| { | |
| "best_metric": 0.8620190620422363, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-2_120-samples_config-1_full_auto/checkpoint-77", | |
| "epoch": 14.0, | |
| "eval_steps": 500, | |
| "global_step": 154, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 0.43156448006629944, | |
| "learning_rate": 1.818181818181818e-06, | |
| "loss": 1.226, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.5782701969146729, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 1.4429, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.4240364730358124, | |
| "learning_rate": 7.272727272727272e-06, | |
| "loss": 1.2635, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.46736571192741394, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 1.297, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.4513182044029236, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 1.3665, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.4108898937702179, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 1.2681, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.2884570360183716, | |
| "eval_runtime": 9.6754, | |
| "eval_samples_per_second": 2.481, | |
| "eval_steps_per_second": 2.481, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.45335185527801514, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 1.3249, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 0.3700883984565735, | |
| "learning_rate": 2.5454545454545454e-05, | |
| "loss": 1.2316, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.4393417239189148, | |
| "learning_rate": 2.909090909090909e-05, | |
| "loss": 1.2719, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 0.5480453372001648, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 1.27, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.3178660571575165, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 1.172, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.25591355562210083, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0715, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.1368393898010254, | |
| "eval_runtime": 9.3888, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.22814896702766418, | |
| "learning_rate": 4.3636363636363636e-05, | |
| "loss": 1.1105, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 0.30155879259109497, | |
| "learning_rate": 4.7272727272727275e-05, | |
| "loss": 1.1054, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.26335978507995605, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 1.0451, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.254153311252594, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 1.0642, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.25019872188568115, | |
| "learning_rate": 5.818181818181818e-05, | |
| "loss": 1.0104, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.0255943536758423, | |
| "eval_runtime": 9.3934, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.2527143061161041, | |
| "learning_rate": 6.181818181818182e-05, | |
| "loss": 1.0364, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.2616344690322876, | |
| "learning_rate": 6.545454545454546e-05, | |
| "loss": 0.9652, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.33973056077957153, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.9506, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.22250650823116302, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.8853, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.28038284182548523, | |
| "learning_rate": 7.636363636363637e-05, | |
| "loss": 0.8553, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.2027643322944641, | |
| "learning_rate": 8e-05, | |
| "loss": 0.9392, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.9240176677703857, | |
| "eval_runtime": 9.3906, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.20338939130306244, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.8421, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.20369814336299896, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.8727, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.18691229820251465, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.827, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.22666899859905243, | |
| "learning_rate": 9.454545454545455e-05, | |
| "loss": 0.8454, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.17996296286582947, | |
| "learning_rate": 9.818181818181818e-05, | |
| "loss": 0.8623, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.8879290223121643, | |
| "eval_runtime": 9.3876, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.17760373651981354, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 0.8081, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.19263812899589539, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.8386, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.1968223601579666, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.7679, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 0.1900816708803177, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.7929, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.19545985758304596, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.7844, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.19075007736682892, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.8299, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.8687613010406494, | |
| "eval_runtime": 9.387, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.21089905500411987, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.7404, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.20160414278507233, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.7853, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.21511638164520264, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.7209, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.23202964663505554, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.76, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.23653540015220642, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 0.7975, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.8620190620422363, | |
| "eval_runtime": 9.3878, | |
| "eval_samples_per_second": 2.557, | |
| "eval_steps_per_second": 2.557, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.090909090909091, | |
| "grad_norm": 0.20377208292484283, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.8173, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.2544645667076111, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.6958, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.454545454545454, | |
| "grad_norm": 0.22579969465732574, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.6893, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.24581699073314667, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.7222, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 7.818181818181818, | |
| "grad_norm": 0.4889249801635742, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.7172, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.2468501627445221, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.7553, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.862983226776123, | |
| "eval_runtime": 9.3927, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 0.2589528262615204, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.7093, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.2995380759239197, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.6612, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 8.545454545454545, | |
| "grad_norm": 0.30884915590286255, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.7043, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.3667954206466675, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.6365, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "grad_norm": 0.34860759973526, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.6918, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.8763653635978699, | |
| "eval_runtime": 9.3944, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.3606255352497101, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.6932, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.272727272727273, | |
| "grad_norm": 0.34833142161369324, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.6079, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.44239556789398193, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.6113, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 9.636363636363637, | |
| "grad_norm": 0.39836251735687256, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.6802, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.4025746285915375, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.6326, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.449406236410141, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.6005, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.9124259352684021, | |
| "eval_runtime": 9.3972, | |
| "eval_samples_per_second": 2.554, | |
| "eval_steps_per_second": 2.554, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.45643338561058044, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.5658, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 10.363636363636363, | |
| "grad_norm": 0.4876292049884796, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.5607, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.5747207999229431, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.563, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 10.727272727272727, | |
| "grad_norm": 0.560712993144989, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.5554, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.5259386897087097, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.6021, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.9519934058189392, | |
| "eval_runtime": 9.3951, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 11.090909090909092, | |
| "grad_norm": 0.6063818335533142, | |
| "learning_rate": 9.554728301876526e-05, | |
| "loss": 0.5445, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.7460753321647644, | |
| "learning_rate": 9.528180468815155e-05, | |
| "loss": 0.5051, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 11.454545454545455, | |
| "grad_norm": 0.6494155526161194, | |
| "learning_rate": 9.50090306530454e-05, | |
| "loss": 0.5097, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 2.544689655303955, | |
| "learning_rate": 9.472900486219769e-05, | |
| "loss": 0.5396, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 11.818181818181818, | |
| "grad_norm": 0.7475159168243408, | |
| "learning_rate": 9.444177243274618e-05, | |
| "loss": 0.5141, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.6875072717666626, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.4832, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.016729712486267, | |
| "eval_runtime": 9.3906, | |
| "eval_samples_per_second": 2.556, | |
| "eval_steps_per_second": 2.556, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.181818181818182, | |
| "grad_norm": 0.7332088351249695, | |
| "learning_rate": 9.384587392471515e-05, | |
| "loss": 0.4737, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.7360023260116577, | |
| "learning_rate": 9.353730385598887e-05, | |
| "loss": 0.4263, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 12.545454545454545, | |
| "grad_norm": 0.7788746356964111, | |
| "learning_rate": 9.322171915289635e-05, | |
| "loss": 0.4503, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 0.8412120938301086, | |
| "learning_rate": 9.289917066174886e-05, | |
| "loss": 0.4398, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "grad_norm": 0.7447380423545837, | |
| "learning_rate": 9.256971035084785e-05, | |
| "loss": 0.5046, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 1.0760546922683716, | |
| "eval_runtime": 9.3947, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.8211064338684082, | |
| "learning_rate": 9.223339130211192e-05, | |
| "loss": 0.3545, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 13.272727272727273, | |
| "grad_norm": 1.4348950386047363, | |
| "learning_rate": 9.189026770252436e-05, | |
| "loss": 0.4168, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.810494601726532, | |
| "learning_rate": 9.154039483540273e-05, | |
| "loss": 0.4436, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 13.636363636363637, | |
| "grad_norm": 0.9057369828224182, | |
| "learning_rate": 9.118382907149165e-05, | |
| "loss": 0.3904, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 1.0466903448104858, | |
| "learning_rate": 9.082062785988049e-05, | |
| "loss": 0.373, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.8568995594978333, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.3353, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.1757985353469849, | |
| "eval_runtime": 9.3941, | |
| "eval_samples_per_second": 2.555, | |
| "eval_steps_per_second": 2.555, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "step": 154, | |
| "total_flos": 4.142690421702656e+16, | |
| "train_loss": 0.7778143480226591, | |
| "train_runtime": 1774.7755, | |
| "train_samples_per_second": 2.479, | |
| "train_steps_per_second": 0.31 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.142690421702656e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |