| { | |
| "best_metric": 0.7095006704330444, | |
| "best_model_checkpoint": "./kd_results/microsoft/beit-base-patch16-384_alpha0.7_temp5.0_t3/checkpoint-1148", | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 3280, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1524390243902439, | |
| "grad_norm": 4.038743495941162, | |
| "learning_rate": 3.8109756097560976e-06, | |
| "loss": 0.6096, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3048780487804878, | |
| "grad_norm": 4.960451126098633, | |
| "learning_rate": 7.621951219512195e-06, | |
| "loss": 0.5777, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4573170731707317, | |
| "grad_norm": 5.872208118438721, | |
| "learning_rate": 1.1432926829268294e-05, | |
| "loss": 0.5769, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6097560975609756, | |
| "grad_norm": 2.5071909427642822, | |
| "learning_rate": 1.524390243902439e-05, | |
| "loss": 0.5671, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7621951219512195, | |
| "grad_norm": 2.9352896213531494, | |
| "learning_rate": 1.9054878048780488e-05, | |
| "loss": 0.5688, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9146341463414634, | |
| "grad_norm": 3.5098843574523926, | |
| "learning_rate": 2.286585365853659e-05, | |
| "loss": 0.5656, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.2613882863340564, | |
| "eval_loss": 1.4240881204605103, | |
| "eval_runtime": 49.9624, | |
| "eval_samples_per_second": 18.454, | |
| "eval_steps_per_second": 0.58, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.0670731707317074, | |
| "grad_norm": 2.3195862770080566, | |
| "learning_rate": 2.6676829268292686e-05, | |
| "loss": 0.572, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2195121951219512, | |
| "grad_norm": 4.526941299438477, | |
| "learning_rate": 3.048780487804878e-05, | |
| "loss": 0.5748, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3719512195121952, | |
| "grad_norm": 5.595492362976074, | |
| "learning_rate": 3.429878048780488e-05, | |
| "loss": 0.5555, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.524390243902439, | |
| "grad_norm": 3.5616228580474854, | |
| "learning_rate": 3.8109756097560976e-05, | |
| "loss": 0.5657, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.6768292682926829, | |
| "grad_norm": 2.869215726852417, | |
| "learning_rate": 4.1920731707317077e-05, | |
| "loss": 0.5663, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.8292682926829267, | |
| "grad_norm": 5.0579609870910645, | |
| "learning_rate": 4.573170731707318e-05, | |
| "loss": 0.5385, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.9817073170731707, | |
| "grad_norm": 2.323190450668335, | |
| "learning_rate": 4.954268292682927e-05, | |
| "loss": 0.5188, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.42950108459869846, | |
| "eval_loss": 1.2839103937149048, | |
| "eval_runtime": 50.2929, | |
| "eval_samples_per_second": 18.333, | |
| "eval_steps_per_second": 0.577, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.1341463414634148, | |
| "grad_norm": 6.725587844848633, | |
| "learning_rate": 4.962737127371274e-05, | |
| "loss": 0.4878, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.2865853658536586, | |
| "grad_norm": 2.7454915046691895, | |
| "learning_rate": 4.92039295392954e-05, | |
| "loss": 0.4825, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.4390243902439024, | |
| "grad_norm": 3.2787322998046875, | |
| "learning_rate": 4.878048780487805e-05, | |
| "loss": 0.4431, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.591463414634146, | |
| "grad_norm": 3.206685781478882, | |
| "learning_rate": 4.8357046070460704e-05, | |
| "loss": 0.4233, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.7439024390243905, | |
| "grad_norm": 2.963148355484009, | |
| "learning_rate": 4.793360433604336e-05, | |
| "loss": 0.3975, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.8963414634146343, | |
| "grad_norm": 3.2302346229553223, | |
| "learning_rate": 4.7510162601626016e-05, | |
| "loss": 0.3829, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7245119305856833, | |
| "eval_loss": 0.8952382802963257, | |
| "eval_runtime": 50.0283, | |
| "eval_samples_per_second": 18.43, | |
| "eval_steps_per_second": 0.58, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 3.048780487804878, | |
| "grad_norm": 2.079411029815674, | |
| "learning_rate": 4.7086720867208675e-05, | |
| "loss": 0.3821, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.201219512195122, | |
| "grad_norm": 3.4006881713867188, | |
| "learning_rate": 4.6663279132791334e-05, | |
| "loss": 0.3457, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.3536585365853657, | |
| "grad_norm": 2.087895631790161, | |
| "learning_rate": 4.6239837398373986e-05, | |
| "loss": 0.3399, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.5060975609756095, | |
| "grad_norm": 4.0018744468688965, | |
| "learning_rate": 4.581639566395664e-05, | |
| "loss": 0.3461, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.658536585365854, | |
| "grad_norm": 3.0642566680908203, | |
| "learning_rate": 4.53929539295393e-05, | |
| "loss": 0.327, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.8109756097560976, | |
| "grad_norm": 2.605104446411133, | |
| "learning_rate": 4.496951219512196e-05, | |
| "loss": 0.3346, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.9634146341463414, | |
| "grad_norm": 3.1627883911132812, | |
| "learning_rate": 4.454607046070461e-05, | |
| "loss": 0.3272, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.710412147505423, | |
| "eval_loss": 0.8815497159957886, | |
| "eval_runtime": 50.9116, | |
| "eval_samples_per_second": 18.11, | |
| "eval_steps_per_second": 0.57, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 4.115853658536586, | |
| "grad_norm": 1.1469324827194214, | |
| "learning_rate": 4.412262872628726e-05, | |
| "loss": 0.3164, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.2682926829268295, | |
| "grad_norm": 1.2427903413772583, | |
| "learning_rate": 4.369918699186992e-05, | |
| "loss": 0.2971, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.420731707317073, | |
| "grad_norm": 2.122042417526245, | |
| "learning_rate": 4.3275745257452574e-05, | |
| "loss": 0.2965, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.573170731707317, | |
| "grad_norm": 1.1339151859283447, | |
| "learning_rate": 4.285230352303523e-05, | |
| "loss": 0.2957, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.725609756097561, | |
| "grad_norm": 1.6679350137710571, | |
| "learning_rate": 4.242886178861789e-05, | |
| "loss": 0.2936, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 4.878048780487805, | |
| "grad_norm": 1.8158940076828003, | |
| "learning_rate": 4.2005420054200545e-05, | |
| "loss": 0.2941, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8004338394793926, | |
| "eval_loss": 0.7333674430847168, | |
| "eval_runtime": 49.4264, | |
| "eval_samples_per_second": 18.654, | |
| "eval_steps_per_second": 0.587, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 5.030487804878049, | |
| "grad_norm": 1.5806382894515991, | |
| "learning_rate": 4.15819783197832e-05, | |
| "loss": 0.2902, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 5.182926829268292, | |
| "grad_norm": 1.114171028137207, | |
| "learning_rate": 4.1158536585365856e-05, | |
| "loss": 0.2703, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.335365853658536, | |
| "grad_norm": 2.182769298553467, | |
| "learning_rate": 4.0735094850948516e-05, | |
| "loss": 0.266, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 5.487804878048781, | |
| "grad_norm": 1.225246548652649, | |
| "learning_rate": 4.031165311653117e-05, | |
| "loss": 0.2817, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.640243902439025, | |
| "grad_norm": 1.086198329925537, | |
| "learning_rate": 3.988821138211382e-05, | |
| "loss": 0.2696, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 5.7926829268292686, | |
| "grad_norm": 2.7490367889404297, | |
| "learning_rate": 3.946476964769648e-05, | |
| "loss": 0.2719, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 5.945121951219512, | |
| "grad_norm": 1.2480428218841553, | |
| "learning_rate": 3.904132791327913e-05, | |
| "loss": 0.2671, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7885032537960954, | |
| "eval_loss": 0.7653427720069885, | |
| "eval_runtime": 50.9303, | |
| "eval_samples_per_second": 18.103, | |
| "eval_steps_per_second": 0.569, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 6.097560975609756, | |
| "grad_norm": 1.6687262058258057, | |
| "learning_rate": 3.861788617886179e-05, | |
| "loss": 0.2634, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 1.3652461767196655, | |
| "learning_rate": 3.8194444444444444e-05, | |
| "loss": 0.2628, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 6.402439024390244, | |
| "grad_norm": 1.3502334356307983, | |
| "learning_rate": 3.77710027100271e-05, | |
| "loss": 0.2565, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 6.554878048780488, | |
| "grad_norm": 1.7233595848083496, | |
| "learning_rate": 3.7347560975609755e-05, | |
| "loss": 0.2611, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 6.7073170731707314, | |
| "grad_norm": 1.5111808776855469, | |
| "learning_rate": 3.6924119241192415e-05, | |
| "loss": 0.2576, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.859756097560975, | |
| "grad_norm": 1.426594853401184, | |
| "learning_rate": 3.6500677506775074e-05, | |
| "loss": 0.2569, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8275488069414316, | |
| "eval_loss": 0.7095006704330444, | |
| "eval_runtime": 49.1007, | |
| "eval_samples_per_second": 18.778, | |
| "eval_steps_per_second": 0.591, | |
| "step": 1148 | |
| }, | |
| { | |
| "epoch": 7.012195121951219, | |
| "grad_norm": 1.0241031646728516, | |
| "learning_rate": 3.6077235772357726e-05, | |
| "loss": 0.2476, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.164634146341464, | |
| "grad_norm": 1.1023404598236084, | |
| "learning_rate": 3.565379403794038e-05, | |
| "loss": 0.2488, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 7.317073170731708, | |
| "grad_norm": 1.48065185546875, | |
| "learning_rate": 3.523035230352303e-05, | |
| "loss": 0.2475, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.469512195121951, | |
| "grad_norm": 2.033147096633911, | |
| "learning_rate": 3.48069105691057e-05, | |
| "loss": 0.2448, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 7.621951219512195, | |
| "grad_norm": 1.1411629915237427, | |
| "learning_rate": 3.438346883468835e-05, | |
| "loss": 0.2445, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 7.774390243902439, | |
| "grad_norm": 2.6337881088256836, | |
| "learning_rate": 3.3960027100271e-05, | |
| "loss": 0.2544, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 7.926829268292683, | |
| "grad_norm": 1.0793572664260864, | |
| "learning_rate": 3.353658536585366e-05, | |
| "loss": 0.2483, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8232104121475055, | |
| "eval_loss": 0.7405510544776917, | |
| "eval_runtime": 49.2945, | |
| "eval_samples_per_second": 18.704, | |
| "eval_steps_per_second": 0.588, | |
| "step": 1312 | |
| }, | |
| { | |
| "epoch": 8.079268292682928, | |
| "grad_norm": 1.5204435586929321, | |
| "learning_rate": 3.3113143631436314e-05, | |
| "loss": 0.2403, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 8.231707317073171, | |
| "grad_norm": 0.900091826915741, | |
| "learning_rate": 3.268970189701897e-05, | |
| "loss": 0.2346, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 8.384146341463415, | |
| "grad_norm": 0.6551163792610168, | |
| "learning_rate": 3.226626016260163e-05, | |
| "loss": 0.2518, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 8.536585365853659, | |
| "grad_norm": 1.1180330514907837, | |
| "learning_rate": 3.1842818428184285e-05, | |
| "loss": 0.2381, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 8.689024390243903, | |
| "grad_norm": 1.2341127395629883, | |
| "learning_rate": 3.141937669376694e-05, | |
| "loss": 0.2417, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 8.841463414634147, | |
| "grad_norm": 0.7426364421844482, | |
| "learning_rate": 3.0995934959349596e-05, | |
| "loss": 0.2388, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 8.99390243902439, | |
| "grad_norm": 0.904787003993988, | |
| "learning_rate": 3.0572493224932255e-05, | |
| "loss": 0.2464, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8177874186550976, | |
| "eval_loss": 0.752011775970459, | |
| "eval_runtime": 49.6362, | |
| "eval_samples_per_second": 18.575, | |
| "eval_steps_per_second": 0.584, | |
| "step": 1476 | |
| }, | |
| { | |
| "epoch": 9.146341463414634, | |
| "grad_norm": 1.2572402954101562, | |
| "learning_rate": 3.0149051490514908e-05, | |
| "loss": 0.2336, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.298780487804878, | |
| "grad_norm": 1.1722904443740845, | |
| "learning_rate": 2.972560975609756e-05, | |
| "loss": 0.2432, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 9.451219512195122, | |
| "grad_norm": 0.7890969514846802, | |
| "learning_rate": 2.930216802168022e-05, | |
| "loss": 0.2384, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 9.603658536585366, | |
| "grad_norm": 1.4712668657302856, | |
| "learning_rate": 2.8878726287262875e-05, | |
| "loss": 0.2402, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 9.75609756097561, | |
| "grad_norm": 0.7594574689865112, | |
| "learning_rate": 2.8455284552845528e-05, | |
| "loss": 0.2398, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 9.908536585365853, | |
| "grad_norm": 0.8254810571670532, | |
| "learning_rate": 2.8031842818428184e-05, | |
| "loss": 0.2312, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8394793926247288, | |
| "eval_loss": 0.7326832413673401, | |
| "eval_runtime": 49.6985, | |
| "eval_samples_per_second": 18.552, | |
| "eval_steps_per_second": 0.584, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 10.060975609756097, | |
| "grad_norm": 1.41318678855896, | |
| "learning_rate": 2.7608401084010843e-05, | |
| "loss": 0.2328, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 10.213414634146341, | |
| "grad_norm": 0.6944267749786377, | |
| "learning_rate": 2.71849593495935e-05, | |
| "loss": 0.2322, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 10.365853658536585, | |
| "grad_norm": 0.8574203252792358, | |
| "learning_rate": 2.676151761517615e-05, | |
| "loss": 0.2361, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 10.518292682926829, | |
| "grad_norm": 0.7547130584716797, | |
| "learning_rate": 2.633807588075881e-05, | |
| "loss": 0.2353, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 10.670731707317072, | |
| "grad_norm": 1.094035029411316, | |
| "learning_rate": 2.5914634146341466e-05, | |
| "loss": 0.231, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 10.823170731707316, | |
| "grad_norm": 0.9883895516395569, | |
| "learning_rate": 2.549119241192412e-05, | |
| "loss": 0.2339, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 10.975609756097562, | |
| "grad_norm": 1.3127468824386597, | |
| "learning_rate": 2.5067750677506774e-05, | |
| "loss": 0.229, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.7960954446854663, | |
| "eval_loss": 0.7743532061576843, | |
| "eval_runtime": 50.2819, | |
| "eval_samples_per_second": 18.337, | |
| "eval_steps_per_second": 0.577, | |
| "step": 1804 | |
| }, | |
| { | |
| "epoch": 11.128048780487806, | |
| "grad_norm": 0.8711831569671631, | |
| "learning_rate": 2.464430894308943e-05, | |
| "loss": 0.2338, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 11.28048780487805, | |
| "grad_norm": 0.6187135577201843, | |
| "learning_rate": 2.422086720867209e-05, | |
| "loss": 0.2268, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 11.432926829268293, | |
| "grad_norm": 1.1254570484161377, | |
| "learning_rate": 2.3797425474254745e-05, | |
| "loss": 0.2281, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 11.585365853658537, | |
| "grad_norm": 0.9288303852081299, | |
| "learning_rate": 2.3373983739837398e-05, | |
| "loss": 0.2267, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 11.737804878048781, | |
| "grad_norm": 1.7284654378890991, | |
| "learning_rate": 2.2950542005420057e-05, | |
| "loss": 0.239, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 11.890243902439025, | |
| "grad_norm": 0.4840218126773834, | |
| "learning_rate": 2.252710027100271e-05, | |
| "loss": 0.2326, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8167028199566161, | |
| "eval_loss": 0.7301051616668701, | |
| "eval_runtime": 49.0383, | |
| "eval_samples_per_second": 18.802, | |
| "eval_steps_per_second": 0.591, | |
| "step": 1968 | |
| }, | |
| { | |
| "epoch": 12.042682926829269, | |
| "grad_norm": 0.895510733127594, | |
| "learning_rate": 2.210365853658537e-05, | |
| "loss": 0.2283, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 12.195121951219512, | |
| "grad_norm": 0.7137194871902466, | |
| "learning_rate": 2.1680216802168024e-05, | |
| "loss": 0.2275, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.347560975609756, | |
| "grad_norm": 0.9961858987808228, | |
| "learning_rate": 2.1256775067750677e-05, | |
| "loss": 0.2333, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.5939064025878906, | |
| "learning_rate": 2.0833333333333336e-05, | |
| "loss": 0.2297, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 12.652439024390244, | |
| "grad_norm": 1.0053822994232178, | |
| "learning_rate": 2.040989159891599e-05, | |
| "loss": 0.2315, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 12.804878048780488, | |
| "grad_norm": 0.6412273645401001, | |
| "learning_rate": 1.9986449864498648e-05, | |
| "loss": 0.2242, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 12.957317073170731, | |
| "grad_norm": 1.381249189376831, | |
| "learning_rate": 1.95630081300813e-05, | |
| "loss": 0.2245, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8145336225596529, | |
| "eval_loss": 0.7386568784713745, | |
| "eval_runtime": 49.1161, | |
| "eval_samples_per_second": 18.772, | |
| "eval_steps_per_second": 0.59, | |
| "step": 2132 | |
| }, | |
| { | |
| "epoch": 13.109756097560975, | |
| "grad_norm": 0.33759984374046326, | |
| "learning_rate": 1.9139566395663956e-05, | |
| "loss": 0.2289, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 13.262195121951219, | |
| "grad_norm": 0.5081386566162109, | |
| "learning_rate": 1.8716124661246615e-05, | |
| "loss": 0.2322, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 13.414634146341463, | |
| "grad_norm": 0.9197372198104858, | |
| "learning_rate": 1.8292682926829268e-05, | |
| "loss": 0.2266, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 13.567073170731707, | |
| "grad_norm": 0.5281209945678711, | |
| "learning_rate": 1.7869241192411927e-05, | |
| "loss": 0.2221, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 13.71951219512195, | |
| "grad_norm": 0.9207494854927063, | |
| "learning_rate": 1.744579945799458e-05, | |
| "loss": 0.223, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 13.871951219512194, | |
| "grad_norm": 0.9660337567329407, | |
| "learning_rate": 1.702235772357724e-05, | |
| "loss": 0.2291, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.824295010845987, | |
| "eval_loss": 0.7216958403587341, | |
| "eval_runtime": 49.8483, | |
| "eval_samples_per_second": 18.496, | |
| "eval_steps_per_second": 0.582, | |
| "step": 2296 | |
| }, | |
| { | |
| "epoch": 14.024390243902438, | |
| "grad_norm": 0.521358847618103, | |
| "learning_rate": 1.659891598915989e-05, | |
| "loss": 0.2263, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 14.176829268292684, | |
| "grad_norm": 0.9041802287101746, | |
| "learning_rate": 1.6175474254742547e-05, | |
| "loss": 0.2294, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 14.329268292682928, | |
| "grad_norm": 0.5227034687995911, | |
| "learning_rate": 1.5752032520325206e-05, | |
| "loss": 0.2268, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 14.481707317073171, | |
| "grad_norm": 0.7264844179153442, | |
| "learning_rate": 1.532859078590786e-05, | |
| "loss": 0.2249, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 14.634146341463415, | |
| "grad_norm": 0.3946647047996521, | |
| "learning_rate": 1.4905149051490516e-05, | |
| "loss": 0.2216, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 14.786585365853659, | |
| "grad_norm": 0.5581846833229065, | |
| "learning_rate": 1.448170731707317e-05, | |
| "loss": 0.2219, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 14.939024390243903, | |
| "grad_norm": 0.4273684322834015, | |
| "learning_rate": 1.4058265582655828e-05, | |
| "loss": 0.2282, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.8058568329718004, | |
| "eval_loss": 0.7337209582328796, | |
| "eval_runtime": 49.4995, | |
| "eval_samples_per_second": 18.626, | |
| "eval_steps_per_second": 0.586, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 15.091463414634147, | |
| "grad_norm": 0.6564639210700989, | |
| "learning_rate": 1.3634823848238485e-05, | |
| "loss": 0.2153, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 15.24390243902439, | |
| "grad_norm": 0.4066160023212433, | |
| "learning_rate": 1.321138211382114e-05, | |
| "loss": 0.2318, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 15.396341463414634, | |
| "grad_norm": 0.6132465600967407, | |
| "learning_rate": 1.2787940379403795e-05, | |
| "loss": 0.2245, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 15.548780487804878, | |
| "grad_norm": 0.9687080979347229, | |
| "learning_rate": 1.2364498644986451e-05, | |
| "loss": 0.2226, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 15.701219512195122, | |
| "grad_norm": 0.7273513674736023, | |
| "learning_rate": 1.1941056910569107e-05, | |
| "loss": 0.2207, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 15.853658536585366, | |
| "grad_norm": 0.521793782711029, | |
| "learning_rate": 1.1517615176151763e-05, | |
| "loss": 0.2236, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8188720173535792, | |
| "eval_loss": 0.7246159315109253, | |
| "eval_runtime": 49.3907, | |
| "eval_samples_per_second": 18.667, | |
| "eval_steps_per_second": 0.587, | |
| "step": 2624 | |
| }, | |
| { | |
| "epoch": 16.00609756097561, | |
| "grad_norm": 0.490955650806427, | |
| "learning_rate": 1.1094173441734418e-05, | |
| "loss": 0.2264, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 16.158536585365855, | |
| "grad_norm": 0.5404984951019287, | |
| "learning_rate": 1.0670731707317074e-05, | |
| "loss": 0.2221, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 16.3109756097561, | |
| "grad_norm": 0.5006304979324341, | |
| "learning_rate": 1.0247289972899728e-05, | |
| "loss": 0.221, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 16.463414634146343, | |
| "grad_norm": 0.524905264377594, | |
| "learning_rate": 9.823848238482384e-06, | |
| "loss": 0.2184, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 16.615853658536587, | |
| "grad_norm": 0.7482074499130249, | |
| "learning_rate": 9.400406504065042e-06, | |
| "loss": 0.2233, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 16.76829268292683, | |
| "grad_norm": 0.49331989884376526, | |
| "learning_rate": 8.976964769647697e-06, | |
| "loss": 0.2173, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 16.920731707317074, | |
| "grad_norm": 0.4380941390991211, | |
| "learning_rate": 8.553523035230353e-06, | |
| "loss": 0.2276, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.824295010845987, | |
| "eval_loss": 0.7213588953018188, | |
| "eval_runtime": 48.669, | |
| "eval_samples_per_second": 18.944, | |
| "eval_steps_per_second": 0.596, | |
| "step": 2788 | |
| }, | |
| { | |
| "epoch": 17.073170731707318, | |
| "grad_norm": 0.4981781840324402, | |
| "learning_rate": 8.130081300813009e-06, | |
| "loss": 0.2244, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 17.225609756097562, | |
| "grad_norm": 0.5120595693588257, | |
| "learning_rate": 7.706639566395663e-06, | |
| "loss": 0.2207, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 17.378048780487806, | |
| "grad_norm": 0.5592207908630371, | |
| "learning_rate": 7.283197831978319e-06, | |
| "loss": 0.228, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 17.53048780487805, | |
| "grad_norm": 0.43658289313316345, | |
| "learning_rate": 6.859756097560977e-06, | |
| "loss": 0.2266, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 17.682926829268293, | |
| "grad_norm": 0.3530956506729126, | |
| "learning_rate": 6.4363143631436324e-06, | |
| "loss": 0.2203, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 17.835365853658537, | |
| "grad_norm": 0.5540438294410706, | |
| "learning_rate": 6.0128726287262874e-06, | |
| "loss": 0.216, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 17.98780487804878, | |
| "grad_norm": 0.5253993272781372, | |
| "learning_rate": 5.589430894308943e-06, | |
| "loss": 0.2178, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8177874186550976, | |
| "eval_loss": 0.7251272201538086, | |
| "eval_runtime": 49.1284, | |
| "eval_samples_per_second": 18.767, | |
| "eval_steps_per_second": 0.59, | |
| "step": 2952 | |
| }, | |
| { | |
| "epoch": 18.140243902439025, | |
| "grad_norm": 0.4079430401325226, | |
| "learning_rate": 5.165989159891599e-06, | |
| "loss": 0.2217, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 18.29268292682927, | |
| "grad_norm": 0.7246861457824707, | |
| "learning_rate": 4.742547425474255e-06, | |
| "loss": 0.2209, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 18.445121951219512, | |
| "grad_norm": 0.522539496421814, | |
| "learning_rate": 4.319105691056911e-06, | |
| "loss": 0.219, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 18.597560975609756, | |
| "grad_norm": 0.34168657660484314, | |
| "learning_rate": 3.8956639566395666e-06, | |
| "loss": 0.2219, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "grad_norm": 0.46584391593933105, | |
| "learning_rate": 3.4722222222222224e-06, | |
| "loss": 0.2182, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 18.902439024390244, | |
| "grad_norm": 0.279475599527359, | |
| "learning_rate": 3.0487804878048782e-06, | |
| "loss": 0.2235, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.8232104121475055, | |
| "eval_loss": 0.7219586372375488, | |
| "eval_runtime": 49.1629, | |
| "eval_samples_per_second": 18.754, | |
| "eval_steps_per_second": 0.59, | |
| "step": 3116 | |
| }, | |
| { | |
| "epoch": 19.054878048780488, | |
| "grad_norm": 0.48366549611091614, | |
| "learning_rate": 2.625338753387534e-06, | |
| "loss": 0.2217, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 19.20731707317073, | |
| "grad_norm": 0.6121975779533386, | |
| "learning_rate": 2.2018970189701894e-06, | |
| "loss": 0.2274, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 19.359756097560975, | |
| "grad_norm": 0.29194459319114685, | |
| "learning_rate": 1.7784552845528455e-06, | |
| "loss": 0.2229, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 19.51219512195122, | |
| "grad_norm": 0.5204634666442871, | |
| "learning_rate": 1.3550135501355015e-06, | |
| "loss": 0.2209, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 19.664634146341463, | |
| "grad_norm": 0.406363308429718, | |
| "learning_rate": 9.315718157181572e-07, | |
| "loss": 0.2216, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 19.817073170731707, | |
| "grad_norm": 0.43061602115631104, | |
| "learning_rate": 5.081300813008131e-07, | |
| "loss": 0.2173, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 19.96951219512195, | |
| "grad_norm": 0.2903171181678772, | |
| "learning_rate": 8.468834688346885e-08, | |
| "loss": 0.2175, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.8221258134490239, | |
| "eval_loss": 0.7220444679260254, | |
| "eval_runtime": 50.0162, | |
| "eval_samples_per_second": 18.434, | |
| "eval_steps_per_second": 0.58, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 3280, | |
| "total_flos": 0.0, | |
| "train_loss": 0.2852184054691617, | |
| "train_runtime": 16462.385, | |
| "train_samples_per_second": 6.347, | |
| "train_steps_per_second": 0.199 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 3280, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "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": 0.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |