{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1088,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01838235294117647,
      "grad_norm": 4.980741500854492,
      "learning_rate": 8.256880733944956e-07,
      "loss": 0.4373,
      "step": 10
    },
    {
      "epoch": 0.03676470588235294,
      "grad_norm": 1.1804918050765991,
      "learning_rate": 1.743119266055046e-06,
      "loss": 0.3921,
      "step": 20
    },
    {
      "epoch": 0.05514705882352941,
      "grad_norm": 0.9609614610671997,
      "learning_rate": 2.6605504587155968e-06,
      "loss": 0.3541,
      "step": 30
    },
    {
      "epoch": 0.07352941176470588,
      "grad_norm": 0.8202292323112488,
      "learning_rate": 3.5779816513761473e-06,
      "loss": 0.3062,
      "step": 40
    },
    {
      "epoch": 0.09191176470588236,
      "grad_norm": 0.839361846446991,
      "learning_rate": 4.4954128440366975e-06,
      "loss": 0.2999,
      "step": 50
    },
    {
      "epoch": 0.11029411764705882,
      "grad_norm": 0.6450493335723877,
      "learning_rate": 5.412844036697248e-06,
      "loss": 0.2592,
      "step": 60
    },
    {
      "epoch": 0.12867647058823528,
      "grad_norm": 0.8965937495231628,
      "learning_rate": 6.330275229357799e-06,
      "loss": 0.3104,
      "step": 70
    },
    {
      "epoch": 0.14705882352941177,
      "grad_norm": 0.7927365899085999,
      "learning_rate": 7.247706422018349e-06,
      "loss": 0.2719,
      "step": 80
    },
    {
      "epoch": 0.16544117647058823,
      "grad_norm": 1.1923980712890625,
      "learning_rate": 8.1651376146789e-06,
      "loss": 0.2705,
      "step": 90
    },
    {
      "epoch": 0.18382352941176472,
      "grad_norm": 0.7955174446105957,
      "learning_rate": 9.08256880733945e-06,
      "loss": 0.276,
      "step": 100
    },
    {
      "epoch": 0.20220588235294118,
      "grad_norm": 0.9915638566017151,
      "learning_rate": 1e-05,
      "loss": 0.2757,
      "step": 110
    },
    {
      "epoch": 0.22058823529411764,
      "grad_norm": 1.160922646522522,
      "learning_rate": 9.99742583072674e-06,
      "loss": 0.2652,
      "step": 120
    },
    {
      "epoch": 0.23897058823529413,
      "grad_norm": 0.896353542804718,
      "learning_rate": 9.98970597344593e-06,
      "loss": 0.2949,
      "step": 130
    },
    {
      "epoch": 0.25735294117647056,
      "grad_norm": 0.6837694048881531,
      "learning_rate": 9.976848377045343e-06,
      "loss": 0.247,
      "step": 140
    },
    {
      "epoch": 0.2757352941176471,
      "grad_norm": 0.7046304941177368,
      "learning_rate": 9.958866280576803e-06,
      "loss": 0.2403,
      "step": 150
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 0.8753547072410583,
      "learning_rate": 9.935778199624394e-06,
      "loss": 0.2597,
      "step": 160
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.8226246237754822,
      "learning_rate": 9.90760790723954e-06,
      "loss": 0.2574,
      "step": 170
    },
    {
      "epoch": 0.33088235294117646,
      "grad_norm": 0.820383608341217,
      "learning_rate": 9.874384409462673e-06,
      "loss": 0.2652,
      "step": 180
    },
    {
      "epoch": 0.3492647058823529,
      "grad_norm": 0.7547650337219238,
      "learning_rate": 9.836141915456646e-06,
      "loss": 0.2522,
      "step": 190
    },
    {
      "epoch": 0.36764705882352944,
      "grad_norm": 0.72496497631073,
      "learning_rate": 9.792919802282656e-06,
      "loss": 0.2536,
      "step": 200
    },
    {
      "epoch": 0.3860294117647059,
      "grad_norm": 0.8042762279510498,
      "learning_rate": 9.744762574354967e-06,
      "loss": 0.2482,
      "step": 210
    },
    {
      "epoch": 0.40441176470588236,
      "grad_norm": 0.7786635756492615,
      "learning_rate": 9.691719817616148e-06,
      "loss": 0.2167,
      "step": 220
    },
    {
      "epoch": 0.4227941176470588,
      "grad_norm": 0.8994752764701843,
      "learning_rate": 9.633846148480024e-06,
      "loss": 0.2359,
      "step": 230
    },
    {
      "epoch": 0.4411764705882353,
      "grad_norm": 0.7100334763526917,
      "learning_rate": 9.571201157594925e-06,
      "loss": 0.231,
      "step": 240
    },
    {
      "epoch": 0.45955882352941174,
      "grad_norm": 0.6736875176429749,
      "learning_rate": 9.503849348485112e-06,
      "loss": 0.2596,
      "step": 250
    },
    {
      "epoch": 0.47794117647058826,
      "grad_norm": 0.8927339315414429,
      "learning_rate": 9.431860071133592e-06,
      "loss": 0.2653,
      "step": 260
    },
    {
      "epoch": 0.4963235294117647,
      "grad_norm": 0.7796552181243896,
      "learning_rate": 9.355307450574666e-06,
      "loss": 0.2423,
      "step": 270
    },
    {
      "epoch": 0.5147058823529411,
      "grad_norm": 0.8039178252220154,
      "learning_rate": 9.27427031056979e-06,
      "loss": 0.2497,
      "step": 280
    },
    {
      "epoch": 0.5330882352941176,
      "grad_norm": 0.8636899590492249,
      "learning_rate": 9.188832092445281e-06,
      "loss": 0.249,
      "step": 290
    },
    {
      "epoch": 0.5514705882352942,
      "grad_norm": 0.6733593940734863,
      "learning_rate": 9.09908076917548e-06,
      "loss": 0.2415,
      "step": 300
    },
    {
      "epoch": 0.5698529411764706,
      "grad_norm": 0.6637346148490906,
      "learning_rate": 9.00510875479983e-06,
      "loss": 0.2207,
      "step": 310
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 0.6064783930778503,
      "learning_rate": 8.907012809267107e-06,
      "loss": 0.2559,
      "step": 320
    },
    {
      "epoch": 0.6066176470588235,
      "grad_norm": 0.8090744018554688,
      "learning_rate": 8.804893938804839e-06,
      "loss": 0.2446,
      "step": 330
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.639586865901947,
      "learning_rate": 8.698857291916456e-06,
      "loss": 0.2277,
      "step": 340
    },
    {
      "epoch": 0.6433823529411765,
      "grad_norm": 0.695817768573761,
      "learning_rate": 8.58901205111326e-06,
      "loss": 0.2669,
      "step": 350
    },
    {
      "epoch": 0.6617647058823529,
      "grad_norm": 0.8129636645317078,
      "learning_rate": 8.475471320492728e-06,
      "loss": 0.2233,
      "step": 360
    },
    {
      "epoch": 0.6801470588235294,
      "grad_norm": 0.6946842074394226,
      "learning_rate": 8.35835200927887e-06,
      "loss": 0.2359,
      "step": 370
    },
    {
      "epoch": 0.6985294117647058,
      "grad_norm": 0.4986434280872345,
      "learning_rate": 8.237774711444575e-06,
      "loss": 0.2144,
      "step": 380
    },
    {
      "epoch": 0.7169117647058824,
      "grad_norm": 0.7395103573799133,
      "learning_rate": 8.113863581539905e-06,
      "loss": 0.231,
      "step": 390
    },
    {
      "epoch": 0.7352941176470589,
      "grad_norm": 0.643147349357605,
      "learning_rate": 7.986746206854143e-06,
      "loss": 0.2324,
      "step": 400
    },
    {
      "epoch": 0.7536764705882353,
      "grad_norm": 0.6714458465576172,
      "learning_rate": 7.856553476043294e-06,
      "loss": 0.2401,
      "step": 410
    },
    {
      "epoch": 0.7720588235294118,
      "grad_norm": 0.8807610273361206,
      "learning_rate": 7.723419444358261e-06,
      "loss": 0.2445,
      "step": 420
    },
    {
      "epoch": 0.7904411764705882,
      "grad_norm": 0.8745494484901428,
      "learning_rate": 7.5874811956124805e-06,
      "loss": 0.2574,
      "step": 430
    },
    {
      "epoch": 0.8088235294117647,
      "grad_norm": 0.6509207487106323,
      "learning_rate": 7.4488787010311425e-06,
      "loss": 0.2179,
      "step": 440
    },
    {
      "epoch": 0.8272058823529411,
      "grad_norm": 0.680579423904419,
      "learning_rate": 7.3077546751273494e-06,
      "loss": 0.2246,
      "step": 450
    },
    {
      "epoch": 0.8455882352941176,
      "grad_norm": 0.6059340834617615,
      "learning_rate": 7.164254428753581e-06,
      "loss": 0.2352,
      "step": 460
    },
    {
      "epoch": 0.8639705882352942,
      "grad_norm": 0.7696027755737305,
      "learning_rate": 7.018525719479805e-06,
      "loss": 0.2517,
      "step": 470
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 0.7263879179954529,
      "learning_rate": 6.870718599452279e-06,
      "loss": 0.2282,
      "step": 480
    },
    {
      "epoch": 0.9007352941176471,
      "grad_norm": 0.7986669540405273,
      "learning_rate": 6.7209852608897005e-06,
      "loss": 0.238,
      "step": 490
    },
    {
      "epoch": 0.9191176470588235,
      "grad_norm": 0.7858214974403381,
      "learning_rate": 6.569479879375795e-06,
      "loss": 0.2461,
      "step": 500
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.5920738577842712,
      "learning_rate": 6.416358455109695e-06,
      "loss": 0.2452,
      "step": 510
    },
    {
      "epoch": 0.9558823529411765,
      "grad_norm": 0.6721693277359009,
      "learning_rate": 6.261778652277565e-06,
      "loss": 0.2225,
      "step": 520
    },
    {
      "epoch": 0.9742647058823529,
      "grad_norm": 0.8532955646514893,
      "learning_rate": 6.105899636710895e-06,
      "loss": 0.2449,
      "step": 530
    },
    {
      "epoch": 0.9926470588235294,
      "grad_norm": 0.7130278944969177,
      "learning_rate": 5.948881911998572e-06,
      "loss": 0.2111,
      "step": 540
    },
    {
      "epoch": 1.0110294117647058,
      "grad_norm": 0.6958662271499634,
      "learning_rate": 5.790887154221521e-06,
      "loss": 0.2243,
      "step": 550
    },
    {
      "epoch": 1.0294117647058822,
      "grad_norm": 0.7343012094497681,
      "learning_rate": 5.632078045480065e-06,
      "loss": 0.2013,
      "step": 560
    },
    {
      "epoch": 1.0477941176470589,
      "grad_norm": 0.7366646528244019,
      "learning_rate": 5.472618106385415e-06,
      "loss": 0.2045,
      "step": 570
    },
    {
      "epoch": 1.0661764705882353,
      "grad_norm": 0.500787615776062,
      "learning_rate": 5.31267152768779e-06,
      "loss": 0.2004,
      "step": 580
    },
    {
      "epoch": 1.0845588235294117,
      "grad_norm": 0.6535488367080688,
      "learning_rate": 5.152403001214483e-06,
      "loss": 0.203,
      "step": 590
    },
    {
      "epoch": 1.1029411764705883,
      "grad_norm": 0.8450496196746826,
      "learning_rate": 4.991977550292028e-06,
      "loss": 0.2002,
      "step": 600
    },
    {
      "epoch": 1.1213235294117647,
      "grad_norm": 0.8866769671440125,
      "learning_rate": 4.831560359826985e-06,
      "loss": 0.2179,
      "step": 610
    },
    {
      "epoch": 1.1397058823529411,
      "grad_norm": 0.8293766975402832,
      "learning_rate": 4.671316606220394e-06,
      "loss": 0.2025,
      "step": 620
    },
    {
      "epoch": 1.1580882352941178,
      "grad_norm": 0.7488539218902588,
      "learning_rate": 4.511411287290964e-06,
      "loss": 0.2119,
      "step": 630
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 0.6722701191902161,
      "learning_rate": 4.35200905238214e-06,
      "loss": 0.2063,
      "step": 640
    },
    {
      "epoch": 1.1948529411764706,
      "grad_norm": 0.7874387502670288,
      "learning_rate": 4.193274032828e-06,
      "loss": 0.1826,
      "step": 650
    },
    {
      "epoch": 1.213235294117647,
      "grad_norm": 0.6973450779914856,
      "learning_rate": 4.035369672952516e-06,
      "loss": 0.1919,
      "step": 660
    },
    {
      "epoch": 1.2316176470588236,
      "grad_norm": 0.5885277390480042,
      "learning_rate": 3.8784585617762084e-06,
      "loss": 0.2193,
      "step": 670
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.8693950772285461,
      "learning_rate": 3.7227022656034873e-06,
      "loss": 0.2069,
      "step": 680
    },
    {
      "epoch": 1.2683823529411764,
      "grad_norm": 0.747329592704773,
      "learning_rate": 3.568261161663042e-06,
      "loss": 0.2188,
      "step": 690
    },
    {
      "epoch": 1.2867647058823528,
      "grad_norm": 0.7728601098060608,
      "learning_rate": 3.4152942729725896e-06,
      "loss": 0.2019,
      "step": 700
    },
    {
      "epoch": 1.3051470588235294,
      "grad_norm": 0.676393449306488,
      "learning_rate": 3.263959104598009e-06,
      "loss": 0.2036,
      "step": 710
    },
    {
      "epoch": 1.3235294117647058,
      "grad_norm": 0.6831977963447571,
      "learning_rate": 3.114411481475455e-06,
      "loss": 0.2052,
      "step": 720
    },
    {
      "epoch": 1.3419117647058822,
      "grad_norm": 0.5936601161956787,
      "learning_rate": 2.966805387963463e-06,
      "loss": 0.2003,
      "step": 730
    },
    {
      "epoch": 1.3602941176470589,
      "grad_norm": 0.6197609901428223,
      "learning_rate": 2.821292809290217e-06,
      "loss": 0.1894,
      "step": 740
    },
    {
      "epoch": 1.3786764705882353,
      "grad_norm": 0.4961455464363098,
      "learning_rate": 2.678023575059274e-06,
      "loss": 0.1873,
      "step": 750
    },
    {
      "epoch": 1.3970588235294117,
      "grad_norm": 0.8226666450500488,
      "learning_rate": 2.5371452049748603e-06,
      "loss": 0.2247,
      "step": 760
    },
    {
      "epoch": 1.4154411764705883,
      "grad_norm": 0.623539388179779,
      "learning_rate": 2.3988027569455895e-06,
      "loss": 0.1846,
      "step": 770
    },
    {
      "epoch": 1.4338235294117647,
      "grad_norm": 0.6834192276000977,
      "learning_rate": 2.2631386777230248e-06,
      "loss": 0.1829,
      "step": 780
    },
    {
      "epoch": 1.4522058823529411,
      "grad_norm": 0.9627503156661987,
      "learning_rate": 2.130292656228856e-06,
      "loss": 0.1813,
      "step": 790
    },
    {
      "epoch": 1.4705882352941178,
      "grad_norm": 0.5110943913459778,
      "learning_rate": 2.0004014797217207e-06,
      "loss": 0.193,
      "step": 800
    },
    {
      "epoch": 1.4889705882352942,
      "grad_norm": 0.6010143756866455,
      "learning_rate": 1.873598892951795e-06,
      "loss": 0.208,
      "step": 810
    },
    {
      "epoch": 1.5073529411764706,
      "grad_norm": 0.8185718655586243,
      "learning_rate": 1.7500154604481312e-06,
      "loss": 0.2006,
      "step": 820
    },
    {
      "epoch": 1.5257352941176472,
      "grad_norm": 0.6126958131790161,
      "learning_rate": 1.629778432080586e-06,
      "loss": 0.1746,
      "step": 830
    },
    {
      "epoch": 1.5441176470588234,
      "grad_norm": 0.5845675468444824,
      "learning_rate": 1.513011612034726e-06,
      "loss": 0.2013,
      "step": 840
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.8226555585861206,
      "learning_rate": 1.3998352313346768e-06,
      "loss": 0.2159,
      "step": 850
    },
    {
      "epoch": 1.5808823529411766,
      "grad_norm": 0.8147032856941223,
      "learning_rate": 1.2903658240450989e-06,
      "loss": 0.2234,
      "step": 860
    },
    {
      "epoch": 1.5992647058823528,
      "grad_norm": 0.7199397683143616,
      "learning_rate": 1.184716107279837e-06,
      "loss": 0.1798,
      "step": 870
    },
    {
      "epoch": 1.6176470588235294,
      "grad_norm": 0.6966183185577393,
      "learning_rate": 1.0829948651407374e-06,
      "loss": 0.1886,
      "step": 880
    },
    {
      "epoch": 1.6360294117647058,
      "grad_norm": 0.614043116569519,
      "learning_rate": 9.85306836706184e-07,
      "loss": 0.1908,
      "step": 890
    },
    {
      "epoch": 1.6544117647058822,
      "grad_norm": 0.8714206218719482,
      "learning_rate": 8.917526081846411e-07,
      "loss": 0.1956,
      "step": 900
    },
    {
      "epoch": 1.6727941176470589,
      "grad_norm": 0.7226178050041199,
      "learning_rate": 8.024285093442874e-07,
      "loss": 0.1845,
      "step": 910
    },
    {
      "epoch": 1.6911764705882353,
      "grad_norm": 0.6284250020980835,
      "learning_rate": 7.17426514325359e-07,
      "loss": 0.204,
      "step": 920
    },
    {
      "epoch": 1.7095588235294117,
      "grad_norm": 0.6917927265167236,
      "learning_rate": 6.36834146937354e-07,
      "loss": 0.1878,
      "step": 930
    },
    {
      "epoch": 1.7279411764705883,
      "grad_norm": 0.5628108382225037,
      "learning_rate": 5.607343905385898e-07,
      "loss": 0.1668,
      "step": 940
    },
    {
      "epoch": 1.7463235294117647,
      "grad_norm": 0.8957297801971436,
      "learning_rate": 4.892056025909148e-07,
      "loss": 0.2103,
      "step": 950
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 0.6452038288116455,
      "learning_rate": 4.2232143397756607e-07,
      "loss": 0.1933,
      "step": 960
    },
    {
      "epoch": 1.7830882352941178,
      "grad_norm": 1.0203959941864014,
      "learning_rate": 3.6015075316722605e-07,
      "loss": 0.21,
      "step": 970
    },
    {
      "epoch": 1.8014705882352942,
      "grad_norm": 0.696893572807312,
      "learning_rate": 3.02757575302392e-07,
      "loss": 0.1955,
      "step": 980
    },
    {
      "epoch": 1.8198529411764706,
      "grad_norm": 0.7462745904922485,
      "learning_rate": 2.5020099628504603e-07,
      "loss": 0.1765,
      "step": 990
    },
    {
      "epoch": 1.8382352941176472,
      "grad_norm": 0.6422279477119446,
      "learning_rate": 2.0253513192751374e-07,
      "loss": 0.1964,
      "step": 1000
    },
    {
      "epoch": 1.8566176470588234,
      "grad_norm": 0.6333200931549072,
      "learning_rate": 1.5980906223115933e-07,
      "loss": 0.1917,
      "step": 1010
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.6655058264732361,
      "learning_rate": 1.220667808502951e-07,
      "loss": 0.197,
      "step": 1020
    },
    {
      "epoch": 1.8933823529411766,
      "grad_norm": 0.4924130141735077,
      "learning_rate": 8.934714979333403e-08,
      "loss": 0.2033,
      "step": 1030
    },
    {
      "epoch": 1.9117647058823528,
      "grad_norm": 0.6515531539916992,
      "learning_rate": 6.168385940783727e-08,
      "loss": 0.1886,
      "step": 1040
    },
    {
      "epoch": 1.9301470588235294,
      "grad_norm": 0.5237254500389099,
      "learning_rate": 3.910539369064603e-08,
      "loss": 0.1924,
      "step": 1050
    },
    {
      "epoch": 1.9485294117647058,
      "grad_norm": 0.8416038751602173,
      "learning_rate": 2.1635000958836748e-08,
      "loss": 0.1904,
      "step": 1060
    },
    {
      "epoch": 1.9669117647058822,
      "grad_norm": 0.604259729385376,
      "learning_rate": 9.290669911672934e-09,
      "loss": 0.194,
      "step": 1070
    },
    {
      "epoch": 1.9852941176470589,
      "grad_norm": 0.6746427416801453,
      "learning_rate": 2.085111108227067e-09,
      "loss": 0.1885,
      "step": 1080
    }
  ],
  "logging_steps": 10,
  "max_steps": 1088,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1729409604948328e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}