| { | |
| "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": 6.487739085997507, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 0.8244, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 6.257010782872823, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.8005, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 6.0499568796484535, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.795, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 5.336074973883894, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.7591, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 4.225684729831529, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.757, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 2.4648598269206192, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7443, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 1.9837709702470825, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.7004, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 3.5039128786145803, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.7013, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 4.267004233786538, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.7496, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0506329113924051, | |
| "grad_norm": 6.390009655169567, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.0467, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1518987341772151, | |
| "grad_norm": 3.8481575964942105, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.7055, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 3.483386642960325, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6902, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3544303797468356, | |
| "grad_norm": 2.0491269459986303, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.6389, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 1.562818895988004, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.6053, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 1.710709037259949, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.6248, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6582278481012658, | |
| "grad_norm": 1.5132825781260761, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.5832, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.759493670886076, | |
| "grad_norm": 1.0478138179186807, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.6072, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 0.9131672793873219, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.5818, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 1.6688149050953398, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.8314, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 0.8178363803961444, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.5352, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.8121062449916002, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.5103, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 0.9370478193496051, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.5434, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 0.6998663270374709, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.5204, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.7313686130024222, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.5108, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 0.8626878184370473, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.5301, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 0.8678070384977639, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.5237, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.7048955390634051, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.5347, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 0.6211734825674653, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.4941, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 1.2186257127731135, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.7523, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.7163249719895798, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.4766, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 0.6167728084773804, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.4715, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 0.5877254693272708, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.4053, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.673114317487297, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.4867, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 0.6047629625040978, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.4526, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 0.5768060173602658, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.455, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.6456448785604184, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.4267, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 0.5395336917689473, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.4134, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 1.016476729451887, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.6812, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.5440357050854963, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.3917, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 0.7178204930607096, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.4073, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 0.5923000027862751, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.3952, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.5312758433075424, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.39, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 0.5481292680986813, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.4016, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 0.5433592998616376, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.3841, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.5169221768248706, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.3739, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.5198192113177972, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.3855, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 0.5037430966881292, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.3636, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.804212479465148, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.561, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.5773209439403371, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.3512, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.5801145503247521, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.3451, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.4766380687877858, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.3095, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 0.6043731919022627, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.3322, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 0.5826259443445325, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.3158, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.5254698673509745, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.3381, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.43198260930804677, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.2936, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.5048637411191237, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.3255, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.9683128691761499, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.5212, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.543079050452297, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.2912, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.5964623589945404, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.2875, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.5365959324885466, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.2851, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.6841574260627378, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.2838, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.5767066827202523, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.2862, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.5369687345202557, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.2762, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.613460922979565, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.2708, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.4972076370281023, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.2538, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.5402631015512322, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.2762, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.8588042109008247, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.3676, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.4711679650181232, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.2414, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.5186059854924637, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.2369, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.5285126867852405, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.2503, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.570558167173894, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.2436, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.47801917015352163, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.2147, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.42211425280473075, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.2216, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.4670149872455072, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.2408, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.4410218475930808, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.2163, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.6700490085033992, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.3386, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.42265333632498664, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.1991, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.4496495131329505, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.1907, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.48240621681785484, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.1961, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.4157845027851205, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.2018, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.4555086067711132, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.1869, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.47102008241653187, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1977, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.4278201643332222, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.198, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.4277331561596734, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.1953, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.41215113208433574, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.1913, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.6501081020554252, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.2902, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.49351299237832125, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.1704, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.45485629042101366, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.1645, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.43594878006143273, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.1648, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.524229117333008, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.1678, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.5762699897421659, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.1797, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.44393882516851846, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.1733, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.40560141070817374, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.1764, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.46367104250146185, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.1715, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.7723363958460977, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.2587, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.5107525831617175, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.1595, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.46771561743202794, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.1514, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.3871064090616983, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.1599, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.40633333182562176, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.1588, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.5179830815968256, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.1556, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.5068373306196533, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.1564, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.4646560863434988, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.1549, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.4470174054593301, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.1593, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.41781726782966944, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.1557, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.6163624530493312, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.2388, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.3779658087921075, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.1564, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.3962142815005405, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.1531, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.37796569420656834, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.1421, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.4190534604243063, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.1396, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.3927060853622339, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.1417, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.3910392543626813, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.152, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.3832903001655876, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.1694, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.38553168802846804, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.142, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.5786623878371228, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.2282, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.3666509583277426, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.1459, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.38747422449556695, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.1477, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.33864628427433874, | |
| "learning_rate": 0.0, | |
| "loss": 0.1503, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "step": 117, | |
| "total_flos": 1.3891141298133402e+17, | |
| "train_loss": 0.36694383532063574, | |
| "train_runtime": 4769.333, | |
| "train_samples_per_second": 0.861, | |
| "train_steps_per_second": 0.025 | |
| } | |
| ], | |
| "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": 1.3891141298133402e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |