| { | |
| "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": 6.409530881944437, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.0131, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 6.496397031544609, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.0298, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 6.412335005791017, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0389, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 5.917924316423909, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0122, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 4.632834220509347, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.968, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 2.835267567638371, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9626, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 2.4284574148184452, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.9582, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 3.725766184826296, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.9335, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 3.992573422771624, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.9354, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0506329113924051, | |
| "grad_norm": 3.863492514421068, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.8951, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1518987341772151, | |
| "grad_norm": 3.3644361158623526, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.8602, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 3.10164545820468, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8677, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3544303797468356, | |
| "grad_norm": 2.1950599468002983, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.8273, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 1.6852378773284569, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.8276, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 1.625838290284311, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.8238, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6582278481012658, | |
| "grad_norm": 1.4457438367252524, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.8024, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.759493670886076, | |
| "grad_norm": 1.205733760546575, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.7873, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 1.1041208073186044, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.796, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 1.2298072338930555, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.7707, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 1.0617115312168655, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.7059, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 1.097768971716664, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.7091, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 1.0548019995600924, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.747, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 0.8795694430542033, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.6889, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 1.0653923019563158, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.6853, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 0.9936491547171582, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.7053, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 0.8742281168827849, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.6957, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 1.1155169354687509, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6752, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 0.9890314717525205, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.6808, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 0.9073752940103923, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.6569, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.987700777431423, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6341, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 0.8263679758711892, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.6165, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 0.7411889779156393, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.5914, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.8344273975690829, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6013, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 0.7866553894367511, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.6268, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 0.7491341842654331, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.6114, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.7103843513107044, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.5991, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 0.7193694854613295, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.613, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 0.7388783765734056, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.5792, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.5794004985757181, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.5515, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 0.6750163325529693, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.5634, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 0.7409541659587603, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.5236, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.6810944429343593, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5293, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 0.6594049663421833, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.5314, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 0.6400540695440019, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.5352, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.7407094921128758, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.5253, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.644338042565549, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.5035, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 0.5944039267838971, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5245, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.7084858130985389, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.4883, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.5744032436210856, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.4705, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.6747460181244268, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.4849, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.627532916913427, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.4446, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 0.5190633889697944, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.4454, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 0.7335501781228257, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.4879, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.5986892548657905, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.4442, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.5310435365272664, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.4493, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.5983489416001115, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.4558, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.5893905768608728, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.4665, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.5780986442279454, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.4175, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.5578228041450422, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.4188, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.6213655908527174, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.3879, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.6039021611126009, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.3654, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.6364489733632237, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.3971, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.6708700503999009, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.3878, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.5776639196380924, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.4223, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.662380953759087, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.3925, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.5530452649913483, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.3827, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.5976565457359024, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.3663, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.6298197865739988, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.3574, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.5267311248401534, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.3197, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.7530751913270037, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.3628, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.522435154208157, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.316, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.5734384509426885, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.3351, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.5474778545359908, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.3213, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.5322058343682415, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.3341, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.5930976275053506, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.3529, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.6411372803643108, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.3371, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.5152316830571662, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.2933, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.5674310972693793, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.2897, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.5340779196327525, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.2994, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.5225818065667289, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.2708, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.48241188299169313, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.3, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.4939782479442797, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.2833, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.5205757669599701, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.3063, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.5093330174902143, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.285, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.5042456557842201, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.3241, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.5569568236175116, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.26, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.5137257458468799, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.2639, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.5270280336212415, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.2902, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.582376363550623, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.2497, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.5428735494836799, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.2959, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.48830303024859717, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.2682, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.6053123846117741, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.2605, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.5925883613554753, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.2403, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.49278344487403086, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.29, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.48537917878875514, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.248, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.4873915776768229, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.262, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.4629810650864849, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.2625, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.5045181377279729, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.2265, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.4708969327546407, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.2416, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.472423130252419, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.2754, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.4540420637241734, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.2487, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.4455907330182406, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.2428, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.4733073919178799, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.2523, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.45156645820184843, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.2167, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.4537688217737921, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.2532, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.47240900523902896, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.2435, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.46159847839078455, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.2585, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.4382621021637263, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.2248, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.4481927882096134, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.2538, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.4287127724969524, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.2289, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.46072512134407945, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.259, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.43081983833015286, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.2365, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.44143190408180266, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.2353, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.42643398741122257, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.2098, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.42418955269328285, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.2057, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.4155172727940168, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.2109, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.4362682445317498, | |
| "learning_rate": 0.0, | |
| "loss": 0.2501, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "step": 117, | |
| "total_flos": 1.4389627161306726e+17, | |
| "train_loss": 0.483334608694427, | |
| "train_runtime": 5712.3452, | |
| "train_samples_per_second": 0.719, | |
| "train_steps_per_second": 0.02 | |
| } | |
| ], | |
| "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": 1.4389627161306726e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |