| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 12.405063291139241, | |
| "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": 6.590982454386708, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.1215, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.20253164556962025, | |
| "grad_norm": 7.29840367445975, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.2616, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3037974683544304, | |
| "grad_norm": 7.56733790889472, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.2829, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4050632911392405, | |
| "grad_norm": 7.101248010112954, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.2749, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5063291139240507, | |
| "grad_norm": 5.674650481238628, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 1.1527, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6075949367088608, | |
| "grad_norm": 3.169285336498066, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1544, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7088607594936709, | |
| "grad_norm": 2.5652767632449067, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 1.058, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.810126582278481, | |
| "grad_norm": 4.034398255081183, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.0969, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9113924050632911, | |
| "grad_norm": 4.464156798873608, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.1044, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.1012658227848102, | |
| "grad_norm": 4.69903982324907, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.0232, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.2025316455696202, | |
| "grad_norm": 4.805922707518498, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 1.0836, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.3037974683544304, | |
| "grad_norm": 4.292865195662355, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9551, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.4050632911392404, | |
| "grad_norm": 2.8739385030420865, | |
| "learning_rate": 9.997762161417517e-06, | |
| "loss": 1.0414, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.5063291139240507, | |
| "grad_norm": 1.9954355598452902, | |
| "learning_rate": 9.991050648838676e-06, | |
| "loss": 1.0685, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.6075949367088609, | |
| "grad_norm": 2.166983715272283, | |
| "learning_rate": 9.979871469976197e-06, | |
| "loss": 1.0862, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.7088607594936709, | |
| "grad_norm": 2.3756090888007817, | |
| "learning_rate": 9.964234631709188e-06, | |
| "loss": 1.028, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.810126582278481, | |
| "grad_norm": 1.7211326830936204, | |
| "learning_rate": 9.944154131125643e-06, | |
| "loss": 0.9118, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.9113924050632911, | |
| "grad_norm": 1.2905785703703883, | |
| "learning_rate": 9.91964794299315e-06, | |
| "loss": 0.98, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.1012658227848102, | |
| "grad_norm": 0.9361691180849981, | |
| "learning_rate": 9.890738003669029e-06, | |
| "loss": 0.9605, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.2025316455696204, | |
| "grad_norm": 1.0756298323387057, | |
| "learning_rate": 9.857450191464337e-06, | |
| "loss": 0.8183, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.3037974683544302, | |
| "grad_norm": 1.1592318343028445, | |
| "learning_rate": 9.819814303479268e-06, | |
| "loss": 0.9042, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.4050632911392404, | |
| "grad_norm": 0.943577389038488, | |
| "learning_rate": 9.777864028930705e-06, | |
| "loss": 0.9048, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.5063291139240507, | |
| "grad_norm": 1.1615684977578948, | |
| "learning_rate": 9.731636918995821e-06, | |
| "loss": 0.9, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.607594936708861, | |
| "grad_norm": 1.178932013568391, | |
| "learning_rate": 9.681174353198687e-06, | |
| "loss": 0.9331, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.708860759493671, | |
| "grad_norm": 0.8299922566249501, | |
| "learning_rate": 9.626521502369984e-06, | |
| "loss": 0.8512, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.810126582278481, | |
| "grad_norm": 0.9420242439489944, | |
| "learning_rate": 9.567727288213005e-06, | |
| "loss": 0.9241, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.911392405063291, | |
| "grad_norm": 0.855735257118179, | |
| "learning_rate": 9.504844339512096e-06, | |
| "loss": 0.8588, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 3.050632911392405, | |
| "grad_norm": 0.8086357463241294, | |
| "learning_rate": 9.437928945022772e-06, | |
| "loss": 0.8256, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.151898734177215, | |
| "grad_norm": 0.7550199789184788, | |
| "learning_rate": 9.36704100308565e-06, | |
| "loss": 0.8077, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.2531645569620253, | |
| "grad_norm": 0.8239216491474649, | |
| "learning_rate": 9.292243968009332e-06, | |
| "loss": 0.8423, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.3544303797468356, | |
| "grad_norm": 0.8703627775981249, | |
| "learning_rate": 9.213604793270196e-06, | |
| "loss": 0.7758, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.4556962025316453, | |
| "grad_norm": 0.8732187111061649, | |
| "learning_rate": 9.131193871579975e-06, | |
| "loss": 0.865, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.5569620253164556, | |
| "grad_norm": 0.7764872775107572, | |
| "learning_rate": 9.045084971874738e-06, | |
| "loss": 0.7932, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.6582278481012658, | |
| "grad_norm": 0.8405346996877903, | |
| "learning_rate": 8.955355173281709e-06, | |
| "loss": 0.8463, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.759493670886076, | |
| "grad_norm": 0.8463774012355435, | |
| "learning_rate": 8.862084796122998e-06, | |
| "loss": 0.8204, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.8607594936708862, | |
| "grad_norm": 0.7495689481741096, | |
| "learning_rate": 8.765357330018056e-06, | |
| "loss": 0.6984, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.962025316455696, | |
| "grad_norm": 0.687017266359812, | |
| "learning_rate": 8.665259359149132e-06, | |
| "loss": 0.8079, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 4.10126582278481, | |
| "grad_norm": 0.76409212630213, | |
| "learning_rate": 8.561880484756726e-06, | |
| "loss": 0.7783, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.2025316455696204, | |
| "grad_norm": 0.6774607390593336, | |
| "learning_rate": 8.455313244934324e-06, | |
| "loss": 0.7406, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.30379746835443, | |
| "grad_norm": 0.6910357934030281, | |
| "learning_rate": 8.345653031794292e-06, | |
| "loss": 0.702, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.405063291139241, | |
| "grad_norm": 0.7559991249978542, | |
| "learning_rate": 8.232998006078998e-06, | |
| "loss": 0.7203, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.506329113924051, | |
| "grad_norm": 0.690558449826033, | |
| "learning_rate": 8.117449009293668e-06, | |
| "loss": 0.6912, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.6075949367088604, | |
| "grad_norm": 0.6582082554609989, | |
| "learning_rate": 7.99910947343957e-06, | |
| "loss": 0.671, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.708860759493671, | |
| "grad_norm": 0.6460730906634552, | |
| "learning_rate": 7.87808532842837e-06, | |
| "loss": 0.7754, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.810126582278481, | |
| "grad_norm": 0.6325596714284567, | |
| "learning_rate": 7.754484907260513e-06, | |
| "loss": 0.6411, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.911392405063291, | |
| "grad_norm": 0.7586355688183598, | |
| "learning_rate": 7.628418849052523e-06, | |
| "loss": 0.6959, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.050632911392405, | |
| "grad_norm": 0.6223949760249545, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5977, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.151898734177215, | |
| "grad_norm": 0.6761682758084384, | |
| "learning_rate": 7.369343312364994e-06, | |
| "loss": 0.5904, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.253164556962025, | |
| "grad_norm": 0.9300469337141741, | |
| "learning_rate": 7.236565741578163e-06, | |
| "loss": 0.7144, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.3544303797468356, | |
| "grad_norm": 0.6859639091201483, | |
| "learning_rate": 7.101786141547829e-06, | |
| "loss": 0.6248, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.455696202531645, | |
| "grad_norm": 0.7117930584112769, | |
| "learning_rate": 6.965125158269619e-06, | |
| "loss": 0.629, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.556962025316456, | |
| "grad_norm": 0.7685569666357732, | |
| "learning_rate": 6.8267051218319766e-06, | |
| "loss": 0.6143, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.658227848101266, | |
| "grad_norm": 0.6374981736631344, | |
| "learning_rate": 6.686649936914151e-06, | |
| "loss": 0.6322, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.759493670886076, | |
| "grad_norm": 0.714171352150345, | |
| "learning_rate": 6.545084971874738e-06, | |
| "loss": 0.5558, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.860759493670886, | |
| "grad_norm": 0.8868394036201642, | |
| "learning_rate": 6.402136946530014e-06, | |
| "loss": 0.7781, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.962025316455696, | |
| "grad_norm": 0.844014142927269, | |
| "learning_rate": 6.257933818722544e-06, | |
| "loss": 0.5786, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 6.10126582278481, | |
| "grad_norm": 0.8753318485515702, | |
| "learning_rate": 6.112604669781572e-06, | |
| "loss": 0.6113, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.2025316455696204, | |
| "grad_norm": 0.7439635107644622, | |
| "learning_rate": 5.9662795889777666e-06, | |
| "loss": 0.6092, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.30379746835443, | |
| "grad_norm": 0.7274076814356746, | |
| "learning_rate": 5.819089557075689e-06, | |
| "loss": 0.5169, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.405063291139241, | |
| "grad_norm": 0.9480384574208381, | |
| "learning_rate": 5.671166329088278e-06, | |
| "loss": 0.5807, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.506329113924051, | |
| "grad_norm": 0.7514912223529391, | |
| "learning_rate": 5.522642316338268e-06, | |
| "loss": 0.5668, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.6075949367088604, | |
| "grad_norm": 0.7693905096530378, | |
| "learning_rate": 5.373650467932122e-06, | |
| "loss": 0.4671, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.708860759493671, | |
| "grad_norm": 0.6629431133809542, | |
| "learning_rate": 5.224324151752575e-06, | |
| "loss": 0.5415, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.810126582278481, | |
| "grad_norm": 0.643491858905086, | |
| "learning_rate": 5.074797035076319e-06, | |
| "loss": 0.6309, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.911392405063291, | |
| "grad_norm": 0.7202823623547724, | |
| "learning_rate": 4.9252029649236835e-06, | |
| "loss": 0.52, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 7.050632911392405, | |
| "grad_norm": 0.7563281400597983, | |
| "learning_rate": 4.775675848247427e-06, | |
| "loss": 0.5096, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.151898734177215, | |
| "grad_norm": 0.5332212294276538, | |
| "learning_rate": 4.626349532067879e-06, | |
| "loss": 0.5238, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 7.253164556962025, | |
| "grad_norm": 0.6382066611893747, | |
| "learning_rate": 4.477357683661734e-06, | |
| "loss": 0.4728, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 7.3544303797468356, | |
| "grad_norm": 0.6656009101566469, | |
| "learning_rate": 4.3288336709117246e-06, | |
| "loss": 0.5005, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.455696202531645, | |
| "grad_norm": 0.6899296913929884, | |
| "learning_rate": 4.180910442924312e-06, | |
| "loss": 0.6106, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.556962025316456, | |
| "grad_norm": 0.7369158371480743, | |
| "learning_rate": 4.033720411022235e-06, | |
| "loss": 0.4696, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.658227848101266, | |
| "grad_norm": 0.6758447506051136, | |
| "learning_rate": 3.887395330218429e-06, | |
| "loss": 0.4597, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.759493670886076, | |
| "grad_norm": 0.623038363526044, | |
| "learning_rate": 3.7420661812774577e-06, | |
| "loss": 0.5037, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.860759493670886, | |
| "grad_norm": 0.7174471147897493, | |
| "learning_rate": 3.5978630534699873e-06, | |
| "loss": 0.4226, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.962025316455696, | |
| "grad_norm": 0.6770886934322002, | |
| "learning_rate": 3.4549150281252635e-06, | |
| "loss": 0.5682, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 8.10126582278481, | |
| "grad_norm": 0.593862901569409, | |
| "learning_rate": 3.3133500630858507e-06, | |
| "loss": 0.4966, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.20253164556962, | |
| "grad_norm": 0.5570560316921678, | |
| "learning_rate": 3.173294878168025e-06, | |
| "loss": 0.4273, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 8.30379746835443, | |
| "grad_norm": 0.7145465704605002, | |
| "learning_rate": 3.0348748417303826e-06, | |
| "loss": 0.4607, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 8.405063291139241, | |
| "grad_norm": 0.6734392413148257, | |
| "learning_rate": 2.8982138584521734e-06, | |
| "loss": 0.4971, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.50632911392405, | |
| "grad_norm": 0.7629467465856415, | |
| "learning_rate": 2.7634342584218364e-06, | |
| "loss": 0.4492, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.60759493670886, | |
| "grad_norm": 0.5668298425456717, | |
| "learning_rate": 2.6306566876350072e-06, | |
| "loss": 0.3769, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.708860759493671, | |
| "grad_norm": 0.584084715420407, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.4417, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.810126582278482, | |
| "grad_norm": 0.6489318137063667, | |
| "learning_rate": 2.371581150947476e-06, | |
| "loss": 0.4059, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.91139240506329, | |
| "grad_norm": 0.6576995736565088, | |
| "learning_rate": 2.245515092739488e-06, | |
| "loss": 0.438, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 9.050632911392405, | |
| "grad_norm": 0.5961066284362792, | |
| "learning_rate": 2.1219146715716332e-06, | |
| "loss": 0.4479, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.151898734177216, | |
| "grad_norm": 0.6831906057171641, | |
| "learning_rate": 2.0008905265604316e-06, | |
| "loss": 0.4449, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 9.253164556962025, | |
| "grad_norm": 0.5641958422223261, | |
| "learning_rate": 1.8825509907063328e-06, | |
| "loss": 0.3953, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 9.354430379746836, | |
| "grad_norm": 0.6423070752948775, | |
| "learning_rate": 1.7670019939210025e-06, | |
| "loss": 0.3775, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 9.455696202531646, | |
| "grad_norm": 0.6417598481473638, | |
| "learning_rate": 1.6543469682057105e-06, | |
| "loss": 0.2979, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.556962025316455, | |
| "grad_norm": 0.768775647364152, | |
| "learning_rate": 1.544686755065677e-06, | |
| "loss": 0.462, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.658227848101266, | |
| "grad_norm": 0.5796444128696349, | |
| "learning_rate": 1.438119515243277e-06, | |
| "loss": 0.433, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.759493670886076, | |
| "grad_norm": 0.596798729407781, | |
| "learning_rate": 1.3347406408508695e-06, | |
| "loss": 0.3968, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.860759493670885, | |
| "grad_norm": 0.6043270358076789, | |
| "learning_rate": 1.234642669981946e-06, | |
| "loss": 0.46, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.962025316455696, | |
| "grad_norm": 0.6198060261705974, | |
| "learning_rate": 1.137915203877003e-06, | |
| "loss": 0.3772, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 10.10126582278481, | |
| "grad_norm": 0.7380908521104141, | |
| "learning_rate": 1.044644826718295e-06, | |
| "loss": 0.3328, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 10.20253164556962, | |
| "grad_norm": 0.6609374759160311, | |
| "learning_rate": 9.549150281252633e-07, | |
| "loss": 0.375, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 10.30379746835443, | |
| "grad_norm": 0.5704674584157859, | |
| "learning_rate": 8.688061284200266e-07, | |
| "loss": 0.4254, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 10.405063291139241, | |
| "grad_norm": 0.5411428652917858, | |
| "learning_rate": 7.863952067298042e-07, | |
| "loss": 0.358, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 10.50632911392405, | |
| "grad_norm": 0.739808139345924, | |
| "learning_rate": 7.077560319906696e-07, | |
| "loss": 0.4264, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.60759493670886, | |
| "grad_norm": 0.6445921556079948, | |
| "learning_rate": 6.329589969143518e-07, | |
| "loss": 0.3381, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.708860759493671, | |
| "grad_norm": 0.6220143759127105, | |
| "learning_rate": 5.620710549772295e-07, | |
| "loss": 0.3823, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.810126582278482, | |
| "grad_norm": 0.6441928904131864, | |
| "learning_rate": 4.951556604879049e-07, | |
| "loss": 0.3828, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.91139240506329, | |
| "grad_norm": 0.617402877570038, | |
| "learning_rate": 4.322727117869951e-07, | |
| "loss": 0.4411, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 11.050632911392405, | |
| "grad_norm": 0.5114079484886208, | |
| "learning_rate": 3.734784976300165e-07, | |
| "loss": 0.463, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 11.151898734177216, | |
| "grad_norm": 0.5216518872901496, | |
| "learning_rate": 3.18825646801314e-07, | |
| "loss": 0.3557, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 11.253164556962025, | |
| "grad_norm": 0.5890165423628578, | |
| "learning_rate": 2.6836308100417874e-07, | |
| "loss": 0.3385, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 11.354430379746836, | |
| "grad_norm": 0.6172447464668225, | |
| "learning_rate": 2.2213597106929608e-07, | |
| "loss": 0.4521, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 11.455696202531646, | |
| "grad_norm": 0.5797254737737666, | |
| "learning_rate": 1.801856965207338e-07, | |
| "loss": 0.3431, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 11.556962025316455, | |
| "grad_norm": 0.5803872796549314, | |
| "learning_rate": 1.4254980853566248e-07, | |
| "loss": 0.3586, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.658227848101266, | |
| "grad_norm": 0.5839062771496589, | |
| "learning_rate": 1.0926199633097156e-07, | |
| "loss": 0.339, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.759493670886076, | |
| "grad_norm": 0.5140164837489923, | |
| "learning_rate": 8.035205700685167e-08, | |
| "loss": 0.3712, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.860759493670885, | |
| "grad_norm": 0.482731241926417, | |
| "learning_rate": 5.584586887435739e-08, | |
| "loss": 0.3831, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.962025316455696, | |
| "grad_norm": 0.49478288049192937, | |
| "learning_rate": 3.576536829081323e-08, | |
| "loss": 0.4202, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 12.10126582278481, | |
| "grad_norm": 0.49260168836035556, | |
| "learning_rate": 2.012853002380466e-08, | |
| "loss": 0.3588, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 12.20253164556962, | |
| "grad_norm": 0.5517193116586361, | |
| "learning_rate": 8.949351161324227e-09, | |
| "loss": 0.3227, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 12.30379746835443, | |
| "grad_norm": 0.49530242616103676, | |
| "learning_rate": 2.237838582483387e-09, | |
| "loss": 0.4346, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 12.405063291139241, | |
| "grad_norm": 0.5053616185363125, | |
| "learning_rate": 0.0, | |
| "loss": 0.3321, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 12.405063291139241, | |
| "step": 117, | |
| "total_flos": 1.0597825494646784e+17, | |
| "train_loss": 0.4767076523385496, | |
| "train_runtime": 4144.2445, | |
| "train_samples_per_second": 0.991, | |
| "train_steps_per_second": 0.028 | |
| } | |
| ], | |
| "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": 1.0597825494646784e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |