kanishka's picture
End of training
e3d39f7 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 20.0,
"eval_steps": 500,
"global_step": 371860,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.05,
"learning_rate": 3.125e-05,
"loss": 6.2331,
"step": 1000
},
{
"epoch": 0.11,
"learning_rate": 6.25e-05,
"loss": 5.0052,
"step": 2000
},
{
"epoch": 0.16,
"learning_rate": 9.375e-05,
"loss": 4.6819,
"step": 3000
},
{
"epoch": 0.22,
"learning_rate": 0.000125,
"loss": 4.4603,
"step": 4000
},
{
"epoch": 0.27,
"learning_rate": 0.00015625,
"loss": 4.2982,
"step": 5000
},
{
"epoch": 0.32,
"learning_rate": 0.0001875,
"loss": 4.1755,
"step": 6000
},
{
"epoch": 0.38,
"learning_rate": 0.00021875,
"loss": 4.0781,
"step": 7000
},
{
"epoch": 0.43,
"learning_rate": 0.00025,
"loss": 3.9849,
"step": 8000
},
{
"epoch": 0.48,
"learning_rate": 0.00028125000000000003,
"loss": 3.9024,
"step": 9000
},
{
"epoch": 0.54,
"learning_rate": 0.00031246875000000003,
"loss": 3.8519,
"step": 10000
},
{
"epoch": 0.59,
"learning_rate": 0.00034371875,
"loss": 3.8,
"step": 11000
},
{
"epoch": 0.65,
"learning_rate": 0.00037490625,
"loss": 3.7622,
"step": 12000
},
{
"epoch": 0.7,
"learning_rate": 0.00040615625,
"loss": 3.7216,
"step": 13000
},
{
"epoch": 0.75,
"learning_rate": 0.00043740625,
"loss": 3.6874,
"step": 14000
},
{
"epoch": 0.81,
"learning_rate": 0.00046865625,
"loss": 3.664,
"step": 15000
},
{
"epoch": 0.86,
"learning_rate": 0.000499875,
"loss": 3.6397,
"step": 16000
},
{
"epoch": 0.91,
"learning_rate": 0.000531125,
"loss": 3.6173,
"step": 17000
},
{
"epoch": 0.97,
"learning_rate": 0.0005623749999999999,
"loss": 3.5992,
"step": 18000
},
{
"epoch": 1.0,
"eval_accuracy": 0.3589761039551882,
"eval_loss": 3.787278413772583,
"eval_runtime": 151.7197,
"eval_samples_per_second": 381.75,
"eval_steps_per_second": 5.965,
"step": 18593
},
{
"epoch": 1.02,
"learning_rate": 0.000593625,
"loss": 3.5759,
"step": 19000
},
{
"epoch": 1.08,
"learning_rate": 0.0006248437500000001,
"loss": 3.5434,
"step": 20000
},
{
"epoch": 1.13,
"learning_rate": 0.00065609375,
"loss": 3.5316,
"step": 21000
},
{
"epoch": 1.18,
"learning_rate": 0.00068734375,
"loss": 3.5214,
"step": 22000
},
{
"epoch": 1.24,
"learning_rate": 0.0007185625,
"loss": 3.5143,
"step": 23000
},
{
"epoch": 1.29,
"learning_rate": 0.0007498125,
"loss": 3.5022,
"step": 24000
},
{
"epoch": 1.34,
"learning_rate": 0.0007810312499999999,
"loss": 3.4899,
"step": 25000
},
{
"epoch": 1.4,
"learning_rate": 0.00081228125,
"loss": 3.4847,
"step": 26000
},
{
"epoch": 1.45,
"learning_rate": 0.00084353125,
"loss": 3.473,
"step": 27000
},
{
"epoch": 1.51,
"learning_rate": 0.0008747500000000001,
"loss": 3.4644,
"step": 28000
},
{
"epoch": 1.56,
"learning_rate": 0.000906,
"loss": 3.4559,
"step": 29000
},
{
"epoch": 1.61,
"learning_rate": 0.00093721875,
"loss": 3.4488,
"step": 30000
},
{
"epoch": 1.67,
"learning_rate": 0.00096846875,
"loss": 3.4366,
"step": 31000
},
{
"epoch": 1.72,
"learning_rate": 0.0009996875,
"loss": 3.4327,
"step": 32000
},
{
"epoch": 1.77,
"learning_rate": 0.0009970870358382863,
"loss": 3.4229,
"step": 33000
},
{
"epoch": 1.83,
"learning_rate": 0.0009941446477961513,
"loss": 3.4124,
"step": 34000
},
{
"epoch": 1.88,
"learning_rate": 0.0009912052021420585,
"loss": 3.4076,
"step": 35000
},
{
"epoch": 1.94,
"learning_rate": 0.0009882628140999235,
"loss": 3.3927,
"step": 36000
},
{
"epoch": 1.99,
"learning_rate": 0.0009853233684458307,
"loss": 3.3841,
"step": 37000
},
{
"epoch": 2.0,
"eval_accuracy": 0.3797898218651393,
"eval_loss": 3.595651149749756,
"eval_runtime": 153.0644,
"eval_samples_per_second": 378.396,
"eval_steps_per_second": 5.913,
"step": 37186
},
{
"epoch": 2.04,
"learning_rate": 0.0009823809804036957,
"loss": 3.3367,
"step": 38000
},
{
"epoch": 2.1,
"learning_rate": 0.0009794415347496028,
"loss": 3.3239,
"step": 39000
},
{
"epoch": 2.15,
"learning_rate": 0.0009764991467074677,
"loss": 3.3239,
"step": 40000
},
{
"epoch": 2.21,
"learning_rate": 0.000973559701053375,
"loss": 3.3189,
"step": 41000
},
{
"epoch": 2.26,
"learning_rate": 0.0009706173130112399,
"loss": 3.3103,
"step": 42000
},
{
"epoch": 2.31,
"learning_rate": 0.000967674924969105,
"loss": 3.3087,
"step": 43000
},
{
"epoch": 2.37,
"learning_rate": 0.0009647325369269699,
"loss": 3.3047,
"step": 44000
},
{
"epoch": 2.42,
"learning_rate": 0.000961793091272877,
"loss": 3.2976,
"step": 45000
},
{
"epoch": 2.47,
"learning_rate": 0.0009588507032307421,
"loss": 3.2965,
"step": 46000
},
{
"epoch": 2.53,
"learning_rate": 0.0009559112575766492,
"loss": 3.2912,
"step": 47000
},
{
"epoch": 2.58,
"learning_rate": 0.0009529688695345143,
"loss": 3.2844,
"step": 48000
},
{
"epoch": 2.64,
"learning_rate": 0.0009500264814923793,
"loss": 3.2795,
"step": 49000
},
{
"epoch": 2.69,
"learning_rate": 0.0009470870358382863,
"loss": 3.2707,
"step": 50000
},
{
"epoch": 2.74,
"learning_rate": 0.0009441446477961514,
"loss": 3.2716,
"step": 51000
},
{
"epoch": 2.8,
"learning_rate": 0.0009412052021420585,
"loss": 3.266,
"step": 52000
},
{
"epoch": 2.85,
"learning_rate": 0.0009382628140999236,
"loss": 3.2643,
"step": 53000
},
{
"epoch": 2.9,
"learning_rate": 0.0009353233684458307,
"loss": 3.2616,
"step": 54000
},
{
"epoch": 2.96,
"learning_rate": 0.0009323809804036956,
"loss": 3.2517,
"step": 55000
},
{
"epoch": 3.0,
"eval_accuracy": 0.39354396919406365,
"eval_loss": 3.4367847442626953,
"eval_runtime": 153.2306,
"eval_samples_per_second": 377.986,
"eval_steps_per_second": 5.906,
"step": 55779
},
{
"epoch": 3.01,
"learning_rate": 0.0009294385923615607,
"loss": 3.2369,
"step": 56000
},
{
"epoch": 3.07,
"learning_rate": 0.0009264962043194257,
"loss": 3.1843,
"step": 57000
},
{
"epoch": 3.12,
"learning_rate": 0.0009235538162772906,
"loss": 3.1894,
"step": 58000
},
{
"epoch": 3.17,
"learning_rate": 0.0009206143706231979,
"loss": 3.1948,
"step": 59000
},
{
"epoch": 3.23,
"learning_rate": 0.0009176719825810627,
"loss": 3.1969,
"step": 60000
},
{
"epoch": 3.28,
"learning_rate": 0.0009147325369269699,
"loss": 3.1937,
"step": 61000
},
{
"epoch": 3.33,
"learning_rate": 0.0009117901488848349,
"loss": 3.1954,
"step": 62000
},
{
"epoch": 3.39,
"learning_rate": 0.0009088477608426999,
"loss": 3.1931,
"step": 63000
},
{
"epoch": 3.44,
"learning_rate": 0.0009059083151886071,
"loss": 3.1939,
"step": 64000
},
{
"epoch": 3.5,
"learning_rate": 0.000902965927146472,
"loss": 3.1922,
"step": 65000
},
{
"epoch": 3.55,
"learning_rate": 0.0009000235391043371,
"loss": 3.1916,
"step": 66000
},
{
"epoch": 3.6,
"learning_rate": 0.0008970811510622021,
"loss": 3.185,
"step": 67000
},
{
"epoch": 3.66,
"learning_rate": 0.0008941417054081092,
"loss": 3.1838,
"step": 68000
},
{
"epoch": 3.71,
"learning_rate": 0.0008911993173659743,
"loss": 3.1852,
"step": 69000
},
{
"epoch": 3.76,
"learning_rate": 0.0008882569293238393,
"loss": 3.1822,
"step": 70000
},
{
"epoch": 3.82,
"learning_rate": 0.0008853145412817042,
"loss": 3.1802,
"step": 71000
},
{
"epoch": 3.87,
"learning_rate": 0.0008823750956276114,
"loss": 3.1848,
"step": 72000
},
{
"epoch": 3.93,
"learning_rate": 0.0008794356499735185,
"loss": 3.1773,
"step": 73000
},
{
"epoch": 3.98,
"learning_rate": 0.0008764932619313836,
"loss": 3.1729,
"step": 74000
},
{
"epoch": 4.0,
"eval_accuracy": 0.39773256398661677,
"eval_loss": 3.415771961212158,
"eval_runtime": 153.1943,
"eval_samples_per_second": 378.075,
"eval_steps_per_second": 5.908,
"step": 74372
},
{
"epoch": 4.03,
"learning_rate": 0.0008735538162772907,
"loss": 3.1341,
"step": 75000
},
{
"epoch": 4.09,
"learning_rate": 0.0008706143706231978,
"loss": 3.1157,
"step": 76000
},
{
"epoch": 4.14,
"learning_rate": 0.0008676719825810628,
"loss": 3.1166,
"step": 77000
},
{
"epoch": 4.2,
"learning_rate": 0.0008647295945389278,
"loss": 3.1151,
"step": 78000
},
{
"epoch": 4.25,
"learning_rate": 0.0008617872064967928,
"loss": 3.1233,
"step": 79000
},
{
"epoch": 4.3,
"learning_rate": 0.0008588448184546578,
"loss": 3.1252,
"step": 80000
},
{
"epoch": 4.36,
"learning_rate": 0.0008559024304125228,
"loss": 3.1251,
"step": 81000
},
{
"epoch": 4.41,
"learning_rate": 0.0008529659271464721,
"loss": 3.1257,
"step": 82000
},
{
"epoch": 4.46,
"learning_rate": 0.000850023539104337,
"loss": 3.1237,
"step": 83000
},
{
"epoch": 4.52,
"learning_rate": 0.0008470840934502443,
"loss": 3.1284,
"step": 84000
},
{
"epoch": 4.57,
"learning_rate": 0.0008441446477961514,
"loss": 3.1222,
"step": 85000
},
{
"epoch": 4.63,
"learning_rate": 0.0008412022597540165,
"loss": 3.1203,
"step": 86000
},
{
"epoch": 4.68,
"learning_rate": 0.0008382598717118813,
"loss": 3.1211,
"step": 87000
},
{
"epoch": 4.73,
"learning_rate": 0.0008353174836697463,
"loss": 3.1236,
"step": 88000
},
{
"epoch": 4.79,
"learning_rate": 0.0008323780380156535,
"loss": 3.1184,
"step": 89000
},
{
"epoch": 4.84,
"learning_rate": 0.0008294356499735185,
"loss": 3.1184,
"step": 90000
},
{
"epoch": 4.89,
"learning_rate": 0.0008264932619313836,
"loss": 3.1263,
"step": 91000
},
{
"epoch": 4.95,
"learning_rate": 0.0008235508738892485,
"loss": 3.1228,
"step": 92000
},
{
"epoch": 5.0,
"eval_accuracy": 0.4017364784833347,
"eval_loss": 3.413193702697754,
"eval_runtime": 153.2543,
"eval_samples_per_second": 377.927,
"eval_steps_per_second": 5.905,
"step": 92965
},
{
"epoch": 5.0,
"learning_rate": 0.0008206114282351556,
"loss": 3.1212,
"step": 93000
},
{
"epoch": 5.06,
"learning_rate": 0.0008176690401930207,
"loss": 3.0496,
"step": 94000
},
{
"epoch": 5.11,
"learning_rate": 0.0008147266521508857,
"loss": 3.0647,
"step": 95000
},
{
"epoch": 5.16,
"learning_rate": 0.0008117872064967929,
"loss": 3.0644,
"step": 96000
},
{
"epoch": 5.22,
"learning_rate": 0.0008088448184546579,
"loss": 3.0695,
"step": 97000
},
{
"epoch": 5.27,
"learning_rate": 0.0008059024304125228,
"loss": 3.068,
"step": 98000
},
{
"epoch": 5.32,
"learning_rate": 0.0008029600423703878,
"loss": 3.0724,
"step": 99000
},
{
"epoch": 5.38,
"learning_rate": 0.0008000235391043371,
"loss": 3.0722,
"step": 100000
},
{
"epoch": 5.43,
"learning_rate": 0.0007970811510622022,
"loss": 3.0746,
"step": 101000
},
{
"epoch": 5.49,
"learning_rate": 0.0007941387630200672,
"loss": 3.0714,
"step": 102000
},
{
"epoch": 5.54,
"learning_rate": 0.0007911993173659742,
"loss": 3.0802,
"step": 103000
},
{
"epoch": 5.59,
"learning_rate": 0.0007882569293238392,
"loss": 3.0739,
"step": 104000
},
{
"epoch": 5.65,
"learning_rate": 0.0007853174836697464,
"loss": 3.0756,
"step": 105000
},
{
"epoch": 5.7,
"learning_rate": 0.0007823750956276114,
"loss": 3.079,
"step": 106000
},
{
"epoch": 5.75,
"learning_rate": 0.0007794327075854764,
"loss": 3.0786,
"step": 107000
},
{
"epoch": 5.81,
"learning_rate": 0.0007764903195433413,
"loss": 3.0766,
"step": 108000
},
{
"epoch": 5.86,
"learning_rate": 0.0007735479315012064,
"loss": 3.0778,
"step": 109000
},
{
"epoch": 5.92,
"learning_rate": 0.0007706055434590714,
"loss": 3.0782,
"step": 110000
},
{
"epoch": 5.97,
"learning_rate": 0.0007676690401930207,
"loss": 3.074,
"step": 111000
},
{
"epoch": 6.0,
"eval_accuracy": 0.4035115994207125,
"eval_loss": 3.3903372287750244,
"eval_runtime": 153.35,
"eval_samples_per_second": 377.691,
"eval_steps_per_second": 5.902,
"step": 111558
},
{
"epoch": 6.02,
"learning_rate": 0.0007647266521508857,
"loss": 3.0485,
"step": 112000
},
{
"epoch": 6.08,
"learning_rate": 0.0007617842641087507,
"loss": 3.012,
"step": 113000
},
{
"epoch": 6.13,
"learning_rate": 0.0007588448184546578,
"loss": 3.0174,
"step": 114000
},
{
"epoch": 6.19,
"learning_rate": 0.0007559024304125228,
"loss": 3.0215,
"step": 115000
},
{
"epoch": 6.24,
"learning_rate": 0.0007529600423703878,
"loss": 3.0252,
"step": 116000
},
{
"epoch": 6.29,
"learning_rate": 0.000750020596716295,
"loss": 3.0277,
"step": 117000
},
{
"epoch": 6.35,
"learning_rate": 0.000747081151062202,
"loss": 3.028,
"step": 118000
},
{
"epoch": 6.4,
"learning_rate": 0.0007441387630200671,
"loss": 3.0358,
"step": 119000
},
{
"epoch": 6.45,
"learning_rate": 0.0007411963749779321,
"loss": 3.0356,
"step": 120000
},
{
"epoch": 6.51,
"learning_rate": 0.0007382539869357971,
"loss": 3.0373,
"step": 121000
},
{
"epoch": 6.56,
"learning_rate": 0.0007353115988936621,
"loss": 3.0347,
"step": 122000
},
{
"epoch": 6.62,
"learning_rate": 0.0007323721532395693,
"loss": 3.0383,
"step": 123000
},
{
"epoch": 6.67,
"learning_rate": 0.0007294297651974343,
"loss": 3.0442,
"step": 124000
},
{
"epoch": 6.72,
"learning_rate": 0.0007264903195433414,
"loss": 3.0388,
"step": 125000
},
{
"epoch": 6.78,
"learning_rate": 0.0007235479315012063,
"loss": 3.0441,
"step": 126000
},
{
"epoch": 6.83,
"learning_rate": 0.0007206084858471136,
"loss": 3.0375,
"step": 127000
},
{
"epoch": 6.88,
"learning_rate": 0.0007176660978049785,
"loss": 3.0415,
"step": 128000
},
{
"epoch": 6.94,
"learning_rate": 0.0007147237097628436,
"loss": 3.0411,
"step": 129000
},
{
"epoch": 6.99,
"learning_rate": 0.0007117842641087506,
"loss": 3.0396,
"step": 130000
},
{
"epoch": 7.0,
"eval_accuracy": 0.40674031350504725,
"eval_loss": 3.373116970062256,
"eval_runtime": 153.1885,
"eval_samples_per_second": 378.09,
"eval_steps_per_second": 5.908,
"step": 130151
},
{
"epoch": 7.05,
"learning_rate": 0.0007088418760666156,
"loss": 2.9861,
"step": 131000
},
{
"epoch": 7.1,
"learning_rate": 0.0007059024304125228,
"loss": 2.9789,
"step": 132000
},
{
"epoch": 7.15,
"learning_rate": 0.0007029600423703878,
"loss": 2.9826,
"step": 133000
},
{
"epoch": 7.21,
"learning_rate": 0.0007000176543282529,
"loss": 2.9891,
"step": 134000
},
{
"epoch": 7.26,
"learning_rate": 0.0006970782086741599,
"loss": 2.9906,
"step": 135000
},
{
"epoch": 7.31,
"learning_rate": 0.0006941358206320249,
"loss": 2.9965,
"step": 136000
},
{
"epoch": 7.37,
"learning_rate": 0.00069119343258989,
"loss": 2.997,
"step": 137000
},
{
"epoch": 7.42,
"learning_rate": 0.0006882539869357971,
"loss": 2.9977,
"step": 138000
},
{
"epoch": 7.48,
"learning_rate": 0.0006853115988936622,
"loss": 3.0007,
"step": 139000
},
{
"epoch": 7.53,
"learning_rate": 0.0006823692108515272,
"loss": 3.0005,
"step": 140000
},
{
"epoch": 7.58,
"learning_rate": 0.0006794297651974342,
"loss": 3.0022,
"step": 141000
},
{
"epoch": 7.64,
"learning_rate": 0.0006764873771552993,
"loss": 3.0106,
"step": 142000
},
{
"epoch": 7.69,
"learning_rate": 0.0006735479315012064,
"loss": 3.0081,
"step": 143000
},
{
"epoch": 7.74,
"learning_rate": 0.0006706084858471136,
"loss": 3.0077,
"step": 144000
},
{
"epoch": 7.8,
"learning_rate": 0.0006676660978049786,
"loss": 3.0103,
"step": 145000
},
{
"epoch": 7.85,
"learning_rate": 0.0006647237097628435,
"loss": 3.0091,
"step": 146000
},
{
"epoch": 7.91,
"learning_rate": 0.0006617813217207086,
"loss": 3.0101,
"step": 147000
},
{
"epoch": 7.96,
"learning_rate": 0.0006588418760666157,
"loss": 3.0136,
"step": 148000
},
{
"epoch": 8.0,
"eval_accuracy": 0.40651351078451636,
"eval_loss": 3.3697192668914795,
"eval_runtime": 153.3608,
"eval_samples_per_second": 377.665,
"eval_steps_per_second": 5.901,
"step": 148744
},
{
"epoch": 8.01,
"learning_rate": 0.0006558994880244807,
"loss": 2.989,
"step": 149000
},
{
"epoch": 8.07,
"learning_rate": 0.0006529600423703879,
"loss": 2.9472,
"step": 150000
},
{
"epoch": 8.12,
"learning_rate": 0.0006500176543282527,
"loss": 2.9482,
"step": 151000
},
{
"epoch": 8.18,
"learning_rate": 0.0006470752662861178,
"loss": 2.9541,
"step": 152000
},
{
"epoch": 8.23,
"learning_rate": 0.0006441358206320249,
"loss": 2.9607,
"step": 153000
},
{
"epoch": 8.28,
"learning_rate": 0.00064119343258989,
"loss": 2.9648,
"step": 154000
},
{
"epoch": 8.34,
"learning_rate": 0.000638251044547755,
"loss": 2.9692,
"step": 155000
},
{
"epoch": 8.39,
"learning_rate": 0.000635311598893662,
"loss": 2.9685,
"step": 156000
},
{
"epoch": 8.44,
"learning_rate": 0.0006323692108515271,
"loss": 2.9732,
"step": 157000
},
{
"epoch": 8.5,
"learning_rate": 0.0006294268228093921,
"loss": 2.9715,
"step": 158000
},
{
"epoch": 8.55,
"learning_rate": 0.0006264873771552993,
"loss": 2.971,
"step": 159000
},
{
"epoch": 8.61,
"learning_rate": 0.0006235479315012064,
"loss": 2.9776,
"step": 160000
},
{
"epoch": 8.66,
"learning_rate": 0.0006206055434590713,
"loss": 2.9726,
"step": 161000
},
{
"epoch": 8.71,
"learning_rate": 0.0006176631554169364,
"loss": 2.9784,
"step": 162000
},
{
"epoch": 8.77,
"learning_rate": 0.0006147207673748014,
"loss": 2.9814,
"step": 163000
},
{
"epoch": 8.82,
"learning_rate": 0.0006117813217207086,
"loss": 2.9793,
"step": 164000
},
{
"epoch": 8.87,
"learning_rate": 0.0006088389336785736,
"loss": 2.9823,
"step": 165000
},
{
"epoch": 8.93,
"learning_rate": 0.0006058965456364386,
"loss": 2.9857,
"step": 166000
},
{
"epoch": 8.98,
"learning_rate": 0.0006029570999823457,
"loss": 2.9841,
"step": 167000
},
{
"epoch": 9.0,
"eval_accuracy": 0.4067498531929369,
"eval_loss": 3.3753931522369385,
"eval_runtime": 153.3674,
"eval_samples_per_second": 377.649,
"eval_steps_per_second": 5.901,
"step": 167337
},
{
"epoch": 9.04,
"learning_rate": 0.0006000147119402107,
"loss": 2.9443,
"step": 168000
},
{
"epoch": 9.09,
"learning_rate": 0.0005970752662861179,
"loss": 2.9196,
"step": 169000
},
{
"epoch": 9.14,
"learning_rate": 0.0005941328782439829,
"loss": 2.9292,
"step": 170000
},
{
"epoch": 9.2,
"learning_rate": 0.0005911904902018478,
"loss": 2.9344,
"step": 171000
},
{
"epoch": 9.25,
"learning_rate": 0.0005882510445477551,
"loss": 2.9343,
"step": 172000
},
{
"epoch": 9.3,
"learning_rate": 0.0005853086565056199,
"loss": 2.9359,
"step": 173000
},
{
"epoch": 9.36,
"learning_rate": 0.0005823692108515272,
"loss": 2.9391,
"step": 174000
},
{
"epoch": 9.41,
"learning_rate": 0.0005794268228093921,
"loss": 2.9477,
"step": 175000
},
{
"epoch": 9.47,
"learning_rate": 0.0005764873771552992,
"loss": 2.9482,
"step": 176000
},
{
"epoch": 9.52,
"learning_rate": 0.0005735449891131643,
"loss": 2.9481,
"step": 177000
},
{
"epoch": 9.57,
"learning_rate": 0.0005706055434590713,
"loss": 2.9467,
"step": 178000
},
{
"epoch": 9.63,
"learning_rate": 0.0005676631554169363,
"loss": 2.9526,
"step": 179000
},
{
"epoch": 9.68,
"learning_rate": 0.0005647237097628435,
"loss": 2.9502,
"step": 180000
},
{
"epoch": 9.73,
"learning_rate": 0.0005617813217207085,
"loss": 2.956,
"step": 181000
},
{
"epoch": 9.79,
"learning_rate": 0.0005588389336785736,
"loss": 2.953,
"step": 182000
},
{
"epoch": 9.84,
"learning_rate": 0.0005558994880244806,
"loss": 2.9529,
"step": 183000
},
{
"epoch": 9.9,
"learning_rate": 0.0005529570999823456,
"loss": 2.9525,
"step": 184000
},
{
"epoch": 9.95,
"learning_rate": 0.0005500176543282528,
"loss": 2.9561,
"step": 185000
},
{
"epoch": 10.0,
"eval_accuracy": 0.4088378355845541,
"eval_loss": 3.376587152481079,
"eval_runtime": 153.4717,
"eval_samples_per_second": 377.392,
"eval_steps_per_second": 5.897,
"step": 185930
},
{
"epoch": 10.0,
"learning_rate": 0.0005470752662861178,
"loss": 2.9518,
"step": 186000
},
{
"epoch": 10.06,
"learning_rate": 0.0005441328782439829,
"loss": 2.8919,
"step": 187000
},
{
"epoch": 10.11,
"learning_rate": 0.0005411904902018479,
"loss": 2.9007,
"step": 188000
},
{
"epoch": 10.17,
"learning_rate": 0.0005382481021597128,
"loss": 2.9042,
"step": 189000
},
{
"epoch": 10.22,
"learning_rate": 0.00053530865650562,
"loss": 2.9049,
"step": 190000
},
{
"epoch": 10.27,
"learning_rate": 0.000532366268463485,
"loss": 2.911,
"step": 191000
},
{
"epoch": 10.33,
"learning_rate": 0.0005294297651974343,
"loss": 2.9177,
"step": 192000
},
{
"epoch": 10.38,
"learning_rate": 0.0005264873771552992,
"loss": 2.9162,
"step": 193000
},
{
"epoch": 10.43,
"learning_rate": 0.0005235449891131642,
"loss": 2.9229,
"step": 194000
},
{
"epoch": 10.49,
"learning_rate": 0.0005206026010710293,
"loss": 2.9241,
"step": 195000
},
{
"epoch": 10.54,
"learning_rate": 0.0005176602130288942,
"loss": 2.9213,
"step": 196000
},
{
"epoch": 10.6,
"learning_rate": 0.0005147207673748015,
"loss": 2.9253,
"step": 197000
},
{
"epoch": 10.65,
"learning_rate": 0.0005117783793326664,
"loss": 2.9293,
"step": 198000
},
{
"epoch": 10.7,
"learning_rate": 0.0005088389336785735,
"loss": 2.9281,
"step": 199000
},
{
"epoch": 10.76,
"learning_rate": 0.0005058965456364385,
"loss": 2.9301,
"step": 200000
},
{
"epoch": 10.81,
"learning_rate": 0.0005029541575943035,
"loss": 2.9346,
"step": 201000
},
{
"epoch": 10.86,
"learning_rate": 0.0005000117695521686,
"loss": 2.9291,
"step": 202000
},
{
"epoch": 10.92,
"learning_rate": 0.0004970723238980757,
"loss": 2.9347,
"step": 203000
},
{
"epoch": 10.97,
"learning_rate": 0.0004941299358559406,
"loss": 2.9356,
"step": 204000
},
{
"epoch": 11.0,
"eval_accuracy": 0.40891509362028,
"eval_loss": 3.383394479751587,
"eval_runtime": 153.4677,
"eval_samples_per_second": 377.402,
"eval_steps_per_second": 5.897,
"step": 204523
},
{
"epoch": 11.03,
"learning_rate": 0.0004911904902018478,
"loss": 2.901,
"step": 205000
},
{
"epoch": 11.08,
"learning_rate": 0.0004882481021597128,
"loss": 2.8769,
"step": 206000
},
{
"epoch": 11.13,
"learning_rate": 0.00048530865650561994,
"loss": 2.8792,
"step": 207000
},
{
"epoch": 11.19,
"learning_rate": 0.00048236626846348497,
"loss": 2.8827,
"step": 208000
},
{
"epoch": 11.24,
"learning_rate": 0.00047942388042135,
"loss": 2.8869,
"step": 209000
},
{
"epoch": 11.29,
"learning_rate": 0.00047648149237921494,
"loss": 2.8893,
"step": 210000
},
{
"epoch": 11.35,
"learning_rate": 0.0004735420467251221,
"loss": 2.8978,
"step": 211000
},
{
"epoch": 11.4,
"learning_rate": 0.0004705996586829871,
"loss": 2.8971,
"step": 212000
},
{
"epoch": 11.46,
"learning_rate": 0.0004676602130288943,
"loss": 2.902,
"step": 213000
},
{
"epoch": 11.51,
"learning_rate": 0.00046471782498675926,
"loss": 2.9047,
"step": 214000
},
{
"epoch": 11.56,
"learning_rate": 0.00046177837933266644,
"loss": 2.9004,
"step": 215000
},
{
"epoch": 11.62,
"learning_rate": 0.00045883599129053137,
"loss": 2.9037,
"step": 216000
},
{
"epoch": 11.67,
"learning_rate": 0.0004558936032483964,
"loss": 2.9057,
"step": 217000
},
{
"epoch": 11.72,
"learning_rate": 0.00045295415759430353,
"loss": 2.9099,
"step": 218000
},
{
"epoch": 11.78,
"learning_rate": 0.00045001176955216857,
"loss": 2.9049,
"step": 219000
},
{
"epoch": 11.83,
"learning_rate": 0.00044706938151003355,
"loss": 2.9094,
"step": 220000
},
{
"epoch": 11.89,
"learning_rate": 0.00044412699346789853,
"loss": 2.9087,
"step": 221000
},
{
"epoch": 11.94,
"learning_rate": 0.0004411875478138057,
"loss": 2.9134,
"step": 222000
},
{
"epoch": 11.99,
"learning_rate": 0.0004382451597716707,
"loss": 2.9099,
"step": 223000
},
{
"epoch": 12.0,
"eval_accuracy": 0.41050405628200876,
"eval_loss": 3.362497568130493,
"eval_runtime": 153.5537,
"eval_samples_per_second": 377.191,
"eval_steps_per_second": 5.894,
"step": 223116
},
{
"epoch": 12.05,
"learning_rate": 0.00043530865650561997,
"loss": 2.8548,
"step": 224000
},
{
"epoch": 12.1,
"learning_rate": 0.000432366268463485,
"loss": 2.8608,
"step": 225000
},
{
"epoch": 12.16,
"learning_rate": 0.00042942388042135,
"loss": 2.8615,
"step": 226000
},
{
"epoch": 12.21,
"learning_rate": 0.0004264844347672571,
"loss": 2.8656,
"step": 227000
},
{
"epoch": 12.26,
"learning_rate": 0.00042354204672512215,
"loss": 2.8624,
"step": 228000
},
{
"epoch": 12.32,
"learning_rate": 0.0004206026010710293,
"loss": 2.8763,
"step": 229000
},
{
"epoch": 12.37,
"learning_rate": 0.0004176631554169364,
"loss": 2.8711,
"step": 230000
},
{
"epoch": 12.42,
"learning_rate": 0.0004147207673748014,
"loss": 2.8768,
"step": 231000
},
{
"epoch": 12.48,
"learning_rate": 0.00041177837933266637,
"loss": 2.8792,
"step": 232000
},
{
"epoch": 12.53,
"learning_rate": 0.00040883893367857355,
"loss": 2.8802,
"step": 233000
},
{
"epoch": 12.59,
"learning_rate": 0.00040589654563643853,
"loss": 2.8804,
"step": 234000
},
{
"epoch": 12.64,
"learning_rate": 0.00040295415759430357,
"loss": 2.8838,
"step": 235000
},
{
"epoch": 12.69,
"learning_rate": 0.00040001176955216855,
"loss": 2.8849,
"step": 236000
},
{
"epoch": 12.75,
"learning_rate": 0.0003970693815100336,
"loss": 2.8849,
"step": 237000
},
{
"epoch": 12.8,
"learning_rate": 0.0003941299358559407,
"loss": 2.8891,
"step": 238000
},
{
"epoch": 12.85,
"learning_rate": 0.0003911875478138057,
"loss": 2.884,
"step": 239000
},
{
"epoch": 12.91,
"learning_rate": 0.0003882451597716707,
"loss": 2.8928,
"step": 240000
},
{
"epoch": 12.96,
"learning_rate": 0.0003853027717295357,
"loss": 2.8924,
"step": 241000
},
{
"epoch": 13.0,
"eval_accuracy": 0.40968881605284935,
"eval_loss": 3.3680169582366943,
"eval_runtime": 153.5248,
"eval_samples_per_second": 377.261,
"eval_steps_per_second": 5.895,
"step": 241709
},
{
"epoch": 13.02,
"learning_rate": 0.00038236332607544284,
"loss": 2.8776,
"step": 242000
},
{
"epoch": 13.07,
"learning_rate": 0.00037942388042134996,
"loss": 2.8353,
"step": 243000
},
{
"epoch": 13.12,
"learning_rate": 0.000376481492379215,
"loss": 2.8328,
"step": 244000
},
{
"epoch": 13.18,
"learning_rate": 0.00037353910433708,
"loss": 2.8419,
"step": 245000
},
{
"epoch": 13.23,
"learning_rate": 0.00037059671629494496,
"loss": 2.8457,
"step": 246000
},
{
"epoch": 13.28,
"learning_rate": 0.00036765432825281,
"loss": 2.8478,
"step": 247000
},
{
"epoch": 13.34,
"learning_rate": 0.000364711940210675,
"loss": 2.8527,
"step": 248000
},
{
"epoch": 13.39,
"learning_rate": 0.0003617724945565821,
"loss": 2.8586,
"step": 249000
},
{
"epoch": 13.45,
"learning_rate": 0.00035883010651444714,
"loss": 2.8535,
"step": 250000
},
{
"epoch": 13.5,
"learning_rate": 0.0003558877184723122,
"loss": 2.8612,
"step": 251000
},
{
"epoch": 13.55,
"learning_rate": 0.0003529453304301771,
"loss": 2.8597,
"step": 252000
},
{
"epoch": 13.61,
"learning_rate": 0.0003500058847760843,
"loss": 2.8663,
"step": 253000
},
{
"epoch": 13.66,
"learning_rate": 0.00034706349673394927,
"loss": 2.8692,
"step": 254000
},
{
"epoch": 13.71,
"learning_rate": 0.0003441211086918143,
"loss": 2.8687,
"step": 255000
},
{
"epoch": 13.77,
"learning_rate": 0.0003411787206496793,
"loss": 2.8653,
"step": 256000
},
{
"epoch": 13.82,
"learning_rate": 0.0003382392749955864,
"loss": 2.868,
"step": 257000
},
{
"epoch": 13.88,
"learning_rate": 0.00033529982934149354,
"loss": 2.8669,
"step": 258000
},
{
"epoch": 13.93,
"learning_rate": 0.0003323574412993586,
"loss": 2.8693,
"step": 259000
},
{
"epoch": 13.98,
"learning_rate": 0.0003294150532572236,
"loss": 2.8738,
"step": 260000
},
{
"epoch": 14.0,
"eval_accuracy": 0.41028229212902523,
"eval_loss": 3.37663197517395,
"eval_runtime": 153.5511,
"eval_samples_per_second": 377.197,
"eval_steps_per_second": 5.894,
"step": 260302
},
{
"epoch": 14.04,
"learning_rate": 0.00032647560760313074,
"loss": 2.8337,
"step": 261000
},
{
"epoch": 14.09,
"learning_rate": 0.00032353321956099567,
"loss": 2.8196,
"step": 262000
},
{
"epoch": 14.15,
"learning_rate": 0.00032059377390690285,
"loss": 2.8247,
"step": 263000
},
{
"epoch": 14.2,
"learning_rate": 0.00031765138586476783,
"loss": 2.8264,
"step": 264000
},
{
"epoch": 14.25,
"learning_rate": 0.000314711940210675,
"loss": 2.835,
"step": 265000
},
{
"epoch": 14.31,
"learning_rate": 0.00031176955216854,
"loss": 2.8342,
"step": 266000
},
{
"epoch": 14.36,
"learning_rate": 0.000308827164126405,
"loss": 2.8358,
"step": 267000
},
{
"epoch": 14.41,
"learning_rate": 0.00030588477608427,
"loss": 2.8364,
"step": 268000
},
{
"epoch": 14.47,
"learning_rate": 0.00030294533043017714,
"loss": 2.8381,
"step": 269000
},
{
"epoch": 14.52,
"learning_rate": 0.0003000029423880422,
"loss": 2.8413,
"step": 270000
},
{
"epoch": 14.58,
"learning_rate": 0.0002970634967339493,
"loss": 2.8438,
"step": 271000
},
{
"epoch": 14.63,
"learning_rate": 0.00029412405107985643,
"loss": 2.8422,
"step": 272000
},
{
"epoch": 14.68,
"learning_rate": 0.0002911816630377214,
"loss": 2.8436,
"step": 273000
},
{
"epoch": 14.74,
"learning_rate": 0.00028823927499558645,
"loss": 2.8482,
"step": 274000
},
{
"epoch": 14.79,
"learning_rate": 0.0002852968869534514,
"loss": 2.8465,
"step": 275000
},
{
"epoch": 14.84,
"learning_rate": 0.0002823544989113164,
"loss": 2.8465,
"step": 276000
},
{
"epoch": 14.9,
"learning_rate": 0.00027941505325722354,
"loss": 2.8479,
"step": 277000
},
{
"epoch": 14.95,
"learning_rate": 0.00027647266521508857,
"loss": 2.8485,
"step": 278000
},
{
"epoch": 15.0,
"eval_accuracy": 0.4107543723177606,
"eval_loss": 3.3745954036712646,
"eval_runtime": 153.3409,
"eval_samples_per_second": 377.714,
"eval_steps_per_second": 5.902,
"step": 278895
},
{
"epoch": 15.01,
"learning_rate": 0.0002735332195609957,
"loss": 2.8453,
"step": 279000
},
{
"epoch": 15.06,
"learning_rate": 0.00027059083151886073,
"loss": 2.7994,
"step": 280000
},
{
"epoch": 15.11,
"learning_rate": 0.0002676484434767257,
"loss": 2.8052,
"step": 281000
},
{
"epoch": 15.17,
"learning_rate": 0.0002647060554345907,
"loss": 2.8091,
"step": 282000
},
{
"epoch": 15.22,
"learning_rate": 0.0002617666097804979,
"loss": 2.8143,
"step": 283000
},
{
"epoch": 15.27,
"learning_rate": 0.000258827164126405,
"loss": 2.8126,
"step": 284000
},
{
"epoch": 15.33,
"learning_rate": 0.00025588477608427004,
"loss": 2.8164,
"step": 285000
},
{
"epoch": 15.38,
"learning_rate": 0.00025294238804213497,
"loss": 2.8154,
"step": 286000
},
{
"epoch": 15.44,
"learning_rate": 0.00025,
"loss": 2.8209,
"step": 287000
},
{
"epoch": 15.49,
"learning_rate": 0.000247057611957865,
"loss": 2.8175,
"step": 288000
},
{
"epoch": 15.54,
"learning_rate": 0.00024411816630377214,
"loss": 2.8195,
"step": 289000
},
{
"epoch": 15.6,
"learning_rate": 0.00024117577826163715,
"loss": 2.8241,
"step": 290000
},
{
"epoch": 15.65,
"learning_rate": 0.0002382363326075443,
"loss": 2.8268,
"step": 291000
},
{
"epoch": 15.7,
"learning_rate": 0.00023529394456540929,
"loss": 2.8289,
"step": 292000
},
{
"epoch": 15.76,
"learning_rate": 0.0002323515565232743,
"loss": 2.8288,
"step": 293000
},
{
"epoch": 15.81,
"learning_rate": 0.0002294091684811393,
"loss": 2.8299,
"step": 294000
},
{
"epoch": 15.87,
"learning_rate": 0.00022646678043900428,
"loss": 2.8288,
"step": 295000
},
{
"epoch": 15.92,
"learning_rate": 0.00022353027717295358,
"loss": 2.8336,
"step": 296000
},
{
"epoch": 15.97,
"learning_rate": 0.00022058788913081857,
"loss": 2.834,
"step": 297000
},
{
"epoch": 16.0,
"eval_accuracy": 0.41068504162830916,
"eval_loss": 3.382263660430908,
"eval_runtime": 153.6919,
"eval_samples_per_second": 376.851,
"eval_steps_per_second": 5.888,
"step": 297488
},
{
"epoch": 16.03,
"learning_rate": 0.00021764550108868357,
"loss": 2.8051,
"step": 298000
},
{
"epoch": 16.08,
"learning_rate": 0.0002147060554345907,
"loss": 2.7898,
"step": 299000
},
{
"epoch": 16.14,
"learning_rate": 0.00021176660978049786,
"loss": 2.7887,
"step": 300000
},
{
"epoch": 16.19,
"learning_rate": 0.00020882422173836286,
"loss": 2.7911,
"step": 301000
},
{
"epoch": 16.24,
"learning_rate": 0.00020588183369622787,
"loss": 2.7958,
"step": 302000
},
{
"epoch": 16.3,
"learning_rate": 0.00020293944565409286,
"loss": 2.7962,
"step": 303000
},
{
"epoch": 16.35,
"learning_rate": 0.00019999705761195786,
"loss": 2.8024,
"step": 304000
},
{
"epoch": 16.4,
"learning_rate": 0.00019705466956982287,
"loss": 2.8012,
"step": 305000
},
{
"epoch": 16.46,
"learning_rate": 0.00019411522391573,
"loss": 2.8034,
"step": 306000
},
{
"epoch": 16.51,
"learning_rate": 0.000191172835873595,
"loss": 2.8037,
"step": 307000
},
{
"epoch": 16.57,
"learning_rate": 0.00018823339021950214,
"loss": 2.808,
"step": 308000
},
{
"epoch": 16.62,
"learning_rate": 0.00018529100217736717,
"loss": 2.8058,
"step": 309000
},
{
"epoch": 16.67,
"learning_rate": 0.0001823515565232743,
"loss": 2.8092,
"step": 310000
},
{
"epoch": 16.73,
"learning_rate": 0.0001794091684811393,
"loss": 2.8117,
"step": 311000
},
{
"epoch": 16.78,
"learning_rate": 0.0001764667804390043,
"loss": 2.809,
"step": 312000
},
{
"epoch": 16.83,
"learning_rate": 0.00017352733478491144,
"loss": 2.8096,
"step": 313000
},
{
"epoch": 16.89,
"learning_rate": 0.00017058494674277645,
"loss": 2.811,
"step": 314000
},
{
"epoch": 16.94,
"learning_rate": 0.00016764550108868358,
"loss": 2.8127,
"step": 315000
},
{
"epoch": 17.0,
"learning_rate": 0.0001647031130465486,
"loss": 2.8108,
"step": 316000
},
{
"epoch": 17.0,
"eval_accuracy": 0.4108320334388902,
"eval_loss": 3.389369010925293,
"eval_runtime": 153.3987,
"eval_samples_per_second": 377.572,
"eval_steps_per_second": 5.9,
"step": 316081
},
{
"epoch": 17.05,
"learning_rate": 0.00016176072500441357,
"loss": 2.7747,
"step": 317000
},
{
"epoch": 17.1,
"learning_rate": 0.0001588183369622786,
"loss": 2.7733,
"step": 318000
},
{
"epoch": 17.16,
"learning_rate": 0.00015587889130818573,
"loss": 2.7751,
"step": 319000
},
{
"epoch": 17.21,
"learning_rate": 0.0001529394456540929,
"loss": 2.7815,
"step": 320000
},
{
"epoch": 17.26,
"learning_rate": 0.00014999705761195787,
"loss": 2.7821,
"step": 321000
},
{
"epoch": 17.32,
"learning_rate": 0.00014705466956982285,
"loss": 2.7855,
"step": 322000
},
{
"epoch": 17.37,
"learning_rate": 0.00014411228152768789,
"loss": 2.7841,
"step": 323000
},
{
"epoch": 17.43,
"learning_rate": 0.000141172835873595,
"loss": 2.7833,
"step": 324000
},
{
"epoch": 17.48,
"learning_rate": 0.00013823044783146002,
"loss": 2.7887,
"step": 325000
},
{
"epoch": 17.53,
"learning_rate": 0.00013529394456540927,
"loss": 2.7903,
"step": 326000
},
{
"epoch": 17.59,
"learning_rate": 0.0001323515565232743,
"loss": 2.7853,
"step": 327000
},
{
"epoch": 17.64,
"learning_rate": 0.00012940916848113928,
"loss": 2.7899,
"step": 328000
},
{
"epoch": 17.69,
"learning_rate": 0.00012646972282704644,
"loss": 2.7899,
"step": 329000
},
{
"epoch": 17.75,
"learning_rate": 0.00012352733478491142,
"loss": 2.7902,
"step": 330000
},
{
"epoch": 17.8,
"learning_rate": 0.00012058494674277644,
"loss": 2.7913,
"step": 331000
},
{
"epoch": 17.86,
"learning_rate": 0.00011764550108868357,
"loss": 2.793,
"step": 332000
},
{
"epoch": 17.91,
"learning_rate": 0.00011470311304654858,
"loss": 2.7972,
"step": 333000
},
{
"epoch": 17.96,
"learning_rate": 0.00011176366739245572,
"loss": 2.7936,
"step": 334000
},
{
"epoch": 18.0,
"eval_accuracy": 0.41010540481766333,
"eval_loss": 3.4001476764678955,
"eval_runtime": 153.4809,
"eval_samples_per_second": 377.37,
"eval_steps_per_second": 5.897,
"step": 334674
},
{
"epoch": 18.02,
"learning_rate": 0.00010882127935032073,
"loss": 2.7829,
"step": 335000
},
{
"epoch": 18.07,
"learning_rate": 0.00010587889130818574,
"loss": 2.7649,
"step": 336000
},
{
"epoch": 18.13,
"learning_rate": 0.00010293650326605072,
"loss": 2.7639,
"step": 337000
},
{
"epoch": 18.18,
"learning_rate": 9.999705761195786e-05,
"loss": 2.7651,
"step": 338000
},
{
"epoch": 18.23,
"learning_rate": 9.705466956982287e-05,
"loss": 2.7705,
"step": 339000
},
{
"epoch": 18.29,
"learning_rate": 9.411522391573001e-05,
"loss": 2.7628,
"step": 340000
},
{
"epoch": 18.34,
"learning_rate": 9.117283587359502e-05,
"loss": 2.7685,
"step": 341000
},
{
"epoch": 18.39,
"learning_rate": 8.823044783146001e-05,
"loss": 2.7684,
"step": 342000
},
{
"epoch": 18.45,
"learning_rate": 8.529100217736715e-05,
"loss": 2.7667,
"step": 343000
},
{
"epoch": 18.5,
"learning_rate": 8.234861413523216e-05,
"loss": 2.7688,
"step": 344000
},
{
"epoch": 18.56,
"learning_rate": 7.940622609309716e-05,
"loss": 2.7708,
"step": 345000
},
{
"epoch": 18.61,
"learning_rate": 7.646678043900428e-05,
"loss": 2.776,
"step": 346000
},
{
"epoch": 18.66,
"learning_rate": 7.35243923968693e-05,
"loss": 2.7729,
"step": 347000
},
{
"epoch": 18.72,
"learning_rate": 7.05820043547343e-05,
"loss": 2.7735,
"step": 348000
},
{
"epoch": 18.77,
"learning_rate": 6.763961631259931e-05,
"loss": 2.7739,
"step": 349000
},
{
"epoch": 18.82,
"learning_rate": 6.470311304654858e-05,
"loss": 2.7743,
"step": 350000
},
{
"epoch": 18.88,
"learning_rate": 6.176072500441358e-05,
"loss": 2.7744,
"step": 351000
},
{
"epoch": 18.93,
"learning_rate": 5.881833696227858e-05,
"loss": 2.7746,
"step": 352000
},
{
"epoch": 18.99,
"learning_rate": 5.587594892014359e-05,
"loss": 2.7783,
"step": 353000
},
{
"epoch": 19.0,
"eval_accuracy": 0.4106592441624668,
"eval_loss": 3.4029664993286133,
"eval_runtime": 153.6736,
"eval_samples_per_second": 376.896,
"eval_steps_per_second": 5.889,
"step": 353267
},
{
"epoch": 19.04,
"learning_rate": 5.293650326605073e-05,
"loss": 2.7604,
"step": 354000
},
{
"epoch": 19.09,
"learning_rate": 4.999411522391573e-05,
"loss": 2.7572,
"step": 355000
},
{
"epoch": 19.15,
"learning_rate": 4.7051727181780734e-05,
"loss": 2.7571,
"step": 356000
},
{
"epoch": 19.2,
"learning_rate": 4.411228152768787e-05,
"loss": 2.7541,
"step": 357000
},
{
"epoch": 19.25,
"learning_rate": 4.1169893485552876e-05,
"loss": 2.755,
"step": 358000
},
{
"epoch": 19.31,
"learning_rate": 3.822750544341788e-05,
"loss": 2.7538,
"step": 359000
},
{
"epoch": 19.36,
"learning_rate": 3.528511740128288e-05,
"loss": 2.7548,
"step": 360000
},
{
"epoch": 19.42,
"learning_rate": 3.234567174719002e-05,
"loss": 2.7545,
"step": 361000
},
{
"epoch": 19.47,
"learning_rate": 2.940622609309716e-05,
"loss": 2.755,
"step": 362000
},
{
"epoch": 19.52,
"learning_rate": 2.6463838050962164e-05,
"loss": 2.7577,
"step": 363000
},
{
"epoch": 19.58,
"learning_rate": 2.35243923968693e-05,
"loss": 2.7627,
"step": 364000
},
{
"epoch": 19.63,
"learning_rate": 2.0582004354734303e-05,
"loss": 2.759,
"step": 365000
},
{
"epoch": 19.68,
"learning_rate": 1.7639616312599305e-05,
"loss": 2.7577,
"step": 366000
},
{
"epoch": 19.74,
"learning_rate": 1.4700170658506444e-05,
"loss": 2.7522,
"step": 367000
},
{
"epoch": 19.79,
"learning_rate": 1.1757782616371448e-05,
"loss": 2.7564,
"step": 368000
},
{
"epoch": 19.85,
"learning_rate": 8.815394574236451e-06,
"loss": 2.7592,
"step": 369000
},
{
"epoch": 19.9,
"learning_rate": 5.8759489201435885e-06,
"loss": 2.7558,
"step": 370000
},
{
"epoch": 19.95,
"learning_rate": 2.933560878008592e-06,
"loss": 2.755,
"step": 371000
},
{
"epoch": 20.0,
"eval_accuracy": 0.4102895476662934,
"eval_loss": 3.412034034729004,
"eval_runtime": 153.6521,
"eval_samples_per_second": 376.949,
"eval_steps_per_second": 5.89,
"step": 371860
},
{
"epoch": 20.0,
"step": 371860,
"total_flos": 1.56669402051072e+18,
"train_loss": 3.0256215252012266,
"train_runtime": 81118.5574,
"train_samples_per_second": 146.686,
"train_steps_per_second": 4.584
}
],
"logging_steps": 1000,
"max_steps": 371860,
"num_input_tokens_seen": 0,
"num_train_epochs": 20,
"save_steps": 5000,
"total_flos": 1.56669402051072e+18,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}