| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 12.30379746835443, | |
| "eval_steps": 500, | |
| "global_step": 117, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10126582278481013, | |
| "grad_norm": 6.072399482609559, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 0.9311, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 6.528417904864116, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.9497, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 6.359051167500462, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.925, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 5.565105862329674, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.844, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 4.727785661318209, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.7805, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 2.939782828489048, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7894, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 2.4973979783489733, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.7919, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 3.4573139987846306, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.6866, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 3.6563089349469426, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7305, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0506329113924051, | |
| "grad_norm": 3.567454282033785, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.6779, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1518987341772151, | |
| "grad_norm": 4.098166086636336, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.6925, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 3.6049015574707237, | |
| "learning_rate": 1e-05, | |
| "loss": 0.638, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3544303797468356, | |
| "grad_norm": 2.649495808700651, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.5902, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 1.903448209503473, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.6156, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 2.3072830102920396, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6304, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6582278481012658, | |
| "grad_norm": 2.5675698669272533, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.5867, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.759493670886076, | |
| "grad_norm": 2.3369229527199322, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.5766, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 1.7502043484517904, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.6034, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 1.3842136352111996, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.6321, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 1.191500677161361, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.5439, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 1.4627345570631196, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.5128, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 1.3954713454302212, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.5064, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 1.149689960616138, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.5034, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.8646967455048312, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.4141, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 1.211177188921121, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.4684, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 1.114994524367215, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.4384, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 1.1034819058451153, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.4866, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 0.9561311333450966, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.4938, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 0.9492493144685268, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.4079, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.8460198523294045, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.3836, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 0.8448002700469955, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.4039, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 0.9979909857423938, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.4142, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 1.008238173973131, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.2943, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 0.8117140188138161, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.3625, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 0.9323524575252192, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.3301, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.7649027000661041, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.3401, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 0.812294103768114, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.3731, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 0.8989937598022456, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.3556, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.6739283815775967, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.3495, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 0.6998338778113444, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.2985, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 0.7834404406433867, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.3125, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.7507220560040153, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.2084, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 0.8027498393678201, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.2719, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 0.6777632208321701, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.2179, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.820932146824711, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.2132, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.7671983330942551, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.2101, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 0.6479792729844295, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.2558, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.6997994303448063, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.2486, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.600614946556084, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.1955, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.6084873052924032, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.186, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.6760586464049322, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.2063, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 0.8388676036575262, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.1844, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 0.6648670967418823, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.2212, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.6913807985993646, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.2044, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.7410214359703502, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.158, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.7098765043481983, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.2431, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.5721423832787186, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.2224, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.5970589230104892, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.1807, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.5675637702182478, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.1659, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.569833648680632, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.1562, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.7751449220394955, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.0946, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.748937902790741, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.1981, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.6309090190435733, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.1124, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.6274934004644721, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.1542, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.6001914889775909, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.1885, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.6404055666945739, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.129, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.5952620582885074, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.0944, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.46524834102802404, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.1655, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.4457965027456963, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.081, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.47788652166072815, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.1226, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.6252950270769206, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.1305, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.579503825540115, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.1636, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.5243943559327514, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.1118, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.5152657410146392, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.1123, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.7529963904638194, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.0736, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.5641636199364488, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.0975, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.5480619305002655, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.0823, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.45454040006743707, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.0993, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.4707710283401514, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.134, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.46013119087249, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.1124, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.3674000562058081, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.068, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.5680967440483309, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1122, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.40616358293863153, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.0526, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.5359486096542745, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.0801, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.5997640716920127, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.097, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.37092137107517437, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.0745, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.3980635601259192, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.1041, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.3493171887225538, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.0496, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.3961182297025844, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.0799, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.44892699199272074, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.0679, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.37258247045248516, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.1011, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.35513862758990106, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.0986, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.37845306175190563, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.0913, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.5739039401971411, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.0863, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.33694694531512864, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.0587, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.2582014853110333, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.0528, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.32586737636444013, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.1201, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.49581423067659225, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.0666, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.372463197361783, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.0978, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.2895426112826978, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.0864, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.3486582961664102, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.0594, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.3073636680406654, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.0587, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.29519395152960587, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.0576, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.3034747099522094, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.0953, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.2529407827790441, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.057, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.24273949649378235, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.044, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.30954761916571744, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.0728, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.24811277009407764, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.0428, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.2989757408965876, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.086, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.22257702361173523, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.0697, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.3009281789737169, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.0781, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.2429827327502092, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.0407, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.2924882374183564, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.09, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.28205168547697096, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.0965, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.2936201996849507, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.0866, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.2535203420104844, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.0495, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.2866817448867829, | |
| "learning_rate": 0.0, | |
| "loss": 0.071, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "step": 117, | |
| "total_flos": 5.164190805262336e+16, | |
| "train_loss": 0.27243100587501484, | |
| "train_runtime": 3122.1661, | |
| "train_samples_per_second": 1.316, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 117, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "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": 5.164190805262336e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |