SLM_vs_LLM_experiments
/
max_seq_length_128_experiments
/distilbert
/distilroberta_base_patent
/trainer_state.json
| { | |
| "best_metric": 1.0022460222244263, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/distilbert/distilroberta_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": 2.0824131965637207, | |
| "learning_rate": 1.9829497016197786e-05, | |
| "loss": 2.183, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 4.778645038604736, | |
| "learning_rate": 1.9658994032395567e-05, | |
| "loss": 1.9562, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 4.916919231414795, | |
| "learning_rate": 1.9488491048593352e-05, | |
| "loss": 1.7366, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.5065836906433105, | |
| "learning_rate": 1.9317988064791137e-05, | |
| "loss": 1.5817, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 3.7337374687194824, | |
| "learning_rate": 1.9147485080988918e-05, | |
| "loss": 1.5474, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_accuracy": 0.4644, | |
| "eval_f1_macro": 0.3007262651897904, | |
| "eval_f1_micro": 0.4644, | |
| "eval_loss": 1.4682279825210571, | |
| "eval_runtime": 2.5172, | |
| "eval_samples_per_second": 1986.333, | |
| "eval_steps_per_second": 31.384, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.306447744369507, | |
| "learning_rate": 1.8976982097186702e-05, | |
| "loss": 1.4614, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 4.347895622253418, | |
| "learning_rate": 1.8806479113384487e-05, | |
| "loss": 1.4045, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 4.505576133728027, | |
| "learning_rate": 1.863597612958227e-05, | |
| "loss": 1.397, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 7.081475734710693, | |
| "learning_rate": 1.8465473145780053e-05, | |
| "loss": 1.3458, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 4.229913711547852, | |
| "learning_rate": 1.8294970161977838e-05, | |
| "loss": 1.2975, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_accuracy": 0.5514, | |
| "eval_f1_macro": 0.385696822491771, | |
| "eval_f1_micro": 0.5514, | |
| "eval_loss": 1.2702211141586304, | |
| "eval_runtime": 2.4835, | |
| "eval_samples_per_second": 2013.318, | |
| "eval_steps_per_second": 31.81, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 6.899994373321533, | |
| "learning_rate": 1.812446717817562e-05, | |
| "loss": 1.3211, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 5.848340034484863, | |
| "learning_rate": 1.7953964194373403e-05, | |
| "loss": 1.2881, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 6.385013103485107, | |
| "learning_rate": 1.7783461210571188e-05, | |
| "loss": 1.205, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 6.531163692474365, | |
| "learning_rate": 1.761295822676897e-05, | |
| "loss": 1.2051, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 6.575864791870117, | |
| "learning_rate": 1.7442455242966754e-05, | |
| "loss": 1.277, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_accuracy": 0.588, | |
| "eval_f1_macro": 0.4212858134893549, | |
| "eval_f1_micro": 0.588, | |
| "eval_loss": 1.1989067792892456, | |
| "eval_runtime": 2.4878, | |
| "eval_samples_per_second": 2009.844, | |
| "eval_steps_per_second": 31.756, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 5.341747760772705, | |
| "learning_rate": 1.727195225916454e-05, | |
| "loss": 1.1411, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 6.646276473999023, | |
| "learning_rate": 1.710144927536232e-05, | |
| "loss": 1.2777, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 7.130493640899658, | |
| "learning_rate": 1.6930946291560104e-05, | |
| "loss": 1.2124, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 6.087169647216797, | |
| "learning_rate": 1.676044330775789e-05, | |
| "loss": 1.2863, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 5.556641101837158, | |
| "learning_rate": 1.658994032395567e-05, | |
| "loss": 1.1483, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_accuracy": 0.6018, | |
| "eval_f1_macro": 0.44326147747487465, | |
| "eval_f1_micro": 0.6018, | |
| "eval_loss": 1.150877833366394, | |
| "eval_runtime": 2.5501, | |
| "eval_samples_per_second": 1960.73, | |
| "eval_steps_per_second": 30.98, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 7.140419960021973, | |
| "learning_rate": 1.6419437340153455e-05, | |
| "loss": 1.1968, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 5.2833147048950195, | |
| "learning_rate": 1.624893435635124e-05, | |
| "loss": 1.1433, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 5.170966148376465, | |
| "learning_rate": 1.607843137254902e-05, | |
| "loss": 1.1196, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 8.696062088012695, | |
| "learning_rate": 1.5907928388746805e-05, | |
| "loss": 1.1489, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 5.4057698249816895, | |
| "learning_rate": 1.573742540494459e-05, | |
| "loss": 1.1909, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.618, | |
| "eval_f1_macro": 0.4784833651456958, | |
| "eval_f1_micro": 0.618, | |
| "eval_loss": 1.1209474802017212, | |
| "eval_runtime": 2.5515, | |
| "eval_samples_per_second": 1959.62, | |
| "eval_steps_per_second": 30.962, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 6.896720886230469, | |
| "learning_rate": 1.556692242114237e-05, | |
| "loss": 1.191, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 6.1380791664123535, | |
| "learning_rate": 1.5396419437340155e-05, | |
| "loss": 1.1447, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 6.051652908325195, | |
| "learning_rate": 1.5225916453537938e-05, | |
| "loss": 1.0847, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 8.247784614562988, | |
| "learning_rate": 1.5055413469735723e-05, | |
| "loss": 1.1495, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 5.845155715942383, | |
| "learning_rate": 1.4884910485933506e-05, | |
| "loss": 1.1243, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_accuracy": 0.622, | |
| "eval_f1_macro": 0.49297464163053206, | |
| "eval_f1_micro": 0.622, | |
| "eval_loss": 1.1127510070800781, | |
| "eval_runtime": 2.5574, | |
| "eval_samples_per_second": 1955.106, | |
| "eval_steps_per_second": 30.891, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 6.277886867523193, | |
| "learning_rate": 1.4714407502131289e-05, | |
| "loss": 1.1136, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 6.175780773162842, | |
| "learning_rate": 1.4543904518329073e-05, | |
| "loss": 1.1245, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 5.58172607421875, | |
| "learning_rate": 1.4373401534526856e-05, | |
| "loss": 1.0753, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 6.277215480804443, | |
| "learning_rate": 1.420289855072464e-05, | |
| "loss": 1.1035, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 6.460805892944336, | |
| "learning_rate": 1.4032395566922424e-05, | |
| "loss": 1.1353, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "eval_accuracy": 0.609, | |
| "eval_f1_macro": 0.49301677161817686, | |
| "eval_f1_micro": 0.609, | |
| "eval_loss": 1.113377571105957, | |
| "eval_runtime": 2.5617, | |
| "eval_samples_per_second": 1951.828, | |
| "eval_steps_per_second": 30.839, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 5.35620641708374, | |
| "learning_rate": 1.3861892583120207e-05, | |
| "loss": 1.1247, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 7.259354591369629, | |
| "learning_rate": 1.369138959931799e-05, | |
| "loss": 1.1226, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 6.781973838806152, | |
| "learning_rate": 1.3520886615515774e-05, | |
| "loss": 1.035, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.575621604919434, | |
| "learning_rate": 1.3350383631713557e-05, | |
| "loss": 1.1588, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 7.666539669036865, | |
| "learning_rate": 1.317988064791134e-05, | |
| "loss": 1.0636, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_accuracy": 0.64, | |
| "eval_f1_macro": 0.5189274075388721, | |
| "eval_f1_micro": 0.64, | |
| "eval_loss": 1.0675694942474365, | |
| "eval_runtime": 2.5643, | |
| "eval_samples_per_second": 1949.816, | |
| "eval_steps_per_second": 30.807, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 4.991756916046143, | |
| "learning_rate": 1.3009377664109125e-05, | |
| "loss": 1.0461, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 6.300774574279785, | |
| "learning_rate": 1.2838874680306906e-05, | |
| "loss": 1.019, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 7.338348388671875, | |
| "learning_rate": 1.2668371696504689e-05, | |
| "loss": 1.0367, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 7.154069423675537, | |
| "learning_rate": 1.2497868712702472e-05, | |
| "loss": 1.094, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 4.677978515625, | |
| "learning_rate": 1.2327365728900256e-05, | |
| "loss": 0.9667, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_accuracy": 0.6404, | |
| "eval_f1_macro": 0.5192652764751173, | |
| "eval_f1_micro": 0.6404, | |
| "eval_loss": 1.0703057050704956, | |
| "eval_runtime": 2.5638, | |
| "eval_samples_per_second": 1950.255, | |
| "eval_steps_per_second": 30.814, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 6.083702564239502, | |
| "learning_rate": 1.215686274509804e-05, | |
| "loss": 1.0621, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 10.16114330291748, | |
| "learning_rate": 1.1986359761295822e-05, | |
| "loss": 1.0172, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 6.390872955322266, | |
| "learning_rate": 1.1815856777493607e-05, | |
| "loss": 0.972, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 7.280271530151367, | |
| "learning_rate": 1.164535379369139e-05, | |
| "loss": 0.9825, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 7.670015335083008, | |
| "learning_rate": 1.1474850809889173e-05, | |
| "loss": 1.0063, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.6386, | |
| "eval_f1_macro": 0.5127563688009887, | |
| "eval_f1_micro": 0.6386, | |
| "eval_loss": 1.049485206604004, | |
| "eval_runtime": 2.5639, | |
| "eval_samples_per_second": 1950.156, | |
| "eval_steps_per_second": 30.812, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 5.522829532623291, | |
| "learning_rate": 1.1304347826086957e-05, | |
| "loss": 1.023, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 9.047003746032715, | |
| "learning_rate": 1.113384484228474e-05, | |
| "loss": 1.0607, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 10.908841133117676, | |
| "learning_rate": 1.0963341858482523e-05, | |
| "loss": 1.0389, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 8.025410652160645, | |
| "learning_rate": 1.0792838874680308e-05, | |
| "loss": 1.0097, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 6.375541687011719, | |
| "learning_rate": 1.062233589087809e-05, | |
| "loss": 0.9521, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "eval_accuracy": 0.6432, | |
| "eval_f1_macro": 0.5185062629959201, | |
| "eval_f1_micro": 0.6432, | |
| "eval_loss": 1.04691743850708, | |
| "eval_runtime": 2.5619, | |
| "eval_samples_per_second": 1951.646, | |
| "eval_steps_per_second": 30.836, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 8.49925708770752, | |
| "learning_rate": 1.0451832907075873e-05, | |
| "loss": 1.0224, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 9.955578804016113, | |
| "learning_rate": 1.0281329923273658e-05, | |
| "loss": 1.0192, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 6.0503363609313965, | |
| "learning_rate": 1.0110826939471441e-05, | |
| "loss": 0.9438, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 7.6630048751831055, | |
| "learning_rate": 9.940323955669226e-06, | |
| "loss": 0.8864, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 6.8778815269470215, | |
| "learning_rate": 9.769820971867009e-06, | |
| "loss": 0.998, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "eval_accuracy": 0.6486, | |
| "eval_f1_macro": 0.5356688160968893, | |
| "eval_f1_micro": 0.6486, | |
| "eval_loss": 1.0358997583389282, | |
| "eval_runtime": 2.5668, | |
| "eval_samples_per_second": 1947.936, | |
| "eval_steps_per_second": 30.777, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 7.481098651885986, | |
| "learning_rate": 9.599317988064793e-06, | |
| "loss": 1.0152, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 7.424659252166748, | |
| "learning_rate": 9.428815004262576e-06, | |
| "loss": 1.01, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 9.956852912902832, | |
| "learning_rate": 9.258312020460359e-06, | |
| "loss": 1.0664, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 7.87490177154541, | |
| "learning_rate": 9.087809036658142e-06, | |
| "loss": 0.924, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 8.608465194702148, | |
| "learning_rate": 8.917306052855925e-06, | |
| "loss": 1.0188, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "eval_accuracy": 0.6418, | |
| "eval_f1_macro": 0.5394812129950733, | |
| "eval_f1_micro": 0.6418, | |
| "eval_loss": 1.0529745817184448, | |
| "eval_runtime": 2.5692, | |
| "eval_samples_per_second": 1946.122, | |
| "eval_steps_per_second": 30.749, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 7.214836597442627, | |
| "learning_rate": 8.74680306905371e-06, | |
| "loss": 0.9906, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 4.831575870513916, | |
| "learning_rate": 8.576300085251492e-06, | |
| "loss": 1.052, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 7.768477916717529, | |
| "learning_rate": 8.405797101449275e-06, | |
| "loss": 1.0451, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 7.0132527351379395, | |
| "learning_rate": 8.23529411764706e-06, | |
| "loss": 0.9187, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 12.03316879272461, | |
| "learning_rate": 8.064791133844843e-06, | |
| "loss": 0.9617, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "eval_accuracy": 0.6526, | |
| "eval_f1_macro": 0.5307017878131918, | |
| "eval_f1_micro": 0.6526, | |
| "eval_loss": 1.0214056968688965, | |
| "eval_runtime": 2.5148, | |
| "eval_samples_per_second": 1988.225, | |
| "eval_steps_per_second": 31.414, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 7.816739559173584, | |
| "learning_rate": 7.894288150042626e-06, | |
| "loss": 1.008, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 9.684741020202637, | |
| "learning_rate": 7.72378516624041e-06, | |
| "loss": 1.0609, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 6.8859453201293945, | |
| "learning_rate": 7.553282182438193e-06, | |
| "loss": 1.0002, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 6.9528350830078125, | |
| "learning_rate": 7.382779198635977e-06, | |
| "loss": 1.0341, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 6.688358783721924, | |
| "learning_rate": 7.21227621483376e-06, | |
| "loss": 1.0234, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_accuracy": 0.6514, | |
| "eval_f1_macro": 0.5495228476194496, | |
| "eval_f1_micro": 0.6514, | |
| "eval_loss": 1.0148464441299438, | |
| "eval_runtime": 2.5658, | |
| "eval_samples_per_second": 1948.681, | |
| "eval_steps_per_second": 30.789, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 7.913070201873779, | |
| "learning_rate": 7.0417732310315436e-06, | |
| "loss": 1.0278, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 7.68366813659668, | |
| "learning_rate": 6.8712702472293265e-06, | |
| "loss": 0.9754, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 7.161551475524902, | |
| "learning_rate": 6.70076726342711e-06, | |
| "loss": 1.0346, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 7.975135803222656, | |
| "learning_rate": 6.530264279624894e-06, | |
| "loss": 0.9265, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 6.6218791007995605, | |
| "learning_rate": 6.359761295822677e-06, | |
| "loss": 0.8914, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "eval_accuracy": 0.6544, | |
| "eval_f1_macro": 0.5603159525538163, | |
| "eval_f1_micro": 0.6544, | |
| "eval_loss": 1.0131620168685913, | |
| "eval_runtime": 2.5635, | |
| "eval_samples_per_second": 1950.467, | |
| "eval_steps_per_second": 30.817, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 11.909503936767578, | |
| "learning_rate": 6.189258312020461e-06, | |
| "loss": 0.9247, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 6.921329975128174, | |
| "learning_rate": 6.018755328218244e-06, | |
| "loss": 0.8752, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 8.646865844726562, | |
| "learning_rate": 5.848252344416027e-06, | |
| "loss": 0.9489, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 6.761762619018555, | |
| "learning_rate": 5.677749360613811e-06, | |
| "loss": 0.9303, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 6.754298210144043, | |
| "learning_rate": 5.507246376811595e-06, | |
| "loss": 0.9269, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "eval_accuracy": 0.6562, | |
| "eval_f1_macro": 0.5646631294942075, | |
| "eval_f1_micro": 0.6562, | |
| "eval_loss": 1.0109927654266357, | |
| "eval_runtime": 2.5641, | |
| "eval_samples_per_second": 1949.972, | |
| "eval_steps_per_second": 30.81, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 7.462425708770752, | |
| "learning_rate": 5.336743393009378e-06, | |
| "loss": 1.0024, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 10.302495002746582, | |
| "learning_rate": 5.1662404092071615e-06, | |
| "loss": 0.932, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 6.7270379066467285, | |
| "learning_rate": 4.995737425404945e-06, | |
| "loss": 1.0035, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "grad_norm": 8.183883666992188, | |
| "learning_rate": 4.825234441602728e-06, | |
| "loss": 0.9479, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 6.861289024353027, | |
| "learning_rate": 4.654731457800512e-06, | |
| "loss": 1.0351, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "eval_accuracy": 0.6528, | |
| "eval_f1_macro": 0.5716748212028093, | |
| "eval_f1_micro": 0.6528, | |
| "eval_loss": 1.0123671293258667, | |
| "eval_runtime": 2.568, | |
| "eval_samples_per_second": 1947.076, | |
| "eval_steps_per_second": 30.764, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 7.027976036071777, | |
| "learning_rate": 4.484228473998296e-06, | |
| "loss": 0.9684, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 7.110782623291016, | |
| "learning_rate": 4.313725490196079e-06, | |
| "loss": 0.9008, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 8.018675804138184, | |
| "learning_rate": 4.143222506393862e-06, | |
| "loss": 0.9205, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 7.683128833770752, | |
| "learning_rate": 3.972719522591646e-06, | |
| "loss": 0.9033, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 8.171706199645996, | |
| "learning_rate": 3.802216538789429e-06, | |
| "loss": 0.9582, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "eval_accuracy": 0.6524, | |
| "eval_f1_macro": 0.555182452599764, | |
| "eval_f1_micro": 0.6524, | |
| "eval_loss": 1.0150039196014404, | |
| "eval_runtime": 2.5637, | |
| "eval_samples_per_second": 1950.307, | |
| "eval_steps_per_second": 30.815, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 8.379620552062988, | |
| "learning_rate": 3.6317135549872124e-06, | |
| "loss": 0.9552, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 7.152894496917725, | |
| "learning_rate": 3.4612105711849957e-06, | |
| "loss": 0.9719, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 6.816888809204102, | |
| "learning_rate": 3.2907075873827795e-06, | |
| "loss": 0.8721, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 7.131276607513428, | |
| "learning_rate": 3.120204603580563e-06, | |
| "loss": 0.9568, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 9.819961547851562, | |
| "learning_rate": 2.949701619778346e-06, | |
| "loss": 0.8959, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_accuracy": 0.659, | |
| "eval_f1_macro": 0.5741291743786698, | |
| "eval_f1_micro": 0.659, | |
| "eval_loss": 1.0068503618240356, | |
| "eval_runtime": 2.5602, | |
| "eval_samples_per_second": 1953.0, | |
| "eval_steps_per_second": 30.857, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 6.770971298217773, | |
| "learning_rate": 2.77919863597613e-06, | |
| "loss": 0.9506, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 6.461827278137207, | |
| "learning_rate": 2.6086956521739132e-06, | |
| "loss": 0.8873, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 5.610044479370117, | |
| "learning_rate": 2.4381926683716966e-06, | |
| "loss": 0.8473, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 7.214883804321289, | |
| "learning_rate": 2.2676896845694803e-06, | |
| "loss": 0.8994, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 10.100293159484863, | |
| "learning_rate": 2.0971867007672637e-06, | |
| "loss": 0.8342, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "eval_accuracy": 0.6596, | |
| "eval_f1_macro": 0.5793678136298913, | |
| "eval_f1_micro": 0.6596, | |
| "eval_loss": 1.00313401222229, | |
| "eval_runtime": 2.5601, | |
| "eval_samples_per_second": 1953.051, | |
| "eval_steps_per_second": 30.858, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 6.870211124420166, | |
| "learning_rate": 1.926683716965047e-06, | |
| "loss": 0.8885, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 10.746337890625, | |
| "learning_rate": 1.7561807331628305e-06, | |
| "loss": 0.9337, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 7.3471221923828125, | |
| "learning_rate": 1.585677749360614e-06, | |
| "loss": 0.9137, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 5.51514196395874, | |
| "learning_rate": 1.4151747655583974e-06, | |
| "loss": 0.8889, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 6.89326810836792, | |
| "learning_rate": 1.2446717817561808e-06, | |
| "loss": 0.883, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "eval_accuracy": 0.6594, | |
| "eval_f1_macro": 0.5767016713811989, | |
| "eval_f1_micro": 0.6594, | |
| "eval_loss": 1.0041770935058594, | |
| "eval_runtime": 2.5645, | |
| "eval_samples_per_second": 1949.663, | |
| "eval_steps_per_second": 30.805, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 5.81862735748291, | |
| "learning_rate": 1.0741687979539643e-06, | |
| "loss": 0.943, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 7.020530700683594, | |
| "learning_rate": 9.036658141517478e-07, | |
| "loss": 0.9289, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 6.579043388366699, | |
| "learning_rate": 7.331628303495311e-07, | |
| "loss": 0.9096, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 6.589502811431885, | |
| "learning_rate": 5.626598465473146e-07, | |
| "loss": 0.9547, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 6.961599349975586, | |
| "learning_rate": 3.921568627450981e-07, | |
| "loss": 0.9377, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "eval_accuracy": 0.6596, | |
| "eval_f1_macro": 0.5724668431238462, | |
| "eval_f1_micro": 0.6596, | |
| "eval_loss": 1.0022460222244263, | |
| "eval_runtime": 2.5601, | |
| "eval_samples_per_second": 1953.074, | |
| "eval_steps_per_second": 30.859, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 6.636003017425537, | |
| "learning_rate": 2.2165387894288152e-07, | |
| "loss": 0.8657, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 8.342982292175293, | |
| "learning_rate": 5.115089514066497e-08, | |
| "loss": 0.9396, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1173, | |
| "total_flos": 2486458469842944.0, | |
| "train_loss": 1.0698413495212564, | |
| "train_runtime": 210.2936, | |
| "train_samples_per_second": 356.644, | |
| "train_steps_per_second": 5.578 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1173, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "total_flos": 2486458469842944.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |