| { | |
| "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": 6.510182371085828, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.0254, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 6.582519572663169, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.0615, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 6.439926344599413, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0233, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 5.972809082275395, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0127, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 4.602197291511798, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 1.0077, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 2.8010233443393284, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9553, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 2.5653132093327526, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.9506, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 3.6845492435295926, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.9288, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 3.8303998000371537, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.9344, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0506329113924051, | |
| "grad_norm": 3.9530795691038776, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.9483, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1518987341772151, | |
| "grad_norm": 3.6754332872546827, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.8686, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 3.17872223589038, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8883, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3544303797468356, | |
| "grad_norm": 2.257441129284527, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.8377, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 1.6781500127358915, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.8419, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 1.5404794138317974, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.7982, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6582278481012658, | |
| "grad_norm": 1.4609784124872685, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.7941, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.759493670886076, | |
| "grad_norm": 1.2668462650796093, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.8057, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 1.1212028554497055, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.7861, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 1.0895633195883203, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.7739, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 1.2184960009998693, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.7816, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.9113579935395382, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.7302, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 0.9613622768105424, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.725, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 0.9117445192159501, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.7069, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.8312878328461633, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.7035, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 0.9916032484533054, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.7033, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 0.9928831167360924, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.6926, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.8409145181859952, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6847, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 0.9775428741516501, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.6755, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 0.9506685442972472, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.651, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.8313006068527614, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6295, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 0.7568739262612543, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.6384, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 0.7271242122850091, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.6182, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.856442566145437, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6218, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 0.7325069492546041, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.6392, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 0.6212352610421124, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.6261, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.6060071335148768, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6191, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 0.7261156434422802, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.6017, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 0.6385843742159706, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.5768, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.6916129236848202, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.5861, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 0.7106132988519652, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.5407, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 0.7480939531051924, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.5468, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.7086578813899572, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5338, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 0.7279031845614357, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.5302, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 0.6690060641268204, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.5381, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.6948889915997236, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.543, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.7208712747592547, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.531, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 0.6854554288239998, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5137, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.6277143980877257, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5208, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.6242139030107925, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.5051, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.6321974143138194, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.4982, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.5990215427199195, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.4771, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 1.7354372122871062, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.4436, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 0.6968566611534831, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.474, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.5981548758044577, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.4713, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.5729606799998556, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.4784, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.6472290900219629, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.4645, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.5948606146175702, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.435, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.6062015199519869, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.4224, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.5678771527556095, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.391, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.5643870535447468, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.3995, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.6505444129923155, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.4284, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.5635394638042144, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.4187, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.560890503698802, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.3896, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.5654352981703814, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.418, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.5508748469379801, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.3848, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.5408750732924548, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.431, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.620416773909927, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.3875, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.5325810639552746, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.3699, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.6317958512380395, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.3672, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.5973649463281437, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.3686, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.640215648359906, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.3808, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.6358373173748053, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.3137, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.516809428787874, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.3268, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.5945930461577688, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.3227, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.6475883531606954, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.3551, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.5223327632443411, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.3877, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.8278278517893417, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.3114, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.6008938273549477, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.2946, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.5456830308100099, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.2876, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.8456374983441585, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.3127, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.6413765752016471, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.304, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.6031440656934118, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.3259, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.7343855532018921, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.31, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.5928960933959891, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.3248, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.5308978538628125, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.3359, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.5773567652739116, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.3083, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.5711312934180898, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.2697, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.5539311689276141, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.294, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.45661281193329273, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.2649, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.5146940575488416, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.2425, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.5595025576450177, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.2892, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.5104504157471689, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.3171, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.5196226513675396, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.2762, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.48804039512189956, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.3063, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.5358391820467538, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.2458, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.5340150707476744, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.266, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.5408390552485048, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.2791, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.4565054778666042, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.261, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.4802683080342892, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.2577, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.582414487366, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.2413, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.5561800241224305, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.2618, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.5065607848893391, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.2802, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.46978401924242125, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.2549, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.4572900759658637, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.2724, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.4597116071430204, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.2767, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.4755583179520235, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.2222, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.5008252269446787, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.2463, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.51617432630608, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.249, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.5072194229856675, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.242, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.5067506460554636, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.2592, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.4361130649466016, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.2628, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.4578810256403724, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.2767, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.4314537549718583, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.2754, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.43931457751789277, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.2452, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.42276342119184573, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.2186, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.4337722002133534, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.2391, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.4278352688615393, | |
| "learning_rate": 0.0, | |
| "loss": 0.2651, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "step": 117, | |
| "total_flos": 1.4604243583369216e+17, | |
| "train_loss": 0.4960343405986444, | |
| "train_runtime": 5639.6715, | |
| "train_samples_per_second": 0.728, | |
| "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": 1.4604243583369216e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |