{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1143,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_loss": 2.1641297340393066,
      "eval_runtime": 0.8244,
      "eval_samples_per_second": 24.261,
      "eval_steps_per_second": 2.426,
      "step": 0
    },
    {
      "epoch": 0.0008750820389411508,
      "grad_norm": 6.372805118560791,
      "learning_rate": 2.915451895043732e-08,
      "loss": 1.6434,
      "step": 1
    },
    {
      "epoch": 0.008750820389411508,
      "grad_norm": 6.278440475463867,
      "learning_rate": 2.915451895043732e-07,
      "loss": 1.7188,
      "step": 10
    },
    {
      "epoch": 0.017501640778823015,
      "grad_norm": 6.023682594299316,
      "learning_rate": 5.830903790087464e-07,
      "loss": 1.7154,
      "step": 20
    },
    {
      "epoch": 0.02625246116823452,
      "grad_norm": 3.7778327465057373,
      "learning_rate": 8.746355685131196e-07,
      "loss": 1.6563,
      "step": 30
    },
    {
      "epoch": 0.03500328155764603,
      "grad_norm": 3.8672306537628174,
      "learning_rate": 1.1661807580174927e-06,
      "loss": 1.5743,
      "step": 40
    },
    {
      "epoch": 0.04375410194705753,
      "grad_norm": 2.53611159324646,
      "learning_rate": 1.4577259475218661e-06,
      "loss": 1.479,
      "step": 50
    },
    {
      "epoch": 0.05250492233646904,
      "grad_norm": 2.019819974899292,
      "learning_rate": 1.7492711370262391e-06,
      "loss": 1.4242,
      "step": 60
    },
    {
      "epoch": 0.06125574272588055,
      "grad_norm": 1.7553459405899048,
      "learning_rate": 2.0408163265306125e-06,
      "loss": 1.3795,
      "step": 70
    },
    {
      "epoch": 0.07000656311529206,
      "grad_norm": 1.4397658109664917,
      "learning_rate": 2.3323615160349855e-06,
      "loss": 1.3745,
      "step": 80
    },
    {
      "epoch": 0.07875738350470357,
      "grad_norm": 1.5793251991271973,
      "learning_rate": 2.6239067055393585e-06,
      "loss": 1.3529,
      "step": 90
    },
    {
      "epoch": 0.08750820389411507,
      "grad_norm": 1.5069832801818848,
      "learning_rate": 2.9154518950437323e-06,
      "loss": 1.3459,
      "step": 100
    },
    {
      "epoch": 0.09625902428352658,
      "grad_norm": 1.4921239614486694,
      "learning_rate": 3.2069970845481052e-06,
      "loss": 1.2743,
      "step": 110
    },
    {
      "epoch": 0.10500984467293809,
      "grad_norm": 1.4761064052581787,
      "learning_rate": 3.4985422740524782e-06,
      "loss": 1.2959,
      "step": 120
    },
    {
      "epoch": 0.1137606650623496,
      "grad_norm": 1.4724879264831543,
      "learning_rate": 3.790087463556852e-06,
      "loss": 1.2773,
      "step": 130
    },
    {
      "epoch": 0.1225114854517611,
      "grad_norm": 1.4954681396484375,
      "learning_rate": 4.081632653061225e-06,
      "loss": 1.2695,
      "step": 140
    },
    {
      "epoch": 0.1312623058411726,
      "grad_norm": 1.4220881462097168,
      "learning_rate": 4.3731778425655976e-06,
      "loss": 1.2668,
      "step": 150
    },
    {
      "epoch": 0.14001312623058412,
      "grad_norm": 1.4039027690887451,
      "learning_rate": 4.664723032069971e-06,
      "loss": 1.2231,
      "step": 160
    },
    {
      "epoch": 0.14876394661999562,
      "grad_norm": 1.468996524810791,
      "learning_rate": 4.956268221574344e-06,
      "loss": 1.221,
      "step": 170
    },
    {
      "epoch": 0.15751476700940714,
      "grad_norm": 1.5780359506607056,
      "learning_rate": 5.247813411078717e-06,
      "loss": 1.2403,
      "step": 180
    },
    {
      "epoch": 0.16626558739881864,
      "grad_norm": 1.4090375900268555,
      "learning_rate": 5.539358600583091e-06,
      "loss": 1.2021,
      "step": 190
    },
    {
      "epoch": 0.17501640778823013,
      "grad_norm": 1.489652395248413,
      "learning_rate": 5.8309037900874645e-06,
      "loss": 1.1731,
      "step": 200
    },
    {
      "epoch": 0.18376722817764166,
      "grad_norm": 1.3121322393417358,
      "learning_rate": 6.122448979591837e-06,
      "loss": 1.1924,
      "step": 210
    },
    {
      "epoch": 0.19251804856705315,
      "grad_norm": 1.3954280614852905,
      "learning_rate": 6.4139941690962105e-06,
      "loss": 1.2121,
      "step": 220
    },
    {
      "epoch": 0.20126886895646467,
      "grad_norm": 1.4191243648529053,
      "learning_rate": 6.705539358600584e-06,
      "loss": 1.1973,
      "step": 230
    },
    {
      "epoch": 0.21001968934587617,
      "grad_norm": 1.4373400211334229,
      "learning_rate": 6.9970845481049564e-06,
      "loss": 1.1701,
      "step": 240
    },
    {
      "epoch": 0.2187705097352877,
      "grad_norm": 1.4669713973999023,
      "learning_rate": 7.28862973760933e-06,
      "loss": 1.1923,
      "step": 250
    },
    {
      "epoch": 0.2275213301246992,
      "grad_norm": 1.4269640445709229,
      "learning_rate": 7.580174927113704e-06,
      "loss": 1.1713,
      "step": 260
    },
    {
      "epoch": 0.23627215051411068,
      "grad_norm": 1.7535545825958252,
      "learning_rate": 7.871720116618077e-06,
      "loss": 1.1891,
      "step": 270
    },
    {
      "epoch": 0.2450229709035222,
      "grad_norm": 1.599094271659851,
      "learning_rate": 8.16326530612245e-06,
      "loss": 1.1827,
      "step": 280
    },
    {
      "epoch": 0.2537737912929337,
      "grad_norm": 1.8278971910476685,
      "learning_rate": 8.454810495626823e-06,
      "loss": 1.1402,
      "step": 290
    },
    {
      "epoch": 0.2625246116823452,
      "grad_norm": 1.5127216577529907,
      "learning_rate": 8.746355685131195e-06,
      "loss": 1.1597,
      "step": 300
    },
    {
      "epoch": 0.27127543207175675,
      "grad_norm": 1.6015137434005737,
      "learning_rate": 9.03790087463557e-06,
      "loss": 1.1398,
      "step": 310
    },
    {
      "epoch": 0.28002625246116825,
      "grad_norm": 1.3944435119628906,
      "learning_rate": 9.329446064139942e-06,
      "loss": 1.1174,
      "step": 320
    },
    {
      "epoch": 0.28877707285057974,
      "grad_norm": 1.2441927194595337,
      "learning_rate": 9.620991253644316e-06,
      "loss": 1.1416,
      "step": 330
    },
    {
      "epoch": 0.29752789323999124,
      "grad_norm": 1.4111624956130981,
      "learning_rate": 9.912536443148689e-06,
      "loss": 1.1352,
      "step": 340
    },
    {
      "epoch": 0.30627871362940273,
      "grad_norm": 1.3007830381393433,
      "learning_rate": 9.999872800049233e-06,
      "loss": 1.1208,
      "step": 350
    },
    {
      "epoch": 0.3150295340188143,
      "grad_norm": 1.3299254179000854,
      "learning_rate": 9.999249795462235e-06,
      "loss": 1.142,
      "step": 360
    },
    {
      "epoch": 0.3237803544082258,
      "grad_norm": 1.2524114847183228,
      "learning_rate": 9.998107687592326e-06,
      "loss": 1.1147,
      "step": 370
    },
    {
      "epoch": 0.3325311747976373,
      "grad_norm": 1.4711254835128784,
      "learning_rate": 9.996446595031691e-06,
      "loss": 1.1182,
      "step": 380
    },
    {
      "epoch": 0.34128199518704877,
      "grad_norm": 1.3600913286209106,
      "learning_rate": 9.994266690261928e-06,
      "loss": 1.1565,
      "step": 390
    },
    {
      "epoch": 0.35003281557646027,
      "grad_norm": 1.3752756118774414,
      "learning_rate": 9.991568199636149e-06,
      "loss": 1.0803,
      "step": 400
    },
    {
      "epoch": 0.3587836359658718,
      "grad_norm": 1.2283236980438232,
      "learning_rate": 9.988351403355455e-06,
      "loss": 1.1371,
      "step": 410
    },
    {
      "epoch": 0.3675344563552833,
      "grad_norm": 1.266762614250183,
      "learning_rate": 9.98461663543987e-06,
      "loss": 1.1027,
      "step": 420
    },
    {
      "epoch": 0.3762852767446948,
      "grad_norm": 1.2771412134170532,
      "learning_rate": 9.980364283693638e-06,
      "loss": 1.0891,
      "step": 430
    },
    {
      "epoch": 0.3850360971341063,
      "grad_norm": 1.2226617336273193,
      "learning_rate": 9.97559478966496e-06,
      "loss": 1.1105,
      "step": 440
    },
    {
      "epoch": 0.39378691752351785,
      "grad_norm": 1.2738267183303833,
      "learning_rate": 9.970308648600142e-06,
      "loss": 1.114,
      "step": 450
    },
    {
      "epoch": 0.40253773791292935,
      "grad_norm": 1.3344817161560059,
      "learning_rate": 9.964506409392178e-06,
      "loss": 1.1178,
      "step": 460
    },
    {
      "epoch": 0.41128855830234085,
      "grad_norm": 1.3076369762420654,
      "learning_rate": 9.958188674523748e-06,
      "loss": 1.1205,
      "step": 470
    },
    {
      "epoch": 0.42003937869175234,
      "grad_norm": 1.251315712928772,
      "learning_rate": 9.951356100004666e-06,
      "loss": 1.0795,
      "step": 480
    },
    {
      "epoch": 0.42879019908116384,
      "grad_norm": 1.1784945726394653,
      "learning_rate": 9.94400939530375e-06,
      "loss": 1.1175,
      "step": 490
    },
    {
      "epoch": 0.4375410194705754,
      "grad_norm": 1.219256043434143,
      "learning_rate": 9.93614932327517e-06,
      "loss": 1.0933,
      "step": 500
    },
    {
      "epoch": 0.4462918398599869,
      "grad_norm": 1.398763656616211,
      "learning_rate": 9.927776700079217e-06,
      "loss": 1.1267,
      "step": 510
    },
    {
      "epoch": 0.4550426602493984,
      "grad_norm": 1.25184166431427,
      "learning_rate": 9.918892395097572e-06,
      "loss": 1.1191,
      "step": 520
    },
    {
      "epoch": 0.4637934806388099,
      "grad_norm": 1.3422456979751587,
      "learning_rate": 9.90949733084303e-06,
      "loss": 1.1003,
      "step": 530
    },
    {
      "epoch": 0.47254430102822137,
      "grad_norm": 1.3312543630599976,
      "learning_rate": 9.8995924828637e-06,
      "loss": 1.1004,
      "step": 540
    },
    {
      "epoch": 0.4812951214176329,
      "grad_norm": 1.2708804607391357,
      "learning_rate": 9.889178879641717e-06,
      "loss": 1.1094,
      "step": 550
    },
    {
      "epoch": 0.4900459418070444,
      "grad_norm": 1.3565181493759155,
      "learning_rate": 9.878257602486445e-06,
      "loss": 1.1252,
      "step": 560
    },
    {
      "epoch": 0.4987967621964559,
      "grad_norm": 1.232021450996399,
      "learning_rate": 9.866829785422201e-06,
      "loss": 1.0976,
      "step": 570
    },
    {
      "epoch": 0.5075475825858674,
      "grad_norm": 1.371229887008667,
      "learning_rate": 9.854896615070495e-06,
      "loss": 1.0941,
      "step": 580
    },
    {
      "epoch": 0.516298402975279,
      "grad_norm": 1.35117769241333,
      "learning_rate": 9.842459330526826e-06,
      "loss": 1.1127,
      "step": 590
    },
    {
      "epoch": 0.5250492233646904,
      "grad_norm": 1.234413504600525,
      "learning_rate": 9.829519223232006e-06,
      "loss": 1.09,
      "step": 600
    },
    {
      "epoch": 0.533800043754102,
      "grad_norm": 1.2576864957809448,
      "learning_rate": 9.816077636838079e-06,
      "loss": 1.1094,
      "step": 610
    },
    {
      "epoch": 0.5425508641435135,
      "grad_norm": 1.340950608253479,
      "learning_rate": 9.802135967068783e-06,
      "loss": 1.0889,
      "step": 620
    },
    {
      "epoch": 0.5513016845329249,
      "grad_norm": 1.350095272064209,
      "learning_rate": 9.787695661574632e-06,
      "loss": 1.0661,
      "step": 630
    },
    {
      "epoch": 0.5600525049223365,
      "grad_norm": 1.2629624605178833,
      "learning_rate": 9.772758219782601e-06,
      "loss": 1.0823,
      "step": 640
    },
    {
      "epoch": 0.5688033253117479,
      "grad_norm": 1.234574556350708,
      "learning_rate": 9.757325192740425e-06,
      "loss": 1.0937,
      "step": 650
    },
    {
      "epoch": 0.5775541457011595,
      "grad_norm": 1.1775031089782715,
      "learning_rate": 9.741398182955548e-06,
      "loss": 1.1066,
      "step": 660
    },
    {
      "epoch": 0.586304966090571,
      "grad_norm": 1.2653073072433472,
      "learning_rate": 9.724978844228717e-06,
      "loss": 1.0964,
      "step": 670
    },
    {
      "epoch": 0.5950557864799825,
      "grad_norm": 1.2665278911590576,
      "learning_rate": 9.70806888148227e-06,
      "loss": 1.1007,
      "step": 680
    },
    {
      "epoch": 0.603806606869394,
      "grad_norm": 1.2909613847732544,
      "learning_rate": 9.690670050583095e-06,
      "loss": 1.0973,
      "step": 690
    },
    {
      "epoch": 0.6125574272588055,
      "grad_norm": 1.2396572828292847,
      "learning_rate": 9.672784158160307e-06,
      "loss": 1.0423,
      "step": 700
    },
    {
      "epoch": 0.621308247648217,
      "grad_norm": 1.2569255828857422,
      "learning_rate": 9.654413061417655e-06,
      "loss": 1.0764,
      "step": 710
    },
    {
      "epoch": 0.6300590680376286,
      "grad_norm": 1.3747756481170654,
      "learning_rate": 9.635558667940683e-06,
      "loss": 1.0563,
      "step": 720
    },
    {
      "epoch": 0.63880988842704,
      "grad_norm": 1.2752923965454102,
      "learning_rate": 9.616222935498647e-06,
      "loss": 1.1084,
      "step": 730
    },
    {
      "epoch": 0.6475607088164516,
      "grad_norm": 1.218320369720459,
      "learning_rate": 9.596407871841229e-06,
      "loss": 1.0674,
      "step": 740
    },
    {
      "epoch": 0.656311529205863,
      "grad_norm": 1.2466013431549072,
      "learning_rate": 9.57611553449006e-06,
      "loss": 1.1048,
      "step": 750
    },
    {
      "epoch": 0.6650623495952745,
      "grad_norm": 1.2956342697143555,
      "learning_rate": 9.555348030525076e-06,
      "loss": 1.0723,
      "step": 760
    },
    {
      "epoch": 0.6738131699846861,
      "grad_norm": 1.29911208152771,
      "learning_rate": 9.534107516365725e-06,
      "loss": 1.0743,
      "step": 770
    },
    {
      "epoch": 0.6825639903740975,
      "grad_norm": 1.2063359022140503,
      "learning_rate": 9.512396197547053e-06,
      "loss": 1.0717,
      "step": 780
    },
    {
      "epoch": 0.6913148107635091,
      "grad_norm": 1.184462547302246,
      "learning_rate": 9.490216328490693e-06,
      "loss": 1.0895,
      "step": 790
    },
    {
      "epoch": 0.7000656311529205,
      "grad_norm": 1.186874270439148,
      "learning_rate": 9.467570212270767e-06,
      "loss": 1.1476,
      "step": 800
    },
    {
      "epoch": 0.7088164515423321,
      "grad_norm": 1.2901315689086914,
      "learning_rate": 9.444460200374747e-06,
      "loss": 1.0538,
      "step": 810
    },
    {
      "epoch": 0.7175672719317436,
      "grad_norm": 1.2175313234329224,
      "learning_rate": 9.42088869245929e-06,
      "loss": 1.079,
      "step": 820
    },
    {
      "epoch": 0.7263180923211551,
      "grad_norm": 1.1771786212921143,
      "learning_rate": 9.396858136101066e-06,
      "loss": 1.0756,
      "step": 830
    },
    {
      "epoch": 0.7350689127105666,
      "grad_norm": 1.2814205884933472,
      "learning_rate": 9.372371026542595e-06,
      "loss": 1.0697,
      "step": 840
    },
    {
      "epoch": 0.7438197330999782,
      "grad_norm": 1.2603764533996582,
      "learning_rate": 9.347429906433175e-06,
      "loss": 1.1079,
      "step": 850
    },
    {
      "epoch": 0.7525705534893896,
      "grad_norm": 1.2228299379348755,
      "learning_rate": 9.322037365564844e-06,
      "loss": 1.0746,
      "step": 860
    },
    {
      "epoch": 0.7613213738788012,
      "grad_norm": 1.1895763874053955,
      "learning_rate": 9.296196040603476e-06,
      "loss": 1.0831,
      "step": 870
    },
    {
      "epoch": 0.7700721942682126,
      "grad_norm": 1.2288453578948975,
      "learning_rate": 9.269908614814995e-06,
      "loss": 1.0649,
      "step": 880
    },
    {
      "epoch": 0.7788230146576242,
      "grad_norm": 1.1580345630645752,
      "learning_rate": 9.243177817786757e-06,
      "loss": 1.1053,
      "step": 890
    },
    {
      "epoch": 0.7875738350470357,
      "grad_norm": 1.1613861322402954,
      "learning_rate": 9.216006425144118e-06,
      "loss": 1.0854,
      "step": 900
    },
    {
      "epoch": 0.7963246554364471,
      "grad_norm": 1.0809310674667358,
      "learning_rate": 9.188397258262221e-06,
      "loss": 1.059,
      "step": 910
    },
    {
      "epoch": 0.8050754758258587,
      "grad_norm": 1.2021125555038452,
      "learning_rate": 9.160353183973045e-06,
      "loss": 1.064,
      "step": 920
    },
    {
      "epoch": 0.8138262962152701,
      "grad_norm": 1.286584496498108,
      "learning_rate": 9.131877114267711e-06,
      "loss": 1.0516,
      "step": 930
    },
    {
      "epoch": 0.8225771166046817,
      "grad_norm": 1.177673578262329,
      "learning_rate": 9.102972005994123e-06,
      "loss": 1.0358,
      "step": 940
    },
    {
      "epoch": 0.8313279369940932,
      "grad_norm": 1.2117091417312622,
      "learning_rate": 9.073640860549931e-06,
      "loss": 1.0701,
      "step": 950
    },
    {
      "epoch": 0.8400787573835047,
      "grad_norm": 1.2018595933914185,
      "learning_rate": 9.043886723570885e-06,
      "loss": 1.0833,
      "step": 960
    },
    {
      "epoch": 0.8488295777729162,
      "grad_norm": 1.1337984800338745,
      "learning_rate": 9.013712684614582e-06,
      "loss": 1.0785,
      "step": 970
    },
    {
      "epoch": 0.8575803981623277,
      "grad_norm": 1.2217832803726196,
      "learning_rate": 8.983121876839658e-06,
      "loss": 1.0623,
      "step": 980
    },
    {
      "epoch": 0.8663312185517392,
      "grad_norm": 1.2249287366867065,
      "learning_rate": 8.95211747668046e-06,
      "loss": 1.0285,
      "step": 990
    },
    {
      "epoch": 0.8750820389411508,
      "grad_norm": 1.1673548221588135,
      "learning_rate": 8.920702703517205e-06,
      "loss": 1.0354,
      "step": 1000
    },
    {
      "epoch": 0.8838328593305622,
      "grad_norm": 1.1298819780349731,
      "learning_rate": 8.888880819341703e-06,
      "loss": 1.0538,
      "step": 1010
    },
    {
      "epoch": 0.8925836797199738,
      "grad_norm": 1.1600337028503418,
      "learning_rate": 8.856655128418638e-06,
      "loss": 1.0527,
      "step": 1020
    },
    {
      "epoch": 0.9013345001093852,
      "grad_norm": 1.2292169332504272,
      "learning_rate": 8.824028976942465e-06,
      "loss": 1.0613,
      "step": 1030
    },
    {
      "epoch": 0.9100853204987968,
      "grad_norm": 1.1878681182861328,
      "learning_rate": 8.791005752689958e-06,
      "loss": 1.0489,
      "step": 1040
    },
    {
      "epoch": 0.9188361408882083,
      "grad_norm": 1.2165477275848389,
      "learning_rate": 8.757588884668435e-06,
      "loss": 1.0675,
      "step": 1050
    },
    {
      "epoch": 0.9275869612776197,
      "grad_norm": 1.1682305335998535,
      "learning_rate": 8.723781842759696e-06,
      "loss": 1.084,
      "step": 1060
    },
    {
      "epoch": 0.9363377816670313,
      "grad_norm": 1.3387882709503174,
      "learning_rate": 8.689588137359731e-06,
      "loss": 1.0564,
      "step": 1070
    },
    {
      "epoch": 0.9450886020564427,
      "grad_norm": 1.2144585847854614,
      "learning_rate": 8.655011319014217e-06,
      "loss": 1.08,
      "step": 1080
    },
    {
      "epoch": 0.9538394224458543,
      "grad_norm": 1.2637611627578735,
      "learning_rate": 8.620054978049828e-06,
      "loss": 1.0716,
      "step": 1090
    },
    {
      "epoch": 0.9625902428352658,
      "grad_norm": 1.230635404586792,
      "learning_rate": 8.584722744201448e-06,
      "loss": 1.0429,
      "step": 1100
    },
    {
      "epoch": 0.9713410632246773,
      "grad_norm": 1.2495629787445068,
      "learning_rate": 8.549018286235255e-06,
      "loss": 1.0464,
      "step": 1110
    },
    {
      "epoch": 0.9800918836140888,
      "grad_norm": 1.1702367067337036,
      "learning_rate": 8.512945311567786e-06,
      "loss": 1.0476,
      "step": 1120
    },
    {
      "epoch": 0.9888427040035004,
      "grad_norm": 1.1367738246917725,
      "learning_rate": 8.476507565880959e-06,
      "loss": 1.0392,
      "step": 1130
    },
    {
      "epoch": 0.9975935243929118,
      "grad_norm": 1.1903005838394165,
      "learning_rate": 8.439708832733142e-06,
      "loss": 1.0414,
      "step": 1140
    }
  ],
  "logging_steps": 10,
  "max_steps": 3426,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.7750297385159885e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}