| { | |
| "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": 6.952715927459061, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.1116, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 6.859193146545483, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.102, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 6.565777639760947, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0465, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 6.108480280141609, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0047, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 4.871617925450606, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.9985, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 3.1999039699037968, | |
| "learning_rate": 5e-06, | |
| "loss": 1.0007, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 2.6806211176978985, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 1.0578, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 3.8486000026761857, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.9204, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 4.105886886960529, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.8604, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0506329113924051, | |
| "grad_norm": 5.082503582675367, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.9304, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1518987341772151, | |
| "grad_norm": 4.467441598381601, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.8958, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 4.240884747894379, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9217, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3544303797468356, | |
| "grad_norm": 3.12028165969881, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.8205, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 2.7458869760234856, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.7993, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 2.657674708823102, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.8035, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6582278481012658, | |
| "grad_norm": 2.6817516300983537, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.8216, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.759493670886076, | |
| "grad_norm": 2.2602091831583033, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.8018, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 2.1544755832064566, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.7327, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 1.997116276728429, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.7324, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 1.6356079781551767, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.6981, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 1.4790667810495146, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.673, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 1.5635540286950511, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.6679, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 1.4702092959948074, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.6823, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 1.0849444291892902, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.654, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 1.2233820329094083, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.6075, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 1.3432547739995635, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.6025, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 1.0984593506495701, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6009, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 1.0839824029546337, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.6277, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 0.9890509503537741, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.557, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 1.0858741654218507, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.5271, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 0.9658067782565661, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.4551, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 1.0407905159149549, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.4715, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 1.4585044872514217, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.5413, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 1.120938323819189, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.4931, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 0.9481215980814871, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.4498, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.9180869565217409, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.4714, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 1.025023355440762, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.4662, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 0.9845064692655155, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.4595, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.9200704088002243, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.4121, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 1.0152238385229302, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.3653, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 1.0864708674809789, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.3247, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 1.3006154670547314, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.4197, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 0.9343933959793107, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.3576, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 0.9022203276290396, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.2823, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.927111049921718, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.3542, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.8657489477802025, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.3507, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 0.9227409541783904, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.3537, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.8264645840258329, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.3114, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.772483148168408, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.2809, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.8448025754998916, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.2551, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.849252745196389, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.2661, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 0.9484138991029252, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.3145, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 0.9138558042285672, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.287, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.8610453729969747, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.2393, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.8789963141765615, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.2291, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.757457197537277, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.1802, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.7527229337854812, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.2452, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.6236266091084942, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.1694, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.7808354030346378, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.242, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.6041758814057354, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.1846, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.805716461556163, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.2197, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.8271535685650299, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.206, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.7729824914975278, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.2011, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.7139313732624798, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.2031, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.7404211648820929, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.1377, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.7853243127909358, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.1594, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.6427428578986121, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.1829, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.5869536423727817, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.1637, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.6290360063402625, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.1785, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.6181798961576288, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.1815, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.6903610351107097, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.1336, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.7800240304772332, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.1147, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.6030791073484069, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.1201, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.5580387579722381, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.1217, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.5836370802326925, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.1196, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.6249797302419956, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.1076, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.5716950038286982, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.1043, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.6257063732501265, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.1175, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.4859152711965658, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.1348, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.5041316931408406, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.1078, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.5779324168457304, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.1119, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.5592787362574003, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1003, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.5307478547514555, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.0719, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.6246977499074344, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.1181, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.5520913554192677, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.1476, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.4762828336737609, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.0903, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.5813749753321199, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.0816, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.5922601920041396, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.1017, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.5665475175534043, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.0833, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.41084743956663994, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.0999, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.3608761095602294, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.0864, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.5028043279664685, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.0808, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.4688226417833508, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.0916, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.523910629923862, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.0998, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.50602258789084, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.1002, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.3502262434093914, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.0937, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.3421654905618719, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.0614, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.32659256044332824, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.0649, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.3969976332986778, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.0803, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.3479668461217749, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.0691, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.3572945993054664, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.0767, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.33703232824024265, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.0921, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.3854610323789469, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.094, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.37034704323826495, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.1082, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.299329283974854, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.0675, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.3295073092390014, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.0806, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.37612848486579653, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.0515, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.3492672733144611, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.092, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.32067686663362877, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.0645, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.3938453273970925, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.0911, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.33351873137488847, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.0807, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.32585236288100555, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.0585, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.34040979624504397, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.0655, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.374831447456687, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.1121, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.2637489376091362, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.0514, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.3091550148647095, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.0518, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.3333224331697071, | |
| "learning_rate": 0.0, | |
| "loss": 0.0753, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "step": 117, | |
| "total_flos": 6.065590739822182e+16, | |
| "train_loss": 0.3474930042448716, | |
| "train_runtime": 5248.1467, | |
| "train_samples_per_second": 0.783, | |
| "train_steps_per_second": 0.022 | |
| } | |
| ], | |
| "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": 6.065590739822182e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |