| { | |
| "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": 6.564193469556872, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.0567, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 6.677965594433443, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.0778, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 6.325784780623696, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0348, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 5.964691370648853, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0284, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 4.750371083675538, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 1.0358, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 2.829469859221562, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9449, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 2.4994325067139123, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.9605, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 3.7198133590725706, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.9355, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 4.218637550073218, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.947, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0506329113924051, | |
| "grad_norm": 3.827201935951685, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.9265, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1518987341772151, | |
| "grad_norm": 3.4596080839693863, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.8887, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2531645569620253, | |
| "grad_norm": 3.178881418196424, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8847, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3544303797468356, | |
| "grad_norm": 2.344113594947447, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 0.8413, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4556962025316456, | |
| "grad_norm": 1.870317354960757, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 0.8678, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5569620253164556, | |
| "grad_norm": 1.7676735201082927, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 0.7923, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6582278481012658, | |
| "grad_norm": 1.7158896399191879, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 0.8005, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.759493670886076, | |
| "grad_norm": 1.2097929490065913, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.8062, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8607594936708862, | |
| "grad_norm": 1.127863594987318, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.7734, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9620253164556962, | |
| "grad_norm": 1.2243998414674488, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.8049, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 1.039831517220381, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.7086, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 0.9880629753694798, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.7076, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 0.9298440686521348, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.7239, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 1.0000344195713171, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.7344, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 0.9594905651577595, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.7267, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 0.8693403181997921, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.7065, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 0.8426372884709951, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.7023, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.7558804423325522, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.6844, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 0.8123261070446662, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.6746, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 1.012776874613277, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.6864, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.7562391219809159, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.6607, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 0.7085406485965592, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.6114, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 1.0082017260284695, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.6295, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.7570364567143146, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.6145, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 0.7423801681185671, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.6228, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 0.8013406417099292, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.617, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.7274350006367865, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.5955, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 0.6869028355791839, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.6043, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 0.7939198305299955, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.5753, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.7060699180953195, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.5818, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 0.7452173145635105, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.5556, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 0.5782410856982204, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.5488, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.8146963765153094, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.5315, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 0.6917260736712701, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.5352, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 0.6207943563630881, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.5263, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.5864622757654318, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.506, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.6597672458518218, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.5295, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 0.5626680123790996, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5371, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.6543557837763462, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5051, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.59854447761141, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.4648, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.5470677504703215, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.4714, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.6795392535749355, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.4745, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 0.6021802608354699, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.4936, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 0.5887761602063247, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.4759, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.6132745878881329, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.4589, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.6193858644728606, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.4768, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.5517080203844341, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.4484, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.7392636737429982, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.45, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.7119641229992989, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.4519, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.6742507430087162, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.3773, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.6306106265944342, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.3955, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.8364092780474184, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.4096, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.5905186275323097, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.4204, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.7924179798082264, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.3834, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.7335907981966016, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.391, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.5595602506675581, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.3905, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.7496572288220894, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.3988, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.6723338689973454, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.3928, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.6033371124500815, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.3562, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.5577232308091906, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.3087, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.6020840115188631, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.3701, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.5878394811499298, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.3567, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.6390663030492089, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.3538, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.5812367380162865, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.3391, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.587912250707191, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.3212, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.5437641883361785, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.346, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.5804381162527165, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.3772, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.5699189025540493, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.3437, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.5335046651896318, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.3, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.694286382853658, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.3309, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.570230674619565, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.3077, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.5722102698259701, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.2901, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.6016229884690351, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.2987, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.5317486606819573, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.3314, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.47296560184365893, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.304, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.5202590805037022, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.2973, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.6003104332379025, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.2696, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.6395034567962172, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.2667, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.4745630087751303, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.2929, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.582918295955013, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.2647, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.7649613187755175, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.3014, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.5706507917411721, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.2478, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.5543949694702526, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.3129, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.6049847228525356, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.2937, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.5398567287244529, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.2965, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.5413623144537589, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.3003, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.5809662518858661, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.3037, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.5334023947007592, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.2305, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.5082306180005118, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.2741, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.5275755551195027, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.2931, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.6241478708839339, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.3018, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.5539539155108638, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.2267, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.5412284430077987, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.2863, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.6816378212120966, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.2528, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.4721832483366706, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.2185, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.5075044237066467, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.2398, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.5041067035838389, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.2785, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.8501672789004305, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.2329, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.4556634659670032, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.2044, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.45031507032413814, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.2909, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.45436615614690357, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.2738, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.4376551253805449, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.2824, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.46302684126690946, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.2305, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.4248519694347001, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.2482, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.40938816207950535, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.2359, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.4362137051978998, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.2313, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.4401456396291026, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.2756, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.430232443827871, | |
| "learning_rate": 0.0, | |
| "loss": 0.2724, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "step": 117, | |
| "total_flos": 1.38189139773227e+17, | |
| "train_loss": 0.49606912755049193, | |
| "train_runtime": 5431.7486, | |
| "train_samples_per_second": 0.756, | |
| "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": 1.38189139773227e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |