| { | |
| "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": 7.910776869268381, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.2102, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 8.068097740119322, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.3049, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 8.767932052986346, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.3678, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 8.57574270903701, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.3404, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 6.336407444734683, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 1.2558, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 3.8435756430110986, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1516, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 3.3689303522600196, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 1.1936, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 3.697830761480994, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.1017, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 4.598516834408057, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.1094, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0506329113924051, | |
| "grad_norm": 5.0331347391052415, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.0965, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1518987341772151, | |
| "grad_norm": 4.416693220117886, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 1.0233, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 4.455003544841934, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0615, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3544303797468356, | |
| "grad_norm": 2.7108881252358006, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.9539, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 2.3781307784075065, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.9062, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 2.7671628682377536, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.9214, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6582278481012658, | |
| "grad_norm": 2.69164548478121, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.9167, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.759493670886076, | |
| "grad_norm": 1.8752826457383043, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.8792, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 1.7629645480935292, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.8826, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 1.8598379801684275, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.8337, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 1.7601902843731678, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.7117, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 1.6688259968463455, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.6983, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 1.324059383773062, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.6972, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 1.5097999286676165, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.6929, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 1.4542453165229061, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6194, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 1.41651566460622, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.6231, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 1.2792142263407915, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.6172, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 1.4125008295953911, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6477, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 1.3129136249171092, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.6234, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 1.3057048645987812, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.5529, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 1.4115007740987349, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.4256, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 1.2836438924827784, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.4343, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 1.2652878607503002, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.4271, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 2.421169431145599, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.407, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 1.6520715456696227, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.4483, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 1.3631097569348252, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.4065, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 1.3352370337105939, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.3704, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 1.282105823307875, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.3583, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 1.386221531298694, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.3774, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 1.3198737253650181, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.2767, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 1.158750195024177, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.2376, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 1.0724191223762354, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.2261, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 1.3525031914915266, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.2602, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 1.332021358804614, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.1984, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 1.2727487563677071, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.2647, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 1.0759580819509384, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.2037, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 1.0176040400298398, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.246, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 1.0680959417024933, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.2231, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.9256879450193155, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.203, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.9589623090756939, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.1167, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.868916396558903, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.1554, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.8692442359134506, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.1287, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 1.1424767308096409, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.1484, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 1.3045907030643125, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.1279, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 1.243853095699175, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.1425, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 1.1062066615401174, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.1665, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.9426574630490809, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.0839, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.9910430987137411, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.1461, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.9533648713102743, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.111, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.7902380193213292, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.0865, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.6696649788093965, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.0769, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.7254238181236121, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.0814, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.8876670262878879, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.0818, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.8556179383508906, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.0887, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.7156050707610626, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.0879, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.818520944666355, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.0627, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.7792278486812971, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.0584, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.7306117515886279, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.0606, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.5661475163044138, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.0598, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.5443456084285492, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.0523, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.5618594391788194, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.054, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.6269556212431191, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.0518, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.530462823349994, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.0549, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.47946808414318354, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.0605, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.5033959589247783, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.0211, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.4521161384306736, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.0539, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.5711802777766155, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.0575, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.3766077953775865, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.0426, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.3533462527788944, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.0284, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.4744403201033981, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.0386, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.4286991622167282, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.0277, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.28266296550971565, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.0317, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.4011884691614344, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0329, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.32495610881138104, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.0285, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.3670908813034359, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.0168, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.4835402040267645, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.0559, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.30931705767215034, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.0378, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.25540668229994884, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.0194, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.20010622740478745, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.0202, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.2842784417897834, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.0517, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.22710293498763928, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.0282, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.3055718760747786, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.0318, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.21669510978249906, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.0298, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.5351222995459937, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.0356, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.5647473673655102, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.0234, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.3001643857339602, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.0191, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.2361063701005521, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.0306, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.17562280493103039, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.0245, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.1887106787055557, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.021, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.19482835139049373, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.0199, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.1359705605289439, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.0069, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.18918775754698086, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.0268, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.23344033112999962, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.0508, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.4228380585885804, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.0326, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.16114680269267959, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.0102, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.1779254687263327, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.01, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.20472340787265206, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.0215, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.17708017610219093, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.0299, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.21217320228855543, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.033, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.33618891343982965, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.0153, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.21084337564007086, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.0158, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.2403111068536703, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.0287, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.13840445454845154, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.0188, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.13818949075674417, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.0164, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.17030750163627423, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.0188, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.2299965005773996, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.044, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.22652915235583296, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.038, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.1665325346779966, | |
| "learning_rate": 0.0, | |
| "loss": 0.0127, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "step": 117, | |
| "total_flos": 2.436858051873997e+16, | |
| "train_loss": 0.3127524615981831, | |
| "train_runtime": 1621.3499, | |
| "train_samples_per_second": 2.534, | |
| "train_steps_per_second": 0.072 | |
| } | |
| ], | |
| "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": 2.436858051873997e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |