": 3703044,
      "step": 3596,
      "train_runtime": 8736.4708,
      "train_tokens_per_second": 423.86
    },
    {
      "epoch": 1.5372368842825088,
      "grad_norm": 2.5404555797576904,
      "learning_rate": 4.637433155080214e-06,
      "loss": 0.10123807191848755,
      "num_input_tokens_seen": 3704752,
      "step": 3597,
      "train_runtime": 8738.7777,
      "train_tokens_per_second": 423.944
    },
    {
      "epoch": 1.5376642803718346,
      "grad_norm": 2.348659038543701,
      "learning_rate": 4.633155080213904e-06,
      "loss": 0.07921864092350006,
      "num_input_tokens_seen": 3705828,
      "step": 3598,
      "train_runtime": 8740.8639,
      "train_tokens_per_second": 423.966
    },
    {
      "epoch": 1.5380916764611605,
      "grad_norm": 2.0623555183410645,
      "learning_rate": 4.628877005347594e-06,
      "loss": 0.07594524323940277,
      "num_input_tokens_seen": 3707008,
      "step": 3599,
      "train_runtime": 8742.9778,
      "train_tokens_per_second": 423.998
    },
    {
      "epoch": 1.538519072550486,
      "grad_norm": 2.5412261486053467,
      "learning_rate": 4.624598930481284e-06,
      "loss": 0.0630202367901802,
      "num_input_tokens_seen": 3707884,
      "step": 3600,
      "train_runtime": 8745.0101,
      "train_tokens_per_second": 424.0
    },
    {
      "epoch": 1.538946468639812,
      "grad_norm": 1.9494917392730713,
      "learning_rate": 4.6203208556149734e-06,
      "loss": 0.09244862198829651,
      "num_input_tokens_seen": 3709478,
      "step": 3601,
      "train_runtime": 8753.5997,
      "train_tokens_per_second": 423.766
    },
    {
      "epoch": 1.5393738647291377,
      "grad_norm": 2.3527469635009766,
      "learning_rate": 4.616042780748664e-06,
      "loss": 0.057154253125190735,
      "num_input_tokens_seen": 3710488,
      "step": 3602,
      "train_runtime": 8755.7033,
      "train_tokens_per_second": 423.78
    },
    {
      "epoch": 1.5398012608184635,
      "grad_norm": 2.1083223819732666,
      "learning_rate": 4.611764705882353e-06,
      "loss": 0.08045226335525513,
      "num_input_tokens_seen": 3711738,
      "step": 3603,
      "train_runtime": 8757.9543,
      "train_tokens_per_second": 423.813
    },
    {
      "epoch": 1.5402286569077894,
      "grad_norm": 1.5975210666656494,
      "learning_rate": 4.6074866310160435e-06,
      "loss": 0.06320622563362122,
      "num_input_tokens_seen": 3713178,
      "step": 3604,
      "train_runtime": 8760.1007,
      "train_tokens_per_second": 423.874
    },
    {
      "epoch": 1.540656052997115,
      "grad_norm": 6.253113269805908,
      "learning_rate": 4.603208556149733e-06,
      "loss": 0.10745090991258621,
      "num_input_tokens_seen": 3713862,
      "step": 3605,
      "train_runtime": 8762.0925,
      "train_tokens_per_second": 423.856
    },
    {
      "epoch": 1.541083449086441,
      "grad_norm": 4.463780403137207,
      "learning_rate": 4.598930481283423e-06,
      "loss": 0.16097380220890045,
      "num_input_tokens_seen": 3715496,
      "step": 3606,
      "train_runtime": 8764.3336,
      "train_tokens_per_second": 423.934
    },
    {
      "epoch": 1.5415108451757666,
      "grad_norm": 2.8171584606170654,
      "learning_rate": 4.594652406417113e-06,
      "loss": 0.11639207601547241,
      "num_input_tokens_seen": 3716606,
      "step": 3607,
      "train_runtime": 8766.4123,
      "train_tokens_per_second": 423.96
    },
    {
      "epoch": 1.5419382412650924,
      "grad_norm": 3.250713348388672,
      "learning_rate": 4.590374331550803e-06,
      "loss": 0.08768874406814575,
      "num_input_tokens_seen": 3717358,
      "step": 3608,
      "train_runtime": 8768.4207,
      "train_tokens_per_second": 423.948
    },
    {
      "epoch": 1.5423656373544183,
      "grad_norm": 2.8987226486206055,
      "learning_rate": 4.586096256684492e-06,
      "loss": 0.08865876495838165,
      "num_input_tokens_seen": 3718354,
      "step": 3609,
      "train_runtime": 8770.4911,
      "train_tokens_per_second": 423.962
    },
    {
      "epoch": 1.5427930334437439,
      "grad_norm": 2.9552831649780273,
      "learning_rate": 4.581818181818183e-06,
      "loss": 0.11212590336799622,
      "num_input_tokens_seen": 3719314,
      "step": 3610,
      "train_runtime": 8772.5412,
      "train_tokens_per_second": 423.972
    },
    {
      "epoch": 1.5432204295330698,
      "grad_norm": 1.9415446519851685,
      "learning_rate": 4.577540106951872e-06,
      "loss": 0.08809635043144226,
      "num_input_tokens_seen": 3720748,
      "step": 3611,
      "train_runtime": 8774.7312,
      "train_tokens_per_second": 424.03
    },
    {
      "epoch": 1.5436478256223956,
      "grad_norm": 2.2836174964904785,
      "learning_rate": 4.573262032085562e-06,
      "loss": 0.09118485450744629,
      "num_input_tokens_seen": 3722006,
      "step": 3612,
      "train_runtime": 8776.8862,
      "train_tokens_per_second": 424.069
    },
    {
      "epoch": 1.5440752217117213,
      "grad_norm": 2.4618706703186035,
      "learning_rate": 4.568983957219252e-06,
      "loss": 0.07965587824583054,
      "num_input_tokens_seen": 3723140,
      "step": 3613,
      "train_runtime": 8779.0024,
      "train_tokens_per_second": 424.096
    },
    {
      "epoch": 1.5445026178010473,
      "grad_norm": 1.9284067153930664,
      "learning_rate": 4.564705882352941e-06,
      "loss": 0.06568905711174011,
      "num_input_tokens_seen": 3724454,
      "step": 3614,
      "train_runtime": 8781.1449,
      "train_tokens_per_second": 424.142
    },
    {
      "epoch": 1.5449300138903728,
      "grad_norm": 2.652906656265259,
      "learning_rate": 4.560427807486632e-06,
      "loss": 0.07547233253717422,
      "num_input_tokens_seen": 3725486,
      "step": 3615,
      "train_runtime": 8783.261,
      "train_tokens_per_second": 424.157
    },
    {
      "epoch": 1.5453574099796987,
      "grad_norm": 2.633727788925171,
      "learning_rate": 4.556149732620321e-06,
      "loss": 0.07704445719718933,
      "num_input_tokens_seen": 3726512,
      "step": 3616,
      "train_runtime": 8785.3309,
      "train_tokens_per_second": 424.174
    },
    {
      "epoch": 1.5457848060690245,
      "grad_norm": 2.2409443855285645,
      "learning_rate": 4.551871657754011e-06,
      "loss": 0.08279399573802948,
      "num_input_tokens_seen": 3727586,
      "step": 3617,
      "train_runtime": 8787.4152,
      "train_tokens_per_second": 424.196
    },
    {
      "epoch": 1.5462122021583502,
      "grad_norm": 1.0659631490707397,
      "learning_rate": 4.547593582887701e-06,
      "loss": 0.025751378387212753,
      "num_input_tokens_seen": 3730588,
      "step": 3618,
      "train_runtime": 8789.9953,
      "train_tokens_per_second": 424.413
    },
    {
      "epoch": 1.5466395982476762,
      "grad_norm": 2.6087591648101807,
      "learning_rate": 4.543315508021391e-06,
      "loss": 0.08426611125469208,
      "num_input_tokens_seen": 3731650,
      "step": 3619,
      "train_runtime": 8792.0707,
      "train_tokens_per_second": 424.434
    },
    {
      "epoch": 1.5470669943370017,
      "grad_norm": 2.809187173843384,
      "learning_rate": 4.5390374331550805e-06,
      "loss": 0.09342777729034424,
      "num_input_tokens_seen": 3732664,
      "step": 3620,
      "train_runtime": 8794.1764,
      "train_tokens_per_second": 424.447
    },
    {
      "epoch": 1.5474943904263276,
      "grad_norm": 2.9087748527526855,
      "learning_rate": 4.534759358288771e-06,
      "loss": 0.07650496810674667,
      "num_input_tokens_seen": 3733644,
      "step": 3621,
      "train_runtime": 8796.2817,
      "train_tokens_per_second": 424.457
    },
    {
      "epoch": 1.5479217865156534,
      "grad_norm": 1.5646504163742065,
      "learning_rate": 4.53048128342246e-06,
      "loss": 0.06139673292636871,
      "num_input_tokens_seen": 3735232,
      "step": 3622,
      "train_runtime": 8798.5094,
      "train_tokens_per_second": 424.53
    },
    {
      "epoch": 1.5483491826049791,
      "grad_norm": 2.5629093647003174,
      "learning_rate": 4.5262032085561506e-06,
      "loss": 0.09025517106056213,
      "num_input_tokens_seen": 3736016,
      "step": 3623,
      "train_runtime": 8800.5451,
      "train_tokens_per_second": 424.521
    },
    {
      "epoch": 1.548776578694305,
      "grad_norm": 2.0147087574005127,
      "learning_rate": 4.52192513368984e-06,
      "loss": 0.07959703356027603,
      "num_input_tokens_seen": 3737764,
      "step": 3624,
      "train_runtime": 8802.8712,
      "train_tokens_per_second": 424.607
    },
    {
      "epoch": 1.5492039747836306,
      "grad_norm": 2.742891550064087,
      "learning_rate": 4.51764705882353e-06,
      "loss": 0.07095218449831009,
      "num_input_tokens_seen": 3738586,
      "step": 3625,
      "train_runtime": 8804.915,
      "train_tokens_per_second": 424.602
    },
    {
      "epoch": 1.5496313708729565,
      "grad_norm": 1.4911378622055054,
      "learning_rate": 4.51336898395722e-06,
      "loss": 0.05005992203950882,
      "num_input_tokens_seen": 3739918,
      "step": 3626,
      "train_runtime": 8807.0623,
      "train_tokens_per_second": 424.65
    },
    {
      "epoch": 1.5500587669622823,
      "grad_norm": 2.6787283420562744,
      "learning_rate": 4.50909090909091e-06,
      "loss": 0.07647174596786499,
      "num_input_tokens_seen": 3740700,
      "step": 3627,
      "train_runtime": 8809.0578,
      "train_tokens_per_second": 424.642
    },
    {
      "epoch": 1.550486163051608,
      "grad_norm": 2.4261646270751953,
      "learning_rate": 4.5048128342245995e-06,
      "loss": 0.06149151921272278,
      "num_input_tokens_seen": 3741490,
      "step": 3628,
      "train_runtime": 8811.0905,
      "train_tokens_per_second": 424.634
    },
    {
      "epoch": 1.550913559140934,
      "grad_norm": 2.779052734375,
      "learning_rate": 4.500534759358289e-06,
      "loss": 0.08920960873365402,
      "num_input_tokens_seen": 3742304,
      "step": 3629,
      "train_runtime": 8813.1086,
      "train_tokens_per_second": 424.629
    },
    {
      "epoch": 1.5513409552302595,
      "grad_norm": 2.4222524166107178,
      "learning_rate": 4.496256684491979e-06,
      "loss": 0.11216238886117935,
      "num_input_tokens_seen": 3743304,
      "step": 3630,
      "train_runtime": 8815.1887,
      "train_tokens_per_second": 424.643
    },
    {
      "epoch": 1.5517683513195855,
      "grad_norm": 3.995540142059326,
      "learning_rate": 4.491978609625669e-06,
      "loss": 0.08351875841617584,
      "num_input_tokens_seen": 3743772,
      "step": 3631,
      "train_runtime": 8823.8094,
      "train_tokens_per_second": 424.281
    },
    {
      "epoch": 1.5521957474089112,
      "grad_norm": 2.49881911277771,
      "learning_rate": 4.487700534759359e-06,
      "loss": 0.07293374836444855,
      "num_input_tokens_seen": 3744968,
      "step": 3632,
      "train_runtime": 8825.9527,
      "train_tokens_per_second": 424.313
    },
    {
      "epoch": 1.552623143498237,
      "grad_norm": 2.3901586532592773,
      "learning_rate": 4.483422459893048e-06,
      "loss": 0.08552029728889465,
      "num_input_tokens_seen": 3746094,
      "step": 3633,
      "train_runtime": 8828.2474,
      "train_tokens_per_second": 424.33
    },
    {
      "epoch": 1.553050539587563,
      "grad_norm": 2.1088755130767822,
      "learning_rate": 4.479144385026739e-06,
      "loss": 0.08151990175247192,
      "num_input_tokens_seen": 3747156,
      "step": 3634,
      "train_runtime": 8830.3601,
      "train_tokens_per_second": 424.349
    },
    {
      "epoch": 1.5534779356768884,
      "grad_norm": 2.7339720726013184,
      "learning_rate": 4.474866310160428e-06,
      "loss": 0.09420132637023926,
      "num_input_tokens_seen": 3748070,
      "step": 3635,
      "train_runtime": 8832.4166,
      "train_tokens_per_second": 424.354
    },
    {
      "epoch": 1.5539053317662144,
      "grad_norm": 3.7338931560516357,
      "learning_rate": 4.4705882352941184e-06,
      "loss": 0.13226300477981567,
      "num_input_tokens_seen": 3748842,
      "step": 3636,
      "train_runtime": 8834.4293,
      "train_tokens_per_second": 424.345
    },
    {
      "epoch": 1.55433272785554,
      "grad_norm": 3.0476105213165283,
      "learning_rate": 4.466310160427808e-06,
      "loss": 0.06818005442619324,
      "num_input_tokens_seen": 3750710,
      "step": 3637,
      "train_runtime": 8836.7178,
      "train_tokens_per_second": 424.446
    },
    {
      "epoch": 1.5547601239448658,
      "grad_norm": 2.1895222663879395,
      "learning_rate": 4.462032085561498e-06,
      "loss": 0.09431041777133942,
      "num_input_tokens_seen": 3751582,
      "step": 3638,
      "train_runtime": 8838.7576,
      "train_tokens_per_second": 424.447
    },
    {
      "epoch": 1.5551875200341918,
      "grad_norm": 2.80151629447937,
      "learning_rate": 4.457754010695188e-06,
      "loss": 0.11778023093938828,
      "num_input_tokens_seen": 3752406,
      "step": 3639,
      "train_runtime": 8840.7865,
      "train_tokens_per_second": 424.443
    },
    {
      "epoch": 1.5556149161235173,
      "grad_norm": 2.352405071258545,
      "learning_rate": 4.453475935828878e-06,
      "loss": 0.06863915920257568,
      "num_input_tokens_seen": 3753206,
      "step": 3640,
      "train_runtime": 8842.8709,
      "train_tokens_per_second": 424.433
    },
    {
      "epoch": 1.5560423122128433,
      "grad_norm": 2.1238908767700195,
      "learning_rate": 4.449197860962567e-06,
      "loss": 0.056637659668922424,
      "num_input_tokens_seen": 3754258,
      "step": 3641,
      "train_runtime": 8844.9687,
      "train_tokens_per_second": 424.451
    },
    {
      "epoch": 1.556469708302169,
      "grad_norm": 2.046593189239502,
      "learning_rate": 4.444919786096258e-06,
      "loss": 0.037819236516952515,
      "num_input_tokens_seen": 3754878,
      "step": 3642,
      "train_runtime": 8846.9281,
      "train_tokens_per_second": 424.427
    },
    {
      "epoch": 1.5568971043914948,
      "grad_norm": 3.086535692214966,
      "learning_rate": 4.440641711229947e-06,
      "loss": 0.09567147493362427,
      "num_input_tokens_seen": 3755606,
      "step": 3643,
      "train_runtime": 8848.9613,
      "train_tokens_per_second": 424.412
    },
    {
      "epoch": 1.5573245004808207,
      "grad_norm": 3.7156472206115723,
      "learning_rate": 4.436363636363637e-06,
      "loss": 0.1002168282866478,
      "num_input_tokens_seen": 3756278,
      "step": 3644,
      "train_runtime": 8850.909,
      "train_tokens_per_second": 424.395
    },
    {
      "epoch": 1.5577518965701462,
      "grad_norm": 3.008974552154541,
      "learning_rate": 4.432085561497327e-06,
      "loss": 0.09877640008926392,
      "num_input_tokens_seen": 3757170,
      "step": 3645,
      "train_runtime": 8852.9649,
      "train_tokens_per_second": 424.397
    },
    {
      "epoch": 1.5581792926594722,
      "grad_norm": 1.9847075939178467,
      "learning_rate": 4.427807486631016e-06,
      "loss": 0.07963381707668304,
      "num_input_tokens_seen": 3758686,
      "step": 3646,
      "train_runtime": 8855.1669,
      "train_tokens_per_second": 424.462
    },
    {
      "epoch": 1.558606688748798,
      "grad_norm": 2.562413215637207,
      "learning_rate": 4.423529411764707e-06,
      "loss": 0.07942426949739456,
      "num_input_tokens_seen": 3759634,
      "step": 3647,
      "train_runtime": 8857.2409,
      "train_tokens_per_second": 424.47
    },
    {
      "epoch": 1.5590340848381237,
      "grad_norm": 2.991321325302124,
      "learning_rate": 4.419251336898396e-06,
      "loss": 0.05702042579650879,
      "num_input_tokens_seen": 3760228,
      "step": 3648,
      "train_runtime": 8859.2119,
      "train_tokens_per_second": 424.443
    },
    {
      "epoch": 1.5594614809274496,
      "grad_norm": 1.9916613101959229,
      "learning_rate": 4.4149732620320855e-06,
      "loss": 0.07082544267177582,
      "num_input_tokens_seen": 3761596,
      "step": 3649,
      "train_runtime": 8861.3961,
      "train_tokens_per_second": 424.492
    },
    {
      "epoch": 1.5598888770167751,
      "grad_norm": 3.1427340507507324,
      "learning_rate": 4.410695187165776e-06,
      "loss": 0.08179493993520737,
      "num_input_tokens_seen": 3762408,
      "step": 3650,
      "train_runtime": 8863.4362,
      "train_tokens_per_second": 424.486
    },
    {
      "epoch": 1.560316273106101,
      "grad_norm": 5.308429718017578,
      "learning_rate": 4.406417112299465e-06,
      "loss": 0.08196905255317688,
      "num_input_tokens_seen": 3764016,
      "step": 3651,
      "train_runtime": 8865.6432,
      "train_tokens_per_second": 424.562
    },
    {
      "epoch": 1.5607436691954268,
      "grad_norm": 2.4211533069610596,
      "learning_rate": 4.4021390374331555e-06,
      "loss": 0.0805337131023407,
      "num_input_tokens_seen": 3765242,
      "step": 3652,
      "train_runtime": 8867.8062,
      "train_tokens_per_second": 424.597
    },
    {
      "epoch": 1.5611710652847526,
      "grad_norm": 2.5551466941833496,
      "learning_rate": 4.397860962566845e-06,
      "loss": 0.0831645280122757,
      "num_input_tokens_seen": 3766200,
      "step": 3653,
      "train_runtime": 8869.8818,
      "train_tokens_per_second": 424.605
    },
    {
      "epoch": 1.5615984613740785,
      "grad_norm": 1.8627734184265137,
      "learning_rate": 4.393582887700535e-06,
      "loss": 0.06860567629337311,
      "num_input_tokens_seen": 3767436,
      "step": 3654,
      "train_runtime": 8872.0258,
      "train_tokens_per_second": 424.642
    },
    {
      "epoch": 1.562025857463404,
      "grad_norm": 3.857524871826172,
      "learning_rate": 4.389304812834225e-06,
      "loss": 0.09804932773113251,
      "num_input_tokens_seen": 3768572,
      "step": 3655,
      "train_runtime": 8874.1348,
      "train_tokens_per_second": 424.669
    },
    {
      "epoch": 1.56245325355273,
      "grad_norm": 2.945697546005249,
      "learning_rate": 4.385026737967915e-06,
      "loss": 0.0862235277891159,
      "num_input_tokens_seen": 3769582,
      "step": 3656,
      "train_runtime": 8876.2154,
      "train_tokens_per_second": 424.683
    },
    {
      "epoch": 1.5628806496420558,
      "grad_norm": 2.284696102142334,
      "learning_rate": 4.3807486631016044e-06,
      "loss": 0.05801151692867279,
      "num_input_tokens_seen": 3770462,
      "step": 3657,
      "train_runtime": 8878.2637,
      "train_tokens_per_second": 424.685
    },
    {
      "epoch": 1.5633080457313815,
      "grad_norm": 2.2022407054901123,
      "learning_rate": 4.376470588235294e-06,
      "loss": 0.06683167815208435,
      "num_input_tokens_seen": 3771968,
      "step": 3658,
      "train_runtime": 8880.4685,
      "train_tokens_per_second": 424.749
    },
    {
      "epoch": 1.5637354418207075,
      "grad_norm": 3.813249349594116,
      "learning_rate": 4.372192513368984e-06,
      "loss": 0.14138197898864746,
      "num_input_tokens_seen": 3773654,
      "step": 3659,
      "train_runtime": 8882.7502,
      "train_tokens_per_second": 424.829
    },
    {
      "epoch": 1.564162837910033,
      "grad_norm": 3.5508217811584473,
      "learning_rate": 4.367914438502674e-06,
      "loss": 0.07752018421888351,
      "num_input_tokens_seen": 3774502,
      "step": 3660,
      "train_runtime": 8884.8096,
      "train_tokens_per_second": 424.826
    },
    {
      "epoch": 1.564590233999359,
      "grad_norm": 1.9456413984298706,
      "learning_rate": 4.363636363636364e-06,
      "loss": 0.07442499697208405,
      "num_input_tokens_seen": 3775572,
      "step": 3661,
      "train_runtime": 8893.6147,
      "train_tokens_per_second": 424.526
    },
    {
      "epoch": 1.5650176300886847,
      "grad_norm": 2.7564380168914795,
      "learning_rate": 4.359358288770053e-06,
      "loss": 0.0837429016828537,
      "num_input_tokens_seen": 3776356,
      "step": 3662,
      "train_runtime": 8895.6281,
      "train_tokens_per_second": 424.518
    },
    {
      "epoch": 1.5654450261780104,
      "grad_norm": 2.5858020782470703,
      "learning_rate": 4.355080213903744e-06,
      "loss": 0.07261614501476288,
      "num_input_tokens_seen": 3777482,
      "step": 3663,
      "train_runtime": 8897.8883,
      "train_tokens_per_second": 424.537
    },
    {
      "epoch": 1.5658724222673364,
      "grad_norm": 2.613537311553955,
      "learning_rate": 4.350802139037433e-06,
      "loss": 0.0808013379573822,
      "num_input_tokens_seen": 3778390,
      "step": 3664,
      "train_runtime": 8899.9412,
      "train_tokens_per_second": 424.541
    },
    {
      "epoch": 1.5662998183566619,
      "grad_norm": 3.2757058143615723,
      "learning_rate": 4.346524064171123e-06,
      "loss": 0.059987809509038925,
      "num_input_tokens_seen": 3779034,
      "step": 3665,
      "train_runtime": 8901.9297,
      "train_tokens_per_second": 424.519
    },
    {
      "epoch": 1.5667272144459878,
      "grad_norm": 2.049184799194336,
      "learning_rate": 4.342245989304813e-06,
      "loss": 0.08657336235046387,
      "num_input_tokens_seen": 3781032,
      "step": 3666,
      "train_runtime": 8904.2584,
      "train_tokens_per_second": 424.632
    },
    {
      "epoch": 1.5671546105353136,
      "grad_norm": 3.0423758029937744,
      "learning_rate": 4.337967914438503e-06,
      "loss": 0.08478708565235138,
      "num_input_tokens_seen": 3781806,
      "step": 3667,
      "train_runtime": 8906.2786,
      "train_tokens_per_second": 424.622
    },
    {
      "epoch": 1.5675820066246393,
      "grad_norm": 1.600178599357605,
      "learning_rate": 4.3336898395721926e-06,
      "loss": 0.058786340057849884,
      "num_input_tokens_seen": 3782970,
      "step": 3668,
      "train_runtime": 8908.3911,
      "train_tokens_per_second": 424.652
    },
    {
      "epoch": 1.5680094027139653,
      "grad_norm": 2.5816683769226074,
      "learning_rate": 4.329411764705883e-06,
      "loss": 0.09872876852750778,
      "num_input_tokens_seen": 3784280,
      "step": 3669,
      "train_runtime": 8910.5328,
      "train_tokens_per_second": 424.697
    },
    {
      "epoch": 1.5684367988032908,
      "grad_norm": 4.025772571563721,
      "learning_rate": 4.325133689839572e-06,
      "loss": 0.09428335726261139,
      "num_input_tokens_seen": 3785344,
      "step": 3670,
      "train_runtime": 8912.6218,
      "train_tokens_per_second": 424.717
    },
    {
      "epoch": 1.5688641948926167,
      "grad_norm": 3.1545445919036865,
      "learning_rate": 4.320855614973263e-06,
      "loss": 0.04300302267074585,
      "num_input_tokens_seen": 3786258,
      "step": 3671,
      "train_runtime": 8914.7065,
      "train_tokens_per_second": 424.72
    },
    {
      "epoch": 1.5692915909819425,
      "grad_norm": 2.8887295722961426,
      "learning_rate": 4.316577540106952e-06,
      "loss": 0.07493719458580017,
      "num_input_tokens_seen": 3787092,
      "step": 3672,
      "train_runtime": 8916.7252,
      "train_tokens_per_second": 424.718
    },
    {
      "epoch": 1.5697189870712682,
      "grad_norm": 2.8340725898742676,
      "learning_rate": 4.312299465240642e-06,
      "loss": 0.12625624239444733,
      "num_input_tokens_seen": 3788700,
      "step": 3673,
      "train_runtime": 8918.955,
      "train_tokens_per_second": 424.792
    },
    {
      "epoch": 1.5701463831605942,
      "grad_norm": 4.254677772521973,
      "learning_rate": 4.308021390374332e-06,
      "loss": 0.10435140877962112,
      "num_input_tokens_seen": 3789418,
      "step": 3674,
      "train_runtime": 8920.9457,
      "train_tokens_per_second": 424.778
    },
    {
      "epoch": 1.5705737792499197,
      "grad_norm": 2.665790557861328,
      "learning_rate": 4.303743315508021e-06,
      "loss": 0.08641651272773743,
      "num_input_tokens_seen": 3790558,
      "step": 3675,
      "train_runtime": 8923.0677,
      "train_tokens_per_second": 424.804
    },
    {
      "epoch": 1.5710011753392457,
      "grad_norm": 2.5489273071289062,
      "learning_rate": 4.2994652406417115e-06,
      "loss": 0.08997465670108795,
      "num_input_tokens_seen": 3791676,
      "step": 3676,
      "train_runtime": 8925.2325,
      "train_tokens_per_second": 424.827
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 1.9713928699493408,
      "learning_rate": 4.295187165775401e-06,
      "loss": 0.04911072179675102,
      "num_input_tokens_seen": 3792800,
      "step": 3677,
      "train_runtime": 8927.3362,
      "train_tokens_per_second": 424.852
    },
    {
      "epoch": 1.5718559675178971,
      "grad_norm": 3.2004568576812744,
      "learning_rate": 4.290909090909091e-06,
      "loss": 0.09403054416179657,
      "num_input_tokens_seen": 3793720,
      "step": 3678,
      "train_runtime": 8929.4043,
      "train_tokens_per_second": 424.857
    },
    {
      "epoch": 1.572283363607223,
      "grad_norm": 2.9870142936706543,
      "learning_rate": 4.286631016042781e-06,
      "loss": 0.09138641506433487,
      "num_input_tokens_seen": 3794584,
      "step": 3679,
      "train_runtime": 8931.4467,
      "train_tokens_per_second": 424.857
    },
    {
      "epoch": 1.5727107596965486,
      "grad_norm": 3.7452101707458496,
      "learning_rate": 4.282352941176471e-06,
      "loss": 0.10864057391881943,
      "num_input_tokens_seen": 3795204,
      "step": 3680,
      "train_runtime": 8933.4038,
      "train_tokens_per_second": 424.833
    },
    {
      "epoch": 1.5731381557858746,
      "grad_norm": 3.3195242881774902,
      "learning_rate": 4.2780748663101604e-06,
      "loss": 0.10429537296295166,
      "num_input_tokens_seen": 3796192,
      "step": 3681,
      "train_runtime": 8935.4692,
      "train_tokens_per_second": 424.845
    },
    {
      "epoch": 1.5735655518752003,
      "grad_norm": 3.5783767700195312,
      "learning_rate": 4.273796791443851e-06,
      "loss": 0.07717061042785645,
      "num_input_tokens_seen": 3797124,
      "step": 3682,
      "train_runtime": 8937.6051,
      "train_tokens_per_second": 424.848
    },
    {
      "epoch": 1.573992947964526,
      "grad_norm": 2.7084686756134033,
      "learning_rate": 4.26951871657754e-06,
      "loss": 0.041866425424814224,
      "num_input_tokens_seen": 3798046,
      "step": 3683,
      "train_runtime": 8939.6566,
      "train_tokens_per_second": 424.854
    },
    {
      "epoch": 1.574420344053852,
      "grad_norm": 3.7582314014434814,
      "learning_rate": 4.2652406417112305e-06,
      "loss": 0.12114076316356659,
      "num_input_tokens_seen": 3798868,
      "step": 3684,
      "train_runtime": 8941.7319,
      "train_tokens_per_second": 424.847
    },
    {
      "epoch": 1.5748477401431777,
      "grad_norm": 3.613131284713745,
      "learning_rate": 4.26096256684492e-06,
      "loss": 0.11577745527029037,
      "num_input_tokens_seen": 3799782,
      "step": 3685,
      "train_runtime": 8943.7806,
      "train_tokens_per_second": 424.852
    },
    {
      "epoch": 1.5752751362325035,
      "grad_norm": 3.2609291076660156,
      "learning_rate": 4.25668449197861e-06,
      "loss": 0.09055764973163605,
      "num_input_tokens_seen": 3800706,
      "step": 3686,
      "train_runtime": 8945.8687,
      "train_tokens_per_second": 424.856
    },
    {
      "epoch": 1.5757025323218292,
      "grad_norm": 2.4341588020324707,
      "learning_rate": 4.2524064171123e-06,
      "loss": 0.09768286347389221,
      "num_input_tokens_seen": 3802162,
      "step": 3687,
      "train_runtime": 8948.0931,
      "train_tokens_per_second": 424.913
    },
    {
      "epoch": 1.576129928411155,
      "grad_norm": 5.3986310958862305,
      "learning_rate": 4.24812834224599e-06,
      "loss": 0.1518974006175995,
      "num_input_tokens_seen": 3802846,
      "step": 3688,
      "train_runtime": 8950.1198,
      "train_tokens_per_second": 424.893
    },
    {
      "epoch": 1.576557324500481,
      "grad_norm": 4.272782325744629,
      "learning_rate": 4.243850267379679e-06,
      "loss": 0.13275907933712006,
      "num_input_tokens_seen": 3803772,
      "step": 3689,
      "train_runtime": 8952.2132,
      "train_tokens_per_second": 424.897
    },
    {
      "epoch": 1.5769847205898067,
      "grad_norm": 3.8578505516052246,
      "learning_rate": 4.23957219251337e-06,
      "loss": 0.10831649601459503,
      "num_input_tokens_seen": 3804470,
      "step": 3690,
      "train_runtime": 8954.2021,
      "train_tokens_per_second": 424.881
    },
    {
      "epoch": 1.5774121166791324,
      "grad_norm": 2.5252091884613037,
      "learning_rate": 4.235294117647059e-06,
      "loss": 0.08624790608882904,
      "num_input_tokens_seen": 3805542,
      "step": 3691,
      "train_runtime": 8963.0058,
      "train_tokens_per_second": 424.583
    },
    {
      "epoch": 1.5778395127684581,
      "grad_norm": 4.458561420440674,
      "learning_rate": 4.2310160427807486e-06,
      "loss": 0.09071683883666992,
      "num_input_tokens_seen": 3806542,
      "step": 3692,
      "train_runtime": 8965.0813,
      "train_tokens_per_second": 424.596
    },
    {
      "epoch": 1.5782669088577839,
      "grad_norm": 2.5220489501953125,
      "learning_rate": 4.226737967914439e-06,
      "loss": 0.06447817385196686,
      "num_input_tokens_seen": 3807480,
      "step": 3693,
      "train_runtime": 8967.2602,
      "train_tokens_per_second": 424.598
    },
    {
      "epoch": 1.5786943049471098,
      "grad_norm": 2.5552539825439453,
      "learning_rate": 4.222459893048128e-06,
      "loss": 0.07284650206565857,
      "num_input_tokens_seen": 3808214,
      "step": 3694,
      "train_runtime": 8969.2391,
      "train_tokens_per_second": 424.586
    },
    {
      "epoch": 1.5791217010364356,
      "grad_norm": 3.8874473571777344,
      "learning_rate": 4.218181818181819e-06,
      "loss": 0.12044481933116913,
      "num_input_tokens_seen": 3809128,
      "step": 3695,
      "train_runtime": 8971.261,
      "train_tokens_per_second": 424.592
    },
    {
      "epoch": 1.5795490971257613,
      "grad_norm": 3.700584888458252,
      "learning_rate": 4.213903743315508e-06,
      "loss": 0.14837245643138885,
      "num_input_tokens_seen": 3810170,
      "step": 3696,
      "train_runtime": 8973.3507,
      "train_tokens_per_second": 424.61
    },
    {
      "epoch": 1.579976493215087,
      "grad_norm": 4.415613651275635,
      "learning_rate": 4.209625668449198e-06,
      "loss": 0.07102113962173462,
      "num_input_tokens_seen": 3811010,
      "step": 3697,
      "train_runtime": 8975.3639,
      "train_tokens_per_second": 424.608
    },
    {
      "epoch": 1.5804038893044128,
      "grad_norm": 3.488647222518921,
      "learning_rate": 4.205347593582888e-06,
      "loss": 0.08090274035930634,
      "num_input_tokens_seen": 3811872,
      "step": 3698,
      "train_runtime": 8977.4098,
      "train_tokens_per_second": 424.607
    },
    {
      "epoch": 1.5808312853937387,
      "grad_norm": 3.9039835929870605,
      "learning_rate": 4.201069518716578e-06,
      "loss": 0.1370634138584137,
      "num_input_tokens_seen": 3812694,
      "step": 3699,
      "train_runtime": 8979.4541,
      "train_tokens_per_second": 424.602
    },
    {
      "epoch": 1.5812586814830645,
      "grad_norm": 2.6942129135131836,
      "learning_rate": 4.1967914438502675e-06,
      "loss": 0.0807517021894455,
      "num_input_tokens_seen": 3813498,
      "step": 3700,
      "train_runtime": 8981.5098,
      "train_tokens_per_second": 424.594
    },
    {
      "epoch": 1.5816860775723902,
      "grad_norm": 3.715888500213623,
      "learning_rate": 4.192513368983958e-06,
      "loss": 0.09637901186943054,
      "num_input_tokens_seen": 3814154,
      "step": 3701,
      "train_runtime": 8983.4601,
      "train_tokens_per_second": 424.575
    },
    {
      "epoch": 1.582113473661716,
      "grad_norm": 2.2719504833221436,
      "learning_rate": 4.188235294117647e-06,
      "loss": 0.06177634745836258,
      "num_input_tokens_seen": 3815062,
      "step": 3702,
      "train_runtime": 8985.5133,
      "train_tokens_per_second": 424.579
    },
    {
      "epoch": 1.5825408697510417,
      "grad_norm": 1.869610071182251,
      "learning_rate": 4.1839572192513376e-06,
      "loss": 0.050009824335575104,
      "num_input_tokens_seen": 3815922,
      "step": 3703,
      "train_runtime": 8987.5679,
      "train_tokens_per_second": 424.578
    },
    {
      "epoch": 1.5829682658403676,
      "grad_norm": 2.434877872467041,
      "learning_rate": 4.179679144385027e-06,
      "loss": 0.0797390341758728,
      "num_input_tokens_seen": 3817030,
      "step": 3704,
      "train_runtime": 8989.652,
      "train_tokens_per_second": 424.603
    },
    {
      "epoch": 1.5833956619296934,
      "grad_norm": 3.3488621711730957,
      "learning_rate": 4.175401069518717e-06,
      "loss": 0.13556364178657532,
      "num_input_tokens_seen": 3818414,
      "step": 3705,
      "train_runtime": 8991.8179,
      "train_tokens_per_second": 424.654
    },
    {
      "epoch": 1.5838230580190191,
      "grad_norm": 2.044520139694214,
      "learning_rate": 4.171122994652407e-06,
      "loss": 0.052254632115364075,
      "num_input_tokens_seen": 3819386,
      "step": 3706,
      "train_runtime": 8993.9044,
      "train_tokens_per_second": 424.664
    },
    {
      "epoch": 1.5842504541083449,
      "grad_norm": 2.247068166732788,
      "learning_rate": 4.166844919786097e-06,
      "loss": 0.0556839257478714,
      "num_input_tokens_seen": 3820526,
      "step": 3707,
      "train_runtime": 8995.9863,
      "train_tokens_per_second": 424.692
    },
    {
      "epoch": 1.5846778501976706,
      "grad_norm": 2.208731174468994,
      "learning_rate": 4.1625668449197865e-06,
      "loss": 0.046598006039857864,
      "num_input_tokens_seen": 3821544,
      "step": 3708,
      "train_runtime": 8998.075,
      "train_tokens_per_second": 424.707
    },
    {
      "epoch": 1.5851052462869966,
      "grad_norm": 3.909510850906372,
      "learning_rate": 4.158288770053476e-06,
      "loss": 0.09692294895648956,
      "num_input_tokens_seen": 3822324,
      "step": 3709,
      "train_runtime": 9000.1669,
      "train_tokens_per_second": 424.695
    },
    {
      "epoch": 1.5855326423763223,
      "grad_norm": 3.59633207321167,
      "learning_rate": 4.154010695187166e-06,
      "loss": 0.08196096122264862,
      "num_input_tokens_seen": 3823002,
      "step": 3710,
      "train_runtime": 9002.2302,
      "train_tokens_per_second": 424.673
    },
    {
      "epoch": 1.585960038465648,
      "grad_norm": 2.2421886920928955,
      "learning_rate": 4.149732620320856e-06,
      "loss": 0.06167696416378021,
      "num_input_tokens_seen": 3824460,
      "step": 3711,
      "train_runtime": 9004.4585,
      "train_tokens_per_second": 424.73
    },
    {
      "epoch": 1.5863874345549738,
      "grad_norm": 3.3259613513946533,
      "learning_rate": 4.145454545454546e-06,
      "loss": 0.06698131561279297,
      "num_input_tokens_seen": 3825000,
      "step": 3712,
      "train_runtime": 9006.442,
      "train_tokens_per_second": 424.696
    },
    {
      "epoch": 1.5868148306442995,
      "grad_norm": 3.280956506729126,
      "learning_rate": 4.141176470588235e-06,
      "loss": 0.09087018668651581,
      "num_input_tokens_seen": 3825930,
      "step": 3713,
      "train_runtime": 9008.5348,
      "train_tokens_per_second": 424.701
    },
    {
      "epoch": 1.5872422267336255,
      "grad_norm": 2.3354268074035645,
      "learning_rate": 4.136898395721926e-06,
      "loss": 0.06623318046331406,
      "num_input_tokens_seen": 3826662,
      "step": 3714,
      "train_runtime": 9010.5644,
      "train_tokens_per_second": 424.686
    },
    {
      "epoch": 1.5876696228229512,
      "grad_norm": 2.9654526710510254,
      "learning_rate": 4.132620320855615e-06,
      "loss": 0.08049989491701126,
      "num_input_tokens_seen": 3827460,
      "step": 3715,
      "train_runtime": 9012.6368,
      "train_tokens_per_second": 424.677
    },
    {
      "epoch": 1.588097018912277,
      "grad_norm": 2.074169397354126,
      "learning_rate": 4.1283422459893054e-06,
      "loss": 0.07111380249261856,
      "num_input_tokens_seen": 3828690,
      "step": 3716,
      "train_runtime": 9014.7895,
      "train_tokens_per_second": 424.712
    },
    {
      "epoch": 1.5885244150016027,
      "grad_norm": 2.4505534172058105,
      "learning_rate": 4.124064171122995e-06,
      "loss": 0.14879193902015686,
      "num_input_tokens_seen": 3830326,
      "step": 3717,
      "train_runtime": 9017.0117,
      "train_tokens_per_second": 424.789
    },
    {
      "epoch": 1.5889518110909284,
      "grad_norm": 2.8902440071105957,
      "learning_rate": 4.119786096256685e-06,
      "loss": 0.07940120249986649,
      "num_input_tokens_seen": 3831206,
      "step": 3718,
      "train_runtime": 9019.068,
      "train_tokens_per_second": 424.79
    },
    {
      "epoch": 1.5893792071802544,
      "grad_norm": 2.028463125228882,
      "learning_rate": 4.115508021390375e-06,
      "loss": 0.08170454949140549,
      "num_input_tokens_seen": 3832776,
      "step": 3719,
      "train_runtime": 9021.2821,
      "train_tokens_per_second": 424.859
    },
    {
      "epoch": 1.5898066032695801,
      "grad_norm": 2.5394840240478516,
      "learning_rate": 4.111229946524065e-06,
      "loss": 0.0726839005947113,
      "num_input_tokens_seen": 3833746,
      "step": 3720,
      "train_runtime": 9023.3669,
      "train_tokens_per_second": 424.869
    },
    {
      "epoch": 1.5902339993589059,
      "grad_norm": 3.7162156105041504,
      "learning_rate": 4.106951871657754e-06,
      "loss": 0.10076382011175156,
      "num_input_tokens_seen": 3834408,
      "step": 3721,
      "train_runtime": 9032.3392,
      "train_tokens_per_second": 424.52
    },
    {
      "epoch": 1.5906613954482316,
      "grad_norm": 2.5366437435150146,
      "learning_rate": 4.102673796791445e-06,
      "loss": 0.06175778806209564,
      "num_input_tokens_seen": 3835440,
      "step": 3722,
      "train_runtime": 9034.4244,
      "train_tokens_per_second": 424.536
    },
    {
      "epoch": 1.5910887915375573,
      "grad_norm": 2.22567081451416,
      "learning_rate": 4.098395721925134e-06,
      "loss": 0.07033620774745941,
      "num_input_tokens_seen": 3836378,
      "step": 3723,
      "train_runtime": 9036.478,
      "train_tokens_per_second": 424.543
    },
    {
      "epoch": 1.5915161876268833,
      "grad_norm": 2.504664182662964,
      "learning_rate": 4.094117647058824e-06,
      "loss": 0.10971106588840485,
      "num_input_tokens_seen": 3837578,
      "step": 3724,
      "train_runtime": 9038.5976,
      "train_tokens_per_second": 424.577
    },
    {
      "epoch": 1.591943583716209,
      "grad_norm": 2.3840789794921875,
      "learning_rate": 4.089839572192514e-06,
      "loss": 0.04765675961971283,
      "num_input_tokens_seen": 3838334,
      "step": 3725,
      "train_runtime": 9040.6102,
      "train_tokens_per_second": 424.566
    },
    {
      "epoch": 1.5923709798055348,
      "grad_norm": 2.7244410514831543,
      "learning_rate": 4.085561497326203e-06,
      "loss": 0.06524491310119629,
      "num_input_tokens_seen": 3839094,
      "step": 3726,
      "train_runtime": 9042.657,
      "train_tokens_per_second": 424.554
    },
    {
      "epoch": 1.5927983758948605,
      "grad_norm": 1.513750672340393,
      "learning_rate": 4.0812834224598936e-06,
      "loss": 0.06865884363651276,
      "num_input_tokens_seen": 3841032,
      "step": 3727,
      "train_runtime": 9045.0115,
      "train_tokens_per_second": 424.658
    },
    {
      "epoch": 1.5932257719841862,
      "grad_norm": 2.969418525695801,
      "learning_rate": 4.077005347593583e-06,
      "loss": 0.11490297317504883,
      "num_input_tokens_seen": 3841950,
      "step": 3728,
      "train_runtime": 9047.0601,
      "train_tokens_per_second": 424.663
    },
    {
      "epoch": 1.5936531680735122,
      "grad_norm": 1.9831045866012573,
      "learning_rate": 4.072727272727273e-06,
      "loss": 0.07737528532743454,
      "num_input_tokens_seen": 3842908,
      "step": 3729,
      "train_runtime": 9049.1077,
      "train_tokens_per_second": 424.673
    },
    {
      "epoch": 1.594080564162838,
      "grad_norm": 5.0819244384765625,
      "learning_rate": 4.068449197860963e-06,
      "loss": 0.08889011293649673,
      "num_input_tokens_seen": 3843884,
      "step": 3730,
      "train_runtime": 9051.1787,
      "train_tokens_per_second": 424.683
    },
    {
      "epoch": 1.5945079602521637,
      "grad_norm": 3.0215320587158203,
      "learning_rate": 4.064171122994653e-06,
      "loss": 0.12201597541570663,
      "num_input_tokens_seen": 3845104,
      "step": 3731,
      "train_runtime": 9053.2866,
      "train_tokens_per_second": 424.719
    },
    {
      "epoch": 1.5949353563414896,
      "grad_norm": 2.506110429763794,
      "learning_rate": 4.0598930481283425e-06,
      "loss": 0.08594998717308044,
      "num_input_tokens_seen": 3846304,
      "step": 3732,
      "train_runtime": 9055.4129,
      "train_tokens_per_second": 424.752
    },
    {
      "epoch": 1.5953627524308152,
      "grad_norm": 1.9650622606277466,
      "learning_rate": 4.055614973262033e-06,
      "loss": 0.06502595543861389,
      "num_input_tokens_seen": 3847416,
      "step": 3733,
      "train_runtime": 9057.5348,
      "train_tokens_per_second": 424.775
    },
    {
      "epoch": 1.5957901485201411,
      "grad_norm": 2.4348790645599365,
      "learning_rate": 4.051336898395722e-06,
      "loss": 0.07998715341091156,
      "num_input_tokens_seen": 3848636,
      "step": 3734,
      "train_runtime": 9059.6656,
      "train_tokens_per_second": 424.81
    },
    {
      "epoch": 1.5962175446094669,
      "grad_norm": 2.49920916557312,
      "learning_rate": 4.0470588235294125e-06,
      "loss": 0.10642632842063904,
      "num_input_tokens_seen": 3849552,
      "step": 3735,
      "train_runtime": 9061.7681,
      "train_tokens_per_second": 424.812
    },
    {
      "epoch": 1.5966449406987926,
      "grad_norm": 3.0594053268432617,
      "learning_rate": 4.042780748663102e-06,
      "loss": 0.14961659908294678,
      "num_input_tokens_seen": 3850674,
      "step": 3736,
      "train_runtime": 9063.8531,
      "train_tokens_per_second": 424.839
    },
    {
      "epoch": 1.5970723367881186,
      "grad_norm": 2.573340654373169,
      "learning_rate": 4.038502673796792e-06,
      "loss": 0.12499786913394928,
      "num_input_tokens_seen": 3851800,
      "step": 3737,
      "train_runtime": 9065.9695,
      "train_tokens_per_second": 424.864
    },
    {
      "epoch": 1.597499732877444,
      "grad_norm": 2.895692825317383,
      "learning_rate": 4.034224598930482e-06,
      "loss": 0.10141438245773315,
      "num_input_tokens_seen": 3852764,
      "step": 3738,
      "train_runtime": 9068.0689,
      "train_tokens_per_second": 424.871
    },
    {
      "epoch": 1.59792712896677,
      "grad_norm": 3.7980387210845947,
      "learning_rate": 4.029946524064172e-06,
      "loss": 0.06430226564407349,
      "num_input_tokens_seen": 3854140,
      "step": 3739,
      "train_runtime": 9070.3188,
      "train_tokens_per_second": 424.918
    },
    {
      "epoch": 1.5983545250560958,
      "grad_norm": 3.8682467937469482,
      "learning_rate": 4.0256684491978615e-06,
      "loss": 0.13706448674201965,
      "num_input_tokens_seen": 3854878,
      "step": 3740,
      "train_runtime": 9072.4271,
      "train_tokens_per_second": 424.9
    },
    {
      "epoch": 1.5987819211454215,
      "grad_norm": 3.0760552883148193,
      "learning_rate": 4.021390374331552e-06,
      "loss": 0.09270555526018143,
      "num_input_tokens_seen": 3855818,
      "step": 3741,
      "train_runtime": 9074.5254,
      "train_tokens_per_second": 424.906
    },
    {
      "epoch": 1.5992093172347475,
      "grad_norm": 2.4270148277282715,
      "learning_rate": 4.017112299465241e-06,
      "loss": 0.06725089997053146,
      "num_input_tokens_seen": 3856612,
      "step": 3742,
      "train_runtime": 9076.5248,
      "train_tokens_per_second": 424.9
    },
    {
      "epoch": 1.599636713324073,
      "grad_norm": 2.623974084854126,
      "learning_rate": 4.012834224598931e-06,
      "loss": 0.0497199185192585,
      "num_input_tokens_seen": 3857720,
      "step": 3743,
      "train_runtime": 9078.6626,
      "train_tokens_per_second": 424.922
    },
    {
      "epoch": 1.600064109413399,
      "grad_norm": 2.8311891555786133,
      "learning_rate": 4.008556149732621e-06,
      "loss": 0.049882713705301285,
      "num_input_tokens_seen": 3858448,
      "step": 3744,
      "train_runtime": 9080.6811,
      "train_tokens_per_second": 424.907
    },
    {
      "epoch": 1.600064109413399,
      "eval_loss": 0.517479419708252,
      "eval_runtime": 58.6866,
      "eval_samples_per_second": 17.006,
      "eval_steps_per_second": 8.503,
      "num_input_tokens_seen": 3858448,
      "step": 3744
    },
    {
      "epoch": 1.6004915055027247,
      "grad_norm": 4.240959644317627,
      "learning_rate": 4.00427807486631e-06,
      "loss": 0.11625102907419205,
      "num_input_tokens_seen": 3859404,
      "step": 3745,
      "train_runtime": 9141.4725,
      "train_tokens_per_second": 422.186
    },
    {
      "epoch": 1.6009189015920504,
      "grad_norm": 3.350600004196167,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.08775855600833893,
      "num_input_tokens_seen": 3860244,
      "step": 3746,
      "train_runtime": 9143.5216,
      "train_tokens_per_second": 422.184
    },
    {
      "epoch": 1.6013462976813764,
      "grad_norm": 2.823941946029663,
      "learning_rate": 3.99572192513369e-06,
      "loss": 0.06984219700098038,
      "num_input_tokens_seen": 3860866,
      "step": 3747,
      "train_runtime": 9145.5095,
      "train_tokens_per_second": 422.16
    },
    {
      "epoch": 1.601773693770702,
      "grad_norm": 2.7791175842285156,
      "learning_rate": 3.99144385026738e-06,
      "loss": 0.07626688480377197,
      "num_input_tokens_seen": 3861616,
      "step": 3748,
      "train_runtime": 9147.5389,
      "train_tokens_per_second": 422.148
    },
    {
      "epoch": 1.6022010898600278,
      "grad_norm": 2.3565356731414795,
      "learning_rate": 3.98716577540107e-06,
      "loss": 0.05657782033085823,
      "num_input_tokens_seen": 3862584,
      "step": 3749,
      "train_runtime": 9149.6331,
      "train_tokens_per_second": 422.157
    },
    {
      "epoch": 1.6026284859493536,
      "grad_norm": 3.3911795616149902,
      "learning_rate": 3.98288770053476e-06,
      "loss": 0.07750269770622253,
      "num_input_tokens_seen": 3863506,
      "step": 3750,
      "train_runtime": 9151.6572,
      "train_tokens_per_second": 422.165
    },
    {
      "epoch": 1.6030558820386793,
      "grad_norm": 3.270345687866211,
      "learning_rate": 3.97860962566845e-06,
      "loss": 0.1226666122674942,
      "num_input_tokens_seen": 3864410,
      "step": 3751,
      "train_runtime": 9160.0275,
      "train_tokens_per_second": 421.878
    },
    {
      "epoch": 1.6034832781280053,
      "grad_norm": 2.3923633098602295,
      "learning_rate": 3.97433155080214e-06,
      "loss": 0.10201317071914673,
      "num_input_tokens_seen": 3866090,
      "step": 3752,
      "train_runtime": 9162.372,
      "train_tokens_per_second": 421.953
    },
    {
      "epoch": 1.6039106742173308,
      "grad_norm": 2.7062480449676514,
      "learning_rate": 3.970053475935829e-06,
      "loss": 0.08417996764183044,
      "num_input_tokens_seen": 3866970,
      "step": 3753,
      "train_runtime": 9164.4497,
      "train_tokens_per_second": 421.953
    },
    {
      "epoch": 1.6043380703066568,
      "grad_norm": 2.8350048065185547,
      "learning_rate": 3.96577540106952e-06,
      "loss": 0.1121244728565216,
      "num_input_tokens_seen": 3867962,
      "step": 3754,
      "train_runtime": 9166.6119,
      "train_tokens_per_second": 421.962
    },
    {
      "epoch": 1.6047654663959825,
      "grad_norm": 2.6301677227020264,
      "learning_rate": 3.961497326203209e-06,
      "loss": 0.10396680235862732,
      "num_input_tokens_seen": 3869128,
      "step": 3755,
      "train_runtime": 9168.7954,
      "train_tokens_per_second": 421.989
    },
    {
      "epoch": 1.6051928624853082,
      "grad_norm": 1.734845757484436,
      "learning_rate": 3.957219251336899e-06,
      "loss": 0.05011291056871414,
      "num_input_tokens_seen": 3870210,
      "step": 3756,
      "train_runtime": 9170.9345,
      "train_tokens_per_second": 422.008
    },
    {
      "epoch": 1.6056202585746342,
      "grad_norm": 2.648728847503662,
      "learning_rate": 3.952941176470588e-06,
      "loss": 0.06750747561454773,
      "num_input_tokens_seen": 3871488,
      "step": 3757,
      "train_runtime": 9173.104,
      "train_tokens_per_second": 422.048
    },
    {
      "epoch": 1.6060476546639597,
      "grad_norm": 3.227602243423462,
      "learning_rate": 3.948663101604278e-06,
      "loss": 0.11042815446853638,
      "num_input_tokens_seen": 3872288,
      "step": 3758,
      "train_runtime": 9175.1457,
      "train_tokens_per_second": 422.041
    },
    {
      "epoch": 1.6064750507532857,
      "grad_norm": 2.364561080932617,
      "learning_rate": 3.944385026737968e-06,
      "loss": 0.09016823768615723,
      "num_input_tokens_seen": 3873428,
      "step": 3759,
      "train_runtime": 9177.2862,
      "train_tokens_per_second": 422.067
    },
    {
      "epoch": 1.6069024468426114,
      "grad_norm": 2.5763111114501953,
      "learning_rate": 3.940106951871658e-06,
      "loss": 0.08186131715774536,
      "num_input_tokens_seen": 3874464,
      "step": 3760,
      "train_runtime": 9179.4236,
      "train_tokens_per_second": 422.081
    },
    {
      "epoch": 1.6073298429319371,
      "grad_norm": 2.8868651390075684,
      "learning_rate": 3.9358288770053474e-06,
      "loss": 0.08451898396015167,
      "num_input_tokens_seen": 3875212,
      "step": 3761,
      "train_runtime": 9181.4308,
      "train_tokens_per_second": 422.071
    },
    {
      "epoch": 1.607757239021263,
      "grad_norm": 1.959585189819336,
      "learning_rate": 3.931550802139038e-06,
      "loss": 0.052592284977436066,
      "num_input_tokens_seen": 3876400,
      "step": 3762,
      "train_runtime": 9183.6352,
      "train_tokens_per_second": 422.099
    },
    {
      "epoch": 1.6081846351105886,
      "grad_norm": 2.130774736404419,
      "learning_rate": 3.927272727272727e-06,
      "loss": 0.07161024212837219,
      "num_input_tokens_seen": 3877194,
      "step": 3763,
      "train_runtime": 9185.7425,
      "train_tokens_per_second": 422.088
    },
    {
      "epoch": 1.6086120311999146,
      "grad_norm": 3.273538827896118,
      "learning_rate": 3.9229946524064175e-06,
      "loss": 0.12302075326442719,
      "num_input_tokens_seen": 3878108,
      "step": 3764,
      "train_runtime": 9187.8694,
      "train_tokens_per_second": 422.09
    },
    {
      "epoch": 1.6090394272892403,
      "grad_norm": 2.040372848510742,
      "learning_rate": 3.918716577540107e-06,
      "loss": 0.06533894687891006,
      "num_input_tokens_seen": 3879234,
      "step": 3765,
      "train_runtime": 9190.0345,
      "train_tokens_per_second": 422.113
    },
    {
      "epoch": 1.609466823378566,
      "grad_norm": 1.4654204845428467,
      "learning_rate": 3.914438502673797e-06,
      "loss": 0.04753059893846512,
      "num_input_tokens_seen": 3880760,
      "step": 3766,
      "train_runtime": 9192.2409,
      "train_tokens_per_second": 422.178
    },
    {
      "epoch": 1.609894219467892,
      "grad_norm": 4.03511381149292,
      "learning_rate": 3.910160427807487e-06,
      "loss": 0.12839317321777344,
      "num_input_tokens_seen": 3881520,
      "step": 3767,
      "train_runtime": 9194.4097,
      "train_tokens_per_second": 422.161
    },
    {
      "epoch": 1.6103216155572175,
      "grad_norm": 2.9209985733032227,
      "learning_rate": 3.905882352941177e-06,
      "loss": 0.08649216592311859,
      "num_input_tokens_seen": 3882328,
      "step": 3768,
      "train_runtime": 9196.4816,
      "train_tokens_per_second": 422.154
    },
    {
      "epoch": 1.6107490116465435,
      "grad_norm": 2.055192708969116,
      "learning_rate": 3.901604278074866e-06,
      "loss": 0.054767996072769165,
      "num_input_tokens_seen": 3883454,
      "step": 3769,
      "train_runtime": 9198.6342,
      "train_tokens_per_second": 422.177
    },
    {
      "epoch": 1.6111764077358692,
      "grad_norm": 2.3946480751037598,
      "learning_rate": 3.897326203208557e-06,
      "loss": 0.0828293189406395,
      "num_input_tokens_seen": 3884428,
      "step": 3770,
      "train_runtime": 9200.7745,
      "train_tokens_per_second": 422.185
    },
    {
      "epoch": 1.611603803825195,
      "grad_norm": 3.2324209213256836,
      "learning_rate": 3.893048128342246e-06,
      "loss": 0.12190607190132141,
      "num_input_tokens_seen": 3885276,
      "step": 3771,
      "train_runtime": 9202.8031,
      "train_tokens_per_second": 422.184
    },
    {
      "epoch": 1.612031199914521,
      "grad_norm": 1.7467137575149536,
      "learning_rate": 3.8887700534759356e-06,
      "loss": 0.06836704909801483,
      "num_input_tokens_seen": 3886900,
      "step": 3772,
      "train_runtime": 9205.0265,
      "train_tokens_per_second": 422.258
    },
    {
      "epoch": 1.6124585960038464,
      "grad_norm": 2.7596757411956787,
      "learning_rate": 3.884491978609626e-06,
      "loss": 0.09314170479774475,
      "num_input_tokens_seen": 3887704,
      "step": 3773,
      "train_runtime": 9207.1557,
      "train_tokens_per_second": 422.248
    },
    {
      "epoch": 1.6128859920931724,
      "grad_norm": 2.2847583293914795,
      "learning_rate": 3.880213903743315e-06,
      "loss": 0.07095928490161896,
      "num_input_tokens_seen": 3888992,
      "step": 3774,
      "train_runtime": 9209.3902,
      "train_tokens_per_second": 422.286
    },
    {
      "epoch": 1.6133133881824981,
      "grad_norm": 4.996047019958496,
      "learning_rate": 3.875935828877006e-06,
      "loss": 0.09208285063505173,
      "num_input_tokens_seen": 3889556,
      "step": 3775,
      "train_runtime": 9211.341,
      "train_tokens_per_second": 422.257
    },
    {
      "epoch": 1.6137407842718239,
      "grad_norm": 2.8608434200286865,
      "learning_rate": 3.871657754010695e-06,
      "loss": 0.09016422182321548,
      "num_input_tokens_seen": 3890232,
      "step": 3776,
      "train_runtime": 9213.3581,
      "train_tokens_per_second": 422.238
    },
    {
      "epoch": 1.6141681803611498,
      "grad_norm": 2.8595383167266846,
      "learning_rate": 3.867379679144385e-06,
      "loss": 0.10831150412559509,
      "num_input_tokens_seen": 3891114,
      "step": 3777,
      "train_runtime": 9215.4275,
      "train_tokens_per_second": 422.239
    },
    {
      "epoch": 1.6145955764504754,
      "grad_norm": 2.205232620239258,
      "learning_rate": 3.863101604278075e-06,
      "loss": 0.078994981944561,
      "num_input_tokens_seen": 3892374,
      "step": 3778,
      "train_runtime": 9217.5667,
      "train_tokens_per_second": 422.278
    },
    {
      "epoch": 1.6150229725398013,
      "grad_norm": 3.1447365283966064,
      "learning_rate": 3.858823529411765e-06,
      "loss": 0.10400420427322388,
      "num_input_tokens_seen": 3893156,
      "step": 3779,
      "train_runtime": 9219.5824,
      "train_tokens_per_second": 422.27
    },
    {
      "epoch": 1.615450368629127,
      "grad_norm": 2.9837634563446045,
      "learning_rate": 3.8545454545454545e-06,
      "loss": 0.08216331899166107,
      "num_input_tokens_seen": 3893842,
      "step": 3780,
      "train_runtime": 9221.5822,
      "train_tokens_per_second": 422.253
    },
    {
      "epoch": 1.6158777647184528,
      "grad_norm": 3.201504707336426,
      "learning_rate": 3.850267379679145e-06,
      "loss": 0.12073357403278351,
      "num_input_tokens_seen": 3895158,
      "step": 3781,
      "train_runtime": 9230.5231,
      "train_tokens_per_second": 421.987
    },
    {
      "epoch": 1.6163051608077788,
      "grad_norm": 3.6015512943267822,
      "learning_rate": 3.845989304812834e-06,
      "loss": 0.17771466076374054,
      "num_input_tokens_seen": 3895982,
      "step": 3782,
      "train_runtime": 9232.5762,
      "train_tokens_per_second": 421.982
    },
    {
      "epoch": 1.6167325568971043,
      "grad_norm": 2.3061437606811523,
      "learning_rate": 3.8417112299465246e-06,
      "loss": 0.05693131685256958,
      "num_input_tokens_seen": 3897070,
      "step": 3783,
      "train_runtime": 9234.7581,
      "train_tokens_per_second": 422.0
    },
    {
      "epoch": 1.6171599529864302,
      "grad_norm": 2.9554741382598877,
      "learning_rate": 3.837433155080214e-06,
      "loss": 0.06877665221691132,
      "num_input_tokens_seen": 3897768,
      "step": 3784,
      "train_runtime": 9236.7563,
      "train_tokens_per_second": 421.985
    },
    {
      "epoch": 1.617587349075756,
      "grad_norm": 2.4159669876098633,
      "learning_rate": 3.833155080213904e-06,
      "loss": 0.061106663197278976,
      "num_input_tokens_seen": 3898448,
      "step": 3785,
      "train_runtime": 9238.7432,
      "train_tokens_per_second": 421.967
    },
    {
      "epoch": 1.6180147451650817,
      "grad_norm": 2.943751811981201,
      "learning_rate": 3.828877005347594e-06,
      "loss": 0.10160140693187714,
      "num_input_tokens_seen": 3899374,
      "step": 3786,
      "train_runtime": 9240.7927,
      "train_tokens_per_second": 421.974
    },
    {
      "epoch": 1.6184421412544077,
      "grad_norm": 7.840415000915527,
      "learning_rate": 3.824598930481284e-06,
      "loss": 0.09496290981769562,
      "num_input_tokens_seen": 3900192,
      "step": 3787,
      "train_runtime": 9242.8317,
      "train_tokens_per_second": 421.969
    },
    {
      "epoch": 1.6188695373437332,
      "grad_norm": 2.1866769790649414,
      "learning_rate": 3.8203208556149735e-06,
      "loss": 0.07331788539886475,
      "num_input_tokens_seen": 3901284,
      "step": 3788,
      "train_runtime": 9244.9524,
      "train_tokens_per_second": 421.991
    },
    {
      "epoch": 1.6192969334330591,
      "grad_norm": 1.3312314748764038,
      "learning_rate": 3.816042780748663e-06,
      "loss": 0.03846209868788719,
      "num_input_tokens_seen": 3902938,
      "step": 3789,
      "train_runtime": 9247.207,
      "train_tokens_per_second": 422.067
    },
    {
      "epoch": 1.6197243295223849,
      "grad_norm": 2.169729709625244,
      "learning_rate": 3.8117647058823532e-06,
      "loss": 0.06322111934423447,
      "num_input_tokens_seen": 3904240,
      "step": 3790,
      "train_runtime": 9249.36,
      "train_tokens_per_second": 422.109
    },
    {
      "epoch": 1.6201517256117106,
      "grad_norm": 3.729789972305298,
      "learning_rate": 3.807486631016043e-06,
      "loss": 0.11125414073467255,
      "num_input_tokens_seen": 3905256,
      "step": 3791,
      "train_runtime": 9251.4417,
      "train_tokens_per_second": 422.124
    },
    {
      "epoch": 1.6205791217010366,
      "grad_norm": 2.740403652191162,
      "learning_rate": 3.803208556149733e-06,
      "loss": 0.07378280907869339,
      "num_input_tokens_seen": 3906224,
      "step": 3792,
      "train_runtime": 9253.5522,
      "train_tokens_per_second": 422.132
    },
    {
      "epoch": 1.621006517790362,
      "grad_norm": 2.981154680252075,
      "learning_rate": 3.798930481283423e-06,
      "loss": 0.0933285653591156,
      "num_input_tokens_seen": 3907026,
      "step": 3793,
      "train_runtime": 9255.6233,
      "train_tokens_per_second": 422.125
    },
    {
      "epoch": 1.621433913879688,
      "grad_norm": 3.3326971530914307,
      "learning_rate": 3.7946524064171127e-06,
      "loss": 0.0761868953704834,
      "num_input_tokens_seen": 3907864,
      "step": 3794,
      "train_runtime": 9257.6446,
      "train_tokens_per_second": 422.123
    },
    {
      "epoch": 1.6218613099690138,
      "grad_norm": 2.3957982063293457,
      "learning_rate": 3.7903743315508026e-06,
      "loss": 0.08792637288570404,
      "num_input_tokens_seen": 3908778,
      "step": 3795,
      "train_runtime": 9259.7343,
      "train_tokens_per_second": 422.126
    },
    {
      "epoch": 1.6222887060583395,
      "grad_norm": 2.007890462875366,
      "learning_rate": 3.786096256684492e-06,
      "loss": 0.047754231840372086,
      "num_input_tokens_seen": 3909858,
      "step": 3796,
      "train_runtime": 9261.826,
      "train_tokens_per_second": 422.148
    },
    {
      "epoch": 1.6227161021476655,
      "grad_norm": 2.265428066253662,
      "learning_rate": 3.781818181818182e-06,
      "loss": 0.09990980476140976,
      "num_input_tokens_seen": 3911210,
      "step": 3797,
      "train_runtime": 9263.9918,
      "train_tokens_per_second": 422.195
    },
    {
      "epoch": 1.623143498236991,
      "grad_norm": 3.427783727645874,
      "learning_rate": 3.7775401069518717e-06,
      "loss": 0.08219851553440094,
      "num_input_tokens_seen": 3912022,
      "step": 3798,
      "train_runtime": 9266.0385,
      "train_tokens_per_second": 422.189
    },
    {
      "epoch": 1.623570894326317,
      "grad_norm": 2.6531877517700195,
      "learning_rate": 3.7732620320855616e-06,
      "loss": 0.08049793541431427,
      "num_input_tokens_seen": 3912882,
      "step": 3799,
      "train_runtime": 9268.1143,
      "train_tokens_per_second": 422.188
    },
    {
      "epoch": 1.6239982904156427,
      "grad_norm": 2.0714240074157715,
      "learning_rate": 3.7689839572192515e-06,
      "loss": 0.05675522983074188,
      "num_input_tokens_seen": 3913696,
      "step": 3800,
      "train_runtime": 9270.1768,
      "train_tokens_per_second": 422.181
    },
    {
      "epoch": 1.6244256865049684,
      "grad_norm": 4.28741979598999,
      "learning_rate": 3.7647058823529414e-06,
      "loss": 0.1655556857585907,
      "num_input_tokens_seen": 3914330,
      "step": 3801,
      "train_runtime": 9272.1305,
      "train_tokens_per_second": 422.161
    },
    {
      "epoch": 1.6248530825942944,
      "grad_norm": 2.941786527633667,
      "learning_rate": 3.7604278074866312e-06,
      "loss": 0.11906811594963074,
      "num_input_tokens_seen": 3915224,
      "step": 3802,
      "train_runtime": 9274.1983,
      "train_tokens_per_second": 422.163
    },
    {
      "epoch": 1.62528047868362,
      "grad_norm": 2.862887382507324,
      "learning_rate": 3.756149732620321e-06,
      "loss": 0.1298823058605194,
      "num_input_tokens_seen": 3916146,
      "step": 3803,
      "train_runtime": 9276.2965,
      "train_tokens_per_second": 422.167
    },
    {
      "epoch": 1.6257078747729459,
      "grad_norm": 2.588500499725342,
      "learning_rate": 3.751871657754011e-06,
      "loss": 0.08678627014160156,
      "num_input_tokens_seen": 3917484,
      "step": 3804,
      "train_runtime": 9278.5016,
      "train_tokens_per_second": 422.211
    },
    {
      "epoch": 1.6261352708622716,
      "grad_norm": 2.484102964401245,
      "learning_rate": 3.747593582887701e-06,
      "loss": 0.05286800488829613,
      "num_input_tokens_seen": 3918170,
      "step": 3805,
      "train_runtime": 9280.5288,
      "train_tokens_per_second": 422.193
    },
    {
      "epoch": 1.6265626669515973,
      "grad_norm": 2.187727212905884,
      "learning_rate": 3.7433155080213907e-06,
      "loss": 0.04978157579898834,
      "num_input_tokens_seen": 3919176,
      "step": 3806,
      "train_runtime": 9282.6062,
      "train_tokens_per_second": 422.206
    },
    {
      "epoch": 1.6269900630409233,
      "grad_norm": 2.7776381969451904,
      "learning_rate": 3.7390374331550806e-06,
      "loss": 0.11812900006771088,
      "num_input_tokens_seen": 3920398,
      "step": 3807,
      "train_runtime": 9284.7168,
      "train_tokens_per_second": 422.242
    },
    {
      "epoch": 1.6274174591302488,
      "grad_norm": 2.3930728435516357,
      "learning_rate": 3.7347593582887704e-06,
      "loss": 0.06416870653629303,
      "num_input_tokens_seen": 3921688,
      "step": 3808,
      "train_runtime": 9286.8595,
      "train_tokens_per_second": 422.284
    },
    {
      "epoch": 1.6278448552195748,
      "grad_norm": 2.4727258682250977,
      "learning_rate": 3.7304812834224603e-06,
      "loss": 0.07441408932209015,
      "num_input_tokens_seen": 3922974,
      "step": 3809,
      "train_runtime": 9289.0377,
      "train_tokens_per_second": 422.323
    },
    {
      "epoch": 1.6282722513089005,
      "grad_norm": 2.7147507667541504,
      "learning_rate": 3.72620320855615e-06,
      "loss": 0.09374475479125977,
      "num_input_tokens_seen": 3924210,
      "step": 3810,
      "train_runtime": 9291.154,
      "train_tokens_per_second": 422.36
    },
    {
      "epoch": 1.6286996473982263,
      "grad_norm": 1.9957373142242432,
      "learning_rate": 3.72192513368984e-06,
      "loss": 0.07390939444303513,
      "num_input_tokens_seen": 3925420,
      "step": 3811,
      "train_runtime": 9299.9155,
      "train_tokens_per_second": 422.092
    },
    {
      "epoch": 1.6291270434875522,
      "grad_norm": 2.818439483642578,
      "learning_rate": 3.71764705882353e-06,
      "loss": 0.11875130981206894,
      "num_input_tokens_seen": 3926212,
      "step": 3812,
      "train_runtime": 9302.1506,
      "train_tokens_per_second": 422.076
    },
    {
      "epoch": 1.6295544395768777,
      "grad_norm": 3.648037910461426,
      "learning_rate": 3.7133689839572194e-06,
      "loss": 0.07872645556926727,
      "num_input_tokens_seen": 3926954,
      "step": 3813,
      "train_runtime": 9304.2217,
      "train_tokens_per_second": 422.062
    },
    {
      "epoch": 1.6299818356662037,
      "grad_norm": 2.7222659587860107,
      "learning_rate": 3.7090909090909092e-06,
      "loss": 0.09119494259357452,
      "num_input_tokens_seen": 3928138,
      "step": 3814,
      "train_runtime": 9306.3209,
      "train_tokens_per_second": 422.094
    },
    {
      "epoch": 1.6304092317555294,
      "grad_norm": 3.381635904312134,
      "learning_rate": 3.704812834224599e-06,
      "loss": 0.10700887441635132,
      "num_input_tokens_seen": 3928964,
      "step": 3815,
      "train_runtime": 9308.3676,
      "train_tokens_per_second": 422.089
    },
    {
      "epoch": 1.6308366278448552,
      "grad_norm": 2.8057944774627686,
      "learning_rate": 3.700534759358289e-06,
      "loss": 0.07540324330329895,
      "num_input_tokens_seen": 3929570,
      "step": 3816,
      "train_runtime": 9310.4234,
      "train_tokens_per_second": 422.061
    },
    {
      "epoch": 1.6312640239341811,
      "grad_norm": 2.2457990646362305,
      "learning_rate": 3.696256684491979e-06,
      "loss": 0.05795879662036896,
      "num_input_tokens_seen": 3930356,
      "step": 3817,
      "train_runtime": 9312.4598,
      "train_tokens_per_second": 422.053
    },
    {
      "epoch": 1.6316914200235066,
      "grad_norm": 2.695540428161621,
      "learning_rate": 3.6919786096256687e-06,
      "loss": 0.0756165012717247,
      "num_input_tokens_seen": 3931126,
      "step": 3818,
      "train_runtime": 9314.5677,
      "train_tokens_per_second": 422.041
    },
    {
      "epoch": 1.6321188161128326,
      "grad_norm": 1.3752598762512207,
      "learning_rate": 3.6877005347593586e-06,
      "loss": 0.03707321360707283,
      "num_input_tokens_seen": 3932444,
      "step": 3819,
      "train_runtime": 9316.7983,
      "train_tokens_per_second": 422.081
    },
    {
      "epoch": 1.6325462122021583,
      "grad_norm": 3.4551901817321777,
      "learning_rate": 3.6834224598930484e-06,
      "loss": 0.08787956088781357,
      "num_input_tokens_seen": 3933232,
      "step": 3820,
      "train_runtime": 9318.8443,
      "train_tokens_per_second": 422.073
    },
    {
      "epoch": 1.632973608291484,
      "grad_norm": 2.727776527404785,
      "learning_rate": 3.6791443850267383e-06,
      "loss": 0.06859318912029266,
      "num_input_tokens_seen": 3934056,
      "step": 3821,
      "train_runtime": 9320.9431,
      "train_tokens_per_second": 422.066
    },
    {
      "epoch": 1.63340100438081,
      "grad_norm": 2.485582113265991,
      "learning_rate": 3.674866310160428e-06,
      "loss": 0.037512075155973434,
      "num_input_tokens_seen": 3935558,
      "step": 3822,
      "train_runtime": 9323.1473,
      "train_tokens_per_second": 422.128
    },
    {
      "epoch": 1.6338284004701356,
      "grad_norm": 1.8260453939437866,
      "learning_rate": 3.670588235294118e-06,
      "loss": 0.050654858350753784,
      "num_input_tokens_seen": 3936478,
      "step": 3823,
      "train_runtime": 9325.231,
      "train_tokens_per_second": 422.132
    },
    {
      "epoch": 1.6342557965594615,
      "grad_norm": 2.0404367446899414,
      "learning_rate": 3.666310160427808e-06,
      "loss": 0.05600399896502495,
      "num_input_tokens_seen": 3937860,
      "step": 3824,
      "train_runtime": 9327.4199,
      "train_tokens_per_second": 422.181
    },
    {
      "epoch": 1.6346831926487873,
      "grad_norm": 2.534850597381592,
      "learning_rate": 3.662032085561498e-06,
      "loss": 0.06724943220615387,
      "num_input_tokens_seen": 3938830,
      "step": 3825,
      "train_runtime": 9329.5297,
      "train_tokens_per_second": 422.19
    },
    {
      "epoch": 1.635110588738113,
      "grad_norm": 2.302006483078003,
      "learning_rate": 3.6577540106951877e-06,
      "loss": 0.06737169623374939,
      "num_input_tokens_seen": 3939720,
      "step": 3826,
      "train_runtime": 9331.594,
      "train_tokens_per_second": 422.192
    },
    {
      "epoch": 1.635537984827439,
      "grad_norm": 3.1940219402313232,
      "learning_rate": 3.6534759358288775e-06,
      "loss": 0.1040285974740982,
      "num_input_tokens_seen": 3940486,
      "step": 3827,
      "train_runtime": 9333.6203,
      "train_tokens_per_second": 422.182
    },
    {
      "epoch": 1.6359653809167645,
      "grad_norm": 2.526012420654297,
      "learning_rate": 3.6491978609625674e-06,
      "loss": 0.07651612162590027,
      "num_input_tokens_seen": 3941660,
      "step": 3828,
      "train_runtime": 9335.7398,
      "train_tokens_per_second": 422.212
    },
    {
      "epoch": 1.6363927770060904,
      "grad_norm": 2.366029977798462,
      "learning_rate": 3.6449197860962573e-06,
      "loss": 0.0706888735294342,
      "num_input_tokens_seen": 3943052,
      "step": 3829,
      "train_runtime": 9337.8894,
      "train_tokens_per_second": 422.264
    },
    {
      "epoch": 1.6368201730954162,
      "grad_norm": 2.3156192302703857,
      "learning_rate": 3.6406417112299467e-06,
      "loss": 0.0577034093439579,
      "num_input_tokens_seen": 3943916,
      "step": 3830,
      "train_runtime": 9339.9091,
      "train_tokens_per_second": 422.265
    },
    {
      "epoch": 1.637247569184742,
      "grad_norm": 2.212792158126831,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.08519941568374634,
      "num_input_tokens_seen": 3945110,
      "step": 3831,
      "train_runtime": 9342.1525,
      "train_tokens_per_second": 422.291
    },
    {
      "epoch": 1.6376749652740679,
      "grad_norm": 2.078310966491699,
      "learning_rate": 3.6320855614973265e-06,
      "loss": 0.06478863209486008,
      "num_input_tokens_seen": 3946710,
      "step": 3832,
      "train_runtime": 9344.3677,
      "train_tokens_per_second": 422.362
    },
    {
      "epoch": 1.6381023613633934,
      "grad_norm": 2.1961634159088135,
      "learning_rate": 3.6278074866310163e-06,
      "loss": 0.06385508924722672,
      "num_input_tokens_seen": 3947828,
      "step": 3833,
      "train_runtime": 9346.5086,
      "train_tokens_per_second": 422.385
    },
    {
      "epoch": 1.6385297574527193,
      "grad_norm": 2.614858627319336,
      "learning_rate": 3.623529411764706e-06,
      "loss": 0.0813913345336914,
      "num_input_tokens_seen": 3948988,
      "step": 3834,
      "train_runtime": 9348.6265,
      "train_tokens_per_second": 422.414
    },
    {
      "epoch": 1.638957153542045,
      "grad_norm": 2.4961130619049072,
      "learning_rate": 3.619251336898396e-06,
      "loss": 0.10427355766296387,
      "num_input_tokens_seen": 3950054,
      "step": 3835,
      "train_runtime": 9350.7429,
      "train_tokens_per_second": 422.432
    },
    {
      "epoch": 1.6393845496313708,
      "grad_norm": 2.771165609359741,
      "learning_rate": 3.614973262032086e-06,
      "loss": 0.09379483759403229,
      "num_input_tokens_seen": 3951030,
      "step": 3836,
      "train_runtime": 9352.8233,
      "train_tokens_per_second": 422.442
    },
    {
      "epoch": 1.6398119457206968,
      "grad_norm": 3.9687018394470215,
      "learning_rate": 3.610695187165776e-06,
      "loss": 0.1353222280740738,
      "num_input_tokens_seen": 3951794,
      "step": 3837,
      "train_runtime": 9354.9149,
      "train_tokens_per_second": 422.43
    },
    {
      "epoch": 1.6402393418100223,
      "grad_norm": 2.26641845703125,
      "learning_rate": 3.6064171122994657e-06,
      "loss": 0.1118968054652214,
      "num_input_tokens_seen": 3953628,
      "step": 3838,
      "train_runtime": 9357.1466,
      "train_tokens_per_second": 422.525
    },
    {
      "epoch": 1.6406667378993482,
      "grad_norm": 2.1277108192443848,
      "learning_rate": 3.6021390374331555e-06,
      "loss": 0.07533089816570282,
      "num_input_tokens_seen": 3954764,
      "step": 3839,
      "train_runtime": 9359.2582,
      "train_tokens_per_second": 422.551
    },
    {
      "epoch": 1.641094133988674,
      "grad_norm": 2.857541799545288,
      "learning_rate": 3.5978609625668454e-06,
      "loss": 0.06564372032880783,
      "num_input_tokens_seen": 3955520,
      "step": 3840,
      "train_runtime": 9361.3426,
      "train_tokens_per_second": 422.538
    },
    {
      "epoch": 1.6415215300779997,
      "grad_norm": 3.0206141471862793,
      "learning_rate": 3.5935828877005353e-06,
      "loss": 0.06101146712899208,
      "num_input_tokens_seen": 3956468,
      "step": 3841,
      "train_runtime": 9369.7666,
      "train_tokens_per_second": 422.259
    },
    {
      "epoch": 1.6419489261673257,
      "grad_norm": 2.0980420112609863,
      "learning_rate": 3.589304812834225e-06,
      "loss": 0.06332190334796906,
      "num_input_tokens_seen": 3957428,
      "step": 3842,
      "train_runtime": 9372.0176,
      "train_tokens_per_second": 422.26
    },
    {
      "epoch": 1.6423763222566512,
      "grad_norm": 2.499056816101074,
      "learning_rate": 3.585026737967915e-06,
      "loss": 0.0762253999710083,
      "num_input_tokens_seen": 3958224,
      "step": 3843,
      "train_runtime": 9374.0415,
      "train_tokens_per_second": 422.254
    },
    {
      "epoch": 1.6428037183459772,
      "grad_norm": 2.2681729793548584,
      "learning_rate": 3.580748663101605e-06,
      "loss": 0.09278064966201782,
      "num_input_tokens_seen": 3959436,
      "step": 3844,
      "train_runtime": 9376.1714,
      "train_tokens_per_second": 422.287
    },
    {
      "epoch": 1.643231114435303,
      "grad_norm": 2.438596248626709,
      "learning_rate": 3.5764705882352948e-06,
      "loss": 0.0931018516421318,
      "num_input_tokens_seen": 3960980,
      "step": 3845,
      "train_runtime": 9378.3748,
      "train_tokens_per_second": 422.352
    },
    {
      "epoch": 1.6436585105246286,
      "grad_norm": 3.4079055786132812,
      "learning_rate": 3.5721925133689846e-06,
      "loss": 0.07901705801486969,
      "num_input_tokens_seen": 3961616,
      "step": 3846,
      "train_runtime": 9380.3306,
      "train_tokens_per_second": 422.332
    },
    {
      "epoch": 1.6440859066139546,
      "grad_norm": 3.087803840637207,
      "learning_rate": 3.567914438502674e-06,
      "loss": 0.08285107463598251,
      "num_input_tokens_seen": 3962210,
      "step": 3847,
      "train_runtime": 9382.3689,
      "train_tokens_per_second": 422.304
    },
    {
      "epoch": 1.64451330270328,
      "grad_norm": 5.08903694152832,
      "learning_rate": 3.563636363636364e-06,
      "loss": 0.19842743873596191,
      "num_input_tokens_seen": 3962874,
      "step": 3848,
      "train_runtime": 9384.3722,
      "train_tokens_per_second": 422.284
    },
    {
      "epoch": 1.644940698792606,
      "grad_norm": 2.4564247131347656,
      "learning_rate": 3.559358288770054e-06,
      "loss": 0.08586368709802628,
      "num_input_tokens_seen": 3964076,
      "step": 3849,
      "train_runtime": 9386.5119,
      "train_tokens_per_second": 422.316
    },
    {
      "epoch": 1.6453680948819318,
      "grad_norm": 2.7422685623168945,
      "learning_rate": 3.5550802139037437e-06,
      "loss": 0.08317172527313232,
      "num_input_tokens_seen": 3964994,
      "step": 3850,
      "train_runtime": 9388.5794,
      "train_tokens_per_second": 422.321
    },
    {
      "epoch": 1.6457954909712575,
      "grad_norm": 3.1364614963531494,
      "learning_rate": 3.5508021390374335e-06,
      "loss": 0.0996050089597702,
      "num_input_tokens_seen": 3965706,
      "step": 3851,
      "train_runtime": 9390.6329,
      "train_tokens_per_second": 422.304
    },
    {
      "epoch": 1.6462228870605835,
      "grad_norm": 3.0305800437927246,
      "learning_rate": 3.5465240641711234e-06,
      "loss": 0.09627687186002731,
      "num_input_tokens_seen": 3966730,
      "step": 3852,
      "train_runtime": 9392.7439,
      "train_tokens_per_second": 422.319
    },
    {
      "epoch": 1.646650283149909,
      "grad_norm": 6.179396152496338,
      "learning_rate": 3.5422459893048133e-06,
      "loss": 0.1773504763841629,
      "num_input_tokens_seen": 3967484,
      "step": 3853,
      "train_runtime": 9394.8014,
      "train_tokens_per_second": 422.306
    },
    {
      "epoch": 1.647077679239235,
      "grad_norm": 2.3203563690185547,
      "learning_rate": 3.537967914438503e-06,
      "loss": 0.06952811032533646,
      "num_input_tokens_seen": 3968286,
      "step": 3854,
      "train_runtime": 9396.8495,
      "train_tokens_per_second": 422.3
    },
    {
      "epoch": 1.6475050753285607,
      "grad_norm": 1.3361934423446655,
      "learning_rate": 3.533689839572193e-06,
      "loss": 0.040817659348249435,
      "num_input_tokens_seen": 3969460,
      "step": 3855,
      "train_runtime": 9399.0104,
      "train_tokens_per_second": 422.327
    },
    {
      "epoch": 1.6479324714178865,
      "grad_norm": 3.130977153778076,
      "learning_rate": 3.529411764705883e-06,
      "loss": 0.10362480580806732,
      "num_input_tokens_seen": 3970320,
      "step": 3856,
      "train_runtime": 9401.0974,
      "train_tokens_per_second": 422.325
    },
    {
      "epoch": 1.6483598675072124,
      "grad_norm": 3.7352306842803955,
      "learning_rate": 3.5251336898395728e-06,
      "loss": 0.09946735203266144,
      "num_input_tokens_seen": 3971452,
      "step": 3857,
      "train_runtime": 9403.2532,
      "train_tokens_per_second": 422.349
    },
    {
      "epoch": 1.648787263596538,
      "grad_norm": 3.034006357192993,
      "learning_rate": 3.5208556149732626e-06,
      "loss": 0.08022941648960114,
      "num_input_tokens_seen": 3972478,
      "step": 3858,
      "train_runtime": 9405.3417,
      "train_tokens_per_second": 422.364
    },
    {
      "epoch": 1.649214659685864,
      "grad_norm": 1.989176869392395,
      "learning_rate": 3.5165775401069525e-06,
      "loss": 0.061091914772987366,
      "num_input_tokens_seen": 3974132,
      "step": 3859,
      "train_runtime": 9407.5469,
      "train_tokens_per_second": 422.441
    },
    {
      "epoch": 1.6496420557751896,
      "grad_norm": 1.645033359527588,
      "learning_rate": 3.5122994652406424e-06,
      "loss": 0.06796301901340485,
      "num_input_tokens_seen": 3976418,
      "step": 3860,
      "train_runtime": 9409.9224,
      "train_tokens_per_second": 422.577
    },
    {
      "epoch": 1.6500694518645154,
      "grad_norm": 2.3346548080444336,
      "learning_rate": 3.5080213903743322e-06,
      "loss": 0.08713055402040482,
      "num_input_tokens_seen": 3977504,
      "step": 3861,
      "train_runtime": 9412.0056,
      "train_tokens_per_second": 422.599
    },
    {
      "epoch": 1.6504968479538413,
      "grad_norm": 2.195277214050293,
      "learning_rate": 3.503743315508022e-06,
      "loss": 0.08690610527992249,
      "num_input_tokens_seen": 3978616,
      "step": 3862,
      "train_runtime": 9414.1236,
      "train_tokens_per_second": 422.622
    },
    {
      "epoch": 1.6509242440431668,
      "grad_norm": 3.190110921859741,
      "learning_rate": 3.499465240641712e-06,
      "loss": 0.11969200521707535,
      "num_input_tokens_seen": 3980158,
      "step": 3863,
      "train_runtime": 9416.3379,
      "train_tokens_per_second": 422.686
    },
    {
      "epoch": 1.6513516401324928,
      "grad_norm": 1.7407569885253906,
      "learning_rate": 3.4951871657754014e-06,
      "loss": 0.047627974301576614,
      "num_input_tokens_seen": 3981222,
      "step": 3864,
      "train_runtime": 9418.4499,
      "train_tokens_per_second": 422.705
    },
    {
      "epoch": 1.6517790362218185,
      "grad_norm": 2.4229302406311035,
      "learning_rate": 3.4909090909090913e-06,
      "loss": 0.08861653506755829,
      "num_input_tokens_seen": 3982374,
      "step": 3865,
      "train_runtime": 9420.6102,
      "train_tokens_per_second": 422.73
    },
    {
      "epoch": 1.6522064323111443,
      "grad_norm": 3.022742748260498,
      "learning_rate": 3.4866310160427807e-06,
      "loss": 0.047502826899290085,
      "num_input_tokens_seen": 3983082,
      "step": 3866,
      "train_runtime": 9422.5942,
      "train_tokens_per_second": 422.716
    },
    {
      "epoch": 1.6526338284004702,
      "grad_norm": 2.184997320175171,
      "learning_rate": 3.4823529411764706e-06,
      "loss": 0.060119304805994034,
      "num_input_tokens_seen": 3983902,
      "step": 3867,
      "train_runtime": 9424.6176,
      "train_tokens_per_second": 422.712
    },
    {
      "epoch": 1.6530612244897958,
      "grad_norm": 2.841846466064453,
      "learning_rate": 3.4780748663101605e-06,
      "loss": 0.10163979232311249,
      "num_input_tokens_seen": 3984910,
      "step": 3868,
      "train_runtime": 9426.6969,
      "train_tokens_per_second": 422.726
    },
    {
      "epoch": 1.6534886205791217,
      "grad_norm": 2.3429675102233887,
      "learning_rate": 3.4737967914438503e-06,
      "loss": 0.08151362091302872,
      "num_input_tokens_seen": 3985990,
      "step": 3869,
      "train_runtime": 9428.8232,
      "train_tokens_per_second": 422.745
    },
    {
      "epoch": 1.6539160166684475,
      "grad_norm": 2.0391833782196045,
      "learning_rate": 3.4695187165775402e-06,
      "loss": 0.057615701109170914,
      "num_input_tokens_seen": 3986918,
      "step": 3870,
      "train_runtime": 9430.9021,
      "train_tokens_per_second": 422.75
    },
    {
      "epoch": 1.6543434127577732,
      "grad_norm": 5.234706401824951,
      "learning_rate": 3.46524064171123e-06,
      "loss": 0.1124400794506073,
      "num_input_tokens_seen": 3987850,
      "step": 3871,
      "train_runtime": 9439.2646,
      "train_tokens_per_second": 422.475
    },
    {
      "epoch": 1.6547708088470992,
      "grad_norm": 2.6989946365356445,
      "learning_rate": 3.46096256684492e-06,
      "loss": 0.07818295806646347,
      "num_input_tokens_seen": 3988636,
      "step": 3872,
      "train_runtime": 9441.5559,
      "train_tokens_per_second": 422.455
    },
    {
      "epoch": 1.6551982049364247,
      "grad_norm": 2.322971820831299,
      "learning_rate": 3.45668449197861e-06,
      "loss": 0.11293534934520721,
      "num_input_tokens_seen": 3989990,
      "step": 3873,
      "train_runtime": 9443.7434,
      "train_tokens_per_second": 422.501
    },
    {
      "epoch": 1.6556256010257506,
      "grad_norm": 2.0886080265045166,
      "learning_rate": 3.4524064171122997e-06,
      "loss": 0.08082932233810425,
      "num_input_tokens_seen": 3991730,
      "step": 3874,
      "train_runtime": 9445.9753,
      "train_tokens_per_second": 422.585
    },
    {
      "epoch": 1.6560529971150764,
      "grad_norm": 2.9542224407196045,
      "learning_rate": 3.4481283422459896e-06,
      "loss": 0.08495029807090759,
      "num_input_tokens_seen": 3992506,
      "step": 3875,
      "train_runtime": 9447.9935,
      "train_tokens_per_second": 422.577
    },
    {
      "epoch": 1.656480393204402,
      "grad_norm": 3.852024793624878,
      "learning_rate": 3.443850267379679e-06,
      "loss": 0.22289283573627472,
      "num_input_tokens_seen": 3993852,
      "step": 3876,
      "train_runtime": 9450.1436,
      "train_tokens_per_second": 422.623
    },
    {
      "epoch": 1.656907789293728,
      "grad_norm": 3.585113525390625,
      "learning_rate": 3.439572192513369e-06,
      "loss": 0.08865998685359955,
      "num_input_tokens_seen": 3994440,
      "step": 3877,
      "train_runtime": 9452.1129,
      "train_tokens_per_second": 422.598
    },
    {
      "epoch": 1.6573351853830538,
      "grad_norm": 1.6091406345367432,
      "learning_rate": 3.4352941176470587e-06,
      "loss": 0.04141169413924217,
      "num_input_tokens_seen": 3995628,
      "step": 3878,
      "train_runtime": 9454.2378,
      "train_tokens_per_second": 422.628
    },
    {
      "epoch": 1.6577625814723795,
      "grad_norm": 2.2071664333343506,
      "learning_rate": 3.4310160427807486e-06,
      "loss": 0.08118504285812378,
      "num_input_tokens_seen": 3996608,
      "step": 3879,
      "train_runtime": 9456.2928,
      "train_tokens_per_second": 422.64
    },
    {
      "epoch": 1.6581899775617053,
      "grad_norm": 1.9391080141067505,
      "learning_rate": 3.4267379679144385e-06,
      "loss": 0.06162365898489952,
      "num_input_tokens_seen": 3997976,
      "step": 3880,
      "train_runtime": 9458.4804,
      "train_tokens_per_second": 422.687
    },
    {
      "epoch": 1.658617373651031,
      "grad_norm": 2.198991060256958,
      "learning_rate": 3.4224598930481284e-06,
      "loss": 0.060698941349983215,
      "num_input_tokens_seen": 3998876,
      "step": 3881,
      "train_runtime": 9460.5308,
      "train_tokens_per_second": 422.69
    },
    {
      "epoch": 1.659044769740357,
      "grad_norm": 2.4187495708465576,
      "learning_rate": 3.4181818181818182e-06,
      "loss": 0.09793850034475327,
      "num_input_tokens_seen": 3999920,
      "step": 3882,
      "train_runtime": 9462.5981,
      "train_tokens_per_second": 422.708
    },
    {
      "epoch": 1.6594721658296827,
      "grad_norm": 2.856333017349243,
      "learning_rate": 3.413903743315508e-06,
      "loss": 0.10958988219499588,
      "num_input_tokens_seen": 4000978,
      "step": 3883,
      "train_runtime": 9464.7062,
      "train_tokens_per_second": 422.726
    },
    {
      "epoch": 1.6598995619190084,
      "grad_norm": 2.926781177520752,
      "learning_rate": 3.409625668449198e-06,
      "loss": 0.07951489090919495,
      "num_input_tokens_seen": 4001788,
      "step": 3884,
      "train_runtime": 9466.7278,
      "train_tokens_per_second": 422.721
    },
    {
      "epoch": 1.6603269580083342,
      "grad_norm": 3.138127326965332,
      "learning_rate": 3.405347593582888e-06,
      "loss": 0.08947142958641052,
      "num_input_tokens_seen": 4002938,
      "step": 3885,
      "train_runtime": 9468.8485,
      "train_tokens_per_second": 422.748
    },
    {
      "epoch": 1.66075435409766,
      "grad_norm": 2.6909360885620117,
      "learning_rate": 3.4010695187165777e-06,
      "loss": 0.05339973419904709,
      "num_input_tokens_seen": 4003722,
      "step": 3886,
      "train_runtime": 9470.857,
      "train_tokens_per_second": 422.741
    },
    {
      "epoch": 1.6611817501869859,
      "grad_norm": 2.3834474086761475,
      "learning_rate": 3.3967914438502676e-06,
      "loss": 0.07068030536174774,
      "num_input_tokens_seen": 4004650,
      "step": 3887,
      "train_runtime": 9472.9129,
      "train_tokens_per_second": 422.747
    },
    {
      "epoch": 1.6616091462763116,
      "grad_norm": 3.6201040744781494,
      "learning_rate": 3.3925133689839574e-06,
      "loss": 0.09548976272344589,
      "num_input_tokens_seen": 4005424,
      "step": 3888,
      "train_runtime": 9474.9266,
      "train_tokens_per_second": 422.739
    },
    {
      "epoch": 1.6620365423656374,
      "grad_norm": 1.813190221786499,
      "learning_rate": 3.3882352941176473e-06,
      "loss": 0.07808499783277512,
      "num_input_tokens_seen": 4007066,
      "step": 3889,
      "train_runtime": 9477.132,
      "train_tokens_per_second": 422.814
    },
    {
      "epoch": 1.662463938454963,
      "grad_norm": 2.406506061553955,
      "learning_rate": 3.383957219251337e-06,
      "loss": 0.08500394970178604,
      "num_input_tokens_seen": 4008118,
      "step": 3890,
      "train_runtime": 9479.213,
      "train_tokens_per_second": 422.832
    },
    {
      "epoch": 1.6628913345442888,
      "grad_norm": 2.3753819465637207,
      "learning_rate": 3.379679144385027e-06,
      "loss": 0.09549464285373688,
      "num_input_tokens_seen": 4009344,
      "step": 3891,
      "train_runtime": 9481.3281,
      "train_tokens_per_second": 422.867
    },
    {
      "epoch": 1.6633187306336148,
      "grad_norm": 2.403844118118286,
      "learning_rate": 3.375401069518717e-06,
      "loss": 0.09369378536939621,
      "num_input_tokens_seen": 4010440,
      "step": 3892,
      "train_runtime": 9483.4402,
      "train_tokens_per_second": 422.889
    },
    {
      "epoch": 1.6637461267229405,
      "grad_norm": 2.5501019954681396,
      "learning_rate": 3.3711229946524064e-06,
      "loss": 0.09688855707645416,
      "num_input_tokens_seen": 4011538,
      "step": 3893,
      "train_runtime": 9485.5187,
      "train_tokens_per_second": 422.912
    },
    {
      "epoch": 1.6641735228122663,
      "grad_norm": 2.1859939098358154,
      "learning_rate": 3.3668449197860962e-06,
      "loss": 0.07070517539978027,
      "num_input_tokens_seen": 4012526,
      "step": 3894,
      "train_runtime": 9487.5658,
      "train_tokens_per_second": 422.925
    },
    {
      "epoch": 1.664600918901592,
      "grad_norm": 2.451632261276245,
      "learning_rate": 3.362566844919786e-06,
      "loss": 0.0956571102142334,
      "num_input_tokens_seen": 4013830,
      "step": 3895,
      "train_runtime": 9489.701,
      "train_tokens_per_second": 422.967
    },
    {
      "epoch": 1.6650283149909177,
      "grad_norm": 2.183400869369507,
      "learning_rate": 3.358288770053476e-06,
      "loss": 0.04365847259759903,
      "num_input_tokens_seen": 4014632,
      "step": 3896,
      "train_runtime": 9491.7222,
      "train_tokens_per_second": 422.961
    },
    {
      "epoch": 1.6654557110802437,
      "grad_norm": 3.0625202655792236,
      "learning_rate": 3.354010695187166e-06,
      "loss": 0.0726744532585144,
      "num_input_tokens_seen": 4015684,
      "step": 3897,
      "train_runtime": 9493.7724,
      "train_tokens_per_second": 422.981
    },
    {
      "epoch": 1.6658831071695694,
      "grad_norm": 2.8540406227111816,
      "learning_rate": 3.3497326203208557e-06,
      "loss": 0.11057975888252258,
      "num_input_tokens_seen": 4016792,
      "step": 3898,
      "train_runtime": 9495.8371,
      "train_tokens_per_second": 423.006
    },
    {
      "epoch": 1.6663105032588952,
      "grad_norm": 2.8264729976654053,
      "learning_rate": 3.3454545454545456e-06,
      "loss": 0.08364143967628479,
      "num_input_tokens_seen": 4017808,
      "step": 3899,
      "train_runtime": 9497.8866,
      "train_tokens_per_second": 423.021
    },
    {
      "epoch": 1.666737899348221,
      "grad_norm": 3.502856969833374,
      "learning_rate": 3.3411764705882354e-06,
      "loss": 0.12757033109664917,
      "num_input_tokens_seen": 4018772,
      "step": 3900,
      "train_runtime": 9499.9886,
      "train_tokens_per_second": 423.029
    },
    {
      "epoch": 1.6671652954375467,
      "grad_norm": 3.1576054096221924,
      "learning_rate": 3.3368983957219253e-06,
      "loss": 0.11345172673463821,
      "num_input_tokens_seen": 4019542,
      "step": 3901,
      "train_runtime": 9508.2656,
      "train_tokens_per_second": 422.742
    },
    {
      "epoch": 1.6675926915268726,
      "grad_norm": 3.2169222831726074,
      "learning_rate": 3.332620320855615e-06,
      "loss": 0.13035550713539124,
      "num_input_tokens_seen": 4020464,
      "step": 3902,
      "train_runtime": 9510.4508,
      "train_tokens_per_second": 422.742
    },
    {
      "epoch": 1.6680200876161984,
      "grad_norm": 2.6613028049468994,
      "learning_rate": 3.328342245989305e-06,
      "loss": 0.1079365611076355,
      "num_input_tokens_seen": 4021362,
      "step": 3903,
      "train_runtime": 9512.5313,
      "train_tokens_per_second": 422.744
    },
    {
      "epoch": 1.668447483705524,
      "grad_norm": 4.194794178009033,
      "learning_rate": 3.324064171122995e-06,
      "loss": 0.10679323971271515,
      "num_input_tokens_seen": 4022122,
      "step": 3904,
      "train_runtime": 9514.5357,
      "train_tokens_per_second": 422.734
    },
    {
      "epoch": 1.6688748797948498,
      "grad_norm": 2.76346492767334,
      "learning_rate": 3.3197860962566848e-06,
      "loss": 0.0850372239947319,
      "num_input_tokens_seen": 4023172,
      "step": 3905,
      "train_runtime": 9516.6246,
      "train_tokens_per_second": 422.752
    },
    {
      "epoch": 1.6693022758841756,
      "grad_norm": 3.2526650428771973,
      "learning_rate": 3.3155080213903747e-06,
      "loss": 0.10961224138736725,
      "num_input_tokens_seen": 4024086,
      "step": 3906,
      "train_runtime": 9518.6463,
      "train_tokens_per_second": 422.758
    },
    {
      "epoch": 1.6697296719735015,
      "grad_norm": 2.1977479457855225,
      "learning_rate": 3.3112299465240645e-06,
      "loss": 0.08637858927249908,
      "num_input_tokens_seen": 4025244,
      "step": 3907,
      "train_runtime": 9520.7575,
      "train_tokens_per_second": 422.786
    },
    {
      "epoch": 1.6701570680628273,
      "grad_norm": 2.0698485374450684,
      "learning_rate": 3.3069518716577544e-06,
      "loss": 0.07969550043344498,
      "num_input_tokens_seen": 4026348,
      "step": 3908,
      "train_runtime": 9522.8813,
      "train_tokens_per_second": 422.808
    },
    {
      "epoch": 1.670584464152153,
      "grad_norm": 1.8329750299453735,
      "learning_rate": 3.3026737967914443e-06,
      "loss": 0.06806891411542892,
      "num_input_tokens_seen": 4027878,
      "step": 3909,
      "train_runtime": 9525.0916,
      "train_tokens_per_second": 422.87
    },
    {
      "epoch": 1.6710118602414787,
      "grad_norm": 4.586109638214111,
      "learning_rate": 3.2983957219251337e-06,
      "loss": 0.10857181251049042,
      "num_input_tokens_seen": 4028804,
      "step": 3910,
      "train_runtime": 9527.1473,
      "train_tokens_per_second": 422.876
    },
    {
      "epoch": 1.6714392563308045,
      "grad_norm": 3.550147533416748,
      "learning_rate": 3.2941176470588236e-06,
      "loss": 0.06560402363538742,
      "num_input_tokens_seen": 4029564,
      "step": 3911,
      "train_runtime": 9529.1975,
      "train_tokens_per_second": 422.865
    },
    {
      "epoch": 1.6718666524201304,
      "grad_norm": 2.4432454109191895,
      "learning_rate": 3.2898395721925134e-06,
      "loss": 0.08360602706670761,
      "num_input_tokens_seen": 4030592,
      "step": 3912,
      "train_runtime": 9531.3195,
      "train_tokens_per_second": 422.879
    },
    {
      "epoch": 1.6722940485094562,
      "grad_norm": 3.205657482147217,
      "learning_rate": 3.2855614973262033e-06,
      "loss": 0.08480235189199448,
      "num_input_tokens_seen": 4031420,
      "step": 3913,
      "train_runtime": 9533.3925,
      "train_tokens_per_second": 422.874
    },
    {
      "epoch": 1.672721444598782,
      "grad_norm": 2.0331385135650635,
      "learning_rate": 3.281283422459893e-06,
      "loss": 0.06084195151925087,
      "num_input_tokens_seen": 4032730,
      "step": 3914,
      "train_runtime": 9535.5692,
      "train_tokens_per_second": 422.914
    },
    {
      "epoch": 1.6731488406881077,
      "grad_norm": 2.2410805225372314,
      "learning_rate": 3.277005347593583e-06,
      "loss": 0.050346631556749344,
      "num_input_tokens_seen": 4033404,
      "step": 3915,
      "train_runtime": 9537.5651,
      "train_tokens_per_second": 422.897
    },
    {
      "epoch": 1.6735762367774334,
      "grad_norm": 3.286407947540283,
      "learning_rate": 3.272727272727273e-06,
      "loss": 0.10637320578098297,
      "num_input_tokens_seen": 4034256,
      "step": 3916,
      "train_runtime": 9539.609,
      "train_tokens_per_second": 422.895
    },
    {
      "epoch": 1.6740036328667594,
      "grad_norm": 2.611971139907837,
      "learning_rate": 3.268449197860963e-06,
      "loss": 0.0768347978591919,
      "num_input_tokens_seen": 4035476,
      "step": 3917,
      "train_runtime": 9541.717,
      "train_tokens_per_second": 422.93
    },
    {
      "epoch": 1.674431028956085,
      "grad_norm": 3.220508098602295,
      "learning_rate": 3.2641711229946527e-06,
      "loss": 0.15087278187274933,
      "num_input_tokens_seen": 4036622,
      "step": 3918,
      "train_runtime": 9543.824,
      "train_tokens_per_second": 422.956
    },
    {
      "epoch": 1.6748584250454108,
      "grad_norm": 1.945664644241333,
      "learning_rate": 3.2598930481283425e-06,
      "loss": 0.0812452957034111,
      "num_input_tokens_seen": 4038000,
      "step": 3919,
      "train_runtime": 9545.9916,
      "train_tokens_per_second": 423.005
    },
    {
      "epoch": 1.6752858211347366,
      "grad_norm": 4.469132423400879,
      "learning_rate": 3.2556149732620324e-06,
      "loss": 0.08381488174200058,
      "num_input_tokens_seen": 4038942,
      "step": 3920,
      "train_runtime": 9548.0248,
      "train_tokens_per_second": 423.013
    },
    {
      "epoch": 1.6757132172240623,
      "grad_norm": 2.264404535293579,
      "learning_rate": 3.2513368983957223e-06,
      "loss": 0.0824776217341423,
      "num_input_tokens_seen": 4040338,
      "step": 3921,
      "train_runtime": 9550.2363,
      "train_tokens_per_second": 423.062
    },
    {
      "epoch": 1.6761406133133883,
      "grad_norm": 2.3763954639434814,
      "learning_rate": 3.247058823529412e-06,
      "loss": 0.07568852603435516,
      "num_input_tokens_seen": 4041320,
      "step": 3922,
      "train_runtime": 9552.3108,
      "train_tokens_per_second": 423.073
    },
    {
      "epoch": 1.676568009402714,
      "grad_norm": 3.5757901668548584,
      "learning_rate": 3.242780748663102e-06,
      "loss": 0.10268846154212952,
      "num_input_tokens_seen": 4042142,
      "step": 3923,
      "train_runtime": 9554.4141,
      "train_tokens_per_second": 423.065
    },
    {
      "epoch": 1.6769954054920397,
      "grad_norm": 2.101426362991333,
      "learning_rate": 3.238502673796792e-06,
      "loss": 0.07353068143129349,
      "num_input_tokens_seen": 4043096,
      "step": 3924,
      "train_runtime": 9556.4642,
      "train_tokens_per_second": 423.074
    },
    {
      "epoch": 1.6774228015813657,
      "grad_norm": 2.1029913425445557,
      "learning_rate": 3.2342245989304817e-06,
      "loss": 0.04573386535048485,
      "num_input_tokens_seen": 4043896,
      "step": 3925,
      "train_runtime": 9558.4895,
      "train_tokens_per_second": 423.069
    },
    {
      "epoch": 1.6778501976706912,
      "grad_norm": 2.0387256145477295,
      "learning_rate": 3.2299465240641716e-06,
      "loss": 0.08806627243757248,
      "num_input_tokens_seen": 4045420,
      "step": 3926,
      "train_runtime": 9560.7051,
      "train_tokens_per_second": 423.13
    },
    {
      "epoch": 1.6782775937600172,
      "grad_norm": 2.8606925010681152,
      "learning_rate": 3.225668449197861e-06,
      "loss": 0.08300386369228363,
      "num_input_tokens_seen": 4046308,
      "step": 3927,
      "train_runtime": 9562.7845,
      "train_tokens_per_second": 423.131
    },
    {
      "epoch": 1.678704989849343,
      "grad_norm": 3.4991629123687744,
      "learning_rate": 3.221390374331551e-06,
      "loss": 0.1041450947523117,
      "num_input_tokens_seen": 4047052,
      "step": 3928,
      "train_runtime": 9564.775,
      "train_tokens_per_second": 423.12
    },
    {
      "epoch": 1.6791323859386686,
      "grad_norm": 2.403409719467163,
      "learning_rate": 3.217112299465241e-06,
      "loss": 0.10034950077533722,
      "num_input_tokens_seen": 4048534,
      "step": 3929,
      "train_runtime": 9566.9845,
      "train_tokens_per_second": 423.178
    },
    {
      "epoch": 1.6795597820279946,
      "grad_norm": 3.0966920852661133,
      "learning_rate": 3.2128342245989307e-06,
      "loss": 0.09005638211965561,
      "num_input_tokens_seen": 4049546,
      "step": 3930,
      "train_runtime": 9569.0847,
      "train_tokens_per_second": 423.191
    },
    {
      "epoch": 1.6799871781173201,
      "grad_norm": 3.459977388381958,
      "learning_rate": 3.2085561497326205e-06,
      "loss": 0.08567871153354645,
      "num_input_tokens_seen": 4050164,
      "step": 3931,
      "train_runtime": 9577.3862,
      "train_tokens_per_second": 422.888
    },
    {
      "epoch": 1.680414574206646,
      "grad_norm": 4.240540504455566,
      "learning_rate": 3.2042780748663104e-06,
      "loss": 0.07416868209838867,
      "num_input_tokens_seen": 4050780,
      "step": 3932,
      "train_runtime": 9579.61,
      "train_tokens_per_second": 422.854
    },
    {
      "epoch": 1.6808419702959718,
      "grad_norm": 5.221320152282715,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.09833960980176926,
      "num_input_tokens_seen": 4051324,
      "step": 3933,
      "train_runtime": 9581.5589,
      "train_tokens_per_second": 422.825
    },
    {
      "epoch": 1.6812693663852976,
      "grad_norm": 2.3435938358306885,
      "learning_rate": 3.19572192513369e-06,
      "loss": 0.06912118196487427,
      "num_input_tokens_seen": 4052426,
      "step": 3934,
      "train_runtime": 9583.677,
      "train_tokens_per_second": 422.847
    },
    {
      "epoch": 1.6816967624746235,
      "grad_norm": 3.06503963470459,
      "learning_rate": 3.19144385026738e-06,
      "loss": 0.0881384015083313,
      "num_input_tokens_seen": 4053116,
      "step": 3935,
      "train_runtime": 9585.6684,
      "train_tokens_per_second": 422.831
    },
    {
      "epoch": 1.682124158563949,
      "grad_norm": 2.6347336769104004,
      "learning_rate": 3.18716577540107e-06,
      "loss": 0.0930546373128891,
      "num_input_tokens_seen": 4053920,
      "step": 3936,
      "train_runtime": 9587.7536,
      "train_tokens_per_second": 422.823
    },
    {
      "epoch": 1.682551554653275,
      "grad_norm": 2.6467833518981934,
      "learning_rate": 3.1828877005347598e-06,
      "loss": 0.09518023580312729,
      "num_input_tokens_seen": 4054992,
      "step": 3937,
      "train_runtime": 9589.9079,
      "train_tokens_per_second": 422.84
    },
    {
      "epoch": 1.6829789507426007,
      "grad_norm": 3.57283616065979,
      "learning_rate": 3.1786096256684496e-06,
      "loss": 0.11834046989679337,
      "num_input_tokens_seen": 4056168,
      "step": 3938,
      "train_runtime": 9592.0435,
      "train_tokens_per_second": 422.868
    },
    {
      "epoch": 1.6834063468319265,
      "grad_norm": 4.581844806671143,
      "learning_rate": 3.1743315508021395e-06,
      "loss": 0.06580094993114471,
      "num_input_tokens_seen": 4057272,
      "step": 3939,
      "train_runtime": 9594.1864,
      "train_tokens_per_second": 422.889
    },
    {
      "epoch": 1.6838337429212524,
      "grad_norm": 2.258781909942627,
      "learning_rate": 3.1700534759358294e-06,
      "loss": 0.06576640158891678,
      "num_input_tokens_seen": 4058336,
      "step": 3940,
      "train_runtime": 9596.2976,
      "train_tokens_per_second": 422.906
    },
    {
      "epoch": 1.684261139010578,
      "grad_norm": 2.5151267051696777,
      "learning_rate": 3.1657754010695192e-06,
      "loss": 0.08555591106414795,
      "num_input_tokens_seen": 4059686,
      "step": 3941,
      "train_runtime": 9598.5049,
      "train_tokens_per_second": 422.95
    },
    {
      "epoch": 1.684688535099904,
      "grad_norm": 4.483311653137207,
      "learning_rate": 3.161497326203209e-06,
      "loss": 0.12450273334980011,
      "num_input_tokens_seen": 4060796,
      "step": 3942,
      "train_runtime": 9600.5856,
      "train_tokens_per_second": 422.974
    },
    {
      "epoch": 1.6851159311892296,
      "grad_norm": 2.560617208480835,
      "learning_rate": 3.157219251336899e-06,
      "loss": 0.10108768939971924,
      "num_input_tokens_seen": 4063020,
      "step": 3943,
      "train_runtime": 9602.947,
      "train_tokens_per_second": 423.101
    },
    {
      "epoch": 1.6855433272785554,
      "grad_norm": 2.902137041091919,
      "learning_rate": 3.1529411764705884e-06,
      "loss": 0.13017889857292175,
      "num_input_tokens_seen": 4064374,
      "step": 3944,
      "train_runtime": 9605.1315,
      "train_tokens_per_second": 423.146
    },
    {
      "epoch": 1.6859707233678813,
      "grad_norm": 3.8050923347473145,
      "learning_rate": 3.1486631016042783e-06,
      "loss": 0.12200959026813507,
      "num_input_tokens_seen": 4065068,
      "step": 3945,
      "train_runtime": 9607.1552,
      "train_tokens_per_second": 423.129
    },
    {
      "epoch": 1.6863981194572069,
      "grad_norm": 4.64519739151001,
      "learning_rate": 3.144385026737968e-06,
      "loss": 0.10243044048547745,
      "num_input_tokens_seen": 4065668,
      "step": 3946,
      "train_runtime": 9609.1522,
      "train_tokens_per_second": 423.104
    },
    {
      "epoch": 1.6868255155465328,
      "grad_norm": 2.8470051288604736,
      "learning_rate": 3.140106951871658e-06,
      "loss": 0.08725105226039886,
      "num_input_tokens_seen": 4067100,
      "step": 3947,
      "train_runtime": 9611.3598,
      "train_tokens_per_second": 423.156
    },
    {
      "epoch": 1.6872529116358586,
      "grad_norm": 2.402822971343994,
      "learning_rate": 3.135828877005348e-06,
      "loss": 0.0632270872592926,
      "num_input_tokens_seen": 4068514,
      "step": 3948,
      "train_runtime": 9613.602,
      "train_tokens_per_second": 423.204
    },
    {
      "epoch": 1.6876803077251843,
      "grad_norm": 1.928742527961731,
      "learning_rate": 3.1315508021390378e-06,
      "loss": 0.06809768080711365,
      "num_input_tokens_seen": 4070026,
      "step": 3949,
      "train_runtime": 9615.8046,
      "train_tokens_per_second": 423.264
    },
    {
      "epoch": 1.6881077038145103,
      "grad_norm": 3.191988229751587,
      "learning_rate": 3.1272727272727276e-06,
      "loss": 0.09078343212604523,
      "num_input_tokens_seen": 4070880,
      "step": 3950,
      "train_runtime": 9617.8258,
      "train_tokens_per_second": 423.264
    },
    {
      "epoch": 1.6885350999038358,
      "grad_norm": 2.2731947898864746,
      "learning_rate": 3.1229946524064175e-06,
      "loss": 0.11136038601398468,
      "num_input_tokens_seen": 4071986,
      "step": 3951,
      "train_runtime": 9620.0628,
      "train_tokens_per_second": 423.281
    },
    {
      "epoch": 1.6889624959931617,
      "grad_norm": 2.486530065536499,
      "learning_rate": 3.1187165775401074e-06,
      "loss": 0.11815786361694336,
      "num_input_tokens_seen": 4072958,
      "step": 3952,
      "train_runtime": 9622.2029,
      "train_tokens_per_second": 423.287
    },
    {
      "epoch": 1.6893898920824875,
      "grad_norm": 2.8962604999542236,
      "learning_rate": 3.1144385026737972e-06,
      "loss": 0.0618472546339035,
      "num_input_tokens_seen": 4073892,
      "step": 3953,
      "train_runtime": 9624.2551,
      "train_tokens_per_second": 423.294
    },
    {
      "epoch": 1.6898172881718132,
      "grad_norm": 1.9911576509475708,
      "learning_rate": 3.110160427807487e-06,
      "loss": 0.06889242678880692,
      "num_input_tokens_seen": 4075672,
      "step": 3954,
      "train_runtime": 9626.4886,
      "train_tokens_per_second": 423.381
    },
    {
      "epoch": 1.6902446842611392,
      "grad_norm": 2.7848258018493652,
      "learning_rate": 3.105882352941177e-06,
      "loss": 0.07982391119003296,
      "num_input_tokens_seen": 4076562,
      "step": 3955,
      "train_runtime": 9628.5727,
      "train_tokens_per_second": 423.382
    },
    {
      "epoch": 1.6906720803504647,
      "grad_norm": 1.9461215734481812,
      "learning_rate": 3.101604278074867e-06,
      "loss": 0.04318079352378845,
      "num_input_tokens_seen": 4077580,
      "step": 3956,
      "train_runtime": 9630.6988,
      "train_tokens_per_second": 423.394
    },
    {
      "epoch": 1.6910994764397906,
      "grad_norm": 3.0274593830108643,
      "learning_rate": 3.0973262032085567e-06,
      "loss": 0.09338752180337906,
      "num_input_tokens_seen": 4078396,
      "step": 3957,
      "train_runtime": 9632.7391,
      "train_tokens_per_second": 423.389
    },
    {
      "epoch": 1.6915268725291164,
      "grad_norm": 3.2393455505371094,
      "learning_rate": 3.0930481283422466e-06,
      "loss": 0.10761010646820068,
      "num_input_tokens_seen": 4079208,
      "step": 3958,
      "train_runtime": 9634.7703,
      "train_tokens_per_second": 423.384
    },
    {
      "epoch": 1.6919542686184421,
      "grad_norm": 3.275947332382202,
      "learning_rate": 3.0887700534759365e-06,
      "loss": 0.07251361012458801,
      "num_input_tokens_seen": 4080156,
      "step": 3959,
      "train_runtime": 9636.8124,
      "train_tokens_per_second": 423.393
    },
    {
      "epoch": 1.692381664707768,
      "grad_norm": 2.4812443256378174,
      "learning_rate": 3.0844919786096263e-06,
      "loss": 0.08334419876337051,
      "num_input_tokens_seen": 4081332,
      "step": 3960,
      "train_runtime": 9638.9275,
      "train_tokens_per_second": 423.422
    },
    {
      "epoch": 1.6928090607970936,
      "grad_norm": 2.866668224334717,
      "learning_rate": 3.0802139037433158e-06,
      "loss": 0.08983352035284042,
      "num_input_tokens_seen": 4082194,
      "step": 3961,
      "train_runtime": 9647.5402,
      "train_tokens_per_second": 423.133
    },
    {
      "epoch": 1.6932364568864195,
      "grad_norm": 3.889932155609131,
      "learning_rate": 3.0759358288770056e-06,
      "loss": 0.10154541581869125,
      "num_input_tokens_seen": 4082842,
      "step": 3962,
      "train_runtime": 9649.5679,
      "train_tokens_per_second": 423.111
    },
    {
      "epoch": 1.6936638529757453,
      "grad_norm": 3.7758491039276123,
      "learning_rate": 3.0716577540106955e-06,
      "loss": 0.10542549937963486,
      "num_input_tokens_seen": 4083728,
      "step": 3963,
      "train_runtime": 9651.6686,
      "train_tokens_per_second": 423.111
    },
    {
      "epoch": 1.694091249065071,
      "grad_norm": 2.9250810146331787,
      "learning_rate": 3.0673796791443854e-06,
      "loss": 0.08727927505970001,
      "num_input_tokens_seen": 4084996,
      "step": 3964,
      "train_runtime": 9653.8209,
      "train_tokens_per_second": 423.148
    },
    {
      "epoch": 1.694518645154397,
      "grad_norm": 2.0858898162841797,
      "learning_rate": 3.0631016042780752e-06,
      "loss": 0.04901638999581337,
      "num_input_tokens_seen": 4085760,
      "step": 3965,
      "train_runtime": 9655.8755,
      "train_tokens_per_second": 423.137
    },
    {
      "epoch": 1.6949460412437225,
      "grad_norm": 2.70337176322937,
      "learning_rate": 3.058823529411765e-06,
      "loss": 0.12117543816566467,
      "num_input_tokens_seen": 4087108,
      "step": 3966,
      "train_runtime": 9658.0868,
      "train_tokens_per_second": 423.18
    },
    {
      "epoch": 1.6953734373330485,
      "grad_norm": 1.875209093093872,
      "learning_rate": 3.054545454545455e-06,
      "loss": 0.12659204006195068,
      "num_input_tokens_seen": 4088596,
      "step": 3967,
      "train_runtime": 9660.3092,
      "train_tokens_per_second": 423.237
    },
    {
      "epoch": 1.6958008334223742,
      "grad_norm": 2.988813877105713,
      "learning_rate": 3.050267379679145e-06,
      "loss": 0.0948270633816719,
      "num_input_tokens_seen": 4089646,
      "step": 3968,
      "train_runtime": 9662.3869,
      "train_tokens_per_second": 423.254
    },
    {
      "epoch": 1.6962282295117,
      "grad_norm": 1.7588655948638916,
      "learning_rate": 3.0459893048128347e-06,
      "loss": 0.05924312770366669,
      "num_input_tokens_seen": 4090980,
      "step": 3969,
      "train_runtime": 9664.5602,
      "train_tokens_per_second": 423.297
    },
    {
      "epoch": 1.696655625601026,
      "grad_norm": 3.538276433944702,
      "learning_rate": 3.0417112299465246e-06,
      "loss": 0.10084767639636993,
      "num_input_tokens_seen": 4091836,
      "step": 3970,
      "train_runtime": 9666.6342,
      "train_tokens_per_second": 423.295
    },
    {
      "epoch": 1.6970830216903514,
      "grad_norm": 2.048569440841675,
      "learning_rate": 3.0374331550802145e-06,
      "loss": 0.052659835666418076,
      "num_input_tokens_seen": 4092640,
      "step": 3971,
      "train_runtime": 9668.6578,
      "train_tokens_per_second": 423.289
    },
    {
      "epoch": 1.6975104177796774,
      "grad_norm": 1.0084600448608398,
      "learning_rate": 3.0331550802139043e-06,
      "loss": 0.03439614176750183,
      "num_input_tokens_seen": 4095382,
      "step": 3972,
      "train_runtime": 9671.2308,
      "train_tokens_per_second": 423.46
    },
    {
      "epoch": 1.697937813869003,
      "grad_norm": 2.9055423736572266,
      "learning_rate": 3.028877005347594e-06,
      "loss": 0.13083118200302124,
      "num_input_tokens_seen": 4096476,
      "step": 3973,
      "train_runtime": 9673.4149,
      "train_tokens_per_second": 423.478
    },
    {
      "epoch": 1.6983652099583288,
      "grad_norm": 3.072705030441284,
      "learning_rate": 3.024598930481284e-06,
      "loss": 0.09290516376495361,
      "num_input_tokens_seen": 4097310,
      "step": 3974,
      "train_runtime": 9675.4882,
      "train_tokens_per_second": 423.473
    },
    {
      "epoch": 1.6987926060476548,
      "grad_norm": 3.6360573768615723,
      "learning_rate": 3.020320855614973e-06,
      "loss": 0.09591491520404816,
      "num_input_tokens_seen": 4098298,
      "step": 3975,
      "train_runtime": 9677.5762,
      "train_tokens_per_second": 423.484
    },
    {
      "epoch": 1.6992200021369803,
      "grad_norm": 3.1038193702697754,
      "learning_rate": 3.016042780748663e-06,
      "loss": 0.0824640691280365,
      "num_input_tokens_seen": 4098958,
      "step": 3976,
      "train_runtime": 9679.5993,
      "train_tokens_per_second": 423.464
    },
    {
      "epoch": 1.6996473982263063,
      "grad_norm": 2.79768705368042,
      "learning_rate": 3.011764705882353e-06,
      "loss": 0.1095389574766159,
      "num_input_tokens_seen": 4100004,
      "step": 3977,
      "train_runtime": 9681.7467,
      "train_tokens_per_second": 423.478
    },
    {
      "epoch": 1.700074794315632,
      "grad_norm": 2.7529211044311523,
      "learning_rate": 3.0074866310160427e-06,
      "loss": 0.08126137405633926,
      "num_input_tokens_seen": 4101046,
      "step": 3978,
      "train_runtime": 9683.8381,
      "train_tokens_per_second": 423.494
    },
    {
      "epoch": 1.7005021904049578,
      "grad_norm": 3.107191562652588,
      "learning_rate": 3.0032085561497326e-06,
      "loss": 0.08961816877126694,
      "num_input_tokens_seen": 4101994,
      "step": 3979,
      "train_runtime": 9685.9528,
      "train_tokens_per_second": 423.499
    },
    {
      "epoch": 1.7009295864942837,
      "grad_norm": 1.573712944984436,
      "learning_rate": 2.9989304812834224e-06,
      "loss": 0.04550402611494064,
      "num_input_tokens_seen": 4102996,
      "step": 3980,
      "train_runtime": 9688.0116,
      "train_tokens_per_second": 423.513
    },
    {
      "epoch": 1.7013569825836092,
      "grad_norm": 2.7542884349823,
      "learning_rate": 2.9946524064171123e-06,
      "loss": 0.0688285231590271,
      "num_input_tokens_seen": 4103766,
      "step": 3981,
      "train_runtime": 9690.0265,
      "train_tokens_per_second": 423.504
    },
    {
      "epoch": 1.7017843786729352,
      "grad_norm": 2.8290889263153076,
      "learning_rate": 2.990374331550802e-06,
      "loss": 0.05508919060230255,
      "num_input_tokens_seen": 4104614,
      "step": 3982,
      "train_runtime": 9692.0843,
      "train_tokens_per_second": 423.502
    },
    {
      "epoch": 1.702211774762261,
      "grad_norm": 2.792149543762207,
      "learning_rate": 2.986096256684492e-06,
      "loss": 0.10512325167655945,
      "num_input_tokens_seen": 4105492,
      "step": 3983,
      "train_runtime": 9694.1483,
      "train_tokens_per_second": 423.502
    },
    {
      "epoch": 1.7026391708515867,
      "grad_norm": 2.788059711456299,
      "learning_rate": 2.981818181818182e-06,
      "loss": 0.08885282278060913,
      "num_input_tokens_seen": 4106256,
      "step": 3984,
      "train_runtime": 9696.2024,
      "train_tokens_per_second": 423.491
    },
    {
      "epoch": 1.7030665669409126,
      "grad_norm": 1.0589274168014526,
      "learning_rate": 2.9775401069518718e-06,
      "loss": 0.02877618744969368,
      "num_input_tokens_seen": 4108050,
      "step": 3985,
      "train_runtime": 9698.5075,
      "train_tokens_per_second": 423.575
    },
    {
      "epoch": 1.7034939630302381,
      "grad_norm": 4.479930400848389,
      "learning_rate": 2.9732620320855617e-06,
      "loss": 0.08620975911617279,
      "num_input_tokens_seen": 4109066,
      "step": 3986,
      "train_runtime": 9700.6266,
      "train_tokens_per_second": 423.588
    },
    {
      "epoch": 1.703921359119564,
      "grad_norm": 2.665229082107544,
      "learning_rate": 2.9689839572192515e-06,
      "loss": 0.07935021817684174,
      "num_input_tokens_seen": 4110052,
      "step": 3987,
      "train_runtime": 9702.7107,
      "train_tokens_per_second": 423.598
    },
    {
      "epoch": 1.7043487552088898,
      "grad_norm": 2.5171425342559814,
      "learning_rate": 2.9647058823529414e-06,
      "loss": 0.06944979727268219,
      "num_input_tokens_seen": 4110898,
      "step": 3988,
      "train_runtime": 9704.7687,
      "train_tokens_per_second": 423.596
    },
    {
      "epoch": 1.7047761512982156,
      "grad_norm": 2.9781835079193115,
      "learning_rate": 2.9604278074866313e-06,
      "loss": 0.08037969470024109,
      "num_input_tokens_seen": 4111684,
      "step": 3989,
      "train_runtime": 9706.8175,
      "train_tokens_per_second": 423.587
    },
    {
      "epoch": 1.7052035473875415,
      "grad_norm": 2.7700400352478027,
      "learning_rate": 2.9561497326203207e-06,
      "loss": 0.08722593635320663,
      "num_input_tokens_seen": 4112802,
      "step": 3990,
      "train_runtime": 9708.9326,
      "train_tokens_per_second": 423.61
    },
    {
      "epoch": 1.705630943476867,
      "grad_norm": 3.2015161514282227,
      "learning_rate": 2.9518716577540106e-06,
      "loss": 0.11456476897001266,
      "num_input_tokens_seen": 4113660,
      "step": 3991,
      "train_runtime": 9717.1854,
      "train_tokens_per_second": 423.339
    },
    {
      "epoch": 1.706058339566193,
      "grad_norm": 2.407200336456299,
      "learning_rate": 2.9475935828877004e-06,
      "loss": 0.06708672642707825,
      "num_input_tokens_seen": 4114290,
      "step": 3992,
      "train_runtime": 9719.3327,
      "train_tokens_per_second": 423.31
    },
    {
      "epoch": 1.7064857356555188,
      "grad_norm": 4.023921012878418,
      "learning_rate": 2.9433155080213903e-06,
      "loss": 0.09772603958845139,
      "num_input_tokens_seen": 4115178,
      "step": 3993,
      "train_runtime": 9721.3956,
      "train_tokens_per_second": 423.311
    },
    {
      "epoch": 1.7069131317448445,
      "grad_norm": 3.6249475479125977,
      "learning_rate": 2.93903743315508e-06,
      "loss": 0.11192765086889267,
      "num_input_tokens_seen": 4116076,
      "step": 3994,
      "train_runtime": 9723.4429,
      "train_tokens_per_second": 423.315
    },
    {
      "epoch": 1.7073405278341705,
      "grad_norm": 2.1237800121307373,
      "learning_rate": 2.93475935828877e-06,
      "loss": 0.06910011917352676,
      "num_input_tokens_seen": 4117060,
      "step": 3995,
      "train_runtime": 9725.5291,
      "train_tokens_per_second": 423.325
    },
    {
      "epoch": 1.707767923923496,
      "grad_norm": 2.8688395023345947,
      "learning_rate": 2.93048128342246e-06,
      "loss": 0.09212521463632584,
      "num_input_tokens_seen": 4118136,
      "step": 3996,
      "train_runtime": 9727.6572,
      "train_tokens_per_second": 423.343
    },
    {
      "epoch": 1.708195320012822,
      "grad_norm": 3.399718761444092,
      "learning_rate": 2.92620320855615e-06,
      "loss": 0.04382210224866867,
      "num_input_tokens_seen": 4118922,
      "step": 3997,
      "train_runtime": 9729.6833,
      "train_tokens_per_second": 423.336
    },
    {
      "epoch": 1.7086227161021477,
      "grad_norm": 2.887514114379883,
      "learning_rate": 2.9219251336898397e-06,
      "loss": 0.12157633900642395,
      "num_input_tokens_seen": 4119792,
      "step": 3998,
      "train_runtime": 9731.7647,
      "train_tokens_per_second": 423.335
    },
    {
      "epoch": 1.7090501121914734,
      "grad_norm": 2.347560167312622,
      "learning_rate": 2.9176470588235295e-06,
      "loss": 0.12193876504898071,
      "num_input_tokens_seen": 4121640,
      "step": 3999,
      "train_runtime": 9734.0305,
      "train_tokens_per_second": 423.426
    },
    {
      "epoch": 1.7094775082807994,
      "grad_norm": 2.65649151802063,
      "learning_rate": 2.9133689839572194e-06,
      "loss": 0.08729573339223862,
      "num_input_tokens_seen": 4122298,
      "step": 4000,
      "train_runtime": 9735.9806,
      "train_tokens_per_second": 423.409
    },
    {
      "epoch": 1.7099049043701249,
      "grad_norm": 2.8900082111358643,
      "learning_rate": 2.9090909090909093e-06,
      "loss": 0.05193738639354706,
      "num_input_tokens_seen": 4123308,
      "step": 4001,
      "train_runtime": 9738.1461,
      "train_tokens_per_second": 423.418
    },
    {
      "epoch": 1.7103323004594508,
      "grad_norm": 3.2903778553009033,
      "learning_rate": 2.904812834224599e-06,
      "loss": 0.09346148371696472,
      "num_input_tokens_seen": 4124276,
      "step": 4002,
      "train_runtime": 9740.209,
      "train_tokens_per_second": 423.428
    },
    {
      "epoch": 1.7107596965487766,
      "grad_norm": 2.147174596786499,
      "learning_rate": 2.900534759358289e-06,
      "loss": 0.04343915730714798,
      "num_input_tokens_seen": 4125396,
      "step": 4003,
      "train_runtime": 9742.3233,
      "train_tokens_per_second": 423.451
    },
    {
      "epoch": 1.7111870926381023,
      "grad_norm": 2.4331154823303223,
      "learning_rate": 2.896256684491979e-06,
      "loss": 0.0731518566608429,
      "num_input_tokens_seen": 4126362,
      "step": 4004,
      "train_runtime": 9744.3706,
      "train_tokens_per_second": 423.461
    },
    {
      "epoch": 1.7116144887274283,
      "grad_norm": 2.9176928997039795,
      "learning_rate": 2.8919786096256687e-06,
      "loss": 0.08682416379451752,
      "num_input_tokens_seen": 4127178,
      "step": 4005,
      "train_runtime": 9746.4738,
      "train_tokens_per_second": 423.453
    },
    {
      "epoch": 1.7120418848167538,
      "grad_norm": 2.6199653148651123,
      "learning_rate": 2.8877005347593586e-06,
      "loss": 0.07903584092855453,
      "num_input_tokens_seen": 4128614,
      "step": 4006,
      "train_runtime": 9748.6771,
      "train_tokens_per_second": 423.505
    },
    {
      "epoch": 1.7124692809060797,
      "grad_norm": 1.868880271911621,
      "learning_rate": 2.883422459893048e-06,
      "loss": 0.052999626845121384,
      "num_input_tokens_seen": 4130058,
      "step": 4007,
      "train_runtime": 9750.844,
      "train_tokens_per_second": 423.559
    },
    {
      "epoch": 1.7128966769954055,
      "grad_norm": 1.91898775100708,
      "learning_rate": 2.879144385026738e-06,
      "loss": 0.0487280897796154,
      "num_input_tokens_seen": 4131012,
      "step": 4008,
      "train_runtime": 9752.9262,
      "train_tokens_per_second": 423.566
    },
    {
      "epoch": 1.7133240730847312,
      "grad_norm": 2.636772871017456,
      "learning_rate": 2.874866310160428e-06,
      "loss": 0.10910525172948837,
      "num_input_tokens_seen": 4132170,
      "step": 4009,
      "train_runtime": 9755.077,
      "train_tokens_per_second": 423.592
    },
    {
      "epoch": 1.7137514691740572,
      "grad_norm": 2.091592311859131,
      "learning_rate": 2.8705882352941177e-06,
      "loss": 0.06968137621879578,
      "num_input_tokens_seen": 4133064,
      "step": 4010,
      "train_runtime": 9757.1401,
      "train_tokens_per_second": 423.594
    },
    {
      "epoch": 1.7141788652633827,
      "grad_norm": 2.232311964035034,
      "learning_rate": 2.8663101604278075e-06,
      "loss": 0.08474753051996231,
      "num_input_tokens_seen": 4134600,
      "step": 4011,
      "train_runtime": 9759.3764,
      "train_tokens_per_second": 423.654
    },
    {
      "epoch": 1.7146062613527087,
      "grad_norm": 2.4938127994537354,
      "learning_rate": 2.8620320855614974e-06,
      "loss": 0.08659147471189499,
      "num_input_tokens_seen": 4135694,
      "step": 4012,
      "train_runtime": 9761.4945,
      "train_tokens_per_second": 423.674
    },
    {
      "epoch": 1.7150336574420344,
      "grad_norm": 1.7887046337127686,
      "learning_rate": 2.8577540106951873e-06,
      "loss": 0.057201042771339417,
      "num_input_tokens_seen": 4136848,
      "step": 4013,
      "train_runtime": 9763.6143,
      "train_tokens_per_second": 423.7
    },
    {
      "epoch": 1.7154610535313601,
      "grad_norm": 2.444875717163086,
      "learning_rate": 2.853475935828877e-06,
      "loss": 0.08035197108983994,
      "num_input_tokens_seen": 4137904,
      "step": 4014,
      "train_runtime": 9765.7148,
      "train_tokens_per_second": 423.717
    },
    {
      "epoch": 1.715888449620686,
      "grad_norm": 2.9307851791381836,
      "learning_rate": 2.849197860962567e-06,
      "loss": 0.08804693073034286,
      "num_input_tokens_seen": 4138772,
      "step": 4015,
      "train_runtime": 9767.7923,
      "train_tokens_per_second": 423.716
    },
    {
      "epoch": 1.7163158457100116,
      "grad_norm": 2.4342997074127197,
      "learning_rate": 2.844919786096257e-06,
      "loss": 0.07923261821269989,
      "num_input_tokens_seen": 4139926,
      "step": 4016,
      "train_runtime": 9769.9015,
      "train_tokens_per_second": 423.743
    },
    {
      "epoch": 1.7167432417993376,
      "grad_norm": 3.785456895828247,
      "learning_rate": 2.8406417112299468e-06,
      "loss": 0.07858654856681824,
      "num_input_tokens_seen": 4140438,
      "step": 4017,
      "train_runtime": 9771.8454,
      "train_tokens_per_second": 423.711
    },
    {
      "epoch": 1.7171706378886633,
      "grad_norm": 3.5055201053619385,
      "learning_rate": 2.8363636363636366e-06,
      "loss": 0.08689583837985992,
      "num_input_tokens_seen": 4141286,
      "step": 4018,
      "train_runtime": 9773.8673,
      "train_tokens_per_second": 423.71
    },
    {
      "epoch": 1.717598033977989,
      "grad_norm": 1.9284234046936035,
      "learning_rate": 2.8320855614973265e-06,
      "loss": 0.04013746604323387,
      "num_input_tokens_seen": 4142434,
      "step": 4019,
      "train_runtime": 9776.0056,
      "train_tokens_per_second": 423.735
    },
    {
      "epoch": 1.718025430067315,
      "grad_norm": 3.628349781036377,
      "learning_rate": 2.8278074866310164e-06,
      "loss": 0.09199301898479462,
      "num_input_tokens_seen": 4143234,
      "step": 4020,
      "train_runtime": 9778.0259,
      "train_tokens_per_second": 423.729
    },
    {
      "epoch": 1.7184528261566405,
      "grad_norm": 2.8210036754608154,
      "learning_rate": 2.8235294117647062e-06,
      "loss": 0.12050800025463104,
      "num_input_tokens_seen": 4144272,
      "step": 4021,
      "train_runtime": 9786.8242,
      "train_tokens_per_second": 423.454
    },
    {
      "epoch": 1.7188802222459665,
      "grad_norm": 6.9073638916015625,
      "learning_rate": 2.819251336898396e-06,
      "loss": 0.10918869078159332,
      "num_input_tokens_seen": 4145398,
      "step": 4022,
      "train_runtime": 9788.9841,
      "train_tokens_per_second": 423.476
    },
    {
      "epoch": 1.7193076183352922,
      "grad_norm": 3.853930950164795,
      "learning_rate": 2.814973262032086e-06,
      "loss": 0.1149483472108841,
      "num_input_tokens_seen": 4147038,
      "step": 4023,
      "train_runtime": 9791.2494,
      "train_tokens_per_second": 423.545
    },
    {
      "epoch": 1.719735014424618,
      "grad_norm": 2.2328014373779297,
      "learning_rate": 2.8106951871657754e-06,
      "loss": 0.06822262704372406,
      "num_input_tokens_seen": 4147982,
      "step": 4024,
      "train_runtime": 9793.3018,
      "train_tokens_per_second": 423.553
    },
    {
      "epoch": 1.720162410513944,
      "grad_norm": 2.3204293251037598,
      "learning_rate": 2.8064171122994653e-06,
      "loss": 0.08057768642902374,
      "num_input_tokens_seen": 4149004,
      "step": 4025,
      "train_runtime": 9795.391,
      "train_tokens_per_second": 423.567
    },
    {
      "epoch": 1.7205898066032694,
      "grad_norm": 2.331454038619995,
      "learning_rate": 2.802139037433155e-06,
      "loss": 0.045126643031835556,
      "num_input_tokens_seen": 4149740,
      "step": 4026,
      "train_runtime": 9797.3776,
      "train_tokens_per_second": 423.556
    },
    {
      "epoch": 1.7210172026925954,
      "grad_norm": 3.34110951423645,
      "learning_rate": 2.797860962566845e-06,
      "loss": 0.11113997548818588,
      "num_input_tokens_seen": 4150614,
      "step": 4027,
      "train_runtime": 9799.4323,
      "train_tokens_per_second": 423.557
    },
    {
      "epoch": 1.7214445987819211,
      "grad_norm": 2.157827615737915,
      "learning_rate": 2.793582887700535e-06,
      "loss": 0.07617693394422531,
      "num_input_tokens_seen": 4152094,
      "step": 4028,
      "train_runtime": 9801.6244,
      "train_tokens_per_second": 423.613
    },
    {
      "epoch": 1.7218719948712469,
      "grad_norm": 2.674255609512329,
      "learning_rate": 2.7893048128342248e-06,
      "loss": 0.05856155976653099,
      "num_input_tokens_seen": 4152830,
      "step": 4029,
      "train_runtime": 9803.6454,
      "train_tokens_per_second": 423.601
    },
    {
      "epoch": 1.7222993909605728,
      "grad_norm": 2.7371299266815186,
      "learning_rate": 2.7850267379679146e-06,
      "loss": 0.09654857218265533,
      "num_input_tokens_seen": 4154218,
      "step": 4030,
      "train_runtime": 9805.842,
      "train_tokens_per_second": 423.647
    },
    {
      "epoch": 1.7227267870498983,
      "grad_norm": 2.5340161323547363,
      "learning_rate": 2.7807486631016045e-06,
      "loss": 0.10056625306606293,
      "num_input_tokens_seen": 4155316,
      "step": 4031,
      "train_runtime": 9808.0092,
      "train_tokens_per_second": 423.666
    },
    {
      "epoch": 1.7231541831392243,
      "grad_norm": 3.9756827354431152,
      "learning_rate": 2.7764705882352944e-06,
      "loss": 0.12368873506784439,
      "num_input_tokens_seen": 4156016,
      "step": 4032,
      "train_runtime": 9810.0293,
      "train_tokens_per_second": 423.65
    },
    {
      "epoch": 1.72358157922855,
      "grad_norm": 2.6393966674804688,
      "learning_rate": 2.7721925133689842e-06,
      "loss": 0.09346798062324524,
      "num_input_tokens_seen": 4157556,
      "step": 4033,
      "train_runtime": 9812.2483,
      "train_tokens_per_second": 423.711
    },
    {
      "epoch": 1.7240089753178758,
      "grad_norm": 2.4262278079986572,
      "learning_rate": 2.767914438502674e-06,
      "loss": 0.08767855167388916,
      "num_input_tokens_seen": 4158576,
      "step": 4034,
      "train_runtime": 9814.4968,
      "train_tokens_per_second": 423.718
    },
    {
      "epoch": 1.7244363714072017,
      "grad_norm": 2.7044572830200195,
      "learning_rate": 2.763636363636364e-06,
      "loss": 0.0657733678817749,
      "num_input_tokens_seen": 4159834,
      "step": 4035,
      "train_runtime": 9816.7045,
      "train_tokens_per_second": 423.751
    },
    {
      "epoch": 1.7248637674965273,
      "grad_norm": 2.0652520656585693,
      "learning_rate": 2.759358288770054e-06,
      "loss": 0.10201188176870346,
      "num_input_tokens_seen": 4161608,
      "step": 4036,
      "train_runtime": 9819.0163,
      "train_tokens_per_second": 423.831
    },
    {
      "epoch": 1.7252911635858532,
      "grad_norm": 2.7821693420410156,
      "learning_rate": 2.7550802139037437e-06,
      "loss": 0.09517615288496017,
      "num_input_tokens_seen": 4162850,
      "step": 4037,
      "train_runtime": 9821.1744,
      "train_tokens_per_second": 423.865
    },
    {
      "epoch": 1.725718559675179,
      "grad_norm": 1.9511961936950684,
      "learning_rate": 2.7508021390374336e-06,
      "loss": 0.07643108069896698,
      "num_input_tokens_seen": 4164050,
      "step": 4038,
      "train_runtime": 9823.3822,
      "train_tokens_per_second": 423.892
    },
    {
      "epoch": 1.7261459557645047,
      "grad_norm": 2.4407958984375,
      "learning_rate": 2.7465240641711235e-06,
      "loss": 0.07977095991373062,
      "num_input_tokens_seen": 4165072,
      "step": 4039,
      "train_runtime": 9825.4732,
      "train_tokens_per_second": 423.905
    },
    {
      "epoch": 1.7265733518538307,
      "grad_norm": 3.0610175132751465,
      "learning_rate": 2.7422459893048133e-06,
      "loss": 0.06689866632223129,
      "num_input_tokens_seen": 4165922,
      "step": 4040,
      "train_runtime": 9827.5314,
      "train_tokens_per_second": 423.903
    },
    {
      "epoch": 1.7270007479431562,
      "grad_norm": 5.814146518707275,
      "learning_rate": 2.7379679144385028e-06,
      "loss": 0.09602482616901398,
      "num_input_tokens_seen": 4166786,
      "step": 4041,
      "train_runtime": 9829.5902,
      "train_tokens_per_second": 423.902
    },
    {
      "epoch": 1.7274281440324821,
      "grad_norm": 3.5886621475219727,
      "learning_rate": 2.7336898395721926e-06,
      "loss": 0.1257716715335846,
      "num_input_tokens_seen": 4167642,
      "step": 4042,
      "train_runtime": 9831.6772,
      "train_tokens_per_second": 423.899
    },
    {
      "epoch": 1.7278555401218079,
      "grad_norm": 2.48893141746521,
      "learning_rate": 2.7294117647058825e-06,
      "loss": 0.1100153923034668,
      "num_input_tokens_seen": 4168700,
      "step": 4043,
      "train_runtime": 9833.7552,
      "train_tokens_per_second": 423.917
    },
    {
      "epoch": 1.7282829362111336,
      "grad_norm": 2.379669427871704,
      "learning_rate": 2.7251336898395724e-06,
      "loss": 0.05183770880103111,
      "num_input_tokens_seen": 4169840,
      "step": 4044,
      "train_runtime": 9835.9603,
      "train_tokens_per_second": 423.938
    },
    {
      "epoch": 1.7287103323004596,
      "grad_norm": 3.1207971572875977,
      "learning_rate": 2.7208556149732622e-06,
      "loss": 0.06831379234790802,
      "num_input_tokens_seen": 4170512,
      "step": 4045,
      "train_runtime": 9837.9654,
      "train_tokens_per_second": 423.92
    },
    {
      "epoch": 1.729137728389785,
      "grad_norm": 2.1130356788635254,
      "learning_rate": 2.716577540106952e-06,
      "loss": 0.058660343289375305,
      "num_input_tokens_seen": 4171454,
      "step": 4046,
      "train_runtime": 9840.0345,
      "train_tokens_per_second": 423.927
    },
    {
      "epoch": 1.729565124479111,
      "grad_norm": 4.324986457824707,
      "learning_rate": 2.712299465240642e-06,
      "loss": 0.10213802009820938,
      "num_input_tokens_seen": 4172116,
      "step": 4047,
      "train_runtime": 9842.1705,
      "train_tokens_per_second": 423.902
    },
    {
      "epoch": 1.7299925205684368,
      "grad_norm": 2.6436688899993896,
      "learning_rate": 2.708021390374332e-06,
      "loss": 0.05921069532632828,
      "num_input_tokens_seen": 4173226,
      "step": 4048,
      "train_runtime": 9844.397,
      "train_tokens_per_second": 423.919
    },
    {
      "epoch": 1.7304199166577625,
      "grad_norm": 2.5965993404388428,
      "learning_rate": 2.7037433155080217e-06,
      "loss": 0.08830476552248001,
      "num_input_tokens_seen": 4174102,
      "step": 4049,
      "train_runtime": 9846.622,
      "train_tokens_per_second": 423.912
    },
    {
      "epoch": 1.7308473127470885,
      "grad_norm": 1.9387271404266357,
      "learning_rate": 2.6994652406417116e-06,
      "loss": 0.04834958165884018,
      "num_input_tokens_seen": 4174824,
      "step": 4050,
      "train_runtime": 9848.6166,
      "train_tokens_per_second": 423.9
    },
    {
      "epoch": 1.731274708836414,
      "grad_norm": 2.625678777694702,
      "learning_rate": 2.6951871657754015e-06,
      "loss": 0.06536933034658432,
      "num_input_tokens_seen": 4176128,
      "step": 4051,
      "train_runtime": 9857.4035,
      "train_tokens_per_second": 423.654
    },
    {
      "epoch": 1.73170210492574,
      "grad_norm": 2.0632362365722656,
      "learning_rate": 2.6909090909090913e-06,
      "loss": 0.07165873795747757,
      "num_input_tokens_seen": 4177458,
      "step": 4052,
      "train_runtime": 9859.5531,
      "train_tokens_per_second": 423.696
    },
    {
      "epoch": 1.7321295010150657,
      "grad_norm": 2.3923587799072266,
      "learning_rate": 2.686631016042781e-06,
      "loss": 0.08612820506095886,
      "num_input_tokens_seen": 4178516,
      "step": 4053,
      "train_runtime": 9861.7402,
      "train_tokens_per_second": 423.71
    },
    {
      "epoch": 1.7325568971043914,
      "grad_norm": 2.4570765495300293,
      "learning_rate": 2.682352941176471e-06,
      "loss": 0.07350552082061768,
      "num_input_tokens_seen": 4179284,
      "step": 4054,
      "train_runtime": 9863.7989,
      "train_tokens_per_second": 423.699
    },
    {
      "epoch": 1.7329842931937174,
      "grad_norm": 3.116313934326172,
      "learning_rate": 2.678074866310161e-06,
      "loss": 0.09485094994306564,
      "num_input_tokens_seen": 4180324,
      "step": 4055,
      "train_runtime": 9865.838,
      "train_tokens_per_second": 423.717
    },
    {
      "epoch": 1.733411689283043,
      "grad_norm": 2.48117995262146,
      "learning_rate": 2.673796791443851e-06,
      "loss": 0.06896473467350006,
      "num_input_tokens_seen": 4181230,
      "step": 4056,
      "train_runtime": 9867.8973,
      "train_tokens_per_second": 423.72
    },
    {
      "epoch": 1.7338390853723689,
      "grad_norm": 2.2488813400268555,
      "learning_rate": 2.6695187165775407e-06,
      "loss": 0.06654829531908035,
      "num_input_tokens_seen": 4182154,
      "step": 4057,
      "train_runtime": 9869.9419,
      "train_tokens_per_second": 423.726
    },
    {
      "epoch": 1.7342664814616946,
      "grad_norm": 2.7358016967773438,
      "learning_rate": 2.66524064171123e-06,
      "loss": 0.08781205117702484,
      "num_input_tokens_seen": 4183822,
      "step": 4058,
      "train_runtime": 9872.1928,
      "train_tokens_per_second": 423.799
    },
    {
      "epoch": 1.7346938775510203,
      "grad_norm": 2.7917540073394775,
      "learning_rate": 2.66096256684492e-06,
      "loss": 0.0921274721622467,
      "num_input_tokens_seen": 4184720,
      "step": 4059,
      "train_runtime": 9874.2316,
      "train_tokens_per_second": 423.802
    },
    {
      "epoch": 1.7351212736403463,
      "grad_norm": 2.315652370452881,
      "learning_rate": 2.65668449197861e-06,
      "loss": 0.07591191679239273,
      "num_input_tokens_seen": 4185892,
      "step": 4060,
      "train_runtime": 9876.3306,
      "train_tokens_per_second": 423.831
    },
    {
      "epoch": 1.7355486697296718,
      "grad_norm": 2.635467290878296,
      "learning_rate": 2.6524064171122997e-06,
      "loss": 0.10073072463274002,
      "num_input_tokens_seen": 4186882,
      "step": 4061,
      "train_runtime": 9878.4004,
      "train_tokens_per_second": 423.842
    },
    {
      "epoch": 1.7359760658189978,
      "grad_norm": 4.159154891967773,
      "learning_rate": 2.6481283422459896e-06,
      "loss": 0.081559918820858,
      "num_input_tokens_seen": 4187578,
      "step": 4062,
      "train_runtime": 9880.3781,
      "train_tokens_per_second": 423.828
    },
    {
      "epoch": 1.7364034619083235,
      "grad_norm": 3.1609933376312256,
      "learning_rate": 2.6438502673796795e-06,
      "loss": 0.09514006227254868,
      "num_input_tokens_seen": 4188628,
      "step": 4063,
      "train_runtime": 9882.4607,
      "train_tokens_per_second": 423.845
    },
    {
      "epoch": 1.7368308579976492,
      "grad_norm": 2.9850661754608154,
      "learning_rate": 2.6395721925133693e-06,
      "loss": 0.09671429544687271,
      "num_input_tokens_seen": 4189492,
      "step": 4064,
      "train_runtime": 9884.4971,
      "train_tokens_per_second": 423.845
    },
    {
      "epoch": 1.7372582540869752,
      "grad_norm": 1.6393934488296509,
      "learning_rate": 2.635294117647059e-06,
      "loss": 0.041917793452739716,
      "num_input_tokens_seen": 4190846,
      "step": 4065,
      "train_runtime": 9886.677,
      "train_tokens_per_second": 423.888
    },
    {
      "epoch": 1.7376856501763007,
      "grad_norm": 2.439072847366333,
      "learning_rate": 2.631016042780749e-06,
      "loss": 0.07730145752429962,
      "num_input_tokens_seen": 4191792,
      "step": 4066,
      "train_runtime": 9888.7218,
      "train_tokens_per_second": 423.896
    },
    {
      "epoch": 1.7381130462656267,
      "grad_norm": 1.4326391220092773,
      "learning_rate": 2.626737967914439e-06,
      "loss": 0.05624020844697952,
      "num_input_tokens_seen": 4193618,
      "step": 4067,
      "train_runtime": 9891.0096,
      "train_tokens_per_second": 423.983
    },
    {
      "epoch": 1.7385404423549524,
      "grad_norm": 2.28743052482605,
      "learning_rate": 2.622459893048129e-06,
      "loss": 0.07754433900117874,
      "num_input_tokens_seen": 4194466,
      "step": 4068,
      "train_runtime": 9893.0257,
      "train_tokens_per_second": 423.982
    },
    {
      "epoch": 1.7389678384442782,
      "grad_norm": 2.206310510635376,
      "learning_rate": 2.6181818181818187e-06,
      "loss": 0.03837336227297783,
      "num_input_tokens_seen": 4195320,
      "step": 4069,
      "train_runtime": 9895.0384,
      "train_tokens_per_second": 423.982
    },
    {
      "epoch": 1.7393952345336041,
      "grad_norm": 2.664803981781006,
      "learning_rate": 2.6139037433155085e-06,
      "loss": 0.11692618578672409,
      "num_input_tokens_seen": 4197270,
      "step": 4070,
      "train_runtime": 9897.3749,
      "train_tokens_per_second": 424.079
    },
    {
      "epoch": 1.7398226306229299,
      "grad_norm": 3.1426937580108643,
      "learning_rate": 2.6096256684491984e-06,
      "loss": 0.09302536398172379,
      "num_input_tokens_seen": 4198056,
      "step": 4071,
      "train_runtime": 9899.4233,
      "train_tokens_per_second": 424.071
    },
    {
      "epoch": 1.7402500267122556,
      "grad_norm": 2.822409152984619,
      "learning_rate": 2.6053475935828883e-06,
      "loss": 0.09082210063934326,
      "num_input_tokens_seen": 4198842,
      "step": 4072,
      "train_runtime": 9901.4388,
      "train_tokens_per_second": 424.064
    },
    {
      "epoch": 1.7406774228015813,
      "grad_norm": 2.297144651412964,
      "learning_rate": 2.601069518716578e-06,
      "loss": 0.10644955188035965,
      "num_input_tokens_seen": 4200016,
      "step": 4073,
      "train_runtime": 9903.5532,
      "train_tokens_per_second": 424.092
    },
    {
      "epoch": 1.741104818890907,
      "grad_norm": 2.7085657119750977,
      "learning_rate": 2.596791443850268e-06,
      "loss": 0.07879453152418137,
      "num_input_tokens_seen": 4200926,
      "step": 4074,
      "train_runtime": 9905.6307,
      "train_tokens_per_second": 424.095
    },
    {
      "epoch": 1.741532214980233,
      "grad_norm": 3.2154722213745117,
      "learning_rate": 2.5925133689839575e-06,
      "loss": 0.10776600986719131,
      "num_input_tokens_seen": 4201996,
      "step": 4075,
      "train_runtime": 9907.7143,
      "train_tokens_per_second": 424.114
    },
    {
      "epoch": 1.7419596110695588,
      "grad_norm": 2.3599696159362793,
      "learning_rate": 2.5882352941176473e-06,
      "loss": 0.08735352754592896,
      "num_input_tokens_seen": 4203100,
      "step": 4076,
      "train_runtime": 9909.8243,
      "train_tokens_per_second": 424.135
    },
    {
      "epoch": 1.7423870071588845,
      "grad_norm": 2.141378879547119,
      "learning_rate": 2.583957219251337e-06,
      "loss": 0.044425349682569504,
      "num_input_tokens_seen": 4204022,
      "step": 4077,
      "train_runtime": 9911.9223,
      "train_tokens_per_second": 424.138
    },
    {
      "epoch": 1.7428144032482102,
      "grad_norm": 3.5553226470947266,
      "learning_rate": 2.579679144385027e-06,
      "loss": 0.16258250176906586,
      "num_input_tokens_seen": 4205362,
      "step": 4078,
      "train_runtime": 9914.0957,
      "train_tokens_per_second": 424.18
    },
    {
      "epoch": 1.743241799337536,
      "grad_norm": 2.341912031173706,
      "learning_rate": 2.575401069518717e-06,
      "loss": 0.08952471613883972,
      "num_input_tokens_seen": 4206674,
      "step": 4079,
      "train_runtime": 9916.2411,
      "train_tokens_per_second": 424.221
    },
    {
      "epoch": 1.743669195426862,
      "grad_norm": 3.0861172676086426,
      "learning_rate": 2.571122994652407e-06,
      "loss": 0.102879099547863,
      "num_input_tokens_seen": 4207948,
      "step": 4080,
      "train_runtime": 9918.3916,
      "train_tokens_per_second": 424.257
    },
    {
      "epoch": 1.7440965915161877,
      "grad_norm": 2.2038893699645996,
      "learning_rate": 2.5668449197860967e-06,
      "loss": 0.0634625032544136,
      "num_input_tokens_seen": 4208712,
      "step": 4081,
      "train_runtime": 9927.2476,
      "train_tokens_per_second": 423.956
    },
    {
      "epoch": 1.7445239876055134,
      "grad_norm": 2.1982100009918213,
      "learning_rate": 2.5625668449197866e-06,
      "loss": 0.05916162207722664,
      "num_input_tokens_seen": 4209686,
      "step": 4082,
      "train_runtime": 9929.3268,
      "train_tokens_per_second": 423.965
    },
    {
      "epoch": 1.7449513836948392,
      "grad_norm": 2.9628520011901855,
      "learning_rate": 2.5582887700534764e-06,
      "loss": 0.1246619001030922,
      "num_input_tokens_seen": 4210652,
      "step": 4083,
      "train_runtime": 9931.4104,
      "train_tokens_per_second": 423.973
    },
    {
      "epoch": 1.745378779784165,
      "grad_norm": 2.510510206222534,
      "learning_rate": 2.554010695187166e-06,
      "loss": 0.08361392468214035,
      "num_input_tokens_seen": 4211788,
      "step": 4084,
      "train_runtime": 9933.5228,
      "train_tokens_per_second": 423.997
    },
    {
      "epoch": 1.7458061758734909,
      "grad_norm": 1.9237068891525269,
      "learning_rate": 2.5497326203208557e-06,
      "loss": 0.06083056330680847,
      "num_input_tokens_seen": 4213396,
      "step": 4085,
      "train_runtime": 9935.7512,
      "train_tokens_per_second": 424.064
    },
    {
      "epoch": 1.7462335719628166,
      "grad_norm": 2.109754800796509,
      "learning_rate": 2.5454545454545456e-06,
      "loss": 0.06658701598644257,
      "num_input_tokens_seen": 4214486,
      "step": 4086,
      "train_runtime": 9937.8984,
      "train_tokens_per_second": 424.082
    },
    {
      "epoch": 1.7466609680521423,
      "grad_norm": 3.3641178607940674,
      "learning_rate": 2.541176470588235e-06,
      "loss": 0.10231752693653107,
      "num_input_tokens_seen": 4215466,
      "step": 4087,
      "train_runtime": 9939.9721,
      "train_tokens_per_second": 424.092
    },
    {
      "epoch": 1.747088364141468,
      "grad_norm": 1.9562214612960815,
      "learning_rate": 2.536898395721925e-06,
      "loss": 0.06954551488161087,
      "num_input_tokens_seen": 4216694,
      "step": 4088,
      "train_runtime": 9942.0867,
      "train_tokens_per_second": 424.126
    },
    {
      "epoch": 1.7475157602307938,
      "grad_norm": 4.043095111846924,
      "learning_rate": 2.532620320855615e-06,
      "loss": 0.08843319118022919,
      "num_input_tokens_seen": 4217480,
      "step": 4089,
      "train_runtime": 9944.131,
      "train_tokens_per_second": 424.118
    },
    {
      "epoch": 1.7479431563201198,
      "grad_norm": 2.893369197845459,
      "learning_rate": 2.5283422459893047e-06,
      "loss": 0.10958456993103027,
      "num_input_tokens_seen": 4218646,
      "step": 4090,
      "train_runtime": 9946.2599,
      "train_tokens_per_second": 424.144
    },
    {
      "epoch": 1.7483705524094455,
      "grad_norm": 2.7690653800964355,
      "learning_rate": 2.5240641711229945e-06,
      "loss": 0.06869184970855713,
      "num_input_tokens_seen": 4219358,
      "step": 4091,
      "train_runtime": 9948.2462,
      "train_tokens_per_second": 424.131
    },
    {
      "epoch": 1.7487979484987712,
      "grad_norm": 2.2930610179901123,
      "learning_rate": 2.5197860962566844e-06,
      "loss": 0.04717502370476723,
      "num_input_tokens_seen": 4220032,
      "step": 4092,
      "train_runtime": 9950.2352,
      "train_tokens_per_second": 424.114
    },
    {
      "epoch": 1.749225344588097,
      "grad_norm": 3.5082364082336426,
      "learning_rate": 2.5155080213903743e-06,
      "loss": 0.08688130974769592,
      "num_input_tokens_seen": 4220818,
      "step": 4093,
      "train_runtime": 9952.4075,
      "train_tokens_per_second": 424.1
    },
    {
      "epoch": 1.7496527406774227,
      "grad_norm": 2.0089828968048096,
      "learning_rate": 2.511229946524064e-06,
      "loss": 0.07756157964468002,
      "num_input_tokens_seen": 4221936,
      "step": 4094,
      "train_runtime": 9954.5544,
      "train_tokens_per_second": 424.121
    },
    {
      "epoch": 1.7500801367667487,
      "grad_norm": 2.7302162647247314,
      "learning_rate": 2.506951871657754e-06,
      "loss": 0.08377182483673096,
      "num_input_tokens_seen": 4223002,
      "step": 4095,
      "train_runtime": 9956.704,
      "train_tokens_per_second": 424.137
    },
    {
      "epoch": 1.7505075328560744,
      "grad_norm": 2.079789638519287,
      "learning_rate": 2.502673796791444e-06,
      "loss": 0.09523649513721466,
      "num_input_tokens_seen": 4224186,
      "step": 4096,
      "train_runtime": 9958.7805,
      "train_tokens_per_second": 424.167
    },
    {
      "epoch": 1.7509349289454001,
      "grad_norm": 2.611577272415161,
      "learning_rate": 2.498395721925134e-06,
      "loss": 0.06579362601041794,
      "num_input_tokens_seen": 4225136,
      "step": 4097,
      "train_runtime": 9960.8575,
      "train_tokens_per_second": 424.174
    },
    {
      "epoch": 1.7513623250347259,
      "grad_norm": 2.561326026916504,
      "learning_rate": 2.4941176470588236e-06,
      "loss": 0.06817270815372467,
      "num_input_tokens_seen": 4226086,
      "step": 4098,
      "train_runtime": 9962.9875,
      "train_tokens_per_second": 424.179
    },
    {
      "epoch": 1.7517897211240516,
      "grad_norm": 2.205995559692383,
      "learning_rate": 2.4898395721925135e-06,
      "loss": 0.08090948313474655,
      "num_input_tokens_seen": 4227154,
      "step": 4099,
      "train_runtime": 9965.069,
      "train_tokens_per_second": 424.197
    },
    {
      "epoch": 1.7522171172133776,
      "grad_norm": 2.5278356075286865,
      "learning_rate": 2.4855614973262034e-06,
      "loss": 0.06919476389884949,
      "num_input_tokens_seen": 4228074,
      "step": 4100,
      "train_runtime": 9967.1196,
      "train_tokens_per_second": 424.202
    },
    {
      "epoch": 1.7526445133027033,
      "grad_norm": 1.5690151453018188,
      "learning_rate": 2.4812834224598932e-06,
      "loss": 0.05812923237681389,
      "num_input_tokens_seen": 4229166,
      "step": 4101,
      "train_runtime": 9969.2039,
      "train_tokens_per_second": 424.223
    },
    {
      "epoch": 1.753071909392029,
      "grad_norm": 2.4146010875701904,
      "learning_rate": 2.477005347593583e-06,
      "loss": 0.09066355973482132,
      "num_input_tokens_seen": 4230220,
      "step": 4102,
      "train_runtime": 9971.328,
      "train_tokens_per_second": 424.238
    },
    {
      "epoch": 1.7534993054813548,
      "grad_norm": 2.908135175704956,
      "learning_rate": 2.472727272727273e-06,
      "loss": 0.0990167185664177,
      "num_input_tokens_seen": 4231336,
      "step": 4103,
      "train_runtime": 9973.4088,
      "train_tokens_per_second": 424.262
    },
    {
      "epoch": 1.7539267015706805,
      "grad_norm": 2.9105708599090576,
      "learning_rate": 2.468449197860963e-06,
      "loss": 0.07659521698951721,
      "num_input_tokens_seen": 4232484,
      "step": 4104,
      "train_runtime": 9975.561,
      "train_tokens_per_second": 424.285
    },
    {
      "epoch": 1.7543540976600065,
      "grad_norm": 0.8447878360748291,
      "learning_rate": 2.4641711229946527e-06,
      "loss": 0.023198440670967102,
      "num_input_tokens_seen": 4235772,
      "step": 4105,
      "train_runtime": 9978.1995,
      "train_tokens_per_second": 424.503
    },
    {
      "epoch": 1.7547814937493322,
      "grad_norm": 1.9445749521255493,
      "learning_rate": 2.4598930481283426e-06,
      "loss": 0.07758910208940506,
      "num_input_tokens_seen": 4237104,
      "step": 4106,
      "train_runtime": 9980.36,
      "train_tokens_per_second": 424.544
    },
    {
      "epoch": 1.755208889838658,
      "grad_norm": 2.8601250648498535,
      "learning_rate": 2.4556149732620324e-06,
      "loss": 0.1165383830666542,
      "num_input_tokens_seen": 4238502,
      "step": 4107,
      "train_runtime": 9982.5433,
      "train_tokens_per_second": 424.591
    },
    {
      "epoch": 1.7556362859279837,
      "grad_norm": 2.2085225582122803,
      "learning_rate": 2.4513368983957223e-06,
      "loss": 0.07841064035892487,
      "num_input_tokens_seen": 4239390,
      "step": 4108,
      "train_runtime": 9984.6094,
      "train_tokens_per_second": 424.592
    },
    {
      "epoch": 1.7560636820173094,
      "grad_norm": 2.3799362182617188,
      "learning_rate": 2.447058823529412e-06,
      "loss": 0.05659405514597893,
      "num_input_tokens_seen": 4240352,
      "step": 4109,
      "train_runtime": 9986.7061,
      "train_tokens_per_second": 424.6
    },
    {
      "epoch": 1.7564910781066354,
      "grad_norm": 3.6929924488067627,
      "learning_rate": 2.442780748663102e-06,
      "loss": 0.11274654418230057,
      "num_input_tokens_seen": 4241222,
      "step": 4110,
      "train_runtime": 9988.7818,
      "train_tokens_per_second": 424.599
    },
    {
      "epoch": 1.7569184741959611,
      "grad_norm": 2.2679834365844727,
      "learning_rate": 2.4385026737967915e-06,
      "loss": 0.059855926781892776,
      "num_input_tokens_seen": 4241998,
      "step": 4111,
      "train_runtime": 9997.4047,
      "train_tokens_per_second": 424.31
    },
    {
      "epoch": 1.7573458702852869,
      "grad_norm": 1.7819002866744995,
      "learning_rate": 2.4342245989304814e-06,
      "loss": 0.06576694548130035,
      "num_input_tokens_seen": 4243686,
      "step": 4112,
      "train_runtime": 9999.723,
      "train_tokens_per_second": 424.38
    },
    {
      "epoch": 1.7577732663746126,
      "grad_norm": 1.7435522079467773,
      "learning_rate": 2.4299465240641712e-06,
      "loss": 0.07108808308839798,
      "num_input_tokens_seen": 4245760,
      "step": 4113,
      "train_runtime": 10002.1461,
      "train_tokens_per_second": 424.485
    },
    {
      "epoch": 1.7582006624639384,
      "grad_norm": 4.905389785766602,
      "learning_rate": 2.425668449197861e-06,
      "loss": 0.13813750445842743,
      "num_input_tokens_seen": 4246502,
      "step": 4114,
      "train_runtime": 10004.1666,
      "train_tokens_per_second": 424.473
    },
    {
      "epoch": 1.7586280585532643,
      "grad_norm": 2.568931818008423,
      "learning_rate": 2.421390374331551e-06,
      "loss": 0.10430927574634552,
      "num_input_tokens_seen": 4247572,
      "step": 4115,
      "train_runtime": 10006.2875,
      "train_tokens_per_second": 424.49
    },
    {
      "epoch": 1.75905545464259,
      "grad_norm": 3.5114471912384033,
      "learning_rate": 2.417112299465241e-06,
      "loss": 0.11384786665439606,
      "num_input_tokens_seen": 4248364,
      "step": 4116,
      "train_runtime": 10008.3559,
      "train_tokens_per_second": 424.482
    },
    {
      "epoch": 1.7594828507319158,
      "grad_norm": 3.654829502105713,
      "learning_rate": 2.4128342245989307e-06,
      "loss": 0.08320469409227371,
      "num_input_tokens_seen": 4249160,
      "step": 4117,
      "train_runtime": 10010.4983,
      "train_tokens_per_second": 424.47
    },
    {
      "epoch": 1.7599102468212415,
      "grad_norm": 2.2291922569274902,
      "learning_rate": 2.4085561497326206e-06,
      "loss": 0.0989278107881546,
      "num_input_tokens_seen": 4250340,
      "step": 4118,
      "train_runtime": 10012.6239,
      "train_tokens_per_second": 424.498
    },
    {
      "epoch": 1.7603376429105673,
      "grad_norm": 3.1687824726104736,
      "learning_rate": 2.4042780748663104e-06,
      "loss": 0.09292609989643097,
      "num_input_tokens_seen": 4250860,
      "step": 4119,
      "train_runtime": 10014.6076,
      "train_tokens_per_second": 424.466
    },
    {
      "epoch": 1.7607650389998932,
      "grad_norm": 4.733868598937988,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.09295797348022461,
      "num_input_tokens_seen": 4251376,
      "step": 4120,
      "train_runtime": 10016.6047,
      "train_tokens_per_second": 424.433
    },
    {
      "epoch": 1.761192435089219,
      "grad_norm": 2.754350185394287,
      "learning_rate": 2.3957219251336898e-06,
      "loss": 0.10203735530376434,
      "num_input_tokens_seen": 4252416,
      "step": 4121,
      "train_runtime": 10018.7618,
      "train_tokens_per_second": 424.445
    },
    {
      "epoch": 1.7616198311785447,
      "grad_norm": 3.1365554332733154,
      "learning_rate": 2.3914438502673796e-06,
      "loss": 0.08412344753742218,
      "num_input_tokens_seen": 4253334,
      "step": 4122,
      "train_runtime": 10020.8355,
      "train_tokens_per_second": 424.449
    },
    {
      "epoch": 1.7620472272678707,
      "grad_norm": 2.3730523586273193,
      "learning_rate": 2.3871657754010695e-06,
      "loss": 0.09628761559724808,
      "num_input_tokens_seen": 4254330,
      "step": 4123,
      "train_runtime": 10022.9015,
      "train_tokens_per_second": 424.461
    },
    {
      "epoch": 1.7624746233571962,
      "grad_norm": 2.632103443145752,
      "learning_rate": 2.3828877005347594e-06,
      "loss": 0.06915301084518433,
      "num_input_tokens_seen": 4255110,
      "step": 4124,
      "train_runtime": 10024.9271,
      "train_tokens_per_second": 424.453
    },
    {
      "epoch": 1.7629020194465221,
      "grad_norm": 2.8245201110839844,
      "learning_rate": 2.3786096256684492e-06,
      "loss": 0.11383970081806183,
      "num_input_tokens_seen": 4256130,
      "step": 4125,
      "train_runtime": 10026.9951,
      "train_tokens_per_second": 424.467
    },
    {
      "epoch": 1.7633294155358479,
      "grad_norm": 1.8119213581085205,
      "learning_rate": 2.374331550802139e-06,
      "loss": 0.06335105001926422,
      "num_input_tokens_seen": 4257098,
      "step": 4126,
      "train_runtime": 10029.1057,
      "train_tokens_per_second": 424.474
    },
    {
      "epoch": 1.7637568116251736,
      "grad_norm": 2.3820314407348633,
      "learning_rate": 2.370053475935829e-06,
      "loss": 0.05974426120519638,
      "num_input_tokens_seen": 4258012,
      "step": 4127,
      "train_runtime": 10031.149,
      "train_tokens_per_second": 424.479
    },
    {
      "epoch": 1.7641842077144996,
      "grad_norm": 2.4769506454467773,
      "learning_rate": 2.365775401069519e-06,
      "loss": 0.0872613787651062,
      "num_input_tokens_seen": 4259294,
      "step": 4128,
      "train_runtime": 10033.33,
      "train_tokens_per_second": 424.514
    },
    {
      "epoch": 1.764611603803825,
      "grad_norm": 3.7077057361602783,
      "learning_rate": 2.3614973262032087e-06,
      "loss": 0.09083158522844315,
      "num_input_tokens_seen": 4259914,
      "step": 4129,
      "train_runtime": 10035.3922,
      "train_tokens_per_second": 424.489
    },
    {
      "epoch": 1.765038999893151,
      "grad_norm": 2.7601659297943115,
      "learning_rate": 2.3572192513368986e-06,
      "loss": 0.09662850201129913,
      "num_input_tokens_seen": 4261544,
      "step": 4130,
      "train_runtime": 10037.5959,
      "train_tokens_per_second": 424.558
    },
    {
      "epoch": 1.7654663959824768,
      "grad_norm": 2.9929752349853516,
      "learning_rate": 2.3529411764705885e-06,
      "loss": 0.11560578644275665,
      "num_input_tokens_seen": 4262430,
      "step": 4131,
      "train_runtime": 10039.6551,
      "train_tokens_per_second": 424.559
    },
    {
      "epoch": 1.7658937920718025,
      "grad_norm": 2.1190741062164307,
      "learning_rate": 2.3486631016042783e-06,
      "loss": 0.07116225361824036,
      "num_input_tokens_seen": 4263670,
      "step": 4132,
      "train_runtime": 10041.8177,
      "train_tokens_per_second": 424.591
    },
    {
      "epoch": 1.7663211881611285,
      "grad_norm": 3.2644152641296387,
      "learning_rate": 2.344385026737968e-06,
      "loss": 0.07705575227737427,
      "num_input_tokens_seen": 4264428,
      "step": 4133,
      "train_runtime": 10043.804,
      "train_tokens_per_second": 424.583
    },
    {
      "epoch": 1.766748584250454,
      "grad_norm": 2.477987766265869,
      "learning_rate": 2.340106951871658e-06,
      "loss": 0.09856512397527695,
      "num_input_tokens_seen": 4265356,
      "step": 4134,
      "train_runtime": 10045.8467,
      "train_tokens_per_second": 424.589
    },
    {
      "epoch": 1.76717598033978,
      "grad_norm": 3.3949313163757324,
      "learning_rate": 2.335828877005348e-06,
      "loss": 0.06697557866573334,
      "num_input_tokens_seen": 4266166,
      "step": 4135,
      "train_runtime": 10047.9783,
      "train_tokens_per_second": 424.58
    },
    {
      "epoch": 1.7676033764291057,
      "grad_norm": 1.8965554237365723,
      "learning_rate": 2.331550802139038e-06,
      "loss": 0.05955762788653374,
      "num_input_tokens_seen": 4267620,
      "step": 4136,
      "train_runtime": 10050.1818,
      "train_tokens_per_second": 424.631
    },
    {
      "epoch": 1.7680307725184314,
      "grad_norm": 2.752488613128662,
      "learning_rate": 2.3272727272727277e-06,
      "loss": 0.11746864020824432,
      "num_input_tokens_seen": 4268660,
      "step": 4137,
      "train_runtime": 10052.2944,
      "train_tokens_per_second": 424.645
    },
    {
      "epoch": 1.7684581686077574,
      "grad_norm": 2.6332907676696777,
      "learning_rate": 2.322994652406417e-06,
      "loss": 0.12384096533060074,
      "num_input_tokens_seen": 4269578,
      "step": 4138,
      "train_runtime": 10054.3228,
      "train_tokens_per_second": 424.651
    },
    {
      "epoch": 1.768885564697083,
      "grad_norm": 3.09279203414917,
      "learning_rate": 2.318716577540107e-06,
      "loss": 0.13293305039405823,
      "num_input_tokens_seen": 4270384,
      "step": 4139,
      "train_runtime": 10056.4416,
      "train_tokens_per_second": 424.642
    },
    {
      "epoch": 1.7693129607864089,
      "grad_norm": 2.9126484394073486,
      "learning_rate": 2.314438502673797e-06,
      "loss": 0.08366966992616653,
      "num_input_tokens_seen": 4271052,
      "step": 4140,
      "train_runtime": 10058.496,
      "train_tokens_per_second": 424.621
    },
    {
      "epoch": 1.7697403568757346,
      "grad_norm": 2.05039119720459,
      "learning_rate": 2.3101604278074867e-06,
      "loss": 0.06966466456651688,
      "num_input_tokens_seen": 4272154,
      "step": 4141,
      "train_runtime": 10067.2058,
      "train_tokens_per_second": 424.363
    },
    {
      "epoch": 1.7701677529650603,
      "grad_norm": 6.64882230758667,
      "learning_rate": 2.3058823529411766e-06,
      "loss": 0.1252056360244751,
      "num_input_tokens_seen": 4273542,
      "step": 4142,
      "train_runtime": 10069.4974,
      "train_tokens_per_second": 424.405
    },
    {
      "epoch": 1.7705951490543863,
      "grad_norm": 2.516887903213501,
      "learning_rate": 2.3016042780748665e-06,
      "loss": 0.08751989901065826,
      "num_input_tokens_seen": 4274804,
      "step": 4143,
      "train_runtime": 10071.6472,
      "train_tokens_per_second": 424.439
    },
    {
      "epoch": 1.7710225451437118,
      "grad_norm": 2.1285414695739746,
      "learning_rate": 2.2973262032085563e-06,
      "loss": 0.053270213305950165,
      "num_input_tokens_seen": 4275670,
      "step": 4144,
      "train_runtime": 10073.7397,
      "train_tokens_per_second": 424.437
    },
    {
      "epoch": 1.7714499412330378,
      "grad_norm": 2.652456760406494,
      "learning_rate": 2.293048128342246e-06,
      "loss": 0.1015603318810463,
      "num_input_tokens_seen": 4276688,
      "step": 4145,
      "train_runtime": 10075.9339,
      "train_tokens_per_second": 424.446
    },
    {
      "epoch": 1.7718773373223635,
      "grad_norm": 3.2553443908691406,
      "learning_rate": 2.288770053475936e-06,
      "loss": 0.045736778527498245,
      "num_input_tokens_seen": 4277754,
      "step": 4146,
      "train_runtime": 10078.0324,
      "train_tokens_per_second": 424.463
    },
    {
      "epoch": 1.7723047334116893,
      "grad_norm": 1.8532472848892212,
      "learning_rate": 2.284491978609626e-06,
      "loss": 0.05074065551161766,
      "num_input_tokens_seen": 4279168,
      "step": 4147,
      "train_runtime": 10080.2279,
      "train_tokens_per_second": 424.511
    },
    {
      "epoch": 1.7727321295010152,
      "grad_norm": 4.018914222717285,
      "learning_rate": 2.280213903743316e-06,
      "loss": 0.09143203496932983,
      "num_input_tokens_seen": 4280070,
      "step": 4148,
      "train_runtime": 10082.3833,
      "train_tokens_per_second": 424.51
    },
    {
      "epoch": 1.7731595255903407,
      "grad_norm": 2.661940336227417,
      "learning_rate": 2.2759358288770057e-06,
      "loss": 0.06518501043319702,
      "num_input_tokens_seen": 4280832,
      "step": 4149,
      "train_runtime": 10084.4795,
      "train_tokens_per_second": 424.497
    },
    {
      "epoch": 1.7735869216796667,
      "grad_norm": 2.9754836559295654,
      "learning_rate": 2.2716577540106955e-06,
      "loss": 0.10202857851982117,
      "num_input_tokens_seen": 4281788,
      "step": 4150,
      "train_runtime": 10086.6366,
      "train_tokens_per_second": 424.501
    },
    {
      "epoch": 1.7740143177689924,
      "grad_norm": 2.1830055713653564,
      "learning_rate": 2.2673796791443854e-06,
      "loss": 0.09176403284072876,
      "num_input_tokens_seen": 4282852,
      "step": 4151,
      "train_runtime": 10088.7505,
      "train_tokens_per_second": 424.518
    },
    {
      "epoch": 1.7744417138583182,
      "grad_norm": 2.747210741043091,
      "learning_rate": 2.2631016042780753e-06,
      "loss": 0.0665617287158966,
      "num_input_tokens_seen": 4283954,
      "step": 4152,
      "train_runtime": 10090.8863,
      "train_tokens_per_second": 424.537
    },
    {
      "epoch": 1.7748691099476441,
      "grad_norm": 2.524003267288208,
      "learning_rate": 2.258823529411765e-06,
      "loss": 0.06814989447593689,
      "num_input_tokens_seen": 4284948,
      "step": 4153,
      "train_runtime": 10092.9774,
      "train_tokens_per_second": 424.547
    },
    {
      "epoch": 1.7752965060369696,
      "grad_norm": 2.7492380142211914,
      "learning_rate": 2.254545454545455e-06,
      "loss": 0.05960290879011154,
      "num_input_tokens_seen": 4285532,
      "step": 4154,
      "train_runtime": 10094.9334,
      "train_tokens_per_second": 424.523
    },
    {
      "epoch": 1.7757239021262956,
      "grad_norm": 3.7199904918670654,
      "learning_rate": 2.2502673796791445e-06,
      "loss": 0.07417763769626617,
      "num_input_tokens_seen": 4286168,
      "step": 4155,
      "train_runtime": 10096.9256,
      "train_tokens_per_second": 424.502
    },
    {
      "epoch": 1.7761512982156213,
      "grad_norm": 3.869108200073242,
      "learning_rate": 2.2459893048128343e-06,
      "loss": 0.11099082976579666,
      "num_input_tokens_seen": 4287070,
      "step": 4156,
      "train_runtime": 10099.0213,
      "train_tokens_per_second": 424.504
    },
    {
      "epoch": 1.776578694304947,
      "grad_norm": 1.6349960565567017,
      "learning_rate": 2.241711229946524e-06,
      "loss": 0.06269954890012741,
      "num_input_tokens_seen": 4288172,
      "step": 4157,
      "train_runtime": 10101.1429,
      "train_tokens_per_second": 424.523
    },
    {
      "epoch": 1.777006090394273,
      "grad_norm": 3.589787244796753,
      "learning_rate": 2.237433155080214e-06,
      "loss": 0.11880078166723251,
      "num_input_tokens_seen": 4289372,
      "step": 4158,
      "train_runtime": 10103.2914,
      "train_tokens_per_second": 424.552
    },
    {
      "epoch": 1.7774334864835986,
      "grad_norm": 2.6562845706939697,
      "learning_rate": 2.233155080213904e-06,
      "loss": 0.08049595355987549,
      "num_input_tokens_seen": 4290176,
      "step": 4159,
      "train_runtime": 10105.3247,
      "train_tokens_per_second": 424.546
    },
    {
      "epoch": 1.7778608825729245,
      "grad_norm": 2.513578176498413,
      "learning_rate": 2.228877005347594e-06,
      "loss": 0.07048610597848892,
      "num_input_tokens_seen": 4291510,
      "step": 4160,
      "train_runtime": 10107.5173,
      "train_tokens_per_second": 424.586
    },
    {
      "epoch": 1.7782882786622503,
      "grad_norm": 2.304647922515869,
      "learning_rate": 2.2245989304812837e-06,
      "loss": 0.08331170678138733,
      "num_input_tokens_seen": 4293014,
      "step": 4161,
      "train_runtime": 10109.7267,
      "train_tokens_per_second": 424.642
    },
    {
      "epoch": 1.778715674751576,
      "grad_norm": 2.37117075920105,
      "learning_rate": 2.2203208556149736e-06,
      "loss": 0.07144568860530853,
      "num_input_tokens_seen": 4294276,
      "step": 4162,
      "train_runtime": 10111.8785,
      "train_tokens_per_second": 424.676
    },
    {
      "epoch": 1.779143070840902,
      "grad_norm": 1.9806787967681885,
      "learning_rate": 2.2160427807486634e-06,
      "loss": 0.09877417981624603,
      "num_input_tokens_seen": 4296278,
      "step": 4163,
      "train_runtime": 10114.2189,
      "train_tokens_per_second": 424.776
    },
    {
      "epoch": 1.7795704669302275,
      "grad_norm": 3.70646071434021,
      "learning_rate": 2.2117647058823533e-06,
      "loss": 0.12689195573329926,
      "num_input_tokens_seen": 4297462,
      "step": 4164,
      "train_runtime": 10116.3629,
      "train_tokens_per_second": 424.803
    },
    {
      "epoch": 1.7799978630195534,
      "grad_norm": 3.1241819858551025,
      "learning_rate": 2.2074866310160427e-06,
      "loss": 0.13027304410934448,
      "num_input_tokens_seen": 4298580,
      "step": 4165,
      "train_runtime": 10118.5039,
      "train_tokens_per_second": 424.824
    },
    {
      "epoch": 1.7804252591088792,
      "grad_norm": 3.3830180168151855,
      "learning_rate": 2.2032085561497326e-06,
      "loss": 0.05383117496967316,
      "num_input_tokens_seen": 4299384,
      "step": 4166,
      "train_runtime": 10120.5879,
      "train_tokens_per_second": 424.816
    },
    {
      "epoch": 1.780852655198205,
      "grad_norm": 2.9454708099365234,
      "learning_rate": 2.1989304812834225e-06,
      "loss": 0.10949993133544922,
      "num_input_tokens_seen": 4300460,
      "step": 4167,
      "train_runtime": 10122.6941,
      "train_tokens_per_second": 424.834
    },
    {
      "epoch": 1.7812800512875309,
      "grad_norm": 2.278724193572998,
      "learning_rate": 2.1946524064171123e-06,
      "loss": 0.05804029852151871,
      "num_input_tokens_seen": 4301430,
      "step": 4168,
      "train_runtime": 10124.7782,
      "train_tokens_per_second": 424.842
    },
    {
      "epoch": 1.7817074473768564,
      "grad_norm": 3.620959520339966,
      "learning_rate": 2.1903743315508022e-06,
      "loss": 0.09964598715305328,
      "num_input_tokens_seen": 4302226,
      "step": 4169,
      "train_runtime": 10126.8278,
      "train_tokens_per_second": 424.835
    },
    {
      "epoch": 1.7821348434661823,
      "grad_norm": 2.205592155456543,
      "learning_rate": 2.186096256684492e-06,
      "loss": 0.06786753237247467,
      "num_input_tokens_seen": 4303312,
      "step": 4170,
      "train_runtime": 10128.9207,
      "train_tokens_per_second": 424.854
    },
    {
      "epoch": 1.782562239555508,
      "grad_norm": 2.3895227909088135,
      "learning_rate": 2.181818181818182e-06,
      "loss": 0.0730641633272171,
      "num_input_tokens_seen": 4304234,
      "step": 4171,
      "train_runtime": 10137.7764,
      "train_tokens_per_second": 424.574
    },
    {
      "epoch": 1.7829896356448338,
      "grad_norm": 2.5390784740448,
      "learning_rate": 2.177540106951872e-06,
      "loss": 0.09327578544616699,
      "num_input_tokens_seen": 4305306,
      "step": 4172,
      "train_runtime": 10139.8546,
      "train_tokens_per_second": 424.592
    },
    {
      "epoch": 1.7834170317341598,
      "grad_norm": 2.7873785495758057,
      "learning_rate": 2.1732620320855617e-06,
      "loss": 0.09051096439361572,
      "num_input_tokens_seen": 4306410,
      "step": 4173,
      "train_runtime": 10142.0027,
      "train_tokens_per_second": 424.611
    },
    {
      "epoch": 1.7838444278234853,
      "grad_norm": 3.297451972961426,
      "learning_rate": 2.1689839572192516e-06,
      "loss": 0.09052055329084396,
      "num_input_tokens_seen": 4307580,
      "step": 4174,
      "train_runtime": 10144.1521,
      "train_tokens_per_second": 424.637
    },
    {
      "epoch": 1.7842718239128113,
      "grad_norm": 2.5635032653808594,
      "learning_rate": 2.1647058823529414e-06,
      "loss": 0.09134936332702637,
      "num_input_tokens_seen": 4308534,
      "step": 4175,
      "train_runtime": 10146.2869,
      "train_tokens_per_second": 424.641
    },
    {
      "epoch": 1.784699220002137,
      "grad_norm": 2.7687714099884033,
      "learning_rate": 2.1604278074866313e-06,
      "loss": 0.06639499962329865,
      "num_input_tokens_seen": 4309642,
      "step": 4176,
      "train_runtime": 10148.471,
      "train_tokens_per_second": 424.659
    },
    {
      "epoch": 1.7851266160914627,
      "grad_norm": 1.0155085325241089,
      "learning_rate": 2.156149732620321e-06,
      "loss": 0.02201678417623043,
      "num_input_tokens_seen": 4311094,
      "step": 4177,
      "train_runtime": 10150.726,
      "train_tokens_per_second": 424.708
    },
    {
      "epoch": 1.7855540121807887,
      "grad_norm": 3.3319475650787354,
      "learning_rate": 2.1518716577540106e-06,
      "loss": 0.08600349724292755,
      "num_input_tokens_seen": 4312188,
      "step": 4178,
      "train_runtime": 10152.8724,
      "train_tokens_per_second": 424.726
    },
    {
      "epoch": 1.7859814082701142,
      "grad_norm": 2.815528154373169,
      "learning_rate": 2.1475935828877005e-06,
      "loss": 0.10754726827144623,
      "num_input_tokens_seen": 4313186,
      "step": 4179,
      "train_runtime": 10155.0108,
      "train_tokens_per_second": 424.735
    },
    {
      "epoch": 1.7864088043594402,
      "grad_norm": 2.539433479309082,
      "learning_rate": 2.1433155080213903e-06,
      "loss": 0.05385895073413849,
      "num_input_tokens_seen": 4314234,
      "step": 4180,
      "train_runtime": 10157.1143,
      "train_tokens_per_second": 424.75
    },
    {
      "epoch": 1.786836200448766,
      "grad_norm": 2.4351701736450195,
      "learning_rate": 2.1390374331550802e-06,
      "loss": 0.06828010082244873,
      "num_input_tokens_seen": 4315652,
      "step": 4181,
      "train_runtime": 10159.2915,
      "train_tokens_per_second": 424.799
    },
    {
      "epoch": 1.7872635965380916,
      "grad_norm": 2.82084584236145,
      "learning_rate": 2.13475935828877e-06,
      "loss": 0.1077570989727974,
      "num_input_tokens_seen": 4316662,
      "step": 4182,
      "train_runtime": 10161.3753,
      "train_tokens_per_second": 424.811
    },
    {
      "epoch": 1.7876909926274176,
      "grad_norm": 2.325443983078003,
      "learning_rate": 2.13048128342246e-06,
      "loss": 0.060035690665245056,
      "num_input_tokens_seen": 4317734,
      "step": 4183,
      "train_runtime": 10163.751,
      "train_tokens_per_second": 424.817
    },
    {
      "epoch": 1.7881183887167431,
      "grad_norm": 2.2599000930786133,
      "learning_rate": 2.12620320855615e-06,
      "loss": 0.07673176378011703,
      "num_input_tokens_seen": 4318806,
      "step": 4184,
      "train_runtime": 10165.9375,
      "train_tokens_per_second": 424.831
    },
    {
      "epoch": 1.788545784806069,
      "grad_norm": 3.37270450592041,
      "learning_rate": 2.1219251336898397e-06,
      "loss": 0.0735541582107544,
      "num_input_tokens_seen": 4319504,
      "step": 4185,
      "train_runtime": 10167.9206,
      "train_tokens_per_second": 424.817
    },
    {
      "epoch": 1.7889731808953948,
      "grad_norm": 2.5903313159942627,
      "learning_rate": 2.1176470588235296e-06,
      "loss": 0.06701052188873291,
      "num_input_tokens_seen": 4320308,
      "step": 4186,
      "train_runtime": 10169.9027,
      "train_tokens_per_second": 424.813
    },
    {
      "epoch": 1.7894005769847205,
      "grad_norm": 2.4194836616516113,
      "learning_rate": 2.1133689839572194e-06,
      "loss": 0.06429250538349152,
      "num_input_tokens_seen": 4321020,
      "step": 4187,
      "train_runtime": 10171.926,
      "train_tokens_per_second": 424.799
    },
    {
      "epoch": 1.7898279730740465,
      "grad_norm": 3.2498037815093994,
      "learning_rate": 2.1090909090909093e-06,
      "loss": 0.0883617103099823,
      "num_input_tokens_seen": 4321884,
      "step": 4188,
      "train_runtime": 10173.9571,
      "train_tokens_per_second": 424.799
    },
    {
      "epoch": 1.790255369163372,
      "grad_norm": 2.4744582176208496,
      "learning_rate": 2.104812834224599e-06,
      "loss": 0.08810482174158096,
      "num_input_tokens_seen": 4323212,
      "step": 4189,
      "train_runtime": 10176.1393,
      "train_tokens_per_second": 424.838
    },
    {
      "epoch": 1.790682765252698,
      "grad_norm": 2.7591638565063477,
      "learning_rate": 2.100534759358289e-06,
      "loss": 0.10325776785612106,
      "num_input_tokens_seen": 4324356,
      "step": 4190,
      "train_runtime": 10178.289,
      "train_tokens_per_second": 424.861
    },
    {
      "epoch": 1.7911101613420237,
      "grad_norm": 2.6523334980010986,
      "learning_rate": 2.096256684491979e-06,
      "loss": 0.06592361629009247,
      "num_input_tokens_seen": 4325268,
      "step": 4191,
      "train_runtime": 10180.3928,
      "train_tokens_per_second": 424.863
    },
    {
      "epoch": 1.7915375574313495,
      "grad_norm": 2.4090828895568848,
      "learning_rate": 2.0919786096256688e-06,
      "loss": 0.056853655725717545,
      "num_input_tokens_seen": 4326220,
      "step": 4192,
      "train_runtime": 10182.5222,
      "train_tokens_per_second": 424.867
    },
    {
      "epoch": 1.7919649535206754,
      "grad_norm": 2.418062448501587,
      "learning_rate": 2.0877005347593586e-06,
      "loss": 0.05139508843421936,
      "num_input_tokens_seen": 4327088,
      "step": 4193,
      "train_runtime": 10184.5981,
      "train_tokens_per_second": 424.866
    },
    {
      "epoch": 1.792392349610001,
      "grad_norm": 2.09745454788208,
      "learning_rate": 2.0834224598930485e-06,
      "loss": 0.05079510807991028,
      "num_input_tokens_seen": 4327884,
      "step": 4194,
      "train_runtime": 10186.6784,
      "train_tokens_per_second": 424.857
    },
    {
      "epoch": 1.792819745699327,
      "grad_norm": 4.581893444061279,
      "learning_rate": 2.079144385026738e-06,
      "loss": 0.08980727195739746,
      "num_input_tokens_seen": 4328774,
      "step": 4195,
      "train_runtime": 10188.7502,
      "train_tokens_per_second": 424.858
    },
    {
      "epoch": 1.7932471417886526,
      "grad_norm": 2.814441680908203,
      "learning_rate": 2.074866310160428e-06,
      "loss": 0.09988869726657867,
      "num_input_tokens_seen": 4329770,
      "step": 4196,
      "train_runtime": 10190.9104,
      "train_tokens_per_second": 424.866
    },
    {
      "epoch": 1.7936745378779784,
      "grad_norm": 4.073714733123779,
      "learning_rate": 2.0705882352941177e-06,
      "loss": 0.08503936231136322,
      "num_input_tokens_seen": 4330684,
      "step": 4197,
      "train_runtime": 10192.9942,
      "train_tokens_per_second": 424.869
    },
    {
      "epoch": 1.7941019339673043,
      "grad_norm": 3.4864861965179443,
      "learning_rate": 2.0663101604278076e-06,
      "loss": 0.10185601562261581,
      "num_input_tokens_seen": 4331352,
      "step": 4198,
      "train_runtime": 10195.0279,
      "train_tokens_per_second": 424.849
    },
    {
      "epoch": 1.7945293300566298,
      "grad_norm": 1.7800720930099487,
      "learning_rate": 2.0620320855614974e-06,
      "loss": 0.06571541726589203,
      "num_input_tokens_seen": 4332626,
      "step": 4199,
      "train_runtime": 10197.1716,
      "train_tokens_per_second": 424.885
    },
    {
      "epoch": 1.7949567261459558,
      "grad_norm": 2.510011672973633,
      "learning_rate": 2.0577540106951873e-06,
      "loss": 0.06940654665231705,
      "num_input_tokens_seen": 4333764,
      "step": 4200,
      "train_runtime": 10199.2971,
      "train_tokens_per_second": 424.908
    },
    {
      "epoch": 1.7953841222352815,
      "grad_norm": 3.0646040439605713,
      "learning_rate": 2.053475935828877e-06,
      "loss": 0.10401661694049835,
      "num_input_tokens_seen": 4334726,
      "step": 4201,
      "train_runtime": 10207.5723,
      "train_tokens_per_second": 424.658
    },
    {
      "epoch": 1.7958115183246073,
      "grad_norm": 1.9562426805496216,
      "learning_rate": 2.049197860962567e-06,
      "loss": 0.0695512443780899,
      "num_input_tokens_seen": 4335492,
      "step": 4202,
      "train_runtime": 10209.6111,
      "train_tokens_per_second": 424.648
    },
    {
      "epoch": 1.7962389144139332,
      "grad_norm": 2.9338138103485107,
      "learning_rate": 2.044919786096257e-06,
      "loss": 0.10441028326749802,
      "num_input_tokens_seen": 4336322,
      "step": 4203,
      "train_runtime": 10211.7639,
      "train_tokens_per_second": 424.64
    },
    {
      "epoch": 1.7966663105032588,
      "grad_norm": 4.342402458190918,
      "learning_rate": 2.0406417112299468e-06,
      "loss": 0.13860847055912018,
      "num_input_tokens_seen": 4337018,
      "step": 4204,
      "train_runtime": 10213.7773,
      "train_tokens_per_second": 424.624
    },
    {
      "epoch": 1.7970937065925847,
      "grad_norm": 2.9212000370025635,
      "learning_rate": 2.0363636363636367e-06,
      "loss": 0.125177264213562,
      "num_input_tokens_seen": 4338166,
      "step": 4205,
      "train_runtime": 10215.8773,
      "train_tokens_per_second": 424.649
    },
    {
      "epoch": 1.7975211026819105,
      "grad_norm": 2.952876329421997,
      "learning_rate": 2.0320855614973265e-06,
      "loss": 0.06996580213308334,
      "num_input_tokens_seen": 4338910,
      "step": 4206,
      "train_runtime": 10217.8616,
      "train_tokens_per_second": 424.64
    },
    {
      "epoch": 1.7979484987712362,
      "grad_norm": 2.904758930206299,
      "learning_rate": 2.0278074866310164e-06,
      "loss": 0.08176860958337784,
      "num_input_tokens_seen": 4339978,
      "step": 4207,
      "train_runtime": 10219.9939,
      "train_tokens_per_second": 424.656
    },
    {
      "epoch": 1.7983758948605622,
      "grad_norm": 2.9267780780792236,
      "learning_rate": 2.0235294117647063e-06,
      "loss": 0.09859760850667953,
      "num_input_tokens_seen": 4340940,
      "step": 4208,
      "train_runtime": 10222.0483,
      "train_tokens_per_second": 424.664
    },
    {
      "epoch": 1.7988032909498877,
      "grad_norm": 1.7343788146972656,
      "learning_rate": 2.019251336898396e-06,
      "loss": 0.07872038334608078,
      "num_input_tokens_seen": 4342710,
      "step": 4209,
      "train_runtime": 10224.2927,
      "train_tokens_per_second": 424.744
    },
    {
      "epoch": 1.7992306870392136,
      "grad_norm": 2.0737686157226562,
      "learning_rate": 2.014973262032086e-06,
      "loss": 0.0616140216588974,
      "num_input_tokens_seen": 4343668,
      "step": 4210,
      "train_runtime": 10226.4916,
      "train_tokens_per_second": 424.747
    },
    {
      "epoch": 1.7996580831285394,
      "grad_norm": 2.372847557067871,
      "learning_rate": 2.010695187165776e-06,
      "loss": 0.06419747322797775,
      "num_input_tokens_seen": 4344736,
      "step": 4211,
      "train_runtime": 10228.587,
      "train_tokens_per_second": 424.764
    },
    {
      "epoch": 1.800085479217865,
      "grad_norm": 3.0257482528686523,
      "learning_rate": 2.0064171122994653e-06,
      "loss": 0.0916258916258812,
      "num_input_tokens_seen": 4345612,
      "step": 4212,
      "train_runtime": 10230.7065,
      "train_tokens_per_second": 424.762
    },
    {
      "epoch": 1.800085479217865,
      "eval_loss": 0.516430139541626,
      "eval_runtime": 58.8158,
      "eval_samples_per_second": 16.968,
      "eval_steps_per_second": 8.484,
      "num_input_tokens_seen": 4345612,
      "step": 4212
    },
    {
      "epoch": 1.800512875307191,
      "grad_norm": 2.602503538131714,
      "learning_rate": 2.002139037433155e-06,
      "loss": 0.07492129504680634,
      "num_input_tokens_seen": 4346522,
      "step": 4213,
      "train_runtime": 10291.6688,
      "train_tokens_per_second": 422.334
    },
    {
      "epoch": 1.8009402713965166,
      "grad_norm": 2.025399923324585,
      "learning_rate": 1.997860962566845e-06,
      "loss": 0.0645725354552269,
      "num_input_tokens_seen": 4347472,
      "step": 4214,
      "train_runtime": 10293.7741,
      "train_tokens_per_second": 422.34
    },
    {
      "epoch": 1.8013676674858425,
      "grad_norm": 2.352571487426758,
      "learning_rate": 1.993582887700535e-06,
      "loss": 0.06522589921951294,
      "num_input_tokens_seen": 4348344,
      "step": 4215,
      "train_runtime": 10295.8063,
      "train_tokens_per_second": 422.341
    },
    {
      "epoch": 1.8017950635751683,
      "grad_norm": 2.581906318664551,
      "learning_rate": 1.989304812834225e-06,
      "loss": 0.08330638706684113,
      "num_input_tokens_seen": 4349248,
      "step": 4216,
      "train_runtime": 10297.8939,
      "train_tokens_per_second": 422.343
    },
    {
      "epoch": 1.802222459664494,
      "grad_norm": 2.9776909351348877,
      "learning_rate": 1.9850267379679147e-06,
      "loss": 0.08715388178825378,
      "num_input_tokens_seen": 4350126,
      "step": 4217,
      "train_runtime": 10300.0,
      "train_tokens_per_second": 422.342
    },
    {
      "epoch": 1.80264985575382,
      "grad_norm": 3.07016921043396,
      "learning_rate": 1.9807486631016045e-06,
      "loss": 0.08147434890270233,
      "num_input_tokens_seen": 4351160,
      "step": 4218,
      "train_runtime": 10302.1379,
      "train_tokens_per_second": 422.355
    },
    {
      "epoch": 1.8030772518431455,
      "grad_norm": 3.6859567165374756,
      "learning_rate": 1.976470588235294e-06,
      "loss": 0.10586757957935333,
      "num_input_tokens_seen": 4352062,
      "step": 4219,
      "train_runtime": 10304.2467,
      "train_tokens_per_second": 422.356
    },
    {
      "epoch": 1.8035046479324714,
      "grad_norm": 3.480611562728882,
      "learning_rate": 1.972192513368984e-06,
      "loss": 0.10214269161224365,
      "num_input_tokens_seen": 4352884,
      "step": 4220,
      "train_runtime": 10306.3509,
      "train_tokens_per_second": 422.35
    },
    {
      "epoch": 1.8039320440217972,
      "grad_norm": 1.4777809381484985,
      "learning_rate": 1.9679144385026737e-06,
      "loss": 0.03467395901679993,
      "num_input_tokens_seen": 4353774,
      "step": 4221,
      "train_runtime": 10308.5142,
      "train_tokens_per_second": 422.347
    },
    {
      "epoch": 1.804359440111123,
      "grad_norm": 3.580695867538452,
      "learning_rate": 1.9636363636363636e-06,
      "loss": 0.15541917085647583,
      "num_input_tokens_seen": 4354676,
      "step": 4222,
      "train_runtime": 10310.6271,
      "train_tokens_per_second": 422.348
    },
    {
      "epoch": 1.8047868362004489,
      "grad_norm": 3.125870704650879,
      "learning_rate": 1.9593582887700535e-06,
      "loss": 0.12012460827827454,
      "num_input_tokens_seen": 4355650,
      "step": 4223,
      "train_runtime": 10312.6946,
      "train_tokens_per_second": 422.358
    },
    {
      "epoch": 1.8052142322897744,
      "grad_norm": 2.353468418121338,
      "learning_rate": 1.9550802139037433e-06,
      "loss": 0.06727476418018341,
      "num_input_tokens_seen": 4356400,
      "step": 4224,
      "train_runtime": 10314.7224,
      "train_tokens_per_second": 422.348
    },
    {
      "epoch": 1.8056416283791004,
      "grad_norm": 2.358220338821411,
      "learning_rate": 1.950802139037433e-06,
      "loss": 0.058052562177181244,
      "num_input_tokens_seen": 4357492,
      "step": 4225,
      "train_runtime": 10316.874,
      "train_tokens_per_second": 422.366
    },
    {
      "epoch": 1.806069024468426,
      "grad_norm": 2.201442003250122,
      "learning_rate": 1.946524064171123e-06,
      "loss": 0.06703916192054749,
      "num_input_tokens_seen": 4358234,
      "step": 4226,
      "train_runtime": 10318.9588,
      "train_tokens_per_second": 422.352
    },
    {
      "epoch": 1.8064964205577518,
      "grad_norm": 3.906421661376953,
      "learning_rate": 1.942245989304813e-06,
      "loss": 0.10453782230615616,
      "num_input_tokens_seen": 4358950,
      "step": 4227,
      "train_runtime": 10320.9959,
      "train_tokens_per_second": 422.338
    },
    {
      "epoch": 1.8069238166470778,
      "grad_norm": 2.7399024963378906,
      "learning_rate": 1.937967914438503e-06,
      "loss": 0.08963947743177414,
      "num_input_tokens_seen": 4360108,
      "step": 4228,
      "train_runtime": 10323.1325,
      "train_tokens_per_second": 422.363
    },
    {
      "epoch": 1.8073512127364033,
      "grad_norm": 2.3357512950897217,
      "learning_rate": 1.9336898395721927e-06,
      "loss": 0.05593331530690193,
      "num_input_tokens_seen": 4361226,
      "step": 4229,
      "train_runtime": 10325.2355,
      "train_tokens_per_second": 422.385
    },
    {
      "epoch": 1.8077786088257293,
      "grad_norm": 2.3317689895629883,
      "learning_rate": 1.9294117647058825e-06,
      "loss": 0.05253561586141586,
      "num_input_tokens_seen": 4361912,
      "step": 4230,
      "train_runtime": 10327.3185,
      "train_tokens_per_second": 422.366
    },
    {
      "epoch": 1.808206004915055,
      "grad_norm": 2.3263156414031982,
      "learning_rate": 1.9251336898395724e-06,
      "loss": 0.07395733892917633,
      "num_input_tokens_seen": 4362986,
      "step": 4231,
      "train_runtime": 10336.0993,
      "train_tokens_per_second": 422.111
    },
    {
      "epoch": 1.8086334010043807,
      "grad_norm": 2.877946376800537,
      "learning_rate": 1.9208556149732623e-06,
      "loss": 0.0581427700817585,
      "num_input_tokens_seen": 4363844,
      "step": 4232,
      "train_runtime": 10338.265,
      "train_tokens_per_second": 422.106
    },
    {
      "epoch": 1.8090607970937067,
      "grad_norm": 2.5502960681915283,
      "learning_rate": 1.916577540106952e-06,
      "loss": 0.06419610232114792,
      "num_input_tokens_seen": 4364784,
      "step": 4233,
      "train_runtime": 10340.4309,
      "train_tokens_per_second": 422.109
    },
    {
      "epoch": 1.8094881931830322,
      "grad_norm": 1.6813851594924927,
      "learning_rate": 1.912299465240642e-06,
      "loss": 0.04561107978224754,
      "num_input_tokens_seen": 4366032,
      "step": 4234,
      "train_runtime": 10342.6218,
      "train_tokens_per_second": 422.14
    },
    {
      "epoch": 1.8099155892723582,
      "grad_norm": 2.87347412109375,
      "learning_rate": 1.9080213903743315e-06,
      "loss": 0.09780064225196838,
      "num_input_tokens_seen": 4366842,
      "step": 4235,
      "train_runtime": 10344.6807,
      "train_tokens_per_second": 422.134
    },
    {
      "epoch": 1.810342985361684,
      "grad_norm": 1.9329286813735962,
      "learning_rate": 1.9037433155080215e-06,
      "loss": 0.09178803861141205,
      "num_input_tokens_seen": 4368430,
      "step": 4236,
      "train_runtime": 10346.8945,
      "train_tokens_per_second": 422.197
    },
    {
      "epoch": 1.8107703814510097,
      "grad_norm": 3.1183297634124756,
      "learning_rate": 1.8994652406417114e-06,
      "loss": 0.11012288928031921,
      "num_input_tokens_seen": 4369198,
      "step": 4237,
      "train_runtime": 10348.9344,
      "train_tokens_per_second": 422.188
    },
    {
      "epoch": 1.8111977775403356,
      "grad_norm": 2.854477882385254,
      "learning_rate": 1.8951871657754013e-06,
      "loss": 0.1509762853384018,
      "num_input_tokens_seen": 4370292,
      "step": 4238,
      "train_runtime": 10351.1221,
      "train_tokens_per_second": 422.205
    },
    {
      "epoch": 1.8116251736296611,
      "grad_norm": 4.804562091827393,
      "learning_rate": 1.890909090909091e-06,
      "loss": 0.1389392614364624,
      "num_input_tokens_seen": 4371172,
      "step": 4239,
      "train_runtime": 10353.2507,
      "train_tokens_per_second": 422.203
    },
    {
      "epoch": 1.812052569718987,
      "grad_norm": 3.8928635120391846,
      "learning_rate": 1.8866310160427808e-06,
      "loss": 0.16309262812137604,
      "num_input_tokens_seen": 4372046,
      "step": 4240,
      "train_runtime": 10355.3311,
      "train_tokens_per_second": 422.202
    },
    {
      "epoch": 1.8124799658083128,
      "grad_norm": 2.530460834503174,
      "learning_rate": 1.8823529411764707e-06,
      "loss": 0.1013408824801445,
      "num_input_tokens_seen": 4373522,
      "step": 4241,
      "train_runtime": 10357.6184,
      "train_tokens_per_second": 422.252
    },
    {
      "epoch": 1.8129073618976386,
      "grad_norm": 4.975642681121826,
      "learning_rate": 1.8780748663101605e-06,
      "loss": 0.09836843609809875,
      "num_input_tokens_seen": 4374056,
      "step": 4242,
      "train_runtime": 10359.6117,
      "train_tokens_per_second": 422.222
    },
    {
      "epoch": 1.8133347579869645,
      "grad_norm": 2.4946558475494385,
      "learning_rate": 1.8737967914438504e-06,
      "loss": 0.11575710773468018,
      "num_input_tokens_seen": 4375190,
      "step": 4243,
      "train_runtime": 10361.7861,
      "train_tokens_per_second": 422.243
    },
    {
      "epoch": 1.81376215407629,
      "grad_norm": 2.6444196701049805,
      "learning_rate": 1.8695187165775403e-06,
      "loss": 0.08153674751520157,
      "num_input_tokens_seen": 4376356,
      "step": 4244,
      "train_runtime": 10363.941,
      "train_tokens_per_second": 422.268
    },
    {
      "epoch": 1.814189550165616,
      "grad_norm": 2.999821662902832,
      "learning_rate": 1.8652406417112302e-06,
      "loss": 0.10564687848091125,
      "num_input_tokens_seen": 4377348,
      "step": 4245,
      "train_runtime": 10366.0595,
      "train_tokens_per_second": 422.277
    },
    {
      "epoch": 1.8146169462549417,
      "grad_norm": 3.17754864692688,
      "learning_rate": 1.86096256684492e-06,
      "loss": 0.061029963195323944,
      "num_input_tokens_seen": 4378734,
      "step": 4246,
      "train_runtime": 10368.2285,
      "train_tokens_per_second": 422.322
    },
    {
      "epoch": 1.8150443423442675,
      "grad_norm": 3.561413526535034,
      "learning_rate": 1.8566844919786097e-06,
      "loss": 0.1227944865822792,
      "num_input_tokens_seen": 4379538,
      "step": 4247,
      "train_runtime": 10370.2522,
      "train_tokens_per_second": 422.317
    },
    {
      "epoch": 1.8154717384335934,
      "grad_norm": 2.7332143783569336,
      "learning_rate": 1.8524064171122996e-06,
      "loss": 0.0694960504770279,
      "num_input_tokens_seen": 4380430,
      "step": 4248,
      "train_runtime": 10372.3526,
      "train_tokens_per_second": 422.318
    },
    {
      "epoch": 1.815899134522919,
      "grad_norm": 2.859757423400879,
      "learning_rate": 1.8481283422459894e-06,
      "loss": 0.10604922473430634,
      "num_input_tokens_seen": 4381210,
      "step": 4249,
      "train_runtime": 10374.3817,
      "train_tokens_per_second": 422.31
    },
    {
      "epoch": 1.816326530612245,
      "grad_norm": 2.461272954940796,
      "learning_rate": 1.8438502673796793e-06,
      "loss": 0.07684335857629776,
      "num_input_tokens_seen": 4382310,
      "step": 4250,
      "train_runtime": 10376.5265,
      "train_tokens_per_second": 422.329
    },
    {
      "epoch": 1.8167539267015707,
      "grad_norm": 1.9117848873138428,
      "learning_rate": 1.8395721925133692e-06,
      "loss": 0.0697866678237915,
      "num_input_tokens_seen": 4383716,
      "step": 4251,
      "train_runtime": 10378.8133,
      "train_tokens_per_second": 422.372
    },
    {
      "epoch": 1.8171813227908964,
      "grad_norm": 3.1455063819885254,
      "learning_rate": 1.835294117647059e-06,
      "loss": 0.10792747884988785,
      "num_input_tokens_seen": 4384510,
      "step": 4252,
      "train_runtime": 10380.8954,
      "train_tokens_per_second": 422.363
    },
    {
      "epoch": 1.8176087188802224,
      "grad_norm": 2.955990791320801,
      "learning_rate": 1.831016042780749e-06,
      "loss": 0.08685193210840225,
      "num_input_tokens_seen": 4385582,
      "step": 4253,
      "train_runtime": 10382.9834,
      "train_tokens_per_second": 422.382
    },
    {
      "epoch": 1.8180361149695479,
      "grad_norm": 1.9731003046035767,
      "learning_rate": 1.8267379679144388e-06,
      "loss": 0.03776971623301506,
      "num_input_tokens_seen": 4386470,
      "step": 4254,
      "train_runtime": 10385.0175,
      "train_tokens_per_second": 422.384
    },
    {
      "epoch": 1.8184635110588738,
      "grad_norm": 3.050600290298462,
      "learning_rate": 1.8224598930481286e-06,
      "loss": 0.07909286022186279,
      "num_input_tokens_seen": 4387808,
      "step": 4255,
      "train_runtime": 10387.2764,
      "train_tokens_per_second": 422.421
    },
    {
      "epoch": 1.8188909071481996,
      "grad_norm": 3.54374098777771,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.14706793427467346,
      "num_input_tokens_seen": 4388772,
      "step": 4256,
      "train_runtime": 10389.3306,
      "train_tokens_per_second": 422.431
    },
    {
      "epoch": 1.8193183032375253,
      "grad_norm": 4.551452159881592,
      "learning_rate": 1.8139037433155082e-06,
      "loss": 0.14810840785503387,
      "num_input_tokens_seen": 4389770,
      "step": 4257,
      "train_runtime": 10391.4106,
      "train_tokens_per_second": 422.442
    },
    {
      "epoch": 1.8197456993268513,
      "grad_norm": 1.8535094261169434,
      "learning_rate": 1.809625668449198e-06,
      "loss": 0.07118307054042816,
      "num_input_tokens_seen": 4391710,
      "step": 4258,
      "train_runtime": 10393.7937,
      "train_tokens_per_second": 422.532
    },
    {
      "epoch": 1.8201730954161768,
      "grad_norm": 2.338582992553711,
      "learning_rate": 1.805347593582888e-06,
      "loss": 0.07584503293037415,
      "num_input_tokens_seen": 4392796,
      "step": 4259,
      "train_runtime": 10395.8823,
      "train_tokens_per_second": 422.552
    },
    {
      "epoch": 1.8206004915055027,
      "grad_norm": 2.635974884033203,
      "learning_rate": 1.8010695187165778e-06,
      "loss": 0.10110776126384735,
      "num_input_tokens_seen": 4393772,
      "step": 4260,
      "train_runtime": 10397.965,
      "train_tokens_per_second": 422.561
    },
    {
      "epoch": 1.8210278875948285,
      "grad_norm": 1.8251999616622925,
      "learning_rate": 1.7967914438502676e-06,
      "loss": 0.07249141484498978,
      "num_input_tokens_seen": 4394920,
      "step": 4261,
      "train_runtime": 10406.9865,
      "train_tokens_per_second": 422.305
    },
    {
      "epoch": 1.8214552836841542,
      "grad_norm": 3.029616594314575,
      "learning_rate": 1.7925133689839575e-06,
      "loss": 0.0901215523481369,
      "num_input_tokens_seen": 4395602,
      "step": 4262,
      "train_runtime": 10409.0475,
      "train_tokens_per_second": 422.287
    },
    {
      "epoch": 1.8218826797734802,
      "grad_norm": 2.4523026943206787,
      "learning_rate": 1.7882352941176474e-06,
      "loss": 0.08679521083831787,
      "num_input_tokens_seen": 4396914,
      "step": 4263,
      "train_runtime": 10411.2343,
      "train_tokens_per_second": 422.324
    },
    {
      "epoch": 1.8223100758628057,
      "grad_norm": 2.0859696865081787,
      "learning_rate": 1.783957219251337e-06,
      "loss": 0.06534142047166824,
      "num_input_tokens_seen": 4397776,
      "step": 4264,
      "train_runtime": 10413.3228,
      "train_tokens_per_second": 422.322
    },
    {
      "epoch": 1.8227374719521316,
      "grad_norm": 3.7017807960510254,
      "learning_rate": 1.779679144385027e-06,
      "loss": 0.09009534120559692,
      "num_input_tokens_seen": 4398380,
      "step": 4265,
      "train_runtime": 10415.3332,
      "train_tokens_per_second": 422.299
    },
    {
      "epoch": 1.8231648680414574,
      "grad_norm": 2.9533321857452393,
      "learning_rate": 1.7754010695187168e-06,
      "loss": 0.046583231538534164,
      "num_input_tokens_seen": 4399132,
      "step": 4266,
      "train_runtime": 10417.3494,
      "train_tokens_per_second": 422.289
    },
    {
      "epoch": 1.8235922641307831,
      "grad_norm": 2.7662792205810547,
      "learning_rate": 1.7711229946524066e-06,
      "loss": 0.12788061797618866,
      "num_input_tokens_seen": 4400352,
      "step": 4267,
      "train_runtime": 10419.4929,
      "train_tokens_per_second": 422.319
    },
    {
      "epoch": 1.824019660220109,
      "grad_norm": 7.752589225769043,
      "learning_rate": 1.7668449197860965e-06,
      "loss": 0.0736791118979454,
      "num_input_tokens_seen": 4400790,
      "step": 4268,
      "train_runtime": 10421.4936,
      "train_tokens_per_second": 422.28
    },
    {
      "epoch": 1.8244470563094348,
      "grad_norm": 2.8798141479492188,
      "learning_rate": 1.7625668449197864e-06,
      "loss": 0.09692294895648956,
      "num_input_tokens_seen": 4402984,
      "step": 4269,
      "train_runtime": 10423.8453,
      "train_tokens_per_second": 422.395
    },
    {
      "epoch": 1.8248744523987606,
      "grad_norm": 2.372035026550293,
      "learning_rate": 1.7582887700534762e-06,
      "loss": 0.09091182053089142,
      "num_input_tokens_seen": 4403918,
      "step": 4270,
      "train_runtime": 10425.9855,
      "train_tokens_per_second": 422.398
    },
    {
      "epoch": 1.8253018484880863,
      "grad_norm": 2.7524266242980957,
      "learning_rate": 1.7540106951871661e-06,
      "loss": 0.08351872861385345,
      "num_input_tokens_seen": 4404562,
      "step": 4271,
      "train_runtime": 10427.9969,
      "train_tokens_per_second": 422.379
    },
    {
      "epoch": 1.825729244577412,
      "grad_norm": 3.4992871284484863,
      "learning_rate": 1.749732620320856e-06,
      "loss": 0.08081549406051636,
      "num_input_tokens_seen": 4405300,
      "step": 4272,
      "train_runtime": 10430.0207,
      "train_tokens_per_second": 422.367
    },
    {
      "epoch": 1.826156640666738,
      "grad_norm": 1.985395073890686,
      "learning_rate": 1.7454545454545456e-06,
      "loss": 0.0659557580947876,
      "num_input_tokens_seen": 4406460,
      "step": 4273,
      "train_runtime": 10432.1349,
      "train_tokens_per_second": 422.393
    },
    {
      "epoch": 1.8265840367560637,
      "grad_norm": 3.024962902069092,
      "learning_rate": 1.7411764705882353e-06,
      "loss": 0.09400507062673569,
      "num_input_tokens_seen": 4407304,
      "step": 4274,
      "train_runtime": 10434.1859,
      "train_tokens_per_second": 422.391
    },
    {
      "epoch": 1.8270114328453895,
      "grad_norm": 3.3022570610046387,
      "learning_rate": 1.7368983957219252e-06,
      "loss": 0.06011106073856354,
      "num_input_tokens_seen": 4407968,
      "step": 4275,
      "train_runtime": 10436.1393,
      "train_tokens_per_second": 422.375
    },
    {
      "epoch": 1.8274388289347152,
      "grad_norm": 3.109867572784424,
      "learning_rate": 1.732620320855615e-06,
      "loss": 0.10541640222072601,
      "num_input_tokens_seen": 4408710,
      "step": 4276,
      "train_runtime": 10438.2918,
      "train_tokens_per_second": 422.359
    },
    {
      "epoch": 1.827866225024041,
      "grad_norm": 2.8406429290771484,
      "learning_rate": 1.728342245989305e-06,
      "loss": 0.07821729779243469,
      "num_input_tokens_seen": 4409334,
      "step": 4277,
      "train_runtime": 10440.2443,
      "train_tokens_per_second": 422.34
    },
    {
      "epoch": 1.828293621113367,
      "grad_norm": 2.863986015319824,
      "learning_rate": 1.7240641711229948e-06,
      "loss": 0.08144165575504303,
      "num_input_tokens_seen": 4409972,
      "step": 4278,
      "train_runtime": 10442.2551,
      "train_tokens_per_second": 422.32
    },
    {
      "epoch": 1.8287210172026926,
      "grad_norm": 3.59318470954895,
      "learning_rate": 1.7197860962566844e-06,
      "loss": 0.12389001995325089,
      "num_input_tokens_seen": 4410774,
      "step": 4279,
      "train_runtime": 10444.2749,
      "train_tokens_per_second": 422.315
    },
    {
      "epoch": 1.8291484132920184,
      "grad_norm": 2.7863714694976807,
      "learning_rate": 1.7155080213903743e-06,
      "loss": 0.07351061701774597,
      "num_input_tokens_seen": 4411776,
      "step": 4280,
      "train_runtime": 10446.3828,
      "train_tokens_per_second": 422.326
    },
    {
      "epoch": 1.8295758093813441,
      "grad_norm": 4.017768383026123,
      "learning_rate": 1.7112299465240642e-06,
      "loss": 0.12098449468612671,
      "num_input_tokens_seen": 4412588,
      "step": 4281,
      "train_runtime": 10448.4174,
      "train_tokens_per_second": 422.321
    },
    {
      "epoch": 1.8300032054706699,
      "grad_norm": 2.2940378189086914,
      "learning_rate": 1.706951871657754e-06,
      "loss": 0.11267227679491043,
      "num_input_tokens_seen": 4414188,
      "step": 4282,
      "train_runtime": 10450.6548,
      "train_tokens_per_second": 422.384
    },
    {
      "epoch": 1.8304306015599958,
      "grad_norm": 3.3319027423858643,
      "learning_rate": 1.702673796791444e-06,
      "loss": 0.13140606880187988,
      "num_input_tokens_seen": 4415524,
      "step": 4283,
      "train_runtime": 10452.9381,
      "train_tokens_per_second": 422.419
    },
    {
      "epoch": 1.8308579976493216,
      "grad_norm": 2.408099889755249,
      "learning_rate": 1.6983957219251338e-06,
      "loss": 0.06313641369342804,
      "num_input_tokens_seen": 4416266,
      "step": 4284,
      "train_runtime": 10454.9555,
      "train_tokens_per_second": 422.409
    },
    {
      "epoch": 1.8312853937386473,
      "grad_norm": 2.946627378463745,
      "learning_rate": 1.6941176470588237e-06,
      "loss": 0.11229752004146576,
      "num_input_tokens_seen": 4417246,
      "step": 4285,
      "train_runtime": 10457.0493,
      "train_tokens_per_second": 422.418
    },
    {
      "epoch": 1.831712789827973,
      "grad_norm": 2.989027261734009,
      "learning_rate": 1.6898395721925135e-06,
      "loss": 0.10097094625234604,
      "num_input_tokens_seen": 4418610,
      "step": 4286,
      "train_runtime": 10459.203,
      "train_tokens_per_second": 422.461
    },
    {
      "epoch": 1.8321401859172988,
      "grad_norm": 3.1083507537841797,
      "learning_rate": 1.6855614973262032e-06,
      "loss": 0.11725413054227829,
      "num_input_tokens_seen": 4419720,
      "step": 4287,
      "train_runtime": 10461.3121,
      "train_tokens_per_second": 422.482
    },
    {
      "epoch": 1.8325675820066247,
      "grad_norm": 3.168668031692505,
      "learning_rate": 1.681283422459893e-06,
      "loss": 0.11754212528467178,
      "num_input_tokens_seen": 4420598,
      "step": 4288,
      "train_runtime": 10463.3595,
      "train_tokens_per_second": 422.484
    },
    {
      "epoch": 1.8329949780959505,
      "grad_norm": 1.998199224472046,
      "learning_rate": 1.677005347593583e-06,
      "loss": 0.06733019649982452,
      "num_input_tokens_seen": 4421472,
      "step": 4289,
      "train_runtime": 10465.4135,
      "train_tokens_per_second": 422.484
    },
    {
      "epoch": 1.8334223741852762,
      "grad_norm": 2.714137554168701,
      "learning_rate": 1.6727272727272728e-06,
      "loss": 0.08577898889780045,
      "num_input_tokens_seen": 4422534,
      "step": 4290,
      "train_runtime": 10467.4972,
      "train_tokens_per_second": 422.502
    },
    {
      "epoch": 1.833849770274602,
      "grad_norm": 2.0508828163146973,
      "learning_rate": 1.6684491978609627e-06,
      "loss": 0.08140479028224945,
      "num_input_tokens_seen": 4423746,
      "step": 4291,
      "train_runtime": 10476.3197,
      "train_tokens_per_second": 422.261
    },
    {
      "epoch": 1.8342771663639277,
      "grad_norm": 2.8936431407928467,
      "learning_rate": 1.6641711229946525e-06,
      "loss": 0.096943199634552,
      "num_input_tokens_seen": 4424922,
      "step": 4292,
      "train_runtime": 10478.4289,
      "train_tokens_per_second": 422.289
    },
    {
      "epoch": 1.8347045624532536,
      "grad_norm": 2.8280794620513916,
      "learning_rate": 1.6598930481283424e-06,
      "loss": 0.48929840326309204,
      "num_input_tokens_seen": 4425968,
      "step": 4293,
      "train_runtime": 10480.5523,
      "train_tokens_per_second": 422.303
    },
    {
      "epoch": 1.8351319585425794,
      "grad_norm": 2.151390790939331,
      "learning_rate": 1.6556149732620323e-06,
      "loss": 0.05391788110136986,
      "num_input_tokens_seen": 4426724,
      "step": 4294,
      "train_runtime": 10482.5623,
      "train_tokens_per_second": 422.294
    },
    {
      "epoch": 1.8355593546319051,
      "grad_norm": 2.100212335586548,
      "learning_rate": 1.6513368983957221e-06,
      "loss": 0.04820049926638603,
      "num_input_tokens_seen": 4427966,
      "step": 4295,
      "train_runtime": 10484.7147,
      "train_tokens_per_second": 422.326
    },
    {
      "epoch": 1.8359867507212309,
      "grad_norm": 2.6540989875793457,
      "learning_rate": 1.6470588235294118e-06,
      "loss": 0.10814224928617477,
      "num_input_tokens_seen": 4429260,
      "step": 4296,
      "train_runtime": 10486.9974,
      "train_tokens_per_second": 422.357
    },
    {
      "epoch": 1.8364141468105566,
      "grad_norm": 2.664121627807617,
      "learning_rate": 1.6427807486631017e-06,
      "loss": 0.11041717231273651,
      "num_input_tokens_seen": 4430404,
      "step": 4297,
      "train_runtime": 10489.1851,
      "train_tokens_per_second": 422.378
    },
    {
      "epoch": 1.8368415428998826,
      "grad_norm": 2.82047700881958,
      "learning_rate": 1.6385026737967915e-06,
      "loss": 0.07266925275325775,
      "num_input_tokens_seen": 4431086,
      "step": 4298,
      "train_runtime": 10491.2078,
      "train_tokens_per_second": 422.362
    },
    {
      "epoch": 1.8372689389892083,
      "grad_norm": 2.561530113220215,
      "learning_rate": 1.6342245989304814e-06,
      "loss": 0.11371771991252899,
      "num_input_tokens_seen": 4432236,
      "step": 4299,
      "train_runtime": 10493.3212,
      "train_tokens_per_second": 422.386
    },
    {
      "epoch": 1.837696335078534,
      "grad_norm": 2.803687572479248,
      "learning_rate": 1.6299465240641713e-06,
      "loss": 0.08940490335226059,
      "num_input_tokens_seen": 4433384,
      "step": 4300,
      "train_runtime": 10495.4313,
      "train_tokens_per_second": 422.411
    },
    {
      "epoch": 1.8381237311678598,
      "grad_norm": 8.537519454956055,
      "learning_rate": 1.6256684491978611e-06,
      "loss": 0.09031785279512405,
      "num_input_tokens_seen": 4434216,
      "step": 4301,
      "train_runtime": 10497.4884,
      "train_tokens_per_second": 422.407
    },
    {
      "epoch": 1.8385511272571855,
      "grad_norm": 2.049527406692505,
      "learning_rate": 1.621390374331551e-06,
      "loss": 0.05839116871356964,
      "num_input_tokens_seen": 4435168,
      "step": 4302,
      "train_runtime": 10499.5612,
      "train_tokens_per_second": 422.415
    },
    {
      "epoch": 1.8389785233465115,
      "grad_norm": 2.960066080093384,
      "learning_rate": 1.6171122994652409e-06,
      "loss": 0.07042038440704346,
      "num_input_tokens_seen": 4435940,
      "step": 4303,
      "train_runtime": 10501.5346,
      "train_tokens_per_second": 422.409
    },
    {
      "epoch": 1.8394059194358372,
      "grad_norm": 3.0962154865264893,
      "learning_rate": 1.6128342245989305e-06,
      "loss": 0.06472747772932053,
      "num_input_tokens_seen": 4436516,
      "step": 4304,
      "train_runtime": 10503.5832,
      "train_tokens_per_second": 422.381
    },
    {
      "epoch": 1.839833315525163,
      "grad_norm": 3.809420108795166,
      "learning_rate": 1.6085561497326204e-06,
      "loss": 0.11174661666154861,
      "num_input_tokens_seen": 4437518,
      "step": 4305,
      "train_runtime": 10505.7084,
      "train_tokens_per_second": 422.391
    },
    {
      "epoch": 1.8402607116144887,
      "grad_norm": 2.252808094024658,
      "learning_rate": 1.6042780748663103e-06,
      "loss": 0.0617913156747818,
      "num_input_tokens_seen": 4438934,
      "step": 4306,
      "train_runtime": 10507.8727,
      "train_tokens_per_second": 422.439
    },
    {
      "epoch": 1.8406881077038144,
      "grad_norm": 2.2713890075683594,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.06388448178768158,
      "num_input_tokens_seen": 4440076,
      "step": 4307,
      "train_runtime": 10509.9813,
      "train_tokens_per_second": 422.463
    },
    {
      "epoch": 1.8411155037931404,
      "grad_norm": 2.956070899963379,
      "learning_rate": 1.59572192513369e-06,
      "loss": 0.05821391940116882,
      "num_input_tokens_seen": 4441162,
      "step": 4308,
      "train_runtime": 10512.0933,
      "train_tokens_per_second": 422.481
    },
    {
      "epoch": 1.8415428998824661,
      "grad_norm": 2.3430144786834717,
      "learning_rate": 1.5914438502673799e-06,
      "loss": 0.07610011100769043,
      "num_input_tokens_seen": 4442044,
      "step": 4309,
      "train_runtime": 10514.131,
      "train_tokens_per_second": 422.483
    },
    {
      "epoch": 1.8419702959717918,
      "grad_norm": 2.7692441940307617,
      "learning_rate": 1.5871657754010697e-06,
      "loss": 0.08628099411725998,
      "num_input_tokens_seen": 4443036,
      "step": 4310,
      "train_runtime": 10516.2088,
      "train_tokens_per_second": 422.494
    },
    {
      "epoch": 1.8423976920611176,
      "grad_norm": 2.8638477325439453,
      "learning_rate": 1.5828877005347596e-06,
      "loss": 0.10645891726016998,
      "num_input_tokens_seen": 4444144,
      "step": 4311,
      "train_runtime": 10518.3899,
      "train_tokens_per_second": 422.512
    },
    {
      "epoch": 1.8428250881504433,
      "grad_norm": 2.6326377391815186,
      "learning_rate": 1.5786096256684495e-06,
      "loss": 0.12054909765720367,
      "num_input_tokens_seen": 4445752,
      "step": 4312,
      "train_runtime": 10520.6662,
      "train_tokens_per_second": 422.573
    },
    {
      "epoch": 1.8432524842397693,
      "grad_norm": 1.775302767753601,
      "learning_rate": 1.5743315508021391e-06,
      "loss": 0.0615677684545517,
      "num_input_tokens_seen": 4446976,
      "step": 4313,
      "train_runtime": 10522.8574,
      "train_tokens_per_second": 422.602
    },
    {
      "epoch": 1.843679880329095,
      "grad_norm": 3.1319146156311035,
      "learning_rate": 1.570053475935829e-06,
      "loss": 0.0874544009566307,
      "num_input_tokens_seen": 4448810,
      "step": 4314,
      "train_runtime": 10525.1773,
      "train_tokens_per_second": 422.683
    },
    {
      "epoch": 1.8441072764184208,
      "grad_norm": 2.8558413982391357,
      "learning_rate": 1.5657754010695189e-06,
      "loss": 0.06850621849298477,
      "num_input_tokens_seen": 4449890,
      "step": 4315,
      "train_runtime": 10527.3344,
      "train_tokens_per_second": 422.699
    },
    {
      "epoch": 1.8445346725077467,
      "grad_norm": 2.847791910171509,
      "learning_rate": 1.5614973262032088e-06,
      "loss": 0.10569768399000168,
      "num_input_tokens_seen": 4451000,
      "step": 4316,
      "train_runtime": 10529.4156,
      "train_tokens_per_second": 422.721
    },
    {
      "epoch": 1.8449620685970722,
      "grad_norm": 3.652818441390991,
      "learning_rate": 1.5572192513368986e-06,
      "loss": 0.10822293162345886,
      "num_input_tokens_seen": 4451860,
      "step": 4317,
      "train_runtime": 10531.4709,
      "train_tokens_per_second": 422.72
    },
    {
      "epoch": 1.8453894646863982,
      "grad_norm": 4.450644493103027,
      "learning_rate": 1.5529411764705885e-06,
      "loss": 0.11765066534280777,
      "num_input_tokens_seen": 4452892,
      "step": 4318,
      "train_runtime": 10533.5389,
      "train_tokens_per_second": 422.735
    },
    {
      "epoch": 1.845816860775724,
      "grad_norm": 2.8366498947143555,
      "learning_rate": 1.5486631016042784e-06,
      "loss": 0.06710223853588104,
      "num_input_tokens_seen": 4453788,
      "step": 4319,
      "train_runtime": 10535.5966,
      "train_tokens_per_second": 422.737
    },
    {
      "epoch": 1.8462442568650497,
      "grad_norm": 2.1052331924438477,
      "learning_rate": 1.5443850267379682e-06,
      "loss": 0.06824570894241333,
      "num_input_tokens_seen": 4454894,
      "step": 4320,
      "train_runtime": 10537.7057,
      "train_tokens_per_second": 422.757
    },
    {
      "epoch": 1.8466716529543756,
      "grad_norm": 2.4886536598205566,
      "learning_rate": 1.5401069518716579e-06,
      "loss": 0.07624855637550354,
      "num_input_tokens_seen": 4455648,
      "step": 4321,
      "train_runtime": 10546.3669,
      "train_tokens_per_second": 422.482
    },
    {
      "epoch": 1.8470990490437011,
      "grad_norm": 3.070173740386963,
      "learning_rate": 1.5358288770053478e-06,
      "loss": 0.1102333813905716,
      "num_input_tokens_seen": 4456670,
      "step": 4322,
      "train_runtime": 10548.4485,
      "train_tokens_per_second": 422.495
    },
    {
      "epoch": 1.847526445133027,
      "grad_norm": 2.2976720333099365,
      "learning_rate": 1.5315508021390376e-06,
      "loss": 0.10118222236633301,
      "num_input_tokens_seen": 4457902,
      "step": 4323,
      "train_runtime": 10550.6793,
      "train_tokens_per_second": 422.523
    },
    {
      "epoch": 1.8479538412223528,
      "grad_norm": 2.7383384704589844,
      "learning_rate": 1.5272727272727275e-06,
      "loss": 0.09638936072587967,
      "num_input_tokens_seen": 4458784,
      "step": 4324,
      "train_runtime": 10552.725,
      "train_tokens_per_second": 422.524
    },
    {
      "epoch": 1.8483812373116786,
      "grad_norm": 2.781740427017212,
      "learning_rate": 1.5229946524064174e-06,
      "loss": 0.07415790110826492,
      "num_input_tokens_seen": 4459534,
      "step": 4325,
      "train_runtime": 10554.7012,
      "train_tokens_per_second": 422.516
    },
    {
      "epoch": 1.8488086334010045,
      "grad_norm": 2.8392863273620605,
      "learning_rate": 1.5187165775401072e-06,
      "loss": 0.11271148920059204,
      "num_input_tokens_seen": 4460384,
      "step": 4326,
      "train_runtime": 10556.7123,
      "train_tokens_per_second": 422.516
    },
    {
      "epoch": 1.84923602949033,
      "grad_norm": 2.6092700958251953,
      "learning_rate": 1.514438502673797e-06,
      "loss": 0.09560420364141464,
      "num_input_tokens_seen": 4461714,
      "step": 4327,
      "train_runtime": 10558.8505,
      "train_tokens_per_second": 422.557
    },
    {
      "epoch": 1.849663425579656,
      "grad_norm": 1.8497010469436646,
      "learning_rate": 1.5101604278074865e-06,
      "loss": 0.06900746375322342,
      "num_input_tokens_seen": 4462744,
      "step": 4328,
      "train_runtime": 10560.914,
      "train_tokens_per_second": 422.572
    },
    {
      "epoch": 1.8500908216689818,
      "grad_norm": 3.0495924949645996,
      "learning_rate": 1.5058823529411764e-06,
      "loss": 0.07236421853303909,
      "num_input_tokens_seen": 4463348,
      "step": 4329,
      "train_runtime": 10562.9353,
      "train_tokens_per_second": 422.548
    },
    {
      "epoch": 1.8505182177583075,
      "grad_norm": 2.3355634212493896,
      "learning_rate": 1.5016042780748663e-06,
      "loss": 0.0714600533246994,
      "num_input_tokens_seen": 4464520,
      "step": 4330,
      "train_runtime": 10565.0512,
      "train_tokens_per_second": 422.574
    },
    {
      "epoch": 1.8509456138476335,
      "grad_norm": 2.9161360263824463,
      "learning_rate": 1.4973262032085562e-06,
      "loss": 0.10887053608894348,
      "num_input_tokens_seen": 4465774,
      "step": 4331,
      "train_runtime": 10567.1926,
      "train_tokens_per_second": 422.607
    },
    {
      "epoch": 1.851373009936959,
      "grad_norm": 2.344881772994995,
      "learning_rate": 1.493048128342246e-06,
      "loss": 0.06462246924638748,
      "num_input_tokens_seen": 4466628,
      "step": 4332,
      "train_runtime": 10569.2426,
      "train_tokens_per_second": 422.606
    },
    {
      "epoch": 1.851800406026285,
      "grad_norm": 3.3946645259857178,
      "learning_rate": 1.4887700534759359e-06,
      "loss": 0.0972798764705658,
      "num_input_tokens_seen": 4467690,
      "step": 4333,
      "train_runtime": 10571.3796,
      "train_tokens_per_second": 422.621
    },
    {
      "epoch": 1.8522278021156107,
      "grad_norm": 1.9606319665908813,
      "learning_rate": 1.4844919786096258e-06,
      "loss": 0.07310633361339569,
      "num_input_tokens_seen": 4468960,
      "step": 4334,
      "train_runtime": 10573.5779,
      "train_tokens_per_second": 422.654
    },
    {
      "epoch": 1.8526551982049364,
      "grad_norm": 5.3527607917785645,
      "learning_rate": 1.4802139037433156e-06,
      "loss": 0.05969268083572388,
      "num_input_tokens_seen": 4470686,
      "step": 4335,
      "train_runtime": 10575.8618,
      "train_tokens_per_second": 422.725
    },
    {
      "epoch": 1.8530825942942624,
      "grad_norm": 3.571293830871582,
      "learning_rate": 1.4759358288770053e-06,
      "loss": 0.13271893560886383,
      "num_input_tokens_seen": 4471520,
      "step": 4336,
      "train_runtime": 10577.9635,
      "train_tokens_per_second": 422.72
    },
    {
      "epoch": 1.8535099903835879,
      "grad_norm": 2.6792490482330322,
      "learning_rate": 1.4716577540106952e-06,
      "loss": 0.10143477469682693,
      "num_input_tokens_seen": 4472562,
      "step": 4337,
      "train_runtime": 10580.1519,
      "train_tokens_per_second": 422.731
    },
    {
      "epoch": 1.8539373864729138,
      "grad_norm": 2.69933819770813,
      "learning_rate": 1.467379679144385e-06,
      "loss": 0.11955171078443527,
      "num_input_tokens_seen": 4474002,
      "step": 4338,
      "train_runtime": 10582.3446,
      "train_tokens_per_second": 422.78
    },
    {
      "epoch": 1.8543647825622396,
      "grad_norm": 2.683866262435913,
      "learning_rate": 1.463101604278075e-06,
      "loss": 0.08492610603570938,
      "num_input_tokens_seen": 4474968,
      "step": 4339,
      "train_runtime": 10584.428,
      "train_tokens_per_second": 422.788
    },
    {
      "epoch": 1.8547921786515653,
      "grad_norm": 3.1592612266540527,
      "learning_rate": 1.4588235294117648e-06,
      "loss": 0.08892189711332321,
      "num_input_tokens_seen": 4475710,
      "step": 4340,
      "train_runtime": 10586.5231,
      "train_tokens_per_second": 422.774
    },
    {
      "epoch": 1.8552195747408913,
      "grad_norm": 2.559145927429199,
      "learning_rate": 1.4545454545454546e-06,
      "loss": 0.061502836644649506,
      "num_input_tokens_seen": 4476448,
      "step": 4341,
      "train_runtime": 10588.513,
      "train_tokens_per_second": 422.765
    },
    {
      "epoch": 1.8556469708302168,
      "grad_norm": 1.2657779455184937,
      "learning_rate": 1.4502673796791445e-06,
      "loss": 0.04046618938446045,
      "num_input_tokens_seen": 4477612,
      "step": 4342,
      "train_runtime": 10590.643,
      "train_tokens_per_second": 422.789
    },
    {
      "epoch": 1.8560743669195428,
      "grad_norm": 2.166637659072876,
      "learning_rate": 1.4459893048128344e-06,
      "loss": 0.09101991355419159,
      "num_input_tokens_seen": 4479194,
      "step": 4343,
      "train_runtime": 10592.8551,
      "train_tokens_per_second": 422.85
    },
    {
      "epoch": 1.8565017630088685,
      "grad_norm": 3.1321773529052734,
      "learning_rate": 1.441711229946524e-06,
      "loss": 0.13200543820858002,
      "num_input_tokens_seen": 4480316,
      "step": 4344,
      "train_runtime": 10594.9903,
      "train_tokens_per_second": 422.871
    },
    {
      "epoch": 1.8569291590981942,
      "grad_norm": 2.6689770221710205,
      "learning_rate": 1.437433155080214e-06,
      "loss": 0.08172488212585449,
      "num_input_tokens_seen": 4481212,
      "step": 4345,
      "train_runtime": 10597.1197,
      "train_tokens_per_second": 422.871
    },
    {
      "epoch": 1.8573565551875202,
      "grad_norm": 3.37274169921875,
      "learning_rate": 1.4331550802139038e-06,
      "loss": 0.07329642027616501,
      "num_input_tokens_seen": 4481994,
      "step": 4346,
      "train_runtime": 10599.2061,
      "train_tokens_per_second": 422.861
    },
    {
      "epoch": 1.8577839512768457,
      "grad_norm": 2.6092402935028076,
      "learning_rate": 1.4288770053475936e-06,
      "loss": 0.06723662465810776,
      "num_input_tokens_seen": 4482730,
      "step": 4347,
      "train_runtime": 10601.2119,
      "train_tokens_per_second": 422.851
    },
    {
      "epoch": 1.8582113473661717,
      "grad_norm": 2.500474214553833,
      "learning_rate": 1.4245989304812835e-06,
      "loss": 0.08050408214330673,
      "num_input_tokens_seen": 4484006,
      "step": 4348,
      "train_runtime": 10603.3625,
      "train_tokens_per_second": 422.885
    },
    {
      "epoch": 1.8586387434554974,
      "grad_norm": 1.4394289255142212,
      "learning_rate": 1.4203208556149734e-06,
      "loss": 0.038653984665870667,
      "num_input_tokens_seen": 4485072,
      "step": 4349,
      "train_runtime": 10605.48,
      "train_tokens_per_second": 422.901
    },
    {
      "epoch": 1.8590661395448231,
      "grad_norm": 4.310345649719238,
      "learning_rate": 1.4160427807486632e-06,
      "loss": 0.15394708514213562,
      "num_input_tokens_seen": 4485804,
      "step": 4350,
      "train_runtime": 10607.5204,
      "train_tokens_per_second": 422.889
    },
    {
      "epoch": 1.859493535634149,
      "grad_norm": 2.608119487762451,
      "learning_rate": 1.4117647058823531e-06,
      "loss": 0.07096138596534729,
      "num_input_tokens_seen": 4486530,
      "step": 4351,
      "train_runtime": 10616.3137,
      "train_tokens_per_second": 422.607
    },
    {
      "epoch": 1.8599209317234746,
      "grad_norm": 3.7449843883514404,
      "learning_rate": 1.407486631016043e-06,
      "loss": 0.1345628947019577,
      "num_input_tokens_seen": 4487486,
      "step": 4352,
      "train_runtime": 10618.5143,
      "train_tokens_per_second": 422.61
    },
    {
      "epoch": 1.8603483278128006,
      "grad_norm": 1.9537577629089355,
      "learning_rate": 1.4032085561497326e-06,
      "loss": 0.048172689974308014,
      "num_input_tokens_seen": 4488192,
      "step": 4353,
      "train_runtime": 10620.5475,
      "train_tokens_per_second": 422.595
    },
    {
      "epoch": 1.8607757239021263,
      "grad_norm": 3.1509320735931396,
      "learning_rate": 1.3989304812834225e-06,
      "loss": 0.1188560426235199,
      "num_input_tokens_seen": 4489022,
      "step": 4354,
      "train_runtime": 10622.624,
      "train_tokens_per_second": 422.591
    },
    {
      "epoch": 1.861203119991452,
      "grad_norm": 3.9425694942474365,
      "learning_rate": 1.3946524064171124e-06,
      "loss": 0.06568313390016556,
      "num_input_tokens_seen": 4490022,
      "step": 4355,
      "train_runtime": 10624.7362,
      "train_tokens_per_second": 422.601
    },
    {
      "epoch": 1.861630516080778,
      "grad_norm": 3.821216106414795,
      "learning_rate": 1.3903743315508022e-06,
      "loss": 0.11505335569381714,
      "num_input_tokens_seen": 4490836,
      "step": 4356,
      "train_runtime": 10626.8472,
      "train_tokens_per_second": 422.593
    },
    {
      "epoch": 1.8620579121701035,
      "grad_norm": 2.295562744140625,
      "learning_rate": 1.3860962566844921e-06,
      "loss": 0.08005987107753754,
      "num_input_tokens_seen": 4491726,
      "step": 4357,
      "train_runtime": 10628.9199,
      "train_tokens_per_second": 422.595
    },
    {
      "epoch": 1.8624853082594295,
      "grad_norm": 3.63199520111084,
      "learning_rate": 1.381818181818182e-06,
      "loss": 0.07291054725646973,
      "num_input_tokens_seen": 4492214,
      "step": 4358,
      "train_runtime": 10630.9016,
      "train_tokens_per_second": 422.562
    },
    {
      "epoch": 1.8629127043487552,
      "grad_norm": 2.9629008769989014,
      "learning_rate": 1.3775401069518719e-06,
      "loss": 0.10023405402898788,
      "num_input_tokens_seen": 4493662,
      "step": 4359,
      "train_runtime": 10633.1476,
      "train_tokens_per_second": 422.609
    },
    {
      "epoch": 1.863340100438081,
      "grad_norm": 3.88024640083313,
      "learning_rate": 1.3732620320855617e-06,
      "loss": 0.2093977928161621,
      "num_input_tokens_seen": 4494776,
      "step": 4360,
      "train_runtime": 10635.2791,
      "train_tokens_per_second": 422.629
    },
    {
      "epoch": 1.863767496527407,
      "grad_norm": 3.2887840270996094,
      "learning_rate": 1.3689839572192514e-06,
      "loss": 0.08741550147533417,
      "num_input_tokens_seen": 4495796,
      "step": 4361,
      "train_runtime": 10637.3795,
      "train_tokens_per_second": 422.641
    },
    {
      "epoch": 1.8641948926167324,
      "grad_norm": 1.6183552742004395,
      "learning_rate": 1.3647058823529413e-06,
      "loss": 0.04639342054724693,
      "num_input_tokens_seen": 4496968,
      "step": 4362,
      "train_runtime": 10639.4666,
      "train_tokens_per_second": 422.669
    },
    {
      "epoch": 1.8646222887060584,
      "grad_norm": 4.033382415771484,
      "learning_rate": 1.3604278074866311e-06,
      "loss": 0.13132628798484802,
      "num_input_tokens_seen": 4497670,
      "step": 4363,
      "train_runtime": 10641.4283,
      "train_tokens_per_second": 422.657
    },
    {
      "epoch": 1.8650496847953841,
      "grad_norm": 2.9869933128356934,
      "learning_rate": 1.356149732620321e-06,
      "loss": 0.1131114587187767,
      "num_input_tokens_seen": 4498990,
      "step": 4364,
      "train_runtime": 10643.6028,
      "train_tokens_per_second": 422.694
    },
    {
      "epoch": 1.8654770808847099,
      "grad_norm": 2.285688638687134,
      "learning_rate": 1.3518716577540109e-06,
      "loss": 0.0976904034614563,
      "num_input_tokens_seen": 4500326,
      "step": 4365,
      "train_runtime": 10645.7592,
      "train_tokens_per_second": 422.734
    },
    {
      "epoch": 1.8659044769740358,
      "grad_norm": 1.824602484703064,
      "learning_rate": 1.3475935828877007e-06,
      "loss": 0.07388639450073242,
      "num_input_tokens_seen": 4501718,
      "step": 4366,
      "train_runtime": 10647.9763,
      "train_tokens_per_second": 422.777
    },
    {
      "epoch": 1.8663318730633613,
      "grad_norm": 2.9529054164886475,
      "learning_rate": 1.3433155080213906e-06,
      "loss": 0.10623446851968765,
      "num_input_tokens_seen": 4502426,
      "step": 4367,
      "train_runtime": 10650.0093,
      "train_tokens_per_second": 422.763
    },
    {
      "epoch": 1.8667592691526873,
      "grad_norm": 3.275920867919922,
      "learning_rate": 1.3390374331550805e-06,
      "loss": 0.10680332034826279,
      "num_input_tokens_seen": 4503306,
      "step": 4368,
      "train_runtime": 10652.0661,
      "train_tokens_per_second": 422.764
    },
    {
      "epoch": 1.867186665242013,
      "grad_norm": 2.381131649017334,
      "learning_rate": 1.3347593582887703e-06,
      "loss": 0.06296646595001221,
      "num_input_tokens_seen": 4504004,
      "step": 4369,
      "train_runtime": 10654.0506,
      "train_tokens_per_second": 422.75
    },
    {
      "epoch": 1.8676140613313388,
      "grad_norm": 3.7544305324554443,
      "learning_rate": 1.33048128342246e-06,
      "loss": 0.10901312530040741,
      "num_input_tokens_seen": 4504794,
      "step": 4370,
      "train_runtime": 10656.0375,
      "train_tokens_per_second": 422.746
    },
    {
      "epoch": 1.8680414574206647,
      "grad_norm": 2.922407388687134,
      "learning_rate": 1.3262032085561499e-06,
      "loss": 0.09354592859745026,
      "num_input_tokens_seen": 4505616,
      "step": 4371,
      "train_runtime": 10658.0775,
      "train_tokens_per_second": 422.742
    },
    {
      "epoch": 1.8684688535099903,
      "grad_norm": 2.0420522689819336,
      "learning_rate": 1.3219251336898397e-06,
      "loss": 0.0932607352733612,
      "num_input_tokens_seen": 4507234,
      "step": 4372,
      "train_runtime": 10660.2814,
      "train_tokens_per_second": 422.806
    },
    {
      "epoch": 1.8688962495993162,
      "grad_norm": 4.0056352615356445,
      "learning_rate": 1.3176470588235296e-06,
      "loss": 0.1315418779850006,
      "num_input_tokens_seen": 4507854,
      "step": 4373,
      "train_runtime": 10662.343,
      "train_tokens_per_second": 422.783
    },
    {
      "epoch": 1.869323645688642,
      "grad_norm": 1.9559130668640137,
      "learning_rate": 1.3133689839572195e-06,
      "loss": 0.06093786284327507,
      "num_input_tokens_seen": 4509030,
      "step": 4374,
      "train_runtime": 10664.4636,
      "train_tokens_per_second": 422.809
    },
    {
      "epoch": 1.8697510417779677,
      "grad_norm": 2.9313979148864746,
      "learning_rate": 1.3090909090909093e-06,
      "loss": 0.07974699139595032,
      "num_input_tokens_seen": 4509846,
      "step": 4375,
      "train_runtime": 10666.5232,
      "train_tokens_per_second": 422.804
    },
    {
      "epoch": 1.8701784378672937,
      "grad_norm": 1.9762990474700928,
      "learning_rate": 1.3048128342245992e-06,
      "loss": 0.05429758131504059,
      "num_input_tokens_seen": 4510804,
      "step": 4376,
      "train_runtime": 10668.615,
      "train_tokens_per_second": 422.811
    },
    {
      "epoch": 1.8706058339566192,
      "grad_norm": 3.1730618476867676,
      "learning_rate": 1.300534759358289e-06,
      "loss": 0.10123877972364426,
      "num_input_tokens_seen": 4511634,
      "step": 4377,
      "train_runtime": 10670.6401,
      "train_tokens_per_second": 422.808
    },
    {
      "epoch": 1.8710332300459451,
      "grad_norm": 3.3578994274139404,
      "learning_rate": 1.2962566844919787e-06,
      "loss": 0.14225836098194122,
      "num_input_tokens_seen": 4512422,
      "step": 4378,
      "train_runtime": 10672.6903,
      "train_tokens_per_second": 422.801
    },
    {
      "epoch": 1.8714606261352709,
      "grad_norm": 3.607334852218628,
      "learning_rate": 1.2919786096256686e-06,
      "loss": 0.06675422191619873,
      "num_input_tokens_seen": 4513500,
      "step": 4379,
      "train_runtime": 10674.8012,
      "train_tokens_per_second": 422.818
    },
    {
      "epoch": 1.8718880222245966,
      "grad_norm": 2.0418055057525635,
      "learning_rate": 1.2877005347593585e-06,
      "loss": 0.04083488881587982,
      "num_input_tokens_seen": 4514482,
      "step": 4380,
      "train_runtime": 10676.8746,
      "train_tokens_per_second": 422.828
    },
    {
      "epoch": 1.8723154183139226,
      "grad_norm": 3.2652392387390137,
      "learning_rate": 1.2834224598930483e-06,
      "loss": 0.12526695430278778,
      "num_input_tokens_seen": 4515452,
      "step": 4381,
      "train_runtime": 10685.7048,
      "train_tokens_per_second": 422.569
    },
    {
      "epoch": 1.872742814403248,
      "grad_norm": 1.8685311079025269,
      "learning_rate": 1.2791443850267382e-06,
      "loss": 0.07261121273040771,
      "num_input_tokens_seen": 4516876,
      "step": 4382,
      "train_runtime": 10687.9702,
      "train_tokens_per_second": 422.613
    },
    {
      "epoch": 1.873170210492574,
      "grad_norm": 3.854722261428833,
      "learning_rate": 1.2748663101604279e-06,
      "loss": 0.08022146672010422,
      "num_input_tokens_seen": 4517666,
      "step": 4383,
      "train_runtime": 10689.9878,
      "train_tokens_per_second": 422.607
    },
    {
      "epoch": 1.8735976065818998,
      "grad_norm": 2.6227288246154785,
      "learning_rate": 1.2705882352941175e-06,
      "loss": 0.0638502910733223,
      "num_input_tokens_seen": 4518722,
      "step": 4384,
      "train_runtime": 10692.0697,
      "train_tokens_per_second": 422.624
    },
    {
      "epoch": 1.8740250026712255,
      "grad_norm": 3.1898860931396484,
      "learning_rate": 1.2663101604278074e-06,
      "loss": 0.1073106899857521,
      "num_input_tokens_seen": 4519492,
      "step": 4385,
      "train_runtime": 10694.2268,
      "train_tokens_per_second": 422.61
    },
    {
      "epoch": 1.8744523987605515,
      "grad_norm": 2.8214521408081055,
      "learning_rate": 1.2620320855614973e-06,
      "loss": 0.07853256165981293,
      "num_input_tokens_seen": 4520456,
      "step": 4386,
      "train_runtime": 10696.3485,
      "train_tokens_per_second": 422.617
    },
    {
      "epoch": 1.874879794849877,
      "grad_norm": 3.138838052749634,
      "learning_rate": 1.2577540106951871e-06,
      "loss": 0.08534351736307144,
      "num_input_tokens_seen": 4521422,
      "step": 4387,
      "train_runtime": 10698.4595,
      "train_tokens_per_second": 422.624
    },
    {
      "epoch": 1.875307190939203,
      "grad_norm": 2.3921470642089844,
      "learning_rate": 1.253475935828877e-06,
      "loss": 0.08733788877725601,
      "num_input_tokens_seen": 4522500,
      "step": 4388,
      "train_runtime": 10700.6864,
      "train_tokens_per_second": 422.636
    },
    {
      "epoch": 1.8757345870285287,
      "grad_norm": 3.278231620788574,
      "learning_rate": 1.249197860962567e-06,
      "loss": 0.11372203379869461,
      "num_input_tokens_seen": 4523376,
      "step": 4389,
      "train_runtime": 10702.7555,
      "train_tokens_per_second": 422.637
    },
    {
      "epoch": 1.8761619831178544,
      "grad_norm": 2.3801357746124268,
      "learning_rate": 1.2449197860962567e-06,
      "loss": 0.09953523427248001,
      "num_input_tokens_seen": 4524336,
      "step": 4390,
      "train_runtime": 10704.8707,
      "train_tokens_per_second": 422.643
    },
    {
      "epoch": 1.8765893792071804,
      "grad_norm": 2.160125494003296,
      "learning_rate": 1.2406417112299466e-06,
      "loss": 0.056954726576805115,
      "num_input_tokens_seen": 4525574,
      "step": 4391,
      "train_runtime": 10707.0061,
      "train_tokens_per_second": 422.674
    },
    {
      "epoch": 1.877016775296506,
      "grad_norm": 2.3410894870758057,
      "learning_rate": 1.2363636363636365e-06,
      "loss": 0.10060116648674011,
      "num_input_tokens_seen": 4527230,
      "step": 4392,
      "train_runtime": 10709.2375,
      "train_tokens_per_second": 422.741
    },
    {
      "epoch": 1.8774441713858319,
      "grad_norm": 2.1308250427246094,
      "learning_rate": 1.2320855614973264e-06,
      "loss": 0.07481782138347626,
      "num_input_tokens_seen": 4528518,
      "step": 4393,
      "train_runtime": 10711.4082,
      "train_tokens_per_second": 422.775
    },
    {
      "epoch": 1.8778715674751576,
      "grad_norm": 2.6850128173828125,
      "learning_rate": 1.2278074866310162e-06,
      "loss": 0.10995124280452728,
      "num_input_tokens_seen": 4529556,
      "step": 4394,
      "train_runtime": 10713.4923,
      "train_tokens_per_second": 422.79
    },
    {
      "epoch": 1.8782989635644833,
      "grad_norm": 3.214914321899414,
      "learning_rate": 1.223529411764706e-06,
      "loss": 0.1323504000902176,
      "num_input_tokens_seen": 4530672,
      "step": 4395,
      "train_runtime": 10715.7274,
      "train_tokens_per_second": 422.806
    },
    {
      "epoch": 1.8787263596538093,
      "grad_norm": 4.051766395568848,
      "learning_rate": 1.2192513368983957e-06,
      "loss": 0.0795101523399353,
      "num_input_tokens_seen": 4531306,
      "step": 4396,
      "train_runtime": 10717.7311,
      "train_tokens_per_second": 422.786
    },
    {
      "epoch": 1.8791537557431348,
      "grad_norm": 3.7544493675231934,
      "learning_rate": 1.2149732620320856e-06,
      "loss": 0.08741353452205658,
      "num_input_tokens_seen": 4532100,
      "step": 4397,
      "train_runtime": 10719.749,
      "train_tokens_per_second": 422.78
    },
    {
      "epoch": 1.8795811518324608,
      "grad_norm": 3.9056150913238525,
      "learning_rate": 1.2106951871657755e-06,
      "loss": 0.06964719295501709,
      "num_input_tokens_seen": 4532796,
      "step": 4398,
      "train_runtime": 10721.702,
      "train_tokens_per_second": 422.768
    },
    {
      "epoch": 1.8800085479217865,
      "grad_norm": 3.321131944656372,
      "learning_rate": 1.2064171122994654e-06,
      "loss": 0.08768081665039062,
      "num_input_tokens_seen": 4533546,
      "step": 4399,
      "train_runtime": 10723.7294,
      "train_tokens_per_second": 422.758
    },
    {
      "epoch": 1.8804359440111122,
      "grad_norm": 2.617940664291382,
      "learning_rate": 1.2021390374331552e-06,
      "loss": 0.06624089181423187,
      "num_input_tokens_seen": 4534492,
      "step": 4400,
      "train_runtime": 10725.8838,
      "train_tokens_per_second": 422.762
    },
    {
      "epoch": 1.8808633401004382,
      "grad_norm": 2.9240293502807617,
      "learning_rate": 1.1978609625668449e-06,
      "loss": 0.0696309357881546,
      "num_input_tokens_seen": 4535452,
      "step": 4401,
      "train_runtime": 10728.0324,
      "train_tokens_per_second": 422.766
    },
    {
      "epoch": 1.8812907361897637,
      "grad_norm": 3.9306929111480713,
      "learning_rate": 1.1935828877005347e-06,
      "loss": 0.08337032794952393,
      "num_input_tokens_seen": 4536142,
      "step": 4402,
      "train_runtime": 10730.039,
      "train_tokens_per_second": 422.752
    },
    {
      "epoch": 1.8817181322790897,
      "grad_norm": 2.114269733428955,
      "learning_rate": 1.1893048128342246e-06,
      "loss": 0.08273835480213165,
      "num_input_tokens_seen": 4537242,
      "step": 4403,
      "train_runtime": 10732.1557,
      "train_tokens_per_second": 422.771
    },
    {
      "epoch": 1.8821455283684154,
      "grad_norm": 3.333991289138794,
      "learning_rate": 1.1850267379679145e-06,
      "loss": 0.09323263168334961,
      "num_input_tokens_seen": 4538114,
      "step": 4404,
      "train_runtime": 10734.3056,
      "train_tokens_per_second": 422.767
    },
    {
      "epoch": 1.8825729244577412,
      "grad_norm": 3.232525110244751,
      "learning_rate": 1.1807486631016044e-06,
      "loss": 0.09853234887123108,
      "num_input_tokens_seen": 4539570,
      "step": 4405,
      "train_runtime": 10736.5325,
      "train_tokens_per_second": 422.815
    },
    {
      "epoch": 1.8830003205470671,
      "grad_norm": 3.4045495986938477,
      "learning_rate": 1.1764705882352942e-06,
      "loss": 0.12907451391220093,
      "num_input_tokens_seen": 4540454,
      "step": 4406,
      "train_runtime": 10738.602,
      "train_tokens_per_second": 422.816
    },
    {
      "epoch": 1.8834277166363926,
      "grad_norm": 3.7330353260040283,
      "learning_rate": 1.172192513368984e-06,
      "loss": 0.07779146730899811,
      "num_input_tokens_seen": 4541300,
      "step": 4407,
      "train_runtime": 10740.6747,
      "train_tokens_per_second": 422.813
    },
    {
      "epoch": 1.8838551127257186,
      "grad_norm": 2.8425981998443604,
      "learning_rate": 1.167914438502674e-06,
      "loss": 0.10518702119588852,
      "num_input_tokens_seen": 4542642,
      "step": 4408,
      "train_runtime": 10742.827,
      "train_tokens_per_second": 422.854
    },
    {
      "epoch": 1.8842825088150443,
      "grad_norm": 2.3275437355041504,
      "learning_rate": 1.1636363636363638e-06,
      "loss": 0.059425994753837585,
      "num_input_tokens_seen": 4543694,
      "step": 4409,
      "train_runtime": 10744.9118,
      "train_tokens_per_second": 422.869
    },
    {
      "epoch": 1.88470990490437,
      "grad_norm": 4.658697128295898,
      "learning_rate": 1.1593582887700535e-06,
      "loss": 0.13761216402053833,
      "num_input_tokens_seen": 4544400,
      "step": 4410,
      "train_runtime": 10746.899,
      "train_tokens_per_second": 422.857
    },
    {
      "epoch": 1.885137300993696,
      "grad_norm": 1.957148790359497,
      "learning_rate": 1.1550802139037434e-06,
      "loss": 0.06264439970254898,
      "num_input_tokens_seen": 4545664,
      "step": 4411,
      "train_runtime": 10755.5374,
      "train_tokens_per_second": 422.635
    },
    {
      "epoch": 1.8855646970830215,
      "grad_norm": 1.3168405294418335,
      "learning_rate": 1.1508021390374332e-06,
      "loss": 0.037496551871299744,
      "num_input_tokens_seen": 4546994,
      "step": 4412,
      "train_runtime": 10757.8631,
      "train_tokens_per_second": 422.667
    },
    {
      "epoch": 1.8859920931723475,
      "grad_norm": 2.067268133163452,
      "learning_rate": 1.146524064171123e-06,
      "loss": 0.0679556205868721,
      "num_input_tokens_seen": 4548262,
      "step": 4413,
      "train_runtime": 10760.0364,
      "train_tokens_per_second": 422.7
    },
    {
      "epoch": 1.8864194892616732,
      "grad_norm": 2.7608087062835693,
      "learning_rate": 1.142245989304813e-06,
      "loss": 0.0793260931968689,
      "num_input_tokens_seen": 4549078,
      "step": 4414,
      "train_runtime": 10762.0967,
      "train_tokens_per_second": 422.694
    },
    {
      "epoch": 1.886846885350999,
      "grad_norm": 1.7697131633758545,
      "learning_rate": 1.1379679144385028e-06,
      "loss": 0.04217218607664108,
      "num_input_tokens_seen": 4549992,
      "step": 4415,
      "train_runtime": 10764.1932,
      "train_tokens_per_second": 422.697
    },
    {
      "epoch": 1.887274281440325,
      "grad_norm": 2.289616584777832,
      "learning_rate": 1.1336898395721927e-06,
      "loss": 0.08312086015939713,
      "num_input_tokens_seen": 4550968,
      "step": 4416,
      "train_runtime": 10766.2911,
      "train_tokens_per_second": 422.705
    },
    {
      "epoch": 1.8877016775296505,
      "grad_norm": 2.1274046897888184,
      "learning_rate": 1.1294117647058826e-06,
      "loss": 0.061488281935453415,
      "num_input_tokens_seen": 4551618,
      "step": 4417,
      "train_runtime": 10768.3105,
      "train_tokens_per_second": 422.686
    },
    {
      "epoch": 1.8881290736189764,
      "grad_norm": 2.6610591411590576,
      "learning_rate": 1.1251336898395722e-06,
      "loss": 0.125120609998703,
      "num_input_tokens_seen": 4552488,
      "step": 4418,
      "train_runtime": 10770.3327,
      "train_tokens_per_second": 422.688
    },
    {
      "epoch": 1.8885564697083022,
      "grad_norm": 2.289487838745117,
      "learning_rate": 1.120855614973262e-06,
      "loss": 0.06735081225633621,
      "num_input_tokens_seen": 4553512,
      "step": 4419,
      "train_runtime": 10772.4039,
      "train_tokens_per_second": 422.702
    },
    {
      "epoch": 1.888983865797628,
      "grad_norm": 2.808692455291748,
      "learning_rate": 1.116577540106952e-06,
      "loss": 0.10418041050434113,
      "num_input_tokens_seen": 4554500,
      "step": 4420,
      "train_runtime": 10774.4607,
      "train_tokens_per_second": 422.713
    },
    {
      "epoch": 1.8894112618869539,
      "grad_norm": 2.686461925506592,
      "learning_rate": 1.1122994652406418e-06,
      "loss": 0.08092980831861496,
      "num_input_tokens_seen": 4555870,
      "step": 4421,
      "train_runtime": 10776.645,
      "train_tokens_per_second": 422.754
    },
    {
      "epoch": 1.8898386579762794,
      "grad_norm": 2.5925116539001465,
      "learning_rate": 1.1080213903743317e-06,
      "loss": 0.0827026292681694,
      "num_input_tokens_seen": 4556762,
      "step": 4422,
      "train_runtime": 10778.7093,
      "train_tokens_per_second": 422.756
    },
    {
      "epoch": 1.8902660540656053,
      "grad_norm": 3.3147714138031006,
      "learning_rate": 1.1037433155080214e-06,
      "loss": 0.06726095825433731,
      "num_input_tokens_seen": 4557706,
      "step": 4423,
      "train_runtime": 10780.791,
      "train_tokens_per_second": 422.762
    },
    {
      "epoch": 1.890693450154931,
      "grad_norm": 3.3284411430358887,
      "learning_rate": 1.0994652406417112e-06,
      "loss": 0.08040409535169601,
      "num_input_tokens_seen": 4558820,
      "step": 4424,
      "train_runtime": 10782.9131,
      "train_tokens_per_second": 422.782
    },
    {
      "epoch": 1.8911208462442568,
      "grad_norm": 2.8569467067718506,
      "learning_rate": 1.0951871657754011e-06,
      "loss": 0.1341744065284729,
      "num_input_tokens_seen": 4560012,
      "step": 4425,
      "train_runtime": 10785.1304,
      "train_tokens_per_second": 422.805
    },
    {
      "epoch": 1.8915482423335828,
      "grad_norm": 3.35448956489563,
      "learning_rate": 1.090909090909091e-06,
      "loss": 0.11207510530948639,
      "num_input_tokens_seen": 4560762,
      "step": 4426,
      "train_runtime": 10787.1654,
      "train_tokens_per_second": 422.795
    },
    {
      "epoch": 1.8919756384229083,
      "grad_norm": 2.942560911178589,
      "learning_rate": 1.0866310160427808e-06,
      "loss": 0.09185966104269028,
      "num_input_tokens_seen": 4561562,
      "step": 4427,
      "train_runtime": 10789.2369,
      "train_tokens_per_second": 422.788
    },
    {
      "epoch": 1.8924030345122342,
      "grad_norm": 3.7005763053894043,
      "learning_rate": 1.0823529411764707e-06,
      "loss": 0.08973820507526398,
      "num_input_tokens_seen": 4562314,
      "step": 4428,
      "train_runtime": 10791.2643,
      "train_tokens_per_second": 422.778
    },
    {
      "epoch": 1.89283043060156,
      "grad_norm": 2.077177047729492,
      "learning_rate": 1.0780748663101606e-06,
      "loss": 0.05421359837055206,
      "num_input_tokens_seen": 4563396,
      "step": 4429,
      "train_runtime": 10793.382,
      "train_tokens_per_second": 422.796
    },
    {
      "epoch": 1.8932578266908857,
      "grad_norm": 2.066877841949463,
      "learning_rate": 1.0737967914438502e-06,
      "loss": 0.0701863244175911,
      "num_input_tokens_seen": 4565338,
      "step": 4430,
      "train_runtime": 10795.6819,
      "train_tokens_per_second": 422.886
    },
    {
      "epoch": 1.8936852227802117,
      "grad_norm": 3.4514458179473877,
      "learning_rate": 1.0695187165775401e-06,
      "loss": 0.08999794721603394,
      "num_input_tokens_seen": 4566132,
      "step": 4431,
      "train_runtime": 10797.7186,
      "train_tokens_per_second": 422.879
    },
    {
      "epoch": 1.8941126188695372,
      "grad_norm": 2.5464253425598145,
      "learning_rate": 1.06524064171123e-06,
      "loss": 0.08568716049194336,
      "num_input_tokens_seen": 4567134,
      "step": 4432,
      "train_runtime": 10799.8723,
      "train_tokens_per_second": 422.888
    },
    {
      "epoch": 1.8945400149588631,
      "grad_norm": 2.3488171100616455,
      "learning_rate": 1.0609625668449198e-06,
      "loss": 0.05926796421408653,
      "num_input_tokens_seen": 4568326,
      "step": 4433,
      "train_runtime": 10802.1148,
      "train_tokens_per_second": 422.91
    },
    {
      "epoch": 1.8949674110481889,
      "grad_norm": 1.9476948976516724,
      "learning_rate": 1.0566844919786097e-06,
      "loss": 0.0564112663269043,
      "num_input_tokens_seen": 4569418,
      "step": 4434,
      "train_runtime": 10804.227,
      "train_tokens_per_second": 422.929
    },
    {
      "epoch": 1.8953948071375146,
      "grad_norm": 3.591588020324707,
      "learning_rate": 1.0524064171122996e-06,
      "loss": 0.09620843827724457,
      "num_input_tokens_seen": 4570096,
      "step": 4435,
      "train_runtime": 10806.2607,
      "train_tokens_per_second": 422.912
    },
    {
      "epoch": 1.8958222032268406,
      "grad_norm": 3.2518348693847656,
      "learning_rate": 1.0481283422459895e-06,
      "loss": 0.07996368408203125,
      "num_input_tokens_seen": 4570922,
      "step": 4436,
      "train_runtime": 10808.3043,
      "train_tokens_per_second": 422.908
    },
    {
      "epoch": 1.896249599316166,
      "grad_norm": 5.111767768859863,
      "learning_rate": 1.0438502673796793e-06,
      "loss": 0.09326605498790741,
      "num_input_tokens_seen": 4571552,
      "step": 4437,
      "train_runtime": 10810.358,
      "train_tokens_per_second": 422.886
    },
    {
      "epoch": 1.896676995405492,
      "grad_norm": 2.164259672164917,
      "learning_rate": 1.039572192513369e-06,
      "loss": 0.05288808420300484,
      "num_input_tokens_seen": 4572550,
      "step": 4438,
      "train_runtime": 10812.442,
      "train_tokens_per_second": 422.897
    },
    {
      "epoch": 1.8971043914948178,
      "grad_norm": 4.1734185218811035,
      "learning_rate": 1.0352941176470589e-06,
      "loss": 0.08563899993896484,
      "num_input_tokens_seen": 4573476,
      "step": 4439,
      "train_runtime": 10814.5183,
      "train_tokens_per_second": 422.901
    },
    {
      "epoch": 1.8975317875841435,
      "grad_norm": 2.8799474239349365,
      "learning_rate": 1.0310160427807487e-06,
      "loss": 0.10343010723590851,
      "num_input_tokens_seen": 4575040,
      "step": 4440,
      "train_runtime": 10816.7109,
      "train_tokens_per_second": 422.96
    },
    {
      "epoch": 1.8979591836734695,
      "grad_norm": 2.5237553119659424,
      "learning_rate": 1.0267379679144386e-06,
      "loss": 0.10083038359880447,
      "num_input_tokens_seen": 4575960,
      "step": 4441,
      "train_runtime": 10825.6237,
      "train_tokens_per_second": 422.697
    },
    {
      "epoch": 1.898386579762795,
      "grad_norm": 2.4181230068206787,
      "learning_rate": 1.0224598930481285e-06,
      "loss": 0.0894622653722763,
      "num_input_tokens_seen": 4576970,
      "step": 4442,
      "train_runtime": 10827.6891,
      "train_tokens_per_second": 422.71
    },
    {
      "epoch": 1.898813975852121,
      "grad_norm": 2.875009059906006,
      "learning_rate": 1.0181818181818183e-06,
      "loss": 0.09652965515851974,
      "num_input_tokens_seen": 4577880,
      "step": 4443,
      "train_runtime": 10829.7334,
      "train_tokens_per_second": 422.714
    },
    {
      "epoch": 1.8992413719414467,
      "grad_norm": 2.7091493606567383,
      "learning_rate": 1.0139037433155082e-06,
      "loss": 0.08103708177804947,
      "num_input_tokens_seen": 4578642,
      "step": 4444,
      "train_runtime": 10831.7101,
      "train_tokens_per_second": 422.707
    },
    {
      "epoch": 1.8996687680307724,
      "grad_norm": 2.3246288299560547,
      "learning_rate": 1.009625668449198e-06,
      "loss": 0.06236422806978226,
      "num_input_tokens_seen": 4579556,
      "step": 4445,
      "train_runtime": 10833.7538,
      "train_tokens_per_second": 422.712
    },
    {
      "epoch": 1.9000961641200984,
      "grad_norm": 2.6732337474823,
      "learning_rate": 1.005347593582888e-06,
      "loss": 0.09596578776836395,
      "num_input_tokens_seen": 4580622,
      "step": 4446,
      "train_runtime": 10835.8395,
      "train_tokens_per_second": 422.729
    },
    {
      "epoch": 1.900523560209424,
      "grad_norm": 2.9142520427703857,
      "learning_rate": 1.0010695187165776e-06,
      "loss": 0.11763831973075867,
      "num_input_tokens_seen": 4581960,
      "step": 4447,
      "train_runtime": 10838.0044,
      "train_tokens_per_second": 422.768
    },
    {
      "epoch": 1.9009509562987499,
      "grad_norm": 2.628542423248291,
      "learning_rate": 9.967914438502675e-07,
      "loss": 0.11464789509773254,
      "num_input_tokens_seen": 4582736,
      "step": 4448,
      "train_runtime": 10840.0139,
      "train_tokens_per_second": 422.761
    },
    {
      "epoch": 1.9013783523880756,
      "grad_norm": 1.5382531881332397,
      "learning_rate": 9.925133689839573e-07,
      "loss": 0.03625532239675522,
      "num_input_tokens_seen": 4584306,
      "step": 4449,
      "train_runtime": 10842.2167,
      "train_tokens_per_second": 422.82
    },
    {
      "epoch": 1.9018057484774014,
      "grad_norm": 2.1141552925109863,
      "learning_rate": 9.88235294117647e-07,
      "loss": 0.06490832567214966,
      "num_input_tokens_seen": 4585572,
      "step": 4450,
      "train_runtime": 10844.3597,
      "train_tokens_per_second": 422.853
    },
    {
      "epoch": 1.9022331445667273,
      "grad_norm": 3.873093366622925,
      "learning_rate": 9.839572192513369e-07,
      "loss": 0.0809621512889862,
      "num_input_tokens_seen": 4586200,
      "step": 4451,
      "train_runtime": 10846.5735,
      "train_tokens_per_second": 422.825
    },
    {
      "epoch": 1.9026605406560528,
      "grad_norm": 2.6607279777526855,
      "learning_rate": 9.796791443850267e-07,
      "loss": 0.10808295756578445,
      "num_input_tokens_seen": 4587598,
      "step": 4452,
      "train_runtime": 10848.7433,
      "train_tokens_per_second": 422.869
    },
    {
      "epoch": 1.9030879367453788,
      "grad_norm": 3.1878325939178467,
      "learning_rate": 9.754010695187166e-07,
      "loss": 0.07286867499351501,
      "num_input_tokens_seen": 4588940,
      "step": 4453,
      "train_runtime": 10850.9056,
      "train_tokens_per_second": 422.908
    },
    {
      "epoch": 1.9035153328347045,
      "grad_norm": 2.5134034156799316,
      "learning_rate": 9.711229946524065e-07,
      "loss": 0.10439202189445496,
      "num_input_tokens_seen": 4590366,
      "step": 4454,
      "train_runtime": 10853.0824,
      "train_tokens_per_second": 422.955
    },
    {
      "epoch": 1.9039427289240303,
      "grad_norm": 2.1706836223602295,
      "learning_rate": 9.668449197860963e-07,
      "loss": 0.07070669531822205,
      "num_input_tokens_seen": 4591482,
      "step": 4455,
      "train_runtime": 10855.1709,
      "train_tokens_per_second": 422.976
    },
    {
      "epoch": 1.9043701250133562,
      "grad_norm": 2.883697748184204,
      "learning_rate": 9.625668449197862e-07,
      "loss": 0.09853550791740417,
      "num_input_tokens_seen": 4592686,
      "step": 4456,
      "train_runtime": 10857.3112,
      "train_tokens_per_second": 423.004
    },
    {
      "epoch": 1.9047975211026817,
      "grad_norm": 2.686236619949341,
      "learning_rate": 9.58288770053476e-07,
      "loss": 0.07270325720310211,
      "num_input_tokens_seen": 4593824,
      "step": 4457,
      "train_runtime": 10859.4302,
      "train_tokens_per_second": 423.026
    },
    {
      "epoch": 1.9052249171920077,
      "grad_norm": 3.183248281478882,
      "learning_rate": 9.540106951871657e-07,
      "loss": 0.1145283430814743,
      "num_input_tokens_seen": 4594684,
      "step": 4458,
      "train_runtime": 10861.4795,
      "train_tokens_per_second": 423.026
    },
    {
      "epoch": 1.9056523132813334,
      "grad_norm": 2.770751953125,
      "learning_rate": 9.497326203208557e-07,
      "loss": 0.12366009503602982,
      "num_input_tokens_seen": 4596346,
      "step": 4459,
      "train_runtime": 10863.7589,
      "train_tokens_per_second": 423.09
    },
    {
      "epoch": 1.9060797093706592,
      "grad_norm": 2.6289772987365723,
      "learning_rate": 9.454545454545455e-07,
      "loss": 0.08646272122859955,
      "num_input_tokens_seen": 4597426,
      "step": 4460,
      "train_runtime": 10865.8989,
      "train_tokens_per_second": 423.106
    },
    {
      "epoch": 1.9065071054599851,
      "grad_norm": 2.8655219078063965,
      "learning_rate": 9.411764705882353e-07,
      "loss": 0.055573564022779465,
      "num_input_tokens_seen": 4598430,
      "step": 4461,
      "train_runtime": 10868.0067,
      "train_tokens_per_second": 423.116
    },
    {
      "epoch": 1.9069345015493109,
      "grad_norm": 3.240882158279419,
      "learning_rate": 9.368983957219252e-07,
      "loss": 0.11798454821109772,
      "num_input_tokens_seen": 4599202,
      "step": 4462,
      "train_runtime": 10870.0356,
      "train_tokens_per_second": 423.108
    },
    {
      "epoch": 1.9073618976386366,
      "grad_norm": 2.6404199600219727,
      "learning_rate": 9.326203208556151e-07,
      "loss": 0.07360601425170898,
      "num_input_tokens_seen": 4600228,
      "step": 4463,
      "train_runtime": 10872.1629,
      "train_tokens_per_second": 423.12
    },
    {
      "epoch": 1.9077892937279624,
      "grad_norm": 3.494415760040283,
      "learning_rate": 9.283422459893048e-07,
      "loss": 0.1317797601222992,
      "num_input_tokens_seen": 4601364,
      "step": 4464,
      "train_runtime": 10874.3183,
      "train_tokens_per_second": 423.14
    },
    {
      "epoch": 1.908216689817288,
      "grad_norm": 2.309741735458374,
      "learning_rate": 9.240641711229947e-07,
      "loss": 0.0942305326461792,
      "num_input_tokens_seen": 4602804,
      "step": 4465,
      "train_runtime": 10876.5232,
      "train_tokens_per_second": 423.187
    },
    {
      "epoch": 1.908644085906614,
      "grad_norm": 2.820481538772583,
      "learning_rate": 9.197860962566846e-07,
      "loss": 0.08808265626430511,
      "num_input_tokens_seen": 4603844,
      "step": 4466,
      "train_runtime": 10878.6275,
      "train_tokens_per_second": 423.201
    },
    {
      "epoch": 1.9090714819959398,
      "grad_norm": 3.6186227798461914,
      "learning_rate": 9.155080213903744e-07,
      "loss": 0.08472894877195358,
      "num_input_tokens_seen": 4604468,
      "step": 4467,
      "train_runtime": 10880.61,
      "train_tokens_per_second": 423.181
    },
    {
      "epoch": 1.9094988780852655,
      "grad_norm": 2.3990604877471924,
      "learning_rate": 9.112299465240643e-07,
      "loss": 0.04658800736069679,
      "num_input_tokens_seen": 4605274,
      "step": 4468,
      "train_runtime": 10882.6271,
      "train_tokens_per_second": 423.177
    },
    {
      "epoch": 1.9099262741745913,
      "grad_norm": 2.6029651165008545,
      "learning_rate": 9.069518716577541e-07,
      "loss": 0.0794363021850586,
      "num_input_tokens_seen": 4606284,
      "step": 4469,
      "train_runtime": 10884.7805,
      "train_tokens_per_second": 423.186
    },
    {
      "epoch": 1.910353670263917,
      "grad_norm": 3.233903169631958,
      "learning_rate": 9.02673796791444e-07,
      "loss": 0.09564993530511856,
      "num_input_tokens_seen": 4607106,
      "step": 4470,
      "train_runtime": 10886.8279,
      "train_tokens_per_second": 423.182
    },
    {
      "epoch": 1.910781066353243,
      "grad_norm": 2.8722848892211914,
      "learning_rate": 8.983957219251338e-07,
      "loss": 0.10308776795864105,
      "num_input_tokens_seen": 4608026,
      "step": 4471,
      "train_runtime": 10895.2446,
      "train_tokens_per_second": 422.939
    },
    {
      "epoch": 1.9112084624425687,
      "grad_norm": 2.3191356658935547,
      "learning_rate": 8.941176470588237e-07,
      "loss": 0.06808208674192429,
      "num_input_tokens_seen": 4609326,
      "step": 4472,
      "train_runtime": 10897.509,
      "train_tokens_per_second": 422.971
    },
    {
      "epoch": 1.9116358585318944,
      "grad_norm": 3.15012788772583,
      "learning_rate": 8.898395721925135e-07,
      "loss": 0.07747022807598114,
      "num_input_tokens_seen": 4610034,
      "step": 4473,
      "train_runtime": 10899.4937,
      "train_tokens_per_second": 422.959
    },
    {
      "epoch": 1.9120632546212202,
      "grad_norm": 2.6855080127716064,
      "learning_rate": 8.855614973262033e-07,
      "loss": 0.08258849382400513,
      "num_input_tokens_seen": 4610946,
      "step": 4474,
      "train_runtime": 10901.5584,
      "train_tokens_per_second": 422.962
    },
    {
      "epoch": 1.912490650710546,
      "grad_norm": 3.6249570846557617,
      "learning_rate": 8.812834224598932e-07,
      "loss": 0.15937760472297668,
      "num_input_tokens_seen": 4611904,
      "step": 4475,
      "train_runtime": 10903.7979,
      "train_tokens_per_second": 422.963
    },
    {
      "epoch": 1.9129180467998719,
      "grad_norm": 3.302855968475342,
      "learning_rate": 8.770053475935831e-07,
      "loss": 0.09469449520111084,
      "num_input_tokens_seen": 4612972,
      "step": 4476,
      "train_runtime": 10905.8861,
      "train_tokens_per_second": 422.98
    },
    {
      "epoch": 1.9133454428891976,
      "grad_norm": 3.47115421295166,
      "learning_rate": 8.727272727272728e-07,
      "loss": 0.12235187739133835,
      "num_input_tokens_seen": 4614188,
      "step": 4477,
      "train_runtime": 10908.0371,
      "train_tokens_per_second": 423.008
    },
    {
      "epoch": 1.9137728389785233,
      "grad_norm": 2.299360752105713,
      "learning_rate": 8.684491978609626e-07,
      "loss": 0.0771092027425766,
      "num_input_tokens_seen": 4615162,
      "step": 4478,
      "train_runtime": 10910.0839,
      "train_tokens_per_second": 423.018
    },
    {
      "epoch": 1.914200235067849,
      "grad_norm": 2.522913932800293,
      "learning_rate": 8.641711229946525e-07,
      "loss": 0.0717511847615242,
      "num_input_tokens_seen": 4616256,
      "step": 4479,
      "train_runtime": 10912.2003,
      "train_tokens_per_second": 423.036
    },
    {
      "epoch": 1.9146276311571748,
      "grad_norm": 2.7510159015655518,
      "learning_rate": 8.598930481283422e-07,
      "loss": 0.1033727303147316,
      "num_input_tokens_seen": 4617298,
      "step": 4480,
      "train_runtime": 10914.3292,
      "train_tokens_per_second": 423.049
    },
    {
      "epoch": 1.9150550272465008,
      "grad_norm": 2.4151244163513184,
      "learning_rate": 8.556149732620321e-07,
      "loss": 0.08595739305019379,
      "num_input_tokens_seen": 4618388,
      "step": 4481,
      "train_runtime": 10916.4773,
      "train_tokens_per_second": 423.066
    },
    {
      "epoch": 1.9154824233358265,
      "grad_norm": 2.7252376079559326,
      "learning_rate": 8.51336898395722e-07,
      "loss": 0.07521509379148483,
      "num_input_tokens_seen": 4619430,
      "step": 4482,
      "train_runtime": 10918.6822,
      "train_tokens_per_second": 423.076
    },
    {
      "epoch": 1.9159098194251523,
      "grad_norm": 2.6735124588012695,
      "learning_rate": 8.470588235294118e-07,
      "loss": 0.10662515461444855,
      "num_input_tokens_seen": 4620298,
      "step": 4483,
      "train_runtime": 10920.7703,
      "train_tokens_per_second": 423.074
    },
    {
      "epoch": 1.916337215514478,
      "grad_norm": 3.219670057296753,
      "learning_rate": 8.427807486631016e-07,
      "loss": 0.07701162248849869,
      "num_input_tokens_seen": 4621030,
      "step": 4484,
      "train_runtime": 10922.8223,
      "train_tokens_per_second": 423.062
    },
    {
      "epoch": 1.9167646116038037,
      "grad_norm": 2.793849468231201,
      "learning_rate": 8.385026737967915e-07,
      "loss": 0.12319433689117432,
      "num_input_tokens_seen": 4622124,
      "step": 4485,
      "train_runtime": 10924.9377,
      "train_tokens_per_second": 423.08
    },
    {
      "epoch": 1.9171920076931297,
      "grad_norm": 2.798619031906128,
      "learning_rate": 8.342245989304813e-07,
      "loss": 0.086573027074337,
      "num_input_tokens_seen": 4622936,
      "step": 4486,
      "train_runtime": 10926.9813,
      "train_tokens_per_second": 423.075
    },
    {
      "epoch": 1.9176194037824554,
      "grad_norm": 4.573685169219971,
      "learning_rate": 8.299465240641712e-07,
      "loss": 0.09519963711500168,
      "num_input_tokens_seen": 4623452,
      "step": 4487,
      "train_runtime": 10928.9411,
      "train_tokens_per_second": 423.047
    },
    {
      "epoch": 1.9180467998717812,
      "grad_norm": 2.2354068756103516,
      "learning_rate": 8.256684491978611e-07,
      "loss": 0.08667831122875214,
      "num_input_tokens_seen": 4624626,
      "step": 4488,
      "train_runtime": 10931.0923,
      "train_tokens_per_second": 423.071
    },
    {
      "epoch": 1.918474195961107,
      "grad_norm": 3.781447172164917,
      "learning_rate": 8.213903743315508e-07,
      "loss": 0.12547264993190765,
      "num_input_tokens_seen": 4625440,
      "step": 4489,
      "train_runtime": 10933.1179,
      "train_tokens_per_second": 423.067
    },
    {
      "epoch": 1.9189015920504326,
      "grad_norm": 2.7677695751190186,
      "learning_rate": 8.171122994652407e-07,
      "loss": 0.0763927474617958,
      "num_input_tokens_seen": 4626682,
      "step": 4490,
      "train_runtime": 10935.3543,
      "train_tokens_per_second": 423.094
    },
    {
      "epoch": 1.9193289881397586,
      "grad_norm": 1.879095435142517,
      "learning_rate": 8.128342245989306e-07,
      "loss": 0.06771612912416458,
      "num_input_tokens_seen": 4628188,
      "step": 4491,
      "train_runtime": 10937.6508,
      "train_tokens_per_second": 423.143
    },
    {
      "epoch": 1.9197563842290843,
      "grad_norm": 4.258627891540527,
      "learning_rate": 8.085561497326204e-07,
      "loss": 0.10876234620809555,
      "num_input_tokens_seen": 4629086,
      "step": 4492,
      "train_runtime": 10939.669,
      "train_tokens_per_second": 423.147
    },
    {
      "epoch": 1.92018378031841,
      "grad_norm": 2.422433614730835,
      "learning_rate": 8.042780748663102e-07,
      "loss": 0.09476907551288605,
      "num_input_tokens_seen": 4630218,
      "step": 4493,
      "train_runtime": 10941.7847,
      "train_tokens_per_second": 423.168
    },
    {
      "epoch": 1.9206111764077358,
      "grad_norm": 2.429797410964966,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.07689294964075089,
      "num_input_tokens_seen": 4631222,
      "step": 4494,
      "train_runtime": 10943.8672,
      "train_tokens_per_second": 423.18
    },
    {
      "epoch": 1.9210385724970616,
      "grad_norm": 2.0165908336639404,
      "learning_rate": 7.957219251336899e-07,
      "loss": 0.05318460613489151,
      "num_input_tokens_seen": 4632586,
      "step": 4495,
      "train_runtime": 10946.084,
      "train_tokens_per_second": 423.219
    },
    {
      "epoch": 1.9214659685863875,
      "grad_norm": 4.20761775970459,
      "learning_rate": 7.914438502673798e-07,
      "loss": 0.07873548567295074,
      "num_input_tokens_seen": 4633310,
      "step": 4496,
      "train_runtime": 10948.1285,
      "train_tokens_per_second": 423.206
    },
    {
      "epoch": 1.9218933646757133,
      "grad_norm": 2.5924205780029297,
      "learning_rate": 7.871657754010696e-07,
      "loss": 0.10081905126571655,
      "num_input_tokens_seen": 4634730,
      "step": 4497,
      "train_runtime": 10950.3803,
      "train_tokens_per_second": 423.248
    },
    {
      "epoch": 1.922320760765039,
      "grad_norm": 3.6822032928466797,
      "learning_rate": 7.828877005347594e-07,
      "loss": 0.09389260411262512,
      "num_input_tokens_seen": 4635824,
      "step": 4498,
      "train_runtime": 10952.5498,
      "train_tokens_per_second": 423.264
    },
    {
      "epoch": 1.9227481568543647,
      "grad_norm": 3.908029079437256,
      "learning_rate": 7.786096256684493e-07,
      "loss": 0.11471231281757355,
      "num_input_tokens_seen": 4636548,
      "step": 4499,
      "train_runtime": 10954.5924,
      "train_tokens_per_second": 423.252
    },
    {
      "epoch": 1.9231755529436905,
      "grad_norm": 2.634157657623291,
      "learning_rate": 7.743315508021392e-07,
      "loss": 0.08247380703687668,
      "num_input_tokens_seen": 4637418,
      "step": 4500,
      "train_runtime": 10956.639,
      "train_tokens_per_second": 423.252
    },
    {
      "epoch": 1.9236029490330164,
      "grad_norm": 2.4601261615753174,
      "learning_rate": 7.700534759358289e-07,
      "loss": 0.08436372131109238,
      "num_input_tokens_seen": 4638588,
      "step": 4501,
      "train_runtime": 10965.4772,
      "train_tokens_per_second": 423.017
    },
    {
      "epoch": 1.9240303451223422,
      "grad_norm": 3.006995677947998,
      "learning_rate": 7.657754010695188e-07,
      "loss": 0.07325327396392822,
      "num_input_tokens_seen": 4639484,
      "step": 4502,
      "train_runtime": 10967.5304,
      "train_tokens_per_second": 423.02
    },
    {
      "epoch": 1.924457741211668,
      "grad_norm": 2.252163887023926,
      "learning_rate": 7.614973262032087e-07,
      "loss": 0.06350468844175339,
      "num_input_tokens_seen": 4640650,
      "step": 4503,
      "train_runtime": 10969.7577,
      "train_tokens_per_second": 423.04
    },
    {
      "epoch": 1.9248851373009936,
      "grad_norm": 4.078907489776611,
      "learning_rate": 7.572192513368986e-07,
      "loss": 0.12562359869480133,
      "num_input_tokens_seen": 4641402,
      "step": 4504,
      "train_runtime": 10971.7872,
      "train_tokens_per_second": 423.031
    },
    {
      "epoch": 1.9253125333903194,
      "grad_norm": 2.573749542236328,
      "learning_rate": 7.529411764705882e-07,
      "loss": 0.09105457365512848,
      "num_input_tokens_seen": 4642344,
      "step": 4505,
      "train_runtime": 10973.9928,
      "train_tokens_per_second": 423.031
    },
    {
      "epoch": 1.9257399294796453,
      "grad_norm": 2.4820497035980225,
      "learning_rate": 7.486631016042781e-07,
      "loss": 0.05052736774086952,
      "num_input_tokens_seen": 4643312,
      "step": 4506,
      "train_runtime": 10976.0944,
      "train_tokens_per_second": 423.039
    },
    {
      "epoch": 1.926167325568971,
      "grad_norm": 2.8915936946868896,
      "learning_rate": 7.443850267379679e-07,
      "loss": 0.12244762480258942,
      "num_input_tokens_seen": 4644386,
      "step": 4507,
      "train_runtime": 10978.178,
      "train_tokens_per_second": 423.056
    },
    {
      "epoch": 1.9265947216582968,
      "grad_norm": 2.169282913208008,
      "learning_rate": 7.401069518716578e-07,
      "loss": 0.05559958890080452,
      "num_input_tokens_seen": 4645144,
      "step": 4508,
      "train_runtime": 10980.181,
      "train_tokens_per_second": 423.048
    },
    {
      "epoch": 1.9270221177476226,
      "grad_norm": 2.6740500926971436,
      "learning_rate": 7.358288770053476e-07,
      "loss": 0.09286932647228241,
      "num_input_tokens_seen": 4646072,
      "step": 4509,
      "train_runtime": 10982.2293,
      "train_tokens_per_second": 423.054
    },
    {
      "epoch": 1.9274495138369483,
      "grad_norm": 6.0752973556518555,
      "learning_rate": 7.315508021390374e-07,
      "loss": 0.06717438995838165,
      "num_input_tokens_seen": 4646864,
      "step": 4510,
      "train_runtime": 10984.2682,
      "train_tokens_per_second": 423.047
    },
    {
      "epoch": 1.9278769099262743,
      "grad_norm": 3.03288197517395,
      "learning_rate": 7.272727272727273e-07,
      "loss": 0.11324767023324966,
      "num_input_tokens_seen": 4647758,
      "step": 4511,
      "train_runtime": 10986.3413,
      "train_tokens_per_second": 423.049
    },
    {
      "epoch": 1.9283043060156,
      "grad_norm": 1.8207811117172241,
      "learning_rate": 7.229946524064172e-07,
      "loss": 0.06085740774869919,
      "num_input_tokens_seen": 4649322,
      "step": 4512,
      "train_runtime": 10988.5814,
      "train_tokens_per_second": 423.105
    },
    {
      "epoch": 1.9287317021049257,
      "grad_norm": 2.4698426723480225,
      "learning_rate": 7.18716577540107e-07,
      "loss": 0.07891259342432022,
      "num_input_tokens_seen": 4650230,
      "step": 4513,
      "train_runtime": 10990.7563,
      "train_tokens_per_second": 423.104
    },
    {
      "epoch": 1.9291590981942517,
      "grad_norm": 3.377413511276245,
      "learning_rate": 7.144385026737968e-07,
      "loss": 0.13608857989311218,
      "num_input_tokens_seen": 4651312,
      "step": 4514,
      "train_runtime": 10992.8335,
      "train_tokens_per_second": 423.122
    },
    {
      "epoch": 1.9295864942835772,
      "grad_norm": 2.521959066390991,
      "learning_rate": 7.101604278074867e-07,
      "loss": 0.08363817632198334,
      "num_input_tokens_seen": 4652318,
      "step": 4515,
      "train_runtime": 10994.9245,
      "train_tokens_per_second": 423.133
    },
    {
      "epoch": 1.9300138903729032,
      "grad_norm": 3.3962671756744385,
      "learning_rate": 7.058823529411766e-07,
      "loss": 0.06874531507492065,
      "num_input_tokens_seen": 4653000,
      "step": 4516,
      "train_runtime": 10996.9724,
      "train_tokens_per_second": 423.116
    },
    {
      "epoch": 1.930441286462229,
      "grad_norm": 2.5928187370300293,
      "learning_rate": 7.016042780748663e-07,
      "loss": 0.067188560962677,
      "num_input_tokens_seen": 4653932,
      "step": 4517,
      "train_runtime": 10999.097,
      "train_tokens_per_second": 423.119
    },
    {
      "epoch": 1.9308686825515546,
      "grad_norm": 2.1606643199920654,
      "learning_rate": 6.973262032085562e-07,
      "loss": 0.05929163843393326,
      "num_input_tokens_seen": 4654536,
      "step": 4518,
      "train_runtime": 11001.1042,
      "train_tokens_per_second": 423.097
    },
    {
      "epoch": 1.9312960786408806,
      "grad_norm": 2.681345224380493,
      "learning_rate": 6.930481283422461e-07,
      "loss": 0.0747632309794426,
      "num_input_tokens_seen": 4655626,
      "step": 4519,
      "train_runtime": 11003.2926,
      "train_tokens_per_second": 423.112
    },
    {
      "epoch": 1.9317234747302061,
      "grad_norm": 3.167240858078003,
      "learning_rate": 6.887700534759359e-07,
      "loss": 0.17355136573314667,
      "num_input_tokens_seen": 4656594,
      "step": 4520,
      "train_runtime": 11005.4029,
      "train_tokens_per_second": 423.119
    },
    {
      "epoch": 1.932150870819532,
      "grad_norm": 2.812575578689575,
      "learning_rate": 6.844919786096257e-07,
      "loss": 0.06282568722963333,
      "num_input_tokens_seen": 4657184,
      "step": 4521,
      "train_runtime": 11007.4516,
      "train_tokens_per_second": 423.094
    },
    {
      "epoch": 1.9325782669088578,
      "grad_norm": 2.9438235759735107,
      "learning_rate": 6.802139037433156e-07,
      "loss": 0.09269047528505325,
      "num_input_tokens_seen": 4658188,
      "step": 4522,
      "train_runtime": 11009.6054,
      "train_tokens_per_second": 423.102
    },
    {
      "epoch": 1.9330056629981835,
      "grad_norm": 3.1238341331481934,
      "learning_rate": 6.759358288770054e-07,
      "loss": 0.09006265550851822,
      "num_input_tokens_seen": 4659012,
      "step": 4523,
      "train_runtime": 11011.675,
      "train_tokens_per_second": 423.097
    },
    {
      "epoch": 1.9334330590875095,
      "grad_norm": 2.5435688495635986,
      "learning_rate": 6.716577540106953e-07,
      "loss": 0.06652931123971939,
      "num_input_tokens_seen": 4659986,
      "step": 4524,
      "train_runtime": 11013.7559,
      "train_tokens_per_second": 423.106
    },
    {
      "epoch": 1.933860455176835,
      "grad_norm": 2.0927414894104004,
      "learning_rate": 6.673796791443852e-07,
      "loss": 0.04113670065999031,
      "num_input_tokens_seen": 4660692,
      "step": 4525,
      "train_runtime": 11015.8221,
      "train_tokens_per_second": 423.091
    },
    {
      "epoch": 1.934287851266161,
      "grad_norm": 2.7394750118255615,
      "learning_rate": 6.631016042780749e-07,
      "loss": 0.08025995641946793,
      "num_input_tokens_seen": 4661526,
      "step": 4526,
      "train_runtime": 11017.8843,
      "train_tokens_per_second": 423.087
    },
    {
      "epoch": 1.9347152473554867,
      "grad_norm": 2.9899301528930664,
      "learning_rate": 6.588235294117648e-07,
      "loss": 0.12163376063108444,
      "num_input_tokens_seen": 4663384,
      "step": 4527,
      "train_runtime": 11020.181,
      "train_tokens_per_second": 423.168
    },
    {
      "epoch": 1.9351426434448125,
      "grad_norm": 2.172370672225952,
      "learning_rate": 6.545454545454547e-07,
      "loss": 0.10206068307161331,
      "num_input_tokens_seen": 4664762,
      "step": 4528,
      "train_runtime": 11022.3904,
      "train_tokens_per_second": 423.208
    },
    {
      "epoch": 1.9355700395341384,
      "grad_norm": 3.474428176879883,
      "learning_rate": 6.502673796791445e-07,
      "loss": 0.11561005562543869,
      "num_input_tokens_seen": 4665776,
      "step": 4529,
      "train_runtime": 11024.5592,
      "train_tokens_per_second": 423.217
    },
    {
      "epoch": 1.935997435623464,
      "grad_norm": 2.96338152885437,
      "learning_rate": 6.459893048128343e-07,
      "loss": 0.07063794136047363,
      "num_input_tokens_seen": 4666948,
      "step": 4530,
      "train_runtime": 11026.7072,
      "train_tokens_per_second": 423.24
    },
    {
      "epoch": 1.93642483171279,
      "grad_norm": 2.158540725708008,
      "learning_rate": 6.417112299465242e-07,
      "loss": 0.05454091727733612,
      "num_input_tokens_seen": 4668144,
      "step": 4531,
      "train_runtime": 11035.5306,
      "train_tokens_per_second": 423.01
    },
    {
      "epoch": 1.9368522278021156,
      "grad_norm": 2.188514232635498,
      "learning_rate": 6.374331550802139e-07,
      "loss": 0.05827419459819794,
      "num_input_tokens_seen": 4669384,
      "step": 4532,
      "train_runtime": 11037.6899,
      "train_tokens_per_second": 423.04
    },
    {
      "epoch": 1.9372796238914414,
      "grad_norm": 2.954690933227539,
      "learning_rate": 6.331550802139037e-07,
      "loss": 0.07166343927383423,
      "num_input_tokens_seen": 4670094,
      "step": 4533,
      "train_runtime": 11039.9946,
      "train_tokens_per_second": 423.016
    },
    {
      "epoch": 1.9377070199807673,
      "grad_norm": 2.2040185928344727,
      "learning_rate": 6.288770053475936e-07,
      "loss": 0.07868187129497528,
      "num_input_tokens_seen": 4671402,
      "step": 4534,
      "train_runtime": 11042.1638,
      "train_tokens_per_second": 423.051
    },
    {
      "epoch": 1.9381344160700928,
      "grad_norm": 2.6908469200134277,
      "learning_rate": 6.245989304812835e-07,
      "loss": 0.08612082898616791,
      "num_input_tokens_seen": 4672470,
      "step": 4535,
      "train_runtime": 11044.2612,
      "train_tokens_per_second": 423.068
    },
    {
      "epoch": 1.9385618121594188,
      "grad_norm": 3.2068042755126953,
      "learning_rate": 6.203208556149733e-07,
      "loss": 0.09664101898670197,
      "num_input_tokens_seen": 4673342,
      "step": 4536,
      "train_runtime": 11046.3287,
      "train_tokens_per_second": 423.067
    },
    {
      "epoch": 1.9389892082487445,
      "grad_norm": 2.803537368774414,
      "learning_rate": 6.160427807486632e-07,
      "loss": 0.10462618619203568,
      "num_input_tokens_seen": 4674028,
      "step": 4537,
      "train_runtime": 11048.3268,
      "train_tokens_per_second": 423.053
    },
    {
      "epoch": 1.9394166043380703,
      "grad_norm": 3.147202730178833,
      "learning_rate": 6.11764705882353e-07,
      "loss": 0.08873769640922546,
      "num_input_tokens_seen": 4674756,
      "step": 4538,
      "train_runtime": 11050.3684,
      "train_tokens_per_second": 423.041
    },
    {
      "epoch": 1.9398440004273962,
      "grad_norm": 3.1199429035186768,
      "learning_rate": 6.074866310160428e-07,
      "loss": 0.09317778795957565,
      "num_input_tokens_seen": 4675534,
      "step": 4539,
      "train_runtime": 11052.4004,
      "train_tokens_per_second": 423.033
    },
    {
      "epoch": 1.9402713965167218,
      "grad_norm": 4.304396629333496,
      "learning_rate": 6.032085561497327e-07,
      "loss": 0.09714939445257187,
      "num_input_tokens_seen": 4676648,
      "step": 4540,
      "train_runtime": 11054.4922,
      "train_tokens_per_second": 423.054
    },
    {
      "epoch": 1.9406987926060477,
      "grad_norm": 2.638195276260376,
      "learning_rate": 5.989304812834224e-07,
      "loss": 0.08140614628791809,
      "num_input_tokens_seen": 4677624,
      "step": 4541,
      "train_runtime": 11056.6009,
      "train_tokens_per_second": 423.062
    },
    {
      "epoch": 1.9411261886953735,
      "grad_norm": 2.2924797534942627,
      "learning_rate": 5.946524064171123e-07,
      "loss": 0.05866424739360809,
      "num_input_tokens_seen": 4678680,
      "step": 4542,
      "train_runtime": 11058.6855,
      "train_tokens_per_second": 423.077
    },
    {
      "epoch": 1.9415535847846992,
      "grad_norm": 3.5593459606170654,
      "learning_rate": 5.903743315508022e-07,
      "loss": 0.11644919216632843,
      "num_input_tokens_seen": 4679828,
      "step": 4543,
      "train_runtime": 11060.7981,
      "train_tokens_per_second": 423.1
    },
    {
      "epoch": 1.9419809808740252,
      "grad_norm": 3.889326810836792,
      "learning_rate": 5.86096256684492e-07,
      "loss": 0.07622142136096954,
      "num_input_tokens_seen": 4680696,
      "step": 4544,
      "train_runtime": 11062.8541,
      "train_tokens_per_second": 423.1
    },
    {
      "epoch": 1.9424083769633507,
      "grad_norm": 3.3499138355255127,
      "learning_rate": 5.818181818181819e-07,
      "loss": 0.08571440726518631,
      "num_input_tokens_seen": 4681318,
      "step": 4545,
      "train_runtime": 11064.8713,
      "train_tokens_per_second": 423.079
    },
    {
      "epoch": 1.9428357730526766,
      "grad_norm": 3.379067897796631,
      "learning_rate": 5.775401069518717e-07,
      "loss": 0.06637506186962128,
      "num_input_tokens_seen": 4682014,
      "step": 4546,
      "train_runtime": 11066.8914,
      "train_tokens_per_second": 423.065
    },
    {
      "epoch": 1.9432631691420024,
      "grad_norm": 2.3626327514648438,
      "learning_rate": 5.732620320855615e-07,
      "loss": 0.07608707249164581,
      "num_input_tokens_seen": 4683086,
      "step": 4547,
      "train_runtime": 11069.0062,
      "train_tokens_per_second": 423.081
    },
    {
      "epoch": 1.943690565231328,
      "grad_norm": 2.6304876804351807,
      "learning_rate": 5.689839572192514e-07,
      "loss": 0.0915163904428482,
      "num_input_tokens_seen": 4684090,
      "step": 4548,
      "train_runtime": 11071.0628,
      "train_tokens_per_second": 423.093
    },
    {
      "epoch": 1.944117961320654,
      "grad_norm": 2.315802574157715,
      "learning_rate": 5.647058823529413e-07,
      "loss": 0.05967039614915848,
      "num_input_tokens_seen": 4685268,
      "step": 4549,
      "train_runtime": 11073.1943,
      "train_tokens_per_second": 423.118
    },
    {
      "epoch": 1.9445453574099796,
      "grad_norm": 3.4916250705718994,
      "learning_rate": 5.60427807486631e-07,
      "loss": 0.07737503945827484,
      "num_input_tokens_seen": 4686228,
      "step": 4550,
      "train_runtime": 11075.251,
      "train_tokens_per_second": 423.126
    },
    {
      "epoch": 1.9449727534993055,
      "grad_norm": 2.7871346473693848,
      "learning_rate": 5.561497326203209e-07,
      "loss": 0.08479427546262741,
      "num_input_tokens_seen": 4687436,
      "step": 4551,
      "train_runtime": 11077.426,
      "train_tokens_per_second": 423.152
    },
    {
      "epoch": 1.9454001495886313,
      "grad_norm": 2.6285719871520996,
      "learning_rate": 5.518716577540107e-07,
      "loss": 0.09612537920475006,
      "num_input_tokens_seen": 4689176,
      "step": 4552,
      "train_runtime": 11079.6737,
      "train_tokens_per_second": 423.223
    },
    {
      "epoch": 1.945827545677957,
      "grad_norm": 3.9141738414764404,
      "learning_rate": 5.475935828877006e-07,
      "loss": 0.08281700313091278,
      "num_input_tokens_seen": 4689724,
      "step": 4553,
      "train_runtime": 11081.7164,
      "train_tokens_per_second": 423.195
    },
    {
      "epoch": 1.946254941767283,
      "grad_norm": 2.7731518745422363,
      "learning_rate": 5.433155080213904e-07,
      "loss": 0.08834431320428848,
      "num_input_tokens_seen": 4690788,
      "step": 4554,
      "train_runtime": 11083.8132,
      "train_tokens_per_second": 423.211
    },
    {
      "epoch": 1.9466823378566085,
      "grad_norm": 3.006840229034424,
      "learning_rate": 5.390374331550803e-07,
      "loss": 0.1024986058473587,
      "num_input_tokens_seen": 4691868,
      "step": 4555,
      "train_runtime": 11085.9127,
      "train_tokens_per_second": 423.228
    },
    {
      "epoch": 1.9471097339459345,
      "grad_norm": 2.871394157409668,
      "learning_rate": 5.347593582887701e-07,
      "loss": 0.08783483505249023,
      "num_input_tokens_seen": 4692568,
      "step": 4556,
      "train_runtime": 11088.0213,
      "train_tokens_per_second": 423.211
    },
    {
      "epoch": 1.9475371300352602,
      "grad_norm": 2.776301383972168,
      "learning_rate": 5.304812834224599e-07,
      "loss": 0.10947674512863159,
      "num_input_tokens_seen": 4693474,
      "step": 4557,
      "train_runtime": 11090.1009,
      "train_tokens_per_second": 423.213
    },
    {
      "epoch": 1.947964526124586,
      "grad_norm": 2.1903395652770996,
      "learning_rate": 5.262032085561498e-07,
      "loss": 0.07248228788375854,
      "num_input_tokens_seen": 4694508,
      "step": 4558,
      "train_runtime": 11092.3078,
      "train_tokens_per_second": 423.222
    },
    {
      "epoch": 1.9483919222139119,
      "grad_norm": 4.3749918937683105,
      "learning_rate": 5.219251336898397e-07,
      "loss": 0.11831034719944,
      "num_input_tokens_seen": 4695362,
      "step": 4559,
      "train_runtime": 11094.3765,
      "train_tokens_per_second": 423.22
    },
    {
      "epoch": 1.9488193183032374,
      "grad_norm": 2.441965103149414,
      "learning_rate": 5.176470588235294e-07,
      "loss": 0.0629957839846611,
      "num_input_tokens_seen": 4696272,
      "step": 4560,
      "train_runtime": 11096.4713,
      "train_tokens_per_second": 423.222
    },
    {
      "epoch": 1.9492467143925634,
      "grad_norm": 4.911445617675781,
      "learning_rate": 5.133689839572193e-07,
      "loss": 0.1619553565979004,
      "num_input_tokens_seen": 4697098,
      "step": 4561,
      "train_runtime": 11104.9215,
      "train_tokens_per_second": 422.974
    },
    {
      "epoch": 1.949674110481889,
      "grad_norm": 2.728515863418579,
      "learning_rate": 5.090909090909092e-07,
      "loss": 0.09682068973779678,
      "num_input_tokens_seen": 4698066,
      "step": 4562,
      "train_runtime": 11107.187,
      "train_tokens_per_second": 422.975
    },
    {
      "epoch": 1.9501015065712148,
      "grad_norm": 1.9767149686813354,
      "learning_rate": 5.04812834224599e-07,
      "loss": 0.05832261964678764,
      "num_input_tokens_seen": 4699038,
      "step": 4563,
      "train_runtime": 11109.3487,
      "train_tokens_per_second": 422.981
    },
    {
      "epoch": 1.9505289026605408,
      "grad_norm": 3.2443628311157227,
      "learning_rate": 5.005347593582888e-07,
      "loss": 0.13833874464035034,
      "num_input_tokens_seen": 4699864,
      "step": 4564,
      "train_runtime": 11111.4391,
      "train_tokens_per_second": 422.975
    },
    {
      "epoch": 1.9509562987498663,
      "grad_norm": 2.99072265625,
      "learning_rate": 4.962566844919787e-07,
      "loss": 0.08302243053913116,
      "num_input_tokens_seen": 4700840,
      "step": 4565,
      "train_runtime": 11113.6501,
      "train_tokens_per_second": 422.979
    },
    {
      "epoch": 1.9513836948391923,
      "grad_norm": 3.036231517791748,
      "learning_rate": 4.919786096256684e-07,
      "loss": 0.09491357952356339,
      "num_input_tokens_seen": 4701826,
      "step": 4566,
      "train_runtime": 11115.8221,
      "train_tokens_per_second": 422.985
    },
    {
      "epoch": 1.951811090928518,
      "grad_norm": 2.312161684036255,
      "learning_rate": 4.877005347593583e-07,
      "loss": 0.08124493062496185,
      "num_input_tokens_seen": 4703148,
      "step": 4567,
      "train_runtime": 11118.0297,
      "train_tokens_per_second": 423.02
    },
    {
      "epoch": 1.9522384870178437,
      "grad_norm": 2.8678829669952393,
      "learning_rate": 4.834224598930482e-07,
      "loss": 0.12963111698627472,
      "num_input_tokens_seen": 4704396,
      "step": 4568,
      "train_runtime": 11120.1772,
      "train_tokens_per_second": 423.05
    },
    {
      "epoch": 1.9526658831071697,
      "grad_norm": 2.644441604614258,
      "learning_rate": 4.79144385026738e-07,
      "loss": 0.10339433699846268,
      "num_input_tokens_seen": 4705448,
      "step": 4569,
      "train_runtime": 11122.3089,
      "train_tokens_per_second": 423.064
    },
    {
      "epoch": 1.9530932791964952,
      "grad_norm": 2.629354476928711,
      "learning_rate": 4.7486631016042785e-07,
      "loss": 0.0813981220126152,
      "num_input_tokens_seen": 4706630,
      "step": 4570,
      "train_runtime": 11124.4823,
      "train_tokens_per_second": 423.088
    },
    {
      "epoch": 1.9535206752858212,
      "grad_norm": 2.3001961708068848,
      "learning_rate": 4.7058823529411767e-07,
      "loss": 0.06940969079732895,
      "num_input_tokens_seen": 4707594,
      "step": 4571,
      "train_runtime": 11126.5829,
      "train_tokens_per_second": 423.094
    },
    {
      "epoch": 1.953948071375147,
      "grad_norm": 3.281641960144043,
      "learning_rate": 4.6631016042780754e-07,
      "loss": 0.14235064387321472,
      "num_input_tokens_seen": 4708502,
      "step": 4572,
      "train_runtime": 11128.7139,
      "train_tokens_per_second": 423.095
    },
    {
      "epoch": 1.9543754674644727,
      "grad_norm": 2.15511155128479,
      "learning_rate": 4.6203208556149735e-07,
      "loss": 0.08907820284366608,
      "num_input_tokens_seen": 4709928,
      "step": 4573,
      "train_runtime": 11130.8898,
      "train_tokens_per_second": 423.14
    },
    {
      "epoch": 1.9548028635537986,
      "grad_norm": 2.717864513397217,
      "learning_rate": 4.577540106951872e-07,
      "loss": 0.08128479868173599,
      "num_input_tokens_seen": 4711620,
      "step": 4574,
      "train_runtime": 11133.1898,
      "train_tokens_per_second": 423.205
    },
    {
      "epoch": 1.9552302596431241,
      "grad_norm": 3.2088775634765625,
      "learning_rate": 4.5347593582887704e-07,
      "loss": 0.10385055840015411,
      "num_input_tokens_seen": 4712450,
      "step": 4575,
      "train_runtime": 11135.2174,
      "train_tokens_per_second": 423.202
    },
    {
      "epoch": 1.95565765573245,
      "grad_norm": 3.097871780395508,
      "learning_rate": 4.491978609625669e-07,
      "loss": 0.06359519809484482,
      "num_input_tokens_seen": 4713134,
      "step": 4576,
      "train_runtime": 11137.2401,
      "train_tokens_per_second": 423.187
    },
    {
      "epoch": 1.9560850518217758,
      "grad_norm": 2.547149896621704,
      "learning_rate": 4.449197860962567e-07,
      "loss": 0.08061902970075607,
      "num_input_tokens_seen": 4714564,
      "step": 4577,
      "train_runtime": 11139.4441,
      "train_tokens_per_second": 423.232
    },
    {
      "epoch": 1.9565124479111016,
      "grad_norm": 1.7962956428527832,
      "learning_rate": 4.406417112299466e-07,
      "loss": 0.0759037435054779,
      "num_input_tokens_seen": 4716236,
      "step": 4578,
      "train_runtime": 11141.7643,
      "train_tokens_per_second": 423.293
    },
    {
      "epoch": 1.9569398440004275,
      "grad_norm": 3.3110880851745605,
      "learning_rate": 4.363636363636364e-07,
      "loss": 0.07120725512504578,
      "num_input_tokens_seen": 4717074,
      "step": 4579,
      "train_runtime": 11143.8633,
      "train_tokens_per_second": 423.289
    },
    {
      "epoch": 1.957367240089753,
      "grad_norm": 2.2016069889068604,
      "learning_rate": 4.3208556149732623e-07,
      "loss": 0.06737089902162552,
      "num_input_tokens_seen": 4717898,
      "step": 4580,
      "train_runtime": 11145.9655,
      "train_tokens_per_second": 423.283
    },
    {
      "epoch": 1.957794636179079,
      "grad_norm": 3.2486517429351807,
      "learning_rate": 4.2780748663101604e-07,
      "loss": 0.1407386064529419,
      "num_input_tokens_seen": 4718616,
      "step": 4581,
      "train_runtime": 11148.0762,
      "train_tokens_per_second": 423.267
    },
    {
      "epoch": 1.9582220322684047,
      "grad_norm": 4.887596130371094,
      "learning_rate": 4.235294117647059e-07,
      "loss": 0.12462133169174194,
      "num_input_tokens_seen": 4719342,
      "step": 4582,
      "train_runtime": 11150.1627,
      "train_tokens_per_second": 423.253
    },
    {
      "epoch": 1.9586494283577305,
      "grad_norm": 10.160305976867676,
      "learning_rate": 4.1925133689839573e-07,
      "loss": 0.1265372633934021,
      "num_input_tokens_seen": 4720046,
      "step": 4583,
      "train_runtime": 11152.2086,
      "train_tokens_per_second": 423.239
    },
    {
      "epoch": 1.9590768244470564,
      "grad_norm": 2.3367831707000732,
      "learning_rate": 4.149732620320856e-07,
      "loss": 0.10565128922462463,
      "num_input_tokens_seen": 4721688,
      "step": 4584,
      "train_runtime": 11154.4968,
      "train_tokens_per_second": 423.299
    },
    {
      "epoch": 1.959504220536382,
      "grad_norm": 2.0345544815063477,
      "learning_rate": 4.106951871657754e-07,
      "loss": 0.0666138231754303,
      "num_input_tokens_seen": 4722754,
      "step": 4585,
      "train_runtime": 11156.6723,
      "train_tokens_per_second": 423.312
    },
    {
      "epoch": 1.959931616625708,
      "grad_norm": 2.5558927059173584,
      "learning_rate": 4.064171122994653e-07,
      "loss": 0.07684600353240967,
      "num_input_tokens_seen": 4723672,
      "step": 4586,
      "train_runtime": 11158.7778,
      "train_tokens_per_second": 423.314
    },
    {
      "epoch": 1.9603590127150337,
      "grad_norm": 2.3063933849334717,
      "learning_rate": 4.021390374331551e-07,
      "loss": 0.11385758966207504,
      "num_input_tokens_seen": 4724836,
      "step": 4587,
      "train_runtime": 11160.9366,
      "train_tokens_per_second": 423.337
    },
    {
      "epoch": 1.9607864088043594,
      "grad_norm": 2.47609281539917,
      "learning_rate": 3.9786096256684497e-07,
      "loss": 0.05028097331523895,
      "num_input_tokens_seen": 4725606,
      "step": 4588,
      "train_runtime": 11163.0262,
      "train_tokens_per_second": 423.327
    },
    {
      "epoch": 1.9612138048936854,
      "grad_norm": 3.7131035327911377,
      "learning_rate": 3.935828877005348e-07,
      "loss": 0.08615584671497345,
      "num_input_tokens_seen": 4726378,
      "step": 4589,
      "train_runtime": 11165.0942,
      "train_tokens_per_second": 423.317
    },
    {
      "epoch": 1.9616412009830109,
      "grad_norm": 2.3089146614074707,
      "learning_rate": 3.8930481283422465e-07,
      "loss": 0.07162284851074219,
      "num_input_tokens_seen": 4727362,
      "step": 4590,
      "train_runtime": 11167.1902,
      "train_tokens_per_second": 423.326
    },
    {
      "epoch": 1.9620685970723368,
      "grad_norm": 2.313880443572998,
      "learning_rate": 3.8502673796791447e-07,
      "loss": 0.06751394271850586,
      "num_input_tokens_seen": 4728912,
      "step": 4591,
      "train_runtime": 11175.7638,
      "train_tokens_per_second": 423.14
    },
    {
      "epoch": 1.9624959931616626,
      "grad_norm": 2.4609978199005127,
      "learning_rate": 3.8074866310160434e-07,
      "loss": 0.08920436352491379,
      "num_input_tokens_seen": 4729916,
      "step": 4592,
      "train_runtime": 11177.919,
      "train_tokens_per_second": 423.148
    },
    {
      "epoch": 1.9629233892509883,
      "grad_norm": 3.2225425243377686,
      "learning_rate": 3.764705882352941e-07,
      "loss": 0.11080729961395264,
      "num_input_tokens_seen": 4730828,
      "step": 4593,
      "train_runtime": 11180.0022,
      "train_tokens_per_second": 423.151
    },
    {
      "epoch": 1.9633507853403143,
      "grad_norm": 2.332368850708008,
      "learning_rate": 3.7219251336898397e-07,
      "loss": 0.09047398716211319,
      "num_input_tokens_seen": 4731948,
      "step": 4594,
      "train_runtime": 11182.1291,
      "train_tokens_per_second": 423.171
    },
    {
      "epoch": 1.9637781814296398,
      "grad_norm": 3.1621766090393066,
      "learning_rate": 3.679144385026738e-07,
      "loss": 0.054371293634176254,
      "num_input_tokens_seen": 4732806,
      "step": 4595,
      "train_runtime": 11184.2209,
      "train_tokens_per_second": 423.168
    },
    {
      "epoch": 1.9642055775189657,
      "grad_norm": 2.657390594482422,
      "learning_rate": 3.6363636363636366e-07,
      "loss": 0.10414905101060867,
      "num_input_tokens_seen": 4733966,
      "step": 4596,
      "train_runtime": 11186.3025,
      "train_tokens_per_second": 423.193
    },
    {
      "epoch": 1.9646329736082915,
      "grad_norm": 1.9876104593276978,
      "learning_rate": 3.593582887700535e-07,
      "loss": 0.05372269079089165,
      "num_input_tokens_seen": 4734720,
      "step": 4597,
      "train_runtime": 11188.3079,
      "train_tokens_per_second": 423.185
    },
    {
      "epoch": 1.9650603696976172,
      "grad_norm": 1.8673558235168457,
      "learning_rate": 3.5508021390374334e-07,
      "loss": 0.0764034241437912,
      "num_input_tokens_seen": 4736286,
      "step": 4598,
      "train_runtime": 11190.5909,
      "train_tokens_per_second": 423.238
    },
    {
      "epoch": 1.9654877657869432,
      "grad_norm": 4.007637023925781,
      "learning_rate": 3.5080213903743316e-07,
      "loss": 0.08312540501356125,
      "num_input_tokens_seen": 4737088,
      "step": 4599,
      "train_runtime": 11192.6458,
      "train_tokens_per_second": 423.232
    },
    {
      "epoch": 1.9659151618762687,
      "grad_norm": 2.573748826980591,
      "learning_rate": 3.4652406417112303e-07,
      "loss": 0.11682626605033875,
      "num_input_tokens_seen": 4738756,
      "step": 4600,
      "train_runtime": 11194.8981,
      "train_tokens_per_second": 423.296
    },
    {
      "epoch": 1.9663425579655947,
      "grad_norm": 2.9857099056243896,
      "learning_rate": 3.4224598930481285e-07,
      "loss": 0.09980595111846924,
      "num_input_tokens_seen": 4739616,
      "step": 4601,
      "train_runtime": 11196.9875,
      "train_tokens_per_second": 423.294
    },
    {
      "epoch": 1.9667699540549204,
      "grad_norm": 2.898076295852661,
      "learning_rate": 3.379679144385027e-07,
      "loss": 0.10526564717292786,
      "num_input_tokens_seen": 4740538,
      "step": 4602,
      "train_runtime": 11199.1576,
      "train_tokens_per_second": 423.294
    },
    {
      "epoch": 1.9671973501442461,
      "grad_norm": 2.2916786670684814,
      "learning_rate": 3.336898395721926e-07,
      "loss": 0.04620625823736191,
      "num_input_tokens_seen": 4741346,
      "step": 4603,
      "train_runtime": 11201.1387,
      "train_tokens_per_second": 423.291
    },
    {
      "epoch": 1.967624746233572,
      "grad_norm": 4.059272289276123,
      "learning_rate": 3.294117647058824e-07,
      "loss": 0.12072921544313431,
      "num_input_tokens_seen": 4742578,
      "step": 4604,
      "train_runtime": 11203.2772,
      "train_tokens_per_second": 423.321
    },
    {
      "epoch": 1.9680521423228976,
      "grad_norm": 2.514986038208008,
      "learning_rate": 3.2513368983957227e-07,
      "loss": 0.0733695849776268,
      "num_input_tokens_seen": 4743362,
      "step": 4605,
      "train_runtime": 11205.3568,
      "train_tokens_per_second": 423.312
    },
    {
      "epoch": 1.9684795384122236,
      "grad_norm": 1.5780837535858154,
      "learning_rate": 3.208556149732621e-07,
      "loss": 0.03916705772280693,
      "num_input_tokens_seen": 4744378,
      "step": 4606,
      "train_runtime": 11207.4514,
      "train_tokens_per_second": 423.324
    },
    {
      "epoch": 1.9689069345015493,
      "grad_norm": 2.7726175785064697,
      "learning_rate": 3.1657754010695185e-07,
      "loss": 0.0669010579586029,
      "num_input_tokens_seen": 4744958,
      "step": 4607,
      "train_runtime": 11209.445,
      "train_tokens_per_second": 423.3
    },
    {
      "epoch": 1.969334330590875,
      "grad_norm": 3.0967917442321777,
      "learning_rate": 3.1229946524064177e-07,
      "loss": 0.06793845444917679,
      "num_input_tokens_seen": 4745776,
      "step": 4608,
      "train_runtime": 11211.5288,
      "train_tokens_per_second": 423.294
    },
    {
      "epoch": 1.969761726680201,
      "grad_norm": 2.4741933345794678,
      "learning_rate": 3.080213903743316e-07,
      "loss": 0.07393999397754669,
      "num_input_tokens_seen": 4746654,
      "step": 4609,
      "train_runtime": 11213.6426,
      "train_tokens_per_second": 423.293
    },
    {
      "epoch": 1.9701891227695265,
      "grad_norm": 2.536719560623169,
      "learning_rate": 3.037433155080214e-07,
      "loss": 0.08613823354244232,
      "num_input_tokens_seen": 4747758,
      "step": 4610,
      "train_runtime": 11215.7791,
      "train_tokens_per_second": 423.311
    },
    {
      "epoch": 1.9706165188588525,
      "grad_norm": 6.438047885894775,
      "learning_rate": 2.994652406417112e-07,
      "loss": 0.10467591881752014,
      "num_input_tokens_seen": 4748890,
      "step": 4611,
      "train_runtime": 11217.9143,
      "train_tokens_per_second": 423.331
    },
    {
      "epoch": 1.9710439149481782,
      "grad_norm": 2.42236590385437,
      "learning_rate": 2.951871657754011e-07,
      "loss": 0.07271308451890945,
      "num_input_tokens_seen": 4749970,
      "step": 4612,
      "train_runtime": 11220.1451,
      "train_tokens_per_second": 423.343
    },
    {
      "epoch": 1.971471311037504,
      "grad_norm": 2.6034903526306152,
      "learning_rate": 2.9090909090909096e-07,
      "loss": 0.06916365027427673,
      "num_input_tokens_seen": 4750858,
      "step": 4613,
      "train_runtime": 11222.1985,
      "train_tokens_per_second": 423.345
    },
    {
      "epoch": 1.97189870712683,
      "grad_norm": 2.8886992931365967,
      "learning_rate": 2.866310160427808e-07,
      "loss": 0.09915220737457275,
      "num_input_tokens_seen": 4752256,
      "step": 4614,
      "train_runtime": 11224.4925,
      "train_tokens_per_second": 423.383
    },
    {
      "epoch": 1.9723261032161554,
      "grad_norm": 2.10439395904541,
      "learning_rate": 2.8235294117647064e-07,
      "loss": 0.06319509446620941,
      "num_input_tokens_seen": 4753160,
      "step": 4615,
      "train_runtime": 11226.5423,
      "train_tokens_per_second": 423.386
    },
    {
      "epoch": 1.9727534993054814,
      "grad_norm": 2.8660531044006348,
      "learning_rate": 2.7807486631016046e-07,
      "loss": 0.11509943753480911,
      "num_input_tokens_seen": 4754174,
      "step": 4616,
      "train_runtime": 11228.6311,
      "train_tokens_per_second": 423.397
    },
    {
      "epoch": 1.9731808953948071,
      "grad_norm": 2.3386032581329346,
      "learning_rate": 2.737967914438503e-07,
      "loss": 0.09830325841903687,
      "num_input_tokens_seen": 4755330,
      "step": 4617,
      "train_runtime": 11230.7667,
      "train_tokens_per_second": 423.42
    },
    {
      "epoch": 1.9736082914841329,
      "grad_norm": 3.7125515937805176,
      "learning_rate": 2.6951871657754015e-07,
      "loss": 0.13523618876934052,
      "num_input_tokens_seen": 4756462,
      "step": 4618,
      "train_runtime": 11232.8546,
      "train_tokens_per_second": 423.442
    },
    {
      "epoch": 1.9740356875734588,
      "grad_norm": 3.787569046020508,
      "learning_rate": 2.6524064171122996e-07,
      "loss": 0.08793338388204575,
      "num_input_tokens_seen": 4757774,
      "step": 4619,
      "train_runtime": 11235.0325,
      "train_tokens_per_second": 423.477
    },
    {
      "epoch": 1.9744630836627843,
      "grad_norm": 2.661362648010254,
      "learning_rate": 2.6096256684491983e-07,
      "loss": 0.08164241909980774,
      "num_input_tokens_seen": 4758754,
      "step": 4620,
      "train_runtime": 11237.1489,
      "train_tokens_per_second": 423.484
    },
    {
      "epoch": 1.9748904797521103,
      "grad_norm": 2.86749005317688,
      "learning_rate": 2.5668449197860965e-07,
      "loss": 0.08064473420381546,
      "num_input_tokens_seen": 4759750,
      "step": 4621,
      "train_runtime": 11245.6235,
      "train_tokens_per_second": 423.254
    },
    {
      "epoch": 1.975317875841436,
      "grad_norm": 2.74626088142395,
      "learning_rate": 2.524064171122995e-07,
      "loss": 0.08887047320604324,
      "num_input_tokens_seen": 4760638,
      "step": 4622,
      "train_runtime": 11247.6435,
      "train_tokens_per_second": 423.256
    },
    {
      "epoch": 1.9757452719307618,
      "grad_norm": 2.620147466659546,
      "learning_rate": 2.4812834224598933e-07,
      "loss": 0.062282875180244446,
      "num_input_tokens_seen": 4761450,
      "step": 4623,
      "train_runtime": 11249.7593,
      "train_tokens_per_second": 423.249
    },
    {
      "epoch": 1.9761726680200877,
      "grad_norm": 2.8994674682617188,
      "learning_rate": 2.4385026737967915e-07,
      "loss": 0.08498292416334152,
      "num_input_tokens_seen": 4762392,
      "step": 4624,
      "train_runtime": 11251.8125,
      "train_tokens_per_second": 423.256
    },
    {
      "epoch": 1.9766000641094132,
      "grad_norm": 3.430529832839966,
      "learning_rate": 2.39572192513369e-07,
      "loss": 0.0841522216796875,
      "num_input_tokens_seen": 4763160,
      "step": 4625,
      "train_runtime": 11253.8374,
      "train_tokens_per_second": 423.248
    },
    {
      "epoch": 1.9770274601987392,
      "grad_norm": 3.1292967796325684,
      "learning_rate": 2.3529411764705883e-07,
      "loss": 0.08050095289945602,
      "num_input_tokens_seen": 4763968,
      "step": 4626,
      "train_runtime": 11255.887,
      "train_tokens_per_second": 423.242
    },
    {
      "epoch": 1.977454856288065,
      "grad_norm": 1.4452532529830933,
      "learning_rate": 2.3101604278074868e-07,
      "loss": 0.08450129628181458,
      "num_input_tokens_seen": 4768176,
      "step": 4627,
      "train_runtime": 11258.714,
      "train_tokens_per_second": 423.51
    },
    {
      "epoch": 1.9778822523773907,
      "grad_norm": 3.6390323638916016,
      "learning_rate": 2.2673796791443852e-07,
      "loss": 0.09862130135297775,
      "num_input_tokens_seen": 4768886,
      "step": 4628,
      "train_runtime": 11260.8063,
      "train_tokens_per_second": 423.494
    },
    {
      "epoch": 1.9783096484667166,
      "grad_norm": 2.7303106784820557,
      "learning_rate": 2.2245989304812836e-07,
      "loss": 0.09639957547187805,
      "num_input_tokens_seen": 4769916,
      "step": 4629,
      "train_runtime": 11262.8929,
      "train_tokens_per_second": 423.507
    },
    {
      "epoch": 1.9787370445560422,
      "grad_norm": 3.2974090576171875,
      "learning_rate": 2.181818181818182e-07,
      "loss": 0.12696050107479095,
      "num_input_tokens_seen": 4771206,
      "step": 4630,
      "train_runtime": 11265.0248,
      "train_tokens_per_second": 423.542
    },
    {
      "epoch": 1.9791644406453681,
      "grad_norm": 4.019373416900635,
      "learning_rate": 2.1390374331550802e-07,
      "loss": 0.0799451470375061,
      "num_input_tokens_seen": 4771916,
      "step": 4631,
      "train_runtime": 11267.0052,
      "train_tokens_per_second": 423.53
    },
    {
      "epoch": 1.9795918367346939,
      "grad_norm": 2.434537887573242,
      "learning_rate": 2.0962566844919786e-07,
      "loss": 0.09077548235654831,
      "num_input_tokens_seen": 4772878,
      "step": 4632,
      "train_runtime": 11269.1042,
      "train_tokens_per_second": 423.537
    },
    {
      "epoch": 1.9800192328240196,
      "grad_norm": 3.5770862102508545,
      "learning_rate": 2.053475935828877e-07,
      "loss": 0.08833125978708267,
      "num_input_tokens_seen": 4773954,
      "step": 4633,
      "train_runtime": 11271.3764,
      "train_tokens_per_second": 423.547
    },
    {
      "epoch": 1.9804466289133456,
      "grad_norm": 2.696756362915039,
      "learning_rate": 2.0106951871657755e-07,
      "loss": 0.0912579596042633,
      "num_input_tokens_seen": 4775096,
      "step": 4634,
      "train_runtime": 11273.5026,
      "train_tokens_per_second": 423.568
    },
    {
      "epoch": 1.980874025002671,
      "grad_norm": 2.3793985843658447,
      "learning_rate": 1.967914438502674e-07,
      "loss": 0.05683675408363342,
      "num_input_tokens_seen": 4775964,
      "step": 4635,
      "train_runtime": 11275.5612,
      "train_tokens_per_second": 423.568
    },
    {
      "epoch": 1.981301421091997,
      "grad_norm": 3.1771419048309326,
      "learning_rate": 1.9251336898395724e-07,
      "loss": 0.07348252832889557,
      "num_input_tokens_seen": 4776628,
      "step": 4636,
      "train_runtime": 11277.5758,
      "train_tokens_per_second": 423.551
    },
    {
      "epoch": 1.9817288171813228,
      "grad_norm": 2.751821756362915,
      "learning_rate": 1.8823529411764705e-07,
      "loss": 0.088255375623703,
      "num_input_tokens_seen": 4777696,
      "step": 4637,
      "train_runtime": 11279.6514,
      "train_tokens_per_second": 423.568
    },
    {
      "epoch": 1.9821562132706485,
      "grad_norm": 4.200915813446045,
      "learning_rate": 1.839572192513369e-07,
      "loss": 0.10017291456460953,
      "num_input_tokens_seen": 4778442,
      "step": 4638,
      "train_runtime": 11281.6723,
      "train_tokens_per_second": 423.558
    },
    {
      "epoch": 1.9825836093599745,
      "grad_norm": 2.176443338394165,
      "learning_rate": 1.7967914438502674e-07,
      "loss": 0.08208013325929642,
      "num_input_tokens_seen": 4779564,
      "step": 4639,
      "train_runtime": 11283.7863,
      "train_tokens_per_second": 423.578
    },
    {
      "epoch": 1.9830110054493,
      "grad_norm": 2.890488386154175,
      "learning_rate": 1.7540106951871658e-07,
      "loss": 0.08005963265895844,
      "num_input_tokens_seen": 4780338,
      "step": 4640,
      "train_runtime": 11285.8429,
      "train_tokens_per_second": 423.569
    },
    {
      "epoch": 1.983438401538626,
      "grad_norm": 1.91335928440094,
      "learning_rate": 1.7112299465240642e-07,
      "loss": 0.05428942292928696,
      "num_input_tokens_seen": 4781518,
      "step": 4641,
      "train_runtime": 11287.9535,
      "train_tokens_per_second": 423.595
    },
    {
      "epoch": 1.9838657976279517,
      "grad_norm": 3.2447330951690674,
      "learning_rate": 1.668449197860963e-07,
      "loss": 0.06744910031557083,
      "num_input_tokens_seen": 4782254,
      "step": 4642,
      "train_runtime": 11289.9807,
      "train_tokens_per_second": 423.584
    },
    {
      "epoch": 1.9842931937172774,
      "grad_norm": 2.5188162326812744,
      "learning_rate": 1.6256684491978613e-07,
      "loss": 0.0847804918885231,
      "num_input_tokens_seen": 4783670,
      "step": 4643,
      "train_runtime": 11292.1256,
      "train_tokens_per_second": 423.629
    },
    {
      "epoch": 1.9847205898066034,
      "grad_norm": 2.0365638732910156,
      "learning_rate": 1.5828877005347592e-07,
      "loss": 0.06964364647865295,
      "num_input_tokens_seen": 4784632,
      "step": 4644,
      "train_runtime": 11294.1707,
      "train_tokens_per_second": 423.637
    },
    {
      "epoch": 1.985147985895929,
      "grad_norm": 3.07765531539917,
      "learning_rate": 1.540106951871658e-07,
      "loss": 0.10357463359832764,
      "num_input_tokens_seen": 4785592,
      "step": 4645,
      "train_runtime": 11296.2335,
      "train_tokens_per_second": 423.645
    },
    {
      "epoch": 1.9855753819852549,
      "grad_norm": 2.394017219543457,
      "learning_rate": 1.497326203208556e-07,
      "loss": 0.10176961123943329,
      "num_input_tokens_seen": 4786746,
      "step": 4646,
      "train_runtime": 11298.385,
      "train_tokens_per_second": 423.666
    },
    {
      "epoch": 1.9860027780745806,
      "grad_norm": 2.5691380500793457,
      "learning_rate": 1.4545454545454548e-07,
      "loss": 0.08107537776231766,
      "num_input_tokens_seen": 4788126,
      "step": 4647,
      "train_runtime": 11300.5954,
      "train_tokens_per_second": 423.706
    },
    {
      "epoch": 1.9864301741639063,
      "grad_norm": 2.3721048831939697,
      "learning_rate": 1.4117647058823532e-07,
      "loss": 0.0785323977470398,
      "num_input_tokens_seen": 4789054,
      "step": 4648,
      "train_runtime": 11302.6587,
      "train_tokens_per_second": 423.71
    },
    {
      "epoch": 1.9868575702532323,
      "grad_norm": 2.7213387489318848,
      "learning_rate": 1.3689839572192514e-07,
      "loss": 0.10906079411506653,
      "num_input_tokens_seen": 4791240,
      "step": 4649,
      "train_runtime": 11305.0102,
      "train_tokens_per_second": 423.816
    },
    {
      "epoch": 1.9872849663425578,
      "grad_norm": 2.8754923343658447,
      "learning_rate": 1.3262032085561498e-07,
      "loss": 0.1952277421951294,
      "num_input_tokens_seen": 4793680,
      "step": 4650,
      "train_runtime": 11307.4275,
      "train_tokens_per_second": 423.941
    },
    {
      "epoch": 1.9877123624318838,
      "grad_norm": 1.8187204599380493,
      "learning_rate": 1.2834224598930482e-07,
      "loss": 0.05892852321267128,
      "num_input_tokens_seen": 4795112,
      "step": 4651,
      "train_runtime": 11316.3663,
      "train_tokens_per_second": 423.732
    },
    {
      "epoch": 1.9881397585212095,
      "grad_norm": 2.653684616088867,
      "learning_rate": 1.2406417112299467e-07,
      "loss": 0.0778639018535614,
      "num_input_tokens_seen": 4796182,
      "step": 4652,
      "train_runtime": 11318.4426,
      "train_tokens_per_second": 423.749
    },
    {
      "epoch": 1.9885671546105352,
      "grad_norm": 2.993579626083374,
      "learning_rate": 1.197860962566845e-07,
      "loss": 0.09557916224002838,
      "num_input_tokens_seen": 4797170,
      "step": 4653,
      "train_runtime": 11320.6146,
      "train_tokens_per_second": 423.755
    },
    {
      "epoch": 1.9889945506998612,
      "grad_norm": 2.8111162185668945,
      "learning_rate": 1.1550802139037434e-07,
      "loss": 0.1099545881152153,
      "num_input_tokens_seen": 4798674,
      "step": 4654,
      "train_runtime": 11322.8392,
      "train_tokens_per_second": 423.805
    },
    {
      "epoch": 1.9894219467891867,
      "grad_norm": 3.348360300064087,
      "learning_rate": 1.1122994652406418e-07,
      "loss": 0.07052014768123627,
      "num_input_tokens_seen": 4799370,
      "step": 4655,
      "train_runtime": 11324.8486,
      "train_tokens_per_second": 423.791
    },
    {
      "epoch": 1.9898493428785127,
      "grad_norm": 2.28330659866333,
      "learning_rate": 1.0695187165775401e-07,
      "loss": 0.07211253046989441,
      "num_input_tokens_seen": 4800314,
      "step": 4656,
      "train_runtime": 11326.9589,
      "train_tokens_per_second": 423.795
    },
    {
      "epoch": 1.9902767389678384,
      "grad_norm": 5.864428997039795,
      "learning_rate": 1.0267379679144385e-07,
      "loss": 0.051632001996040344,
      "num_input_tokens_seen": 4801288,
      "step": 4657,
      "train_runtime": 11329.0195,
      "train_tokens_per_second": 423.804
    },
    {
      "epoch": 1.9907041350571641,
      "grad_norm": 1.5526164770126343,
      "learning_rate": 9.83957219251337e-08,
      "loss": 0.07412509620189667,
      "num_input_tokens_seen": 4802798,
      "step": 4658,
      "train_runtime": 11331.2324,
      "train_tokens_per_second": 423.855
    },
    {
      "epoch": 1.99113153114649,
      "grad_norm": 4.0224785804748535,
      "learning_rate": 9.411764705882353e-08,
      "loss": 0.1350272297859192,
      "num_input_tokens_seen": 4803440,
      "step": 4659,
      "train_runtime": 11333.2069,
      "train_tokens_per_second": 423.838
    },
    {
      "epoch": 1.9915589272358158,
      "grad_norm": 2.5184638500213623,
      "learning_rate": 8.983957219251337e-08,
      "loss": 0.06275865435600281,
      "num_input_tokens_seen": 4804572,
      "step": 4660,
      "train_runtime": 11335.3584,
      "train_tokens_per_second": 423.857
    },
    {
      "epoch": 1.9919863233251416,
      "grad_norm": 2.1118130683898926,
      "learning_rate": 8.556149732620321e-08,
      "loss": 0.08507050573825836,
      "num_input_tokens_seen": 4806040,
      "step": 4661,
      "train_runtime": 11337.5289,
      "train_tokens_per_second": 423.905
    },
    {
      "epoch": 1.9924137194144673,
      "grad_norm": 2.8188984394073486,
      "learning_rate": 8.128342245989307e-08,
      "loss": 0.053010810166597366,
      "num_input_tokens_seen": 4806868,
      "step": 4662,
      "train_runtime": 11339.5499,
      "train_tokens_per_second": 423.903
    },
    {
      "epoch": 1.992841115503793,
      "grad_norm": 4.274598121643066,
      "learning_rate": 7.70053475935829e-08,
      "loss": 0.1370730996131897,
      "num_input_tokens_seen": 4808028,
      "step": 4663,
      "train_runtime": 11341.7035,
      "train_tokens_per_second": 423.925
    },
    {
      "epoch": 1.993268511593119,
      "grad_norm": 3.228985071182251,
      "learning_rate": 7.272727272727274e-08,
      "loss": 0.07455558329820633,
      "num_input_tokens_seen": 4808608,
      "step": 4664,
      "train_runtime": 11343.7136,
      "train_tokens_per_second": 423.901
    },
    {
      "epoch": 1.9936959076824448,
      "grad_norm": 3.326261281967163,
      "learning_rate": 6.844919786096257e-08,
      "loss": 0.11473308503627777,
      "num_input_tokens_seen": 4809524,
      "step": 4665,
      "train_runtime": 11345.7453,
      "train_tokens_per_second": 423.906
    },
    {
      "epoch": 1.9941233037717705,
      "grad_norm": 2.3843088150024414,
      "learning_rate": 6.417112299465241e-08,
      "loss": 0.09405253827571869,
      "num_input_tokens_seen": 4810664,
      "step": 4666,
      "train_runtime": 11347.8896,
      "train_tokens_per_second": 423.926
    },
    {
      "epoch": 1.9945506998610962,
      "grad_norm": 2.6884829998016357,
      "learning_rate": 5.989304812834225e-08,
      "loss": 0.09168808907270432,
      "num_input_tokens_seen": 4811590,
      "step": 4667,
      "train_runtime": 11349.9908,
      "train_tokens_per_second": 423.929
    },
    {
      "epoch": 1.994978095950422,
      "grad_norm": 3.2621068954467773,
      "learning_rate": 5.561497326203209e-08,
      "loss": 0.10863994807004929,
      "num_input_tokens_seen": 4812744,
      "step": 4668,
      "train_runtime": 11352.1075,
      "train_tokens_per_second": 423.952
    },
    {
      "epoch": 1.995405492039748,
      "grad_norm": 3.2397677898406982,
      "learning_rate": 5.133689839572193e-08,
      "loss": 0.08679312467575073,
      "num_input_tokens_seen": 4813772,
      "step": 4669,
      "train_runtime": 11354.214,
      "train_tokens_per_second": 423.963
    },
    {
      "epoch": 1.9958328881290737,
      "grad_norm": 2.595773458480835,
      "learning_rate": 4.705882352941176e-08,
      "loss": 0.06904657185077667,
      "num_input_tokens_seen": 4814570,
      "step": 4670,
      "train_runtime": 11356.275,
      "train_tokens_per_second": 423.957
    },
    {
      "epoch": 1.9962602842183994,
      "grad_norm": 2.263497829437256,
      "learning_rate": 4.2780748663101606e-08,
      "loss": 0.06095545366406441,
      "num_input_tokens_seen": 4815478,
      "step": 4671,
      "train_runtime": 11358.3398,
      "train_tokens_per_second": 423.96
    },
    {
      "epoch": 1.9966876803077251,
      "grad_norm": 3.5418701171875,
      "learning_rate": 3.850267379679145e-08,
      "loss": 0.10343067348003387,
      "num_input_tokens_seen": 4816704,
      "step": 4672,
      "train_runtime": 11360.544,
      "train_tokens_per_second": 423.985
    },
    {
      "epoch": 1.9971150763970509,
      "grad_norm": 2.61396861076355,
      "learning_rate": 3.4224598930481285e-08,
      "loss": 0.07814645022153854,
      "num_input_tokens_seen": 4817578,
      "step": 4673,
      "train_runtime": 11362.7083,
      "train_tokens_per_second": 423.981
    },
    {
      "epoch": 1.9975424724863768,
      "grad_norm": 2.797137975692749,
      "learning_rate": 2.994652406417113e-08,
      "loss": 0.07945340126752853,
      "num_input_tokens_seen": 4818504,
      "step": 4674,
      "train_runtime": 11364.7756,
      "train_tokens_per_second": 423.986
    },
    {
      "epoch": 1.9979698685757026,
      "grad_norm": 1.979058027267456,
      "learning_rate": 2.5668449197860963e-08,
      "loss": 0.07385670393705368,
      "num_input_tokens_seen": 4819624,
      "step": 4675,
      "train_runtime": 11366.89,
      "train_tokens_per_second": 424.006
    },
    {
      "epoch": 1.9983972646650283,
      "grad_norm": 2.5515716075897217,
      "learning_rate": 2.1390374331550803e-08,
      "loss": 0.07762759178876877,
      "num_input_tokens_seen": 4820900,
      "step": 4676,
      "train_runtime": 11369.0116,
      "train_tokens_per_second": 424.039
    },
    {
      "epoch": 1.998824660754354,
      "grad_norm": 3.5036349296569824,
      "learning_rate": 1.7112299465240642e-08,
      "loss": 0.1370440274477005,
      "num_input_tokens_seen": 4822008,
      "step": 4677,
      "train_runtime": 11371.0541,
      "train_tokens_per_second": 424.06
    },
    {
      "epoch": 1.9992520568436798,
      "grad_norm": 2.773545503616333,
      "learning_rate": 1.2834224598930482e-08,
      "loss": 0.10351832956075668,
      "num_input_tokens_seen": 4823118,
      "step": 4678,
      "train_runtime": 11373.1772,
      "train_tokens_per_second": 424.078
    },
    {
      "epoch": 1.9996794529330058,
      "grad_norm": 2.52895450592041,
      "learning_rate": 8.556149732620321e-09,
      "loss": 0.0622512549161911,
      "num_input_tokens_seen": 4823998,
      "step": 4679,
      "train_runtime": 11375.2703,
      "train_tokens_per_second": 424.078
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.6760764122009277,
      "learning_rate": 4.2780748663101606e-09,
      "loss": 0.08774103969335556,
      "num_input_tokens_seen": 4824912,
      "step": 4680,
      "train_runtime": 11376.8826,
      "train_tokens_per_second": 424.098
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5158717632293701,
      "eval_runtime": 58.6594,
      "eval_samples_per_second": 17.013,
      "eval_steps_per_second": 8.507,
      "num_input_tokens_seen": 4824912,
      "step": 4680
    }
  ],
  "logging_steps": 1,
  "max_steps": 4680,
  "num_input_tokens_seen": 4824912,
  "num_train_epochs": 2,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7281004250986496e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}