| { | |
| "best_metric": 0.939655172413793, | |
| "best_model_checkpoint": "/scratch/camembertv2/runs/results/ftb_ner/camembertav2-base-bf16-p2-17000/max_seq_length-192-gradient_accumulation_steps-2-precision-fp32-learning_rate-5.000000000000001e-05-epochs-8-lr_scheduler-linear-warmup_steps-0.1/SEED-42/checkpoint-4944", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 4944, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16181229773462782, | |
| "grad_norm": 0.7081506848335266, | |
| "learning_rate": 1.0101010101010103e-05, | |
| "loss": 1.713, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32362459546925565, | |
| "grad_norm": 0.6198400855064392, | |
| "learning_rate": 2.0202020202020206e-05, | |
| "loss": 0.1665, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4854368932038835, | |
| "grad_norm": 0.813007652759552, | |
| "learning_rate": 3.030303030303031e-05, | |
| "loss": 0.0816, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6472491909385113, | |
| "grad_norm": 0.9836126565933228, | |
| "learning_rate": 4.040404040404041e-05, | |
| "loss": 0.0487, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8090614886731392, | |
| "grad_norm": 0.2043960988521576, | |
| "learning_rate": 4.9943807597212865e-05, | |
| "loss": 0.0415, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.970873786407767, | |
| "grad_norm": 0.6549485921859741, | |
| "learning_rate": 4.8819959541470004e-05, | |
| "loss": 0.0435, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.9896461049632957, | |
| "eval_f1": 0.8642940946421588, | |
| "eval_loss": 0.04151414707303047, | |
| "eval_precision": 0.8599221789883269, | |
| "eval_recall": 0.8687106918238994, | |
| "eval_runtime": 4.2561, | |
| "eval_samples_per_second": 290.171, | |
| "eval_steps_per_second": 36.418, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.132686084142395, | |
| "grad_norm": 0.905681848526001, | |
| "learning_rate": 4.7696111485727136e-05, | |
| "loss": 0.0221, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2944983818770226, | |
| "grad_norm": 0.4660435616970062, | |
| "learning_rate": 4.6572263429984275e-05, | |
| "loss": 0.029, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.4563106796116505, | |
| "grad_norm": 0.4658781886100769, | |
| "learning_rate": 4.544841537424141e-05, | |
| "loss": 0.0265, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.6181229773462782, | |
| "grad_norm": 1.3315564393997192, | |
| "learning_rate": 4.4324567318498546e-05, | |
| "loss": 0.022, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.779935275080906, | |
| "grad_norm": 0.08498021215200424, | |
| "learning_rate": 4.3200719262755685e-05, | |
| "loss": 0.025, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.941747572815534, | |
| "grad_norm": 0.4140647351741791, | |
| "learning_rate": 4.207687120701282e-05, | |
| "loss": 0.0386, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9920017530404295, | |
| "eval_f1": 0.8975457732761979, | |
| "eval_loss": 0.035175956785678864, | |
| "eval_precision": 0.8895752895752895, | |
| "eval_recall": 0.9056603773584906, | |
| "eval_runtime": 3.7654, | |
| "eval_samples_per_second": 327.989, | |
| "eval_steps_per_second": 41.165, | |
| "step": 1236 | |
| }, | |
| { | |
| "epoch": 2.103559870550162, | |
| "grad_norm": 0.43421319127082825, | |
| "learning_rate": 4.0953023151269956e-05, | |
| "loss": 0.0249, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.26537216828479, | |
| "grad_norm": 0.2701377272605896, | |
| "learning_rate": 3.9829175095527095e-05, | |
| "loss": 0.0191, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.4271844660194173, | |
| "grad_norm": 0.11492807418107986, | |
| "learning_rate": 3.870532703978423e-05, | |
| "loss": 0.0136, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.588996763754045, | |
| "grad_norm": 0.582848072052002, | |
| "learning_rate": 3.758147898404136e-05, | |
| "loss": 0.0258, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.750809061488673, | |
| "grad_norm": 0.2515222132205963, | |
| "learning_rate": 3.6457630928298505e-05, | |
| "loss": 0.0169, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.912621359223301, | |
| "grad_norm": 0.36368367075920105, | |
| "learning_rate": 3.533378287255564e-05, | |
| "loss": 0.0123, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9929330557685987, | |
| "eval_f1": 0.9168309026409145, | |
| "eval_loss": 0.030728859826922417, | |
| "eval_precision": 0.9193675889328063, | |
| "eval_recall": 0.914308176100629, | |
| "eval_runtime": 3.7477, | |
| "eval_samples_per_second": 329.532, | |
| "eval_steps_per_second": 41.358, | |
| "step": 1854 | |
| }, | |
| { | |
| "epoch": 3.074433656957929, | |
| "grad_norm": 0.3048587739467621, | |
| "learning_rate": 3.420993481681277e-05, | |
| "loss": 0.0128, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.236245954692557, | |
| "grad_norm": 0.01976470835506916, | |
| "learning_rate": 3.308608676106991e-05, | |
| "loss": 0.0145, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.3980582524271843, | |
| "grad_norm": 0.17214246094226837, | |
| "learning_rate": 3.196223870532705e-05, | |
| "loss": 0.0092, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.559870550161812, | |
| "grad_norm": 0.2264121174812317, | |
| "learning_rate": 3.083839064958418e-05, | |
| "loss": 0.0111, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.72168284789644, | |
| "grad_norm": 0.13608065247535706, | |
| "learning_rate": 2.971454259384132e-05, | |
| "loss": 0.0124, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.883495145631068, | |
| "grad_norm": 0.027897851541638374, | |
| "learning_rate": 2.8590694538098453e-05, | |
| "loss": 0.0136, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9914265366495015, | |
| "eval_f1": 0.9107497024990083, | |
| "eval_loss": 0.03449518606066704, | |
| "eval_precision": 0.9191353082465973, | |
| "eval_recall": 0.9025157232704403, | |
| "eval_runtime": 3.7618, | |
| "eval_samples_per_second": 328.3, | |
| "eval_steps_per_second": 41.204, | |
| "step": 2472 | |
| }, | |
| { | |
| "epoch": 4.0453074433656955, | |
| "grad_norm": 0.18220163881778717, | |
| "learning_rate": 2.746684648235559e-05, | |
| "loss": 0.0088, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.207119741100324, | |
| "grad_norm": 0.016186678782105446, | |
| "learning_rate": 2.6342998426612728e-05, | |
| "loss": 0.0087, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.368932038834951, | |
| "grad_norm": 0.18213553726673126, | |
| "learning_rate": 2.5219150370869863e-05, | |
| "loss": 0.0051, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.53074433656958, | |
| "grad_norm": 0.012032161466777325, | |
| "learning_rate": 2.4095302315127e-05, | |
| "loss": 0.0081, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.692556634304207, | |
| "grad_norm": 0.0591534860432148, | |
| "learning_rate": 2.2971454259384134e-05, | |
| "loss": 0.0042, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.854368932038835, | |
| "grad_norm": 0.3236534893512726, | |
| "learning_rate": 2.1847606203641273e-05, | |
| "loss": 0.009, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9935904459296593, | |
| "eval_f1": 0.9192352711666016, | |
| "eval_loss": 0.029976682737469673, | |
| "eval_precision": 0.9124709527498064, | |
| "eval_recall": 0.9261006289308176, | |
| "eval_runtime": 3.7424, | |
| "eval_samples_per_second": 330.006, | |
| "eval_steps_per_second": 41.418, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 5.016181229773463, | |
| "grad_norm": 0.006900902837514877, | |
| "learning_rate": 2.072375814789841e-05, | |
| "loss": 0.0089, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 5.17799352750809, | |
| "grad_norm": 0.10268358141183853, | |
| "learning_rate": 1.9599910092155544e-05, | |
| "loss": 0.003, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.339805825242719, | |
| "grad_norm": 0.008188015781342983, | |
| "learning_rate": 1.847606203641268e-05, | |
| "loss": 0.003, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 5.501618122977346, | |
| "grad_norm": 0.159342423081398, | |
| "learning_rate": 1.735221398066982e-05, | |
| "loss": 0.0027, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.663430420711974, | |
| "grad_norm": 0.4539357125759125, | |
| "learning_rate": 1.622836592492695e-05, | |
| "loss": 0.0042, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.825242718446602, | |
| "grad_norm": 0.038250233978033066, | |
| "learning_rate": 1.510451786918409e-05, | |
| "loss": 0.0039, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.9870550161812295, | |
| "grad_norm": 0.004649197217077017, | |
| "learning_rate": 1.3980669813441227e-05, | |
| "loss": 0.0038, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9935630546729484, | |
| "eval_f1": 0.9324960753532182, | |
| "eval_loss": 0.033860765397548676, | |
| "eval_precision": 0.9310344827586207, | |
| "eval_recall": 0.9339622641509434, | |
| "eval_runtime": 3.7552, | |
| "eval_samples_per_second": 328.879, | |
| "eval_steps_per_second": 41.276, | |
| "step": 3708 | |
| }, | |
| { | |
| "epoch": 6.148867313915858, | |
| "grad_norm": 0.20126421749591827, | |
| "learning_rate": 1.285682175769836e-05, | |
| "loss": 0.0016, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.310679611650485, | |
| "grad_norm": 0.006144899874925613, | |
| "learning_rate": 1.1732973701955498e-05, | |
| "loss": 0.0024, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 6.472491909385114, | |
| "grad_norm": 1.5297178030014038, | |
| "learning_rate": 1.0609125646212633e-05, | |
| "loss": 0.0022, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.634304207119741, | |
| "grad_norm": 0.04670717939734459, | |
| "learning_rate": 9.48527759046977e-06, | |
| "loss": 0.002, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 6.796116504854369, | |
| "grad_norm": 0.028466541320085526, | |
| "learning_rate": 8.361429534726907e-06, | |
| "loss": 0.0018, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.957928802588997, | |
| "grad_norm": 0.09033439308404922, | |
| "learning_rate": 7.237581478984042e-06, | |
| "loss": 0.0019, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9936726196997918, | |
| "eval_f1": 0.9328125, | |
| "eval_loss": 0.03374877944588661, | |
| "eval_precision": 0.9270186335403726, | |
| "eval_recall": 0.9386792452830188, | |
| "eval_runtime": 3.7709, | |
| "eval_samples_per_second": 327.506, | |
| "eval_steps_per_second": 41.104, | |
| "step": 4326 | |
| }, | |
| { | |
| "epoch": 7.119741100323624, | |
| "grad_norm": 0.020341284573078156, | |
| "learning_rate": 6.113733423241179e-06, | |
| "loss": 0.0017, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.281553398058253, | |
| "grad_norm": 0.026211928576231003, | |
| "learning_rate": 4.989885367498316e-06, | |
| "loss": 0.0009, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.44336569579288, | |
| "grad_norm": 0.0029867186676710844, | |
| "learning_rate": 3.866037311755451e-06, | |
| "loss": 0.0014, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 7.605177993527509, | |
| "grad_norm": 0.014907300472259521, | |
| "learning_rate": 2.742189256012588e-06, | |
| "loss": 0.0014, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 7.766990291262136, | |
| "grad_norm": 0.12841272354125977, | |
| "learning_rate": 1.6183412002697239e-06, | |
| "loss": 0.0013, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.9288025889967635, | |
| "grad_norm": 0.0018621513154357672, | |
| "learning_rate": 4.9449314452686e-07, | |
| "loss": 0.0013, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9941108798071655, | |
| "eval_f1": 0.939655172413793, | |
| "eval_loss": 0.03547634929418564, | |
| "eval_precision": 0.93671875, | |
| "eval_recall": 0.9426100628930818, | |
| "eval_runtime": 3.7628, | |
| "eval_samples_per_second": 328.211, | |
| "eval_steps_per_second": 41.192, | |
| "step": 4944 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 4944, | |
| "total_flos": 2834249641269840.0, | |
| "train_loss": 0.05133420664178129, | |
| "train_runtime": 849.5258, | |
| "train_samples_per_second": 93.05, | |
| "train_steps_per_second": 5.82 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 4944, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "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": 2834249641269840.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |