deep-haiku-gpt-j-6b-8bit / trainer_state.json
fabianmmueller's picture
First model version
5588683
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 10.0,
"global_step": 157480,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.06,
"learning_rate": 4.97140678612276e-05,
"loss": 2.3365,
"step": 1000
},
{
"epoch": 0.13,
"learning_rate": 4.939636548481383e-05,
"loss": 1.9007,
"step": 2000
},
{
"epoch": 0.19,
"learning_rate": 4.907866310840005e-05,
"loss": 1.8121,
"step": 3000
},
{
"epoch": 0.25,
"learning_rate": 4.8760960731986276e-05,
"loss": 1.7655,
"step": 4000
},
{
"epoch": 0.32,
"learning_rate": 4.84432583555725e-05,
"loss": 1.7269,
"step": 5000
},
{
"epoch": 0.38,
"learning_rate": 4.8125555979158726e-05,
"loss": 1.6977,
"step": 6000
},
{
"epoch": 0.44,
"learning_rate": 4.780785360274495e-05,
"loss": 1.678,
"step": 7000
},
{
"epoch": 0.51,
"learning_rate": 4.7490151226331175e-05,
"loss": 1.6612,
"step": 8000
},
{
"epoch": 0.57,
"learning_rate": 4.71724488499174e-05,
"loss": 1.6365,
"step": 9000
},
{
"epoch": 0.64,
"learning_rate": 4.6854746473503625e-05,
"loss": 1.6223,
"step": 10000
},
{
"epoch": 0.7,
"learning_rate": 4.653704409708985e-05,
"loss": 1.6127,
"step": 11000
},
{
"epoch": 0.76,
"learning_rate": 4.6219341720676074e-05,
"loss": 1.5877,
"step": 12000
},
{
"epoch": 0.83,
"learning_rate": 4.59016393442623e-05,
"loss": 1.5809,
"step": 13000
},
{
"epoch": 0.89,
"learning_rate": 4.5583936967848524e-05,
"loss": 1.5737,
"step": 14000
},
{
"epoch": 0.95,
"learning_rate": 4.526623459143475e-05,
"loss": 1.5611,
"step": 15000
},
{
"epoch": 1.02,
"learning_rate": 4.494853221502097e-05,
"loss": 1.5297,
"step": 16000
},
{
"epoch": 1.08,
"learning_rate": 4.46308298386072e-05,
"loss": 1.4783,
"step": 17000
},
{
"epoch": 1.14,
"learning_rate": 4.431312746219342e-05,
"loss": 1.472,
"step": 18000
},
{
"epoch": 1.21,
"learning_rate": 4.399542508577965e-05,
"loss": 1.4821,
"step": 19000
},
{
"epoch": 1.27,
"learning_rate": 4.367772270936587e-05,
"loss": 1.4681,
"step": 20000
},
{
"epoch": 1.33,
"learning_rate": 4.33600203329521e-05,
"loss": 1.4564,
"step": 21000
},
{
"epoch": 1.4,
"learning_rate": 4.304231795653832e-05,
"loss": 1.4509,
"step": 22000
},
{
"epoch": 1.46,
"learning_rate": 4.2724615580124546e-05,
"loss": 1.4478,
"step": 23000
},
{
"epoch": 1.52,
"learning_rate": 4.2406913203710764e-05,
"loss": 1.4446,
"step": 24000
},
{
"epoch": 1.59,
"learning_rate": 4.208921082729699e-05,
"loss": 1.439,
"step": 25000
},
{
"epoch": 1.65,
"learning_rate": 4.1771508450883214e-05,
"loss": 1.4367,
"step": 26000
},
{
"epoch": 1.71,
"learning_rate": 4.145380607446944e-05,
"loss": 1.4181,
"step": 27000
},
{
"epoch": 1.78,
"learning_rate": 4.113610369805566e-05,
"loss": 1.417,
"step": 28000
},
{
"epoch": 1.84,
"learning_rate": 4.081840132164189e-05,
"loss": 1.4056,
"step": 29000
},
{
"epoch": 1.91,
"learning_rate": 4.050069894522811e-05,
"loss": 1.4006,
"step": 30000
},
{
"epoch": 1.97,
"learning_rate": 4.018299656881434e-05,
"loss": 1.4053,
"step": 31000
},
{
"epoch": 2.03,
"learning_rate": 3.986529419240056e-05,
"loss": 1.3491,
"step": 32000
},
{
"epoch": 2.1,
"learning_rate": 3.954759181598679e-05,
"loss": 1.3028,
"step": 33000
},
{
"epoch": 2.16,
"learning_rate": 3.922988943957301e-05,
"loss": 1.3012,
"step": 34000
},
{
"epoch": 2.22,
"learning_rate": 3.8912187063159236e-05,
"loss": 1.2896,
"step": 35000
},
{
"epoch": 2.29,
"learning_rate": 3.859448468674546e-05,
"loss": 1.2955,
"step": 36000
},
{
"epoch": 2.35,
"learning_rate": 3.8276782310331686e-05,
"loss": 1.2843,
"step": 37000
},
{
"epoch": 2.41,
"learning_rate": 3.7959079933917904e-05,
"loss": 1.2876,
"step": 38000
},
{
"epoch": 2.48,
"learning_rate": 3.764137755750413e-05,
"loss": 1.2826,
"step": 39000
},
{
"epoch": 2.54,
"learning_rate": 3.732367518109035e-05,
"loss": 1.2706,
"step": 40000
},
{
"epoch": 2.6,
"learning_rate": 3.700597280467658e-05,
"loss": 1.2743,
"step": 41000
},
{
"epoch": 2.67,
"learning_rate": 3.66882704282628e-05,
"loss": 1.2783,
"step": 42000
},
{
"epoch": 2.73,
"learning_rate": 3.637056805184903e-05,
"loss": 1.264,
"step": 43000
},
{
"epoch": 2.79,
"learning_rate": 3.605286567543525e-05,
"loss": 1.2614,
"step": 44000
},
{
"epoch": 2.86,
"learning_rate": 3.573516329902148e-05,
"loss": 1.2549,
"step": 45000
},
{
"epoch": 2.92,
"learning_rate": 3.54174609226077e-05,
"loss": 1.2453,
"step": 46000
},
{
"epoch": 2.98,
"learning_rate": 3.5099758546193926e-05,
"loss": 1.2425,
"step": 47000
},
{
"epoch": 3.05,
"learning_rate": 3.478205616978015e-05,
"loss": 1.1538,
"step": 48000
},
{
"epoch": 3.11,
"learning_rate": 3.4464353793366376e-05,
"loss": 1.1307,
"step": 49000
},
{
"epoch": 3.18,
"learning_rate": 3.41466514169526e-05,
"loss": 1.1375,
"step": 50000
},
{
"epoch": 3.24,
"learning_rate": 3.3828949040538825e-05,
"loss": 1.1258,
"step": 51000
},
{
"epoch": 3.3,
"learning_rate": 3.351124666412505e-05,
"loss": 1.1237,
"step": 52000
},
{
"epoch": 3.37,
"learning_rate": 3.3193544287711275e-05,
"loss": 1.1226,
"step": 53000
},
{
"epoch": 3.43,
"learning_rate": 3.28758419112975e-05,
"loss": 1.1246,
"step": 54000
},
{
"epoch": 3.49,
"learning_rate": 3.2558139534883724e-05,
"loss": 1.1197,
"step": 55000
},
{
"epoch": 3.56,
"learning_rate": 3.224043715846995e-05,
"loss": 1.117,
"step": 56000
},
{
"epoch": 3.62,
"learning_rate": 3.1922734782056174e-05,
"loss": 1.1206,
"step": 57000
},
{
"epoch": 3.68,
"learning_rate": 3.160503240564239e-05,
"loss": 1.1023,
"step": 58000
},
{
"epoch": 3.75,
"learning_rate": 3.1287330029228617e-05,
"loss": 1.1076,
"step": 59000
},
{
"epoch": 3.81,
"learning_rate": 3.096962765281484e-05,
"loss": 1.1061,
"step": 60000
},
{
"epoch": 3.87,
"learning_rate": 3.0651925276401066e-05,
"loss": 1.1017,
"step": 61000
},
{
"epoch": 3.94,
"learning_rate": 3.033422289998729e-05,
"loss": 1.1005,
"step": 62000
},
{
"epoch": 4.0,
"learning_rate": 3.0016520523573516e-05,
"loss": 1.0891,
"step": 63000
},
{
"epoch": 4.06,
"learning_rate": 2.969881814715974e-05,
"loss": 0.9693,
"step": 64000
},
{
"epoch": 4.13,
"learning_rate": 2.9381115770745965e-05,
"loss": 0.9703,
"step": 65000
},
{
"epoch": 4.19,
"learning_rate": 2.906341339433219e-05,
"loss": 0.9693,
"step": 66000
},
{
"epoch": 4.25,
"learning_rate": 2.8745711017918414e-05,
"loss": 0.9698,
"step": 67000
},
{
"epoch": 4.32,
"learning_rate": 2.842800864150464e-05,
"loss": 0.9715,
"step": 68000
},
{
"epoch": 4.38,
"learning_rate": 2.8110306265090864e-05,
"loss": 0.9734,
"step": 69000
},
{
"epoch": 4.45,
"learning_rate": 2.779260388867709e-05,
"loss": 0.9796,
"step": 70000
},
{
"epoch": 4.51,
"learning_rate": 2.7474901512263313e-05,
"loss": 0.9729,
"step": 71000
},
{
"epoch": 4.57,
"learning_rate": 2.7157199135849538e-05,
"loss": 0.9727,
"step": 72000
},
{
"epoch": 4.64,
"learning_rate": 2.6839496759435763e-05,
"loss": 0.9721,
"step": 73000
},
{
"epoch": 4.7,
"learning_rate": 2.6521794383021988e-05,
"loss": 0.9643,
"step": 74000
},
{
"epoch": 4.76,
"learning_rate": 2.6204092006608212e-05,
"loss": 0.9599,
"step": 75000
},
{
"epoch": 4.83,
"learning_rate": 2.5886389630194434e-05,
"loss": 0.9651,
"step": 76000
},
{
"epoch": 4.89,
"learning_rate": 2.556868725378066e-05,
"loss": 0.9587,
"step": 77000
},
{
"epoch": 4.95,
"learning_rate": 2.5250984877366883e-05,
"loss": 0.9568,
"step": 78000
},
{
"epoch": 5.02,
"learning_rate": 2.4933282500953108e-05,
"loss": 0.9187,
"step": 79000
},
{
"epoch": 5.08,
"learning_rate": 2.4615580124539333e-05,
"loss": 0.8357,
"step": 80000
},
{
"epoch": 5.14,
"learning_rate": 2.4297877748125557e-05,
"loss": 0.8324,
"step": 81000
},
{
"epoch": 5.21,
"learning_rate": 2.3980175371711782e-05,
"loss": 0.84,
"step": 82000
},
{
"epoch": 5.27,
"learning_rate": 2.3662472995298007e-05,
"loss": 0.8392,
"step": 83000
},
{
"epoch": 5.33,
"learning_rate": 2.3344770618884228e-05,
"loss": 0.8439,
"step": 84000
},
{
"epoch": 5.4,
"learning_rate": 2.3027068242470453e-05,
"loss": 0.8375,
"step": 85000
},
{
"epoch": 5.46,
"learning_rate": 2.2709365866056678e-05,
"loss": 0.8363,
"step": 86000
},
{
"epoch": 5.52,
"learning_rate": 2.2391663489642903e-05,
"loss": 0.8387,
"step": 87000
},
{
"epoch": 5.59,
"learning_rate": 2.2073961113229127e-05,
"loss": 0.8458,
"step": 88000
},
{
"epoch": 5.65,
"learning_rate": 2.1756258736815352e-05,
"loss": 0.84,
"step": 89000
},
{
"epoch": 5.72,
"learning_rate": 2.1438556360401577e-05,
"loss": 0.833,
"step": 90000
},
{
"epoch": 5.78,
"learning_rate": 2.11208539839878e-05,
"loss": 0.8356,
"step": 91000
},
{
"epoch": 5.84,
"learning_rate": 2.0803151607574026e-05,
"loss": 0.8367,
"step": 92000
},
{
"epoch": 5.91,
"learning_rate": 2.048544923116025e-05,
"loss": 0.8323,
"step": 93000
},
{
"epoch": 5.97,
"learning_rate": 2.0167746854746476e-05,
"loss": 0.8304,
"step": 94000
},
{
"epoch": 6.03,
"learning_rate": 1.98500444783327e-05,
"loss": 0.7779,
"step": 95000
},
{
"epoch": 6.1,
"learning_rate": 1.9532342101918925e-05,
"loss": 0.7192,
"step": 96000
},
{
"epoch": 6.16,
"learning_rate": 1.921463972550515e-05,
"loss": 0.7189,
"step": 97000
},
{
"epoch": 6.22,
"learning_rate": 1.8896937349091375e-05,
"loss": 0.7193,
"step": 98000
},
{
"epoch": 6.29,
"learning_rate": 1.85792349726776e-05,
"loss": 0.7281,
"step": 99000
},
{
"epoch": 6.35,
"learning_rate": 1.826153259626382e-05,
"loss": 0.7303,
"step": 100000
},
{
"epoch": 6.41,
"learning_rate": 1.7943830219850045e-05,
"loss": 0.7308,
"step": 101000
},
{
"epoch": 6.48,
"learning_rate": 1.762612784343627e-05,
"loss": 0.7285,
"step": 102000
},
{
"epoch": 6.54,
"learning_rate": 1.730842546702249e-05,
"loss": 0.7279,
"step": 103000
},
{
"epoch": 6.6,
"learning_rate": 1.6990723090608716e-05,
"loss": 0.7323,
"step": 104000
},
{
"epoch": 6.67,
"learning_rate": 1.667302071419494e-05,
"loss": 0.7309,
"step": 105000
},
{
"epoch": 6.73,
"learning_rate": 1.6355318337781166e-05,
"loss": 0.729,
"step": 106000
},
{
"epoch": 6.79,
"learning_rate": 1.603761596136739e-05,
"loss": 0.7285,
"step": 107000
},
{
"epoch": 6.86,
"learning_rate": 1.5719913584953615e-05,
"loss": 0.7269,
"step": 108000
},
{
"epoch": 6.92,
"learning_rate": 1.540221120853984e-05,
"loss": 0.7236,
"step": 109000
},
{
"epoch": 6.99,
"learning_rate": 1.5084508832126065e-05,
"loss": 0.7238,
"step": 110000
},
{
"epoch": 7.05,
"learning_rate": 1.476680645571229e-05,
"loss": 0.6498,
"step": 111000
},
{
"epoch": 7.11,
"learning_rate": 1.4449104079298514e-05,
"loss": 0.6289,
"step": 112000
},
{
"epoch": 7.18,
"learning_rate": 1.4131401702884739e-05,
"loss": 0.629,
"step": 113000
},
{
"epoch": 7.24,
"learning_rate": 1.3813699326470964e-05,
"loss": 0.6331,
"step": 114000
},
{
"epoch": 7.3,
"learning_rate": 1.3495996950057188e-05,
"loss": 0.6354,
"step": 115000
},
{
"epoch": 7.37,
"learning_rate": 1.3178294573643413e-05,
"loss": 0.634,
"step": 116000
},
{
"epoch": 7.43,
"learning_rate": 1.2860592197229638e-05,
"loss": 0.636,
"step": 117000
},
{
"epoch": 7.49,
"learning_rate": 1.254288982081586e-05,
"loss": 0.6422,
"step": 118000
},
{
"epoch": 7.56,
"learning_rate": 1.2225187444402086e-05,
"loss": 0.6371,
"step": 119000
},
{
"epoch": 7.62,
"learning_rate": 1.1907485067988309e-05,
"loss": 0.6383,
"step": 120000
},
{
"epoch": 7.68,
"learning_rate": 1.1589782691574534e-05,
"loss": 0.6401,
"step": 121000
},
{
"epoch": 7.75,
"learning_rate": 1.1272080315160757e-05,
"loss": 0.6332,
"step": 122000
},
{
"epoch": 7.81,
"learning_rate": 1.0954377938746981e-05,
"loss": 0.6408,
"step": 123000
},
{
"epoch": 7.87,
"learning_rate": 1.0636675562333206e-05,
"loss": 0.6337,
"step": 124000
},
{
"epoch": 7.94,
"learning_rate": 1.031897318591943e-05,
"loss": 0.6406,
"step": 125000
},
{
"epoch": 8.0,
"learning_rate": 1.0001270809505656e-05,
"loss": 0.6372,
"step": 126000
},
{
"epoch": 8.06,
"learning_rate": 9.68356843309188e-06,
"loss": 0.5542,
"step": 127000
},
{
"epoch": 8.13,
"learning_rate": 9.365866056678105e-06,
"loss": 0.5541,
"step": 128000
},
{
"epoch": 8.19,
"learning_rate": 9.04816368026433e-06,
"loss": 0.565,
"step": 129000
},
{
"epoch": 8.26,
"learning_rate": 8.730461303850553e-06,
"loss": 0.5644,
"step": 130000
},
{
"epoch": 8.32,
"learning_rate": 8.412758927436778e-06,
"loss": 0.5651,
"step": 131000
},
{
"epoch": 8.38,
"learning_rate": 8.095056551023002e-06,
"loss": 0.5661,
"step": 132000
},
{
"epoch": 8.45,
"learning_rate": 7.777354174609227e-06,
"loss": 0.5704,
"step": 133000
},
{
"epoch": 8.51,
"learning_rate": 7.459651798195451e-06,
"loss": 0.5669,
"step": 134000
},
{
"epoch": 8.57,
"learning_rate": 7.141949421781676e-06,
"loss": 0.5694,
"step": 135000
},
{
"epoch": 8.64,
"learning_rate": 6.8242470453679e-06,
"loss": 0.5705,
"step": 136000
},
{
"epoch": 8.7,
"learning_rate": 6.506544668954124e-06,
"loss": 0.5682,
"step": 137000
},
{
"epoch": 8.76,
"learning_rate": 6.188842292540348e-06,
"loss": 0.5676,
"step": 138000
},
{
"epoch": 8.83,
"learning_rate": 5.871139916126573e-06,
"loss": 0.5674,
"step": 139000
},
{
"epoch": 8.89,
"learning_rate": 5.553437539712798e-06,
"loss": 0.5681,
"step": 140000
},
{
"epoch": 8.95,
"learning_rate": 5.2357351632990216e-06,
"loss": 0.5636,
"step": 141000
},
{
"epoch": 9.02,
"learning_rate": 4.918032786885246e-06,
"loss": 0.5495,
"step": 142000
},
{
"epoch": 9.08,
"learning_rate": 4.60033041047147e-06,
"loss": 0.5124,
"step": 143000
},
{
"epoch": 9.14,
"learning_rate": 4.282628034057695e-06,
"loss": 0.514,
"step": 144000
},
{
"epoch": 9.21,
"learning_rate": 3.964925657643919e-06,
"loss": 0.517,
"step": 145000
},
{
"epoch": 9.27,
"learning_rate": 3.6472232812301436e-06,
"loss": 0.5173,
"step": 146000
},
{
"epoch": 9.33,
"learning_rate": 3.3295209048163683e-06,
"loss": 0.5164,
"step": 147000
},
{
"epoch": 9.4,
"learning_rate": 3.0118185284025926e-06,
"loss": 0.5202,
"step": 148000
},
{
"epoch": 9.46,
"learning_rate": 2.694116151988817e-06,
"loss": 0.5183,
"step": 149000
},
{
"epoch": 9.53,
"learning_rate": 2.3764137755750413e-06,
"loss": 0.517,
"step": 150000
},
{
"epoch": 9.59,
"learning_rate": 2.058711399161266e-06,
"loss": 0.5131,
"step": 151000
},
{
"epoch": 9.65,
"learning_rate": 1.7410090227474903e-06,
"loss": 0.516,
"step": 152000
},
{
"epoch": 9.72,
"learning_rate": 1.4233066463337146e-06,
"loss": 0.5135,
"step": 153000
},
{
"epoch": 9.78,
"learning_rate": 1.105604269919939e-06,
"loss": 0.5121,
"step": 154000
},
{
"epoch": 9.84,
"learning_rate": 7.879018935061634e-07,
"loss": 0.514,
"step": 155000
},
{
"epoch": 9.91,
"learning_rate": 4.701995170923879e-07,
"loss": 0.5134,
"step": 156000
},
{
"epoch": 9.97,
"learning_rate": 1.5249714067861227e-07,
"loss": 0.5109,
"step": 157000
}
],
"max_steps": 157480,
"num_train_epochs": 10,
"total_flos": 1.68490001031168e+16,
"trial_name": null,
"trial_params": null
}