| { | |
| "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": 6.71834945708344, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.0542, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 6.714659479739678, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.0371, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 6.538347981539867, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0522, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 6.057420443703975, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0048, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 4.727346404713889, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.9911, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 2.977063854467385, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9531, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 2.626289852537077, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.9542, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 3.8074880086876273, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.9254, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 4.10056424618358, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.9318, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0506329113924051, | |
| "grad_norm": 3.9981371485737975, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.948, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1518987341772151, | |
| "grad_norm": 3.410026719114527, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.877, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 3.1763153314410584, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8772, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3544303797468356, | |
| "grad_norm": 2.3506927738700654, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.8455, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 1.7212067035140353, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.834, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 1.586919178693919, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.8086, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6582278481012658, | |
| "grad_norm": 1.4914241740919658, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.8147, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.759493670886076, | |
| "grad_norm": 1.2243745427240722, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.7868, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 1.2092398178758617, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.791, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 1.1117824298189902, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.7732, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 1.0804521826962254, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.7537, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.9963530275824865, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.7332, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 0.7876083390579625, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.7179, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 0.9495130255701153, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.6899, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.8502487990399321, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.719, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 0.9991525820194059, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.7048, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 0.7909486047117417, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.6848, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.9196150065428615, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6766, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 0.9173156913093092, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.676, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 0.8217318840304639, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.6733, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.74670828231198, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6223, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 0.7685607157823121, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.6005, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 0.8279658050914768, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.6483, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.766685566237162, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6221, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 0.8888745053090412, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.5996, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 0.8978133556608923, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.608, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.8078932401731329, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6089, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 0.7769297534125119, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.5842, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 0.6778913448585069, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.6174, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.6142738474706112, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.5844, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 0.6755984263229494, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.5802, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 0.7082946922182762, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.5567, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.609660419654126, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.4813, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 0.6472669876637741, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.5414, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 0.627004874358507, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.5246, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.6490696706201671, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5524, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.6288159897812808, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.5176, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 0.5258686116464694, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5137, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.6620191216713934, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.4797, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.6159727334750635, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.4778, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.5907229707972482, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.4827, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.62743913161887, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.4633, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 0.6405882775236793, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.4606, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 0.642266855554924, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.464, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.5685454702189753, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.4553, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.6288730659930968, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.4424, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.5545182465818295, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.4795, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.570794952025768, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.4453, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.5945530015199985, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.4455, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.6055968476231539, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.4074, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.6566239763607999, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.4345, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.6266008428283812, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.3824, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.5824604863664419, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.378, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.6157195855794587, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.4026, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.6577181070903069, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.3375, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.6960774589216226, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.4026, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.6850622105424243, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.3868, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.6383503994608305, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.3419, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.6174230804639452, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.3686, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.60151444172394, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.3663, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.6475980733169446, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.3257, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.6316912079236687, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.347, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.5538373695222134, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.3446, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.6592613322990627, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.3229, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.6168709271929764, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.3285, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.5819619207165834, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.3491, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.7355175776936779, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.3603, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.556818969666461, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.303, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.5346718398584845, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.319, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.5338068518037049, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.3095, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.549269192878999, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.2692, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.5768724755682354, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.2742, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.5545906135216732, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.299, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.49075948256913554, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.3247, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.557708338026568, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.3225, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.5419812377466001, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.2728, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.6475324596250226, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.2549, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.6055398672493939, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.243, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.638933520352964, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.262, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.894842467985003, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.2517, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.682021791675799, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.2387, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.5326965565790988, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.2754, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.7013623296758206, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.2813, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.7290133960155052, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.2731, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.6644751944316775, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.2852, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.4975157588124464, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.2698, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.5373050938849909, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.2338, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.5071978683985747, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.2536, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.5523916787178953, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.2404, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.6138370887200673, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.2137, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.700874243471082, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.2631, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.5030801253377872, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.2187, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.5763494730110765, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.2783, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.4684946499165857, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.2618, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.4850870393561313, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.2373, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.5279141522444547, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.2367, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.5847267842763526, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.2331, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.5726835215039042, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.2455, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.5656765601737138, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.2268, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.6109681446104426, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.215, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.45587164085029397, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.2527, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.4810604411567113, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.24, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.440896408526604, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.2804, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.46550921703628506, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.2412, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.44045357520064476, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.2402, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.44135518150741937, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.2584, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.42335163425515143, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.2459, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.41693789099627687, | |
| "learning_rate": 0.0, | |
| "loss": 0.2168, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "step": 117, | |
| "total_flos": 1.4162939924919091e+17, | |
| "train_loss": 0.4871606370832166, | |
| "train_runtime": 5590.4328, | |
| "train_samples_per_second": 0.735, | |
| "train_steps_per_second": 0.021 | |
| } | |
| ], | |
| "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": 1.4162939924919091e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |