| { | |
| "best_metric": 0.9373559355735779, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/FacebookAI/roberta_base_patent/checkpoint-1150", | |
| "epoch": 3.0, | |
| "eval_steps": 50, | |
| "global_step": 1173, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 3.06965708732605, | |
| "learning_rate": 1.9829497016197786e-05, | |
| "loss": 2.1487, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.080266952514648, | |
| "learning_rate": 1.9658994032395567e-05, | |
| "loss": 1.9493, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.19921875, | |
| "learning_rate": 1.9488491048593352e-05, | |
| "loss": 1.7343, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 7.48213529586792, | |
| "learning_rate": 1.9317988064791137e-05, | |
| "loss": 1.5789, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 4.417103290557861, | |
| "learning_rate": 1.9147485080988918e-05, | |
| "loss": 1.5014, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_accuracy": 0.5094, | |
| "eval_f1_macro": 0.3518682769080962, | |
| "eval_f1_micro": 0.5094, | |
| "eval_loss": 1.424868106842041, | |
| "eval_runtime": 4.5754, | |
| "eval_samples_per_second": 1092.79, | |
| "eval_steps_per_second": 17.266, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 6.253727436065674, | |
| "learning_rate": 1.8976982097186702e-05, | |
| "loss": 1.432, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 7.414198398590088, | |
| "learning_rate": 1.8806479113384487e-05, | |
| "loss": 1.4074, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 5.386297225952148, | |
| "learning_rate": 1.863597612958227e-05, | |
| "loss": 1.3724, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 7.517615795135498, | |
| "learning_rate": 1.8465473145780053e-05, | |
| "loss": 1.3282, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 5.478610515594482, | |
| "learning_rate": 1.8294970161977838e-05, | |
| "loss": 1.2516, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_accuracy": 0.5716, | |
| "eval_f1_macro": 0.41097137925827465, | |
| "eval_f1_micro": 0.5716, | |
| "eval_loss": 1.235276222229004, | |
| "eval_runtime": 4.6509, | |
| "eval_samples_per_second": 1075.062, | |
| "eval_steps_per_second": 16.986, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 9.781042098999023, | |
| "learning_rate": 1.812446717817562e-05, | |
| "loss": 1.284, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 8.909000396728516, | |
| "learning_rate": 1.7953964194373403e-05, | |
| "loss": 1.2797, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 9.288195610046387, | |
| "learning_rate": 1.7783461210571188e-05, | |
| "loss": 1.1562, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 9.174208641052246, | |
| "learning_rate": 1.761295822676897e-05, | |
| "loss": 1.1898, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 8.350323677062988, | |
| "learning_rate": 1.7442455242966754e-05, | |
| "loss": 1.2231, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_accuracy": 0.6184, | |
| "eval_f1_macro": 0.47063282197492184, | |
| "eval_f1_micro": 0.6184, | |
| "eval_loss": 1.1279376745224, | |
| "eval_runtime": 4.6588, | |
| "eval_samples_per_second": 1073.247, | |
| "eval_steps_per_second": 16.957, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 6.855159759521484, | |
| "learning_rate": 1.727195225916454e-05, | |
| "loss": 1.0786, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 8.450016021728516, | |
| "learning_rate": 1.710144927536232e-05, | |
| "loss": 1.2323, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 9.08232307434082, | |
| "learning_rate": 1.6930946291560104e-05, | |
| "loss": 1.1601, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 6.015944480895996, | |
| "learning_rate": 1.676044330775789e-05, | |
| "loss": 1.2161, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 6.851879119873047, | |
| "learning_rate": 1.658994032395567e-05, | |
| "loss": 1.1169, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_accuracy": 0.6346, | |
| "eval_f1_macro": 0.501600026681922, | |
| "eval_f1_micro": 0.6346, | |
| "eval_loss": 1.0773311853408813, | |
| "eval_runtime": 4.675, | |
| "eval_samples_per_second": 1069.526, | |
| "eval_steps_per_second": 16.899, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 9.108482360839844, | |
| "learning_rate": 1.6419437340153455e-05, | |
| "loss": 1.1574, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 5.855172157287598, | |
| "learning_rate": 1.624893435635124e-05, | |
| "loss": 1.0912, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 6.393794536590576, | |
| "learning_rate": 1.607843137254902e-05, | |
| "loss": 1.0653, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 11.836044311523438, | |
| "learning_rate": 1.5907928388746805e-05, | |
| "loss": 1.0897, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 6.626937389373779, | |
| "learning_rate": 1.573742540494459e-05, | |
| "loss": 1.1195, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.6398, | |
| "eval_f1_macro": 0.5181922250775877, | |
| "eval_f1_micro": 0.6398, | |
| "eval_loss": 1.068590521812439, | |
| "eval_runtime": 4.683, | |
| "eval_samples_per_second": 1067.686, | |
| "eval_steps_per_second": 16.869, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 9.058147430419922, | |
| "learning_rate": 1.556692242114237e-05, | |
| "loss": 1.1408, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 7.55896520614624, | |
| "learning_rate": 1.5396419437340155e-05, | |
| "loss": 1.1246, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 7.467323303222656, | |
| "learning_rate": 1.5225916453537938e-05, | |
| "loss": 1.0448, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 8.626157760620117, | |
| "learning_rate": 1.5055413469735723e-05, | |
| "loss": 1.0834, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 7.654249668121338, | |
| "learning_rate": 1.4884910485933506e-05, | |
| "loss": 1.0737, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_accuracy": 0.6426, | |
| "eval_f1_macro": 0.5232106803544677, | |
| "eval_f1_micro": 0.6426, | |
| "eval_loss": 1.0547648668289185, | |
| "eval_runtime": 4.6943, | |
| "eval_samples_per_second": 1065.121, | |
| "eval_steps_per_second": 16.829, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 8.641997337341309, | |
| "learning_rate": 1.4714407502131289e-05, | |
| "loss": 1.0827, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 7.739094257354736, | |
| "learning_rate": 1.4543904518329073e-05, | |
| "loss": 1.0868, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 6.531001091003418, | |
| "learning_rate": 1.4373401534526856e-05, | |
| "loss": 1.038, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 5.194310188293457, | |
| "learning_rate": 1.420289855072464e-05, | |
| "loss": 1.064, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.974400997161865, | |
| "learning_rate": 1.4032395566922424e-05, | |
| "loss": 1.0981, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_accuracy": 0.6376, | |
| "eval_f1_macro": 0.5604579153144441, | |
| "eval_f1_micro": 0.6376, | |
| "eval_loss": 1.043789267539978, | |
| "eval_runtime": 4.6993, | |
| "eval_samples_per_second": 1063.984, | |
| "eval_steps_per_second": 16.811, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 8.46001148223877, | |
| "learning_rate": 1.3861892583120207e-05, | |
| "loss": 1.102, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 8.373648643493652, | |
| "learning_rate": 1.369138959931799e-05, | |
| "loss": 1.0584, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 7.385919094085693, | |
| "learning_rate": 1.3520886615515774e-05, | |
| "loss": 1.0176, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 7.325735092163086, | |
| "learning_rate": 1.3350383631713557e-05, | |
| "loss": 1.092, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 7.1516008377075195, | |
| "learning_rate": 1.317988064791134e-05, | |
| "loss": 1.0147, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.6606, | |
| "eval_f1_macro": 0.5852129283356313, | |
| "eval_f1_micro": 0.6606, | |
| "eval_loss": 0.9970415830612183, | |
| "eval_runtime": 4.6968, | |
| "eval_samples_per_second": 1064.548, | |
| "eval_steps_per_second": 16.82, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 6.275845050811768, | |
| "learning_rate": 1.3009377664109125e-05, | |
| "loss": 0.9861, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 8.04443645477295, | |
| "learning_rate": 1.2838874680306906e-05, | |
| "loss": 0.9872, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 8.21890926361084, | |
| "learning_rate": 1.2668371696504689e-05, | |
| "loss": 0.9606, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 7.705150604248047, | |
| "learning_rate": 1.2497868712702472e-05, | |
| "loss": 1.0163, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 6.80154275894165, | |
| "learning_rate": 1.2327365728900256e-05, | |
| "loss": 0.9049, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_accuracy": 0.6572, | |
| "eval_f1_macro": 0.5804469606073931, | |
| "eval_f1_micro": 0.6572, | |
| "eval_loss": 1.0097591876983643, | |
| "eval_runtime": 4.6955, | |
| "eval_samples_per_second": 1064.848, | |
| "eval_steps_per_second": 16.825, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 8.022605895996094, | |
| "learning_rate": 1.215686274509804e-05, | |
| "loss": 0.99, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 11.028081893920898, | |
| "learning_rate": 1.1986359761295822e-05, | |
| "loss": 0.9621, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 9.201847076416016, | |
| "learning_rate": 1.1815856777493607e-05, | |
| "loss": 0.9033, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 9.63035774230957, | |
| "learning_rate": 1.164535379369139e-05, | |
| "loss": 0.9062, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 9.333333969116211, | |
| "learning_rate": 1.1474850809889173e-05, | |
| "loss": 0.945, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.662, | |
| "eval_f1_macro": 0.5872747578923891, | |
| "eval_f1_micro": 0.662, | |
| "eval_loss": 0.9906530380249023, | |
| "eval_runtime": 4.6972, | |
| "eval_samples_per_second": 1064.474, | |
| "eval_steps_per_second": 16.819, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 6.602180004119873, | |
| "learning_rate": 1.1304347826086957e-05, | |
| "loss": 0.9584, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 10.108154296875, | |
| "learning_rate": 1.113384484228474e-05, | |
| "loss": 0.9882, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 9.116514205932617, | |
| "learning_rate": 1.0963341858482523e-05, | |
| "loss": 0.9376, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 8.690980911254883, | |
| "learning_rate": 1.0792838874680308e-05, | |
| "loss": 0.944, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 7.5809831619262695, | |
| "learning_rate": 1.062233589087809e-05, | |
| "loss": 0.9206, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "eval_accuracy": 0.6636, | |
| "eval_f1_macro": 0.5777404076748278, | |
| "eval_f1_micro": 0.6636, | |
| "eval_loss": 0.9864922761917114, | |
| "eval_runtime": 4.6971, | |
| "eval_samples_per_second": 1064.478, | |
| "eval_steps_per_second": 16.819, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 9.885359764099121, | |
| "learning_rate": 1.0451832907075873e-05, | |
| "loss": 0.9316, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 9.082917213439941, | |
| "learning_rate": 1.0281329923273658e-05, | |
| "loss": 0.9554, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 8.62850284576416, | |
| "learning_rate": 1.0110826939471441e-05, | |
| "loss": 0.8987, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 9.194178581237793, | |
| "learning_rate": 9.940323955669226e-06, | |
| "loss": 0.8146, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 8.861343383789062, | |
| "learning_rate": 9.769820971867009e-06, | |
| "loss": 0.9263, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "eval_accuracy": 0.6664, | |
| "eval_f1_macro": 0.5968165647393973, | |
| "eval_f1_micro": 0.6664, | |
| "eval_loss": 0.9686366319656372, | |
| "eval_runtime": 4.6944, | |
| "eval_samples_per_second": 1065.103, | |
| "eval_steps_per_second": 16.829, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 8.089481353759766, | |
| "learning_rate": 9.599317988064793e-06, | |
| "loss": 0.9515, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 8.640591621398926, | |
| "learning_rate": 9.428815004262576e-06, | |
| "loss": 0.9406, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 12.582009315490723, | |
| "learning_rate": 9.258312020460359e-06, | |
| "loss": 1.0249, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 10.242859840393066, | |
| "learning_rate": 9.087809036658142e-06, | |
| "loss": 0.838, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 9.301600456237793, | |
| "learning_rate": 8.917306052855925e-06, | |
| "loss": 0.9629, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "eval_accuracy": 0.666, | |
| "eval_f1_macro": 0.5941368953862343, | |
| "eval_f1_micro": 0.666, | |
| "eval_loss": 0.9790616035461426, | |
| "eval_runtime": 4.693, | |
| "eval_samples_per_second": 1065.419, | |
| "eval_steps_per_second": 16.834, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 7.933264255523682, | |
| "learning_rate": 8.74680306905371e-06, | |
| "loss": 0.9275, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 6.248515605926514, | |
| "learning_rate": 8.576300085251492e-06, | |
| "loss": 0.9821, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 9.65717887878418, | |
| "learning_rate": 8.405797101449275e-06, | |
| "loss": 0.9896, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 8.450819969177246, | |
| "learning_rate": 8.23529411764706e-06, | |
| "loss": 0.8661, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 10.561975479125977, | |
| "learning_rate": 8.064791133844843e-06, | |
| "loss": 0.8913, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_accuracy": 0.6746, | |
| "eval_f1_macro": 0.600228948901954, | |
| "eval_f1_micro": 0.6746, | |
| "eval_loss": 0.9578554630279541, | |
| "eval_runtime": 4.6945, | |
| "eval_samples_per_second": 1065.084, | |
| "eval_steps_per_second": 16.828, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 8.024334907531738, | |
| "learning_rate": 7.894288150042626e-06, | |
| "loss": 0.9376, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 8.777924537658691, | |
| "learning_rate": 7.72378516624041e-06, | |
| "loss": 0.995, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 8.383976936340332, | |
| "learning_rate": 7.553282182438193e-06, | |
| "loss": 0.9258, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 7.384089469909668, | |
| "learning_rate": 7.382779198635977e-06, | |
| "loss": 0.982, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 7.771151542663574, | |
| "learning_rate": 7.21227621483376e-06, | |
| "loss": 0.96, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_accuracy": 0.6696, | |
| "eval_f1_macro": 0.6024537798111538, | |
| "eval_f1_micro": 0.6696, | |
| "eval_loss": 0.9523646235466003, | |
| "eval_runtime": 4.6912, | |
| "eval_samples_per_second": 1065.829, | |
| "eval_steps_per_second": 16.84, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 9.032242774963379, | |
| "learning_rate": 7.0417732310315436e-06, | |
| "loss": 0.972, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 8.851130485534668, | |
| "learning_rate": 6.8712702472293265e-06, | |
| "loss": 0.9154, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 8.968002319335938, | |
| "learning_rate": 6.70076726342711e-06, | |
| "loss": 0.96, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 8.527153015136719, | |
| "learning_rate": 6.530264279624894e-06, | |
| "loss": 0.8241, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 7.565133571624756, | |
| "learning_rate": 6.359761295822677e-06, | |
| "loss": 0.8284, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "eval_accuracy": 0.6738, | |
| "eval_f1_macro": 0.607303709282453, | |
| "eval_f1_micro": 0.6738, | |
| "eval_loss": 0.9539583921432495, | |
| "eval_runtime": 4.6905, | |
| "eval_samples_per_second": 1065.982, | |
| "eval_steps_per_second": 16.843, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 13.65481948852539, | |
| "learning_rate": 6.189258312020461e-06, | |
| "loss": 0.8675, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 6.871121883392334, | |
| "learning_rate": 6.018755328218244e-06, | |
| "loss": 0.7866, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 11.004075050354004, | |
| "learning_rate": 5.848252344416027e-06, | |
| "loss": 0.8634, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 6.931361198425293, | |
| "learning_rate": 5.677749360613811e-06, | |
| "loss": 0.8855, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 8.874567985534668, | |
| "learning_rate": 5.507246376811595e-06, | |
| "loss": 0.8558, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_accuracy": 0.6742, | |
| "eval_f1_macro": 0.594933677189279, | |
| "eval_f1_micro": 0.6742, | |
| "eval_loss": 0.9495751857757568, | |
| "eval_runtime": 4.6887, | |
| "eval_samples_per_second": 1066.398, | |
| "eval_steps_per_second": 16.849, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 9.37259292602539, | |
| "learning_rate": 5.336743393009378e-06, | |
| "loss": 0.8929, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 10.430830955505371, | |
| "learning_rate": 5.1662404092071615e-06, | |
| "loss": 0.8427, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 6.501585960388184, | |
| "learning_rate": 4.995737425404945e-06, | |
| "loss": 0.9074, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 9.332304000854492, | |
| "learning_rate": 4.825234441602728e-06, | |
| "loss": 0.8769, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 8.06771183013916, | |
| "learning_rate": 4.654731457800512e-06, | |
| "loss": 0.9643, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "eval_accuracy": 0.6748, | |
| "eval_f1_macro": 0.6074089654591388, | |
| "eval_f1_micro": 0.6748, | |
| "eval_loss": 0.9519968628883362, | |
| "eval_runtime": 4.6886, | |
| "eval_samples_per_second": 1066.415, | |
| "eval_steps_per_second": 16.849, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 9.928189277648926, | |
| "learning_rate": 4.484228473998296e-06, | |
| "loss": 0.8591, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 7.7598772048950195, | |
| "learning_rate": 4.313725490196079e-06, | |
| "loss": 0.8091, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 9.674939155578613, | |
| "learning_rate": 4.143222506393862e-06, | |
| "loss": 0.7889, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 10.04906177520752, | |
| "learning_rate": 3.972719522591646e-06, | |
| "loss": 0.8251, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 10.514945983886719, | |
| "learning_rate": 3.802216538789429e-06, | |
| "loss": 0.873, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "eval_accuracy": 0.6732, | |
| "eval_f1_macro": 0.596296648251256, | |
| "eval_f1_micro": 0.6732, | |
| "eval_loss": 0.9520734548568726, | |
| "eval_runtime": 4.6894, | |
| "eval_samples_per_second": 1066.228, | |
| "eval_steps_per_second": 16.846, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 10.718672752380371, | |
| "learning_rate": 3.6317135549872124e-06, | |
| "loss": 0.8836, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 10.7012357711792, | |
| "learning_rate": 3.4612105711849957e-06, | |
| "loss": 0.8958, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 8.067261695861816, | |
| "learning_rate": 3.2907075873827795e-06, | |
| "loss": 0.784, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 8.40015697479248, | |
| "learning_rate": 3.120204603580563e-06, | |
| "loss": 0.8807, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 9.301580429077148, | |
| "learning_rate": 2.949701619778346e-06, | |
| "loss": 0.8271, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_accuracy": 0.6782, | |
| "eval_f1_macro": 0.6122813087810713, | |
| "eval_f1_micro": 0.6782, | |
| "eval_loss": 0.9399372935295105, | |
| "eval_runtime": 4.6876, | |
| "eval_samples_per_second": 1066.637, | |
| "eval_steps_per_second": 16.853, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 7.235937118530273, | |
| "learning_rate": 2.77919863597613e-06, | |
| "loss": 0.8525, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 7.0617523193359375, | |
| "learning_rate": 2.6086956521739132e-06, | |
| "loss": 0.8082, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 7.261152744293213, | |
| "learning_rate": 2.4381926683716966e-06, | |
| "loss": 0.7973, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 10.145052909851074, | |
| "learning_rate": 2.2676896845694803e-06, | |
| "loss": 0.8335, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 10.394144058227539, | |
| "learning_rate": 2.0971867007672637e-06, | |
| "loss": 0.7572, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "eval_accuracy": 0.6788, | |
| "eval_f1_macro": 0.6075649716220995, | |
| "eval_f1_micro": 0.6788, | |
| "eval_loss": 0.940003514289856, | |
| "eval_runtime": 4.6862, | |
| "eval_samples_per_second": 1066.96, | |
| "eval_steps_per_second": 16.858, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 9.629010200500488, | |
| "learning_rate": 1.926683716965047e-06, | |
| "loss": 0.8284, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 11.781661033630371, | |
| "learning_rate": 1.7561807331628305e-06, | |
| "loss": 0.8437, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 9.535713195800781, | |
| "learning_rate": 1.585677749360614e-06, | |
| "loss": 0.8281, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 12.147420883178711, | |
| "learning_rate": 1.4151747655583974e-06, | |
| "loss": 0.7891, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 8.067042350769043, | |
| "learning_rate": 1.2446717817561808e-06, | |
| "loss": 0.7949, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "eval_accuracy": 0.6808, | |
| "eval_f1_macro": 0.6098706235096221, | |
| "eval_f1_micro": 0.6808, | |
| "eval_loss": 0.9386488199234009, | |
| "eval_runtime": 4.687, | |
| "eval_samples_per_second": 1066.791, | |
| "eval_steps_per_second": 16.855, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 7.676300525665283, | |
| "learning_rate": 1.0741687979539643e-06, | |
| "loss": 0.8565, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 9.350565910339355, | |
| "learning_rate": 9.036658141517478e-07, | |
| "loss": 0.8433, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 7.868011474609375, | |
| "learning_rate": 7.331628303495311e-07, | |
| "loss": 0.8412, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 9.809996604919434, | |
| "learning_rate": 5.626598465473146e-07, | |
| "loss": 0.8546, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 10.371634483337402, | |
| "learning_rate": 3.921568627450981e-07, | |
| "loss": 0.8183, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "eval_accuracy": 0.679, | |
| "eval_f1_macro": 0.6097649958285241, | |
| "eval_f1_micro": 0.679, | |
| "eval_loss": 0.9373559355735779, | |
| "eval_runtime": 4.6836, | |
| "eval_samples_per_second": 1067.561, | |
| "eval_steps_per_second": 16.867, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 8.789111137390137, | |
| "learning_rate": 2.2165387894288152e-07, | |
| "loss": 0.7812, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 9.823261260986328, | |
| "learning_rate": 5.115089514066497e-08, | |
| "loss": 0.8309, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1173, | |
| "total_flos": 4938378618339328.0, | |
| "train_loss": 1.0076844364989037, | |
| "train_runtime": 372.6672, | |
| "train_samples_per_second": 201.252, | |
| "train_steps_per_second": 3.148 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1173, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "total_flos": 4938378618339328.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |