SLM_vs_LLM_experiments
/
max_seq_length_128_experiments
/LoRA
/Qwen
/Qwen1.5_7B_LoRA_MAdAiLab
/amazon_attrprompt
/checkpoint-350
/trainer_state.json
| { | |
| "best_metric": 0.42495009303092957, | |
| "best_model_checkpoint": "../experiments_checkpoints/LoRA/Qwen/Qwen1.5_7B_LoRA_MAdAiLab/amazon_attrprompt/checkpoint-350", | |
| "epoch": 0.9210526315789473, | |
| "eval_steps": 50, | |
| "global_step": 350, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 88.5622329711914, | |
| "learning_rate": 4.956140350877193e-05, | |
| "loss": 9.8375, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 76.16983032226562, | |
| "learning_rate": 4.912280701754386e-05, | |
| "loss": 6.3422, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 68.94097137451172, | |
| "learning_rate": 4.868421052631579e-05, | |
| "loss": 3.866, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 72.53247833251953, | |
| "learning_rate": 4.824561403508772e-05, | |
| "loss": 2.1033, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 67.55592346191406, | |
| "learning_rate": 4.780701754385965e-05, | |
| "loss": 1.6833, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_accuracy": 0.6640316205533597, | |
| "eval_f1_macro": 0.5878963576218975, | |
| "eval_f1_micro": 0.6640316205533597, | |
| "eval_loss": 1.2279417514801025, | |
| "eval_runtime": 19.0551, | |
| "eval_samples_per_second": 79.664, | |
| "eval_steps_per_second": 2.519, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 34.1900520324707, | |
| "learning_rate": 4.736842105263158e-05, | |
| "loss": 1.0584, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 48.36394119262695, | |
| "learning_rate": 4.6929824561403515e-05, | |
| "loss": 1.0168, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 46.12382888793945, | |
| "learning_rate": 4.649122807017544e-05, | |
| "loss": 0.8845, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 43.66932678222656, | |
| "learning_rate": 4.605263157894737e-05, | |
| "loss": 0.8027, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 28.10245132446289, | |
| "learning_rate": 4.56140350877193e-05, | |
| "loss": 0.6531, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_accuracy": 0.8155467720685112, | |
| "eval_f1_macro": 0.7766571443103144, | |
| "eval_f1_micro": 0.8155467720685112, | |
| "eval_loss": 0.6577733755111694, | |
| "eval_runtime": 19.1186, | |
| "eval_samples_per_second": 79.399, | |
| "eval_steps_per_second": 2.511, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 59.69050216674805, | |
| "learning_rate": 4.517543859649123e-05, | |
| "loss": 0.8677, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 56.882625579833984, | |
| "learning_rate": 4.473684210526316e-05, | |
| "loss": 0.7174, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 29.96848487854004, | |
| "learning_rate": 4.429824561403509e-05, | |
| "loss": 0.5984, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 53.101444244384766, | |
| "learning_rate": 4.3859649122807014e-05, | |
| "loss": 0.5318, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 30.550228118896484, | |
| "learning_rate": 4.342105263157895e-05, | |
| "loss": 0.6075, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "eval_accuracy": 0.8326745718050066, | |
| "eval_f1_macro": 0.8113232244484544, | |
| "eval_f1_micro": 0.8326745718050066, | |
| "eval_loss": 0.5934926867485046, | |
| "eval_runtime": 19.1445, | |
| "eval_samples_per_second": 79.292, | |
| "eval_steps_per_second": 2.507, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 47.42914581298828, | |
| "learning_rate": 4.298245614035088e-05, | |
| "loss": 0.7186, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 34.166526794433594, | |
| "learning_rate": 4.254385964912281e-05, | |
| "loss": 0.7026, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 28.214841842651367, | |
| "learning_rate": 4.210526315789474e-05, | |
| "loss": 0.6292, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 45.511444091796875, | |
| "learning_rate": 4.166666666666667e-05, | |
| "loss": 0.7115, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 18.114608764648438, | |
| "learning_rate": 4.12280701754386e-05, | |
| "loss": 0.5646, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_accuracy": 0.8379446640316206, | |
| "eval_f1_macro": 0.8193854450487876, | |
| "eval_f1_micro": 0.8379446640316206, | |
| "eval_loss": 0.5659688115119934, | |
| "eval_runtime": 19.1468, | |
| "eval_samples_per_second": 79.282, | |
| "eval_steps_per_second": 2.507, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 29.09646987915039, | |
| "learning_rate": 4.078947368421053e-05, | |
| "loss": 0.4866, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 44.29309844970703, | |
| "learning_rate": 4.0350877192982455e-05, | |
| "loss": 0.4314, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 42.93623733520508, | |
| "learning_rate": 3.991228070175439e-05, | |
| "loss": 0.7159, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 33.71025848388672, | |
| "learning_rate": 3.9473684210526316e-05, | |
| "loss": 0.6136, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 42.2908821105957, | |
| "learning_rate": 3.9035087719298244e-05, | |
| "loss": 0.6148, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "eval_accuracy": 0.8425559947299077, | |
| "eval_f1_macro": 0.8319363334583545, | |
| "eval_f1_micro": 0.8425559947299077, | |
| "eval_loss": 0.531785249710083, | |
| "eval_runtime": 19.145, | |
| "eval_samples_per_second": 79.29, | |
| "eval_steps_per_second": 2.507, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 38.71971130371094, | |
| "learning_rate": 3.859649122807018e-05, | |
| "loss": 0.582, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 21.194923400878906, | |
| "learning_rate": 3.815789473684211e-05, | |
| "loss": 0.5039, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 47.179935455322266, | |
| "learning_rate": 3.771929824561404e-05, | |
| "loss": 0.5564, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 34.193275451660156, | |
| "learning_rate": 3.728070175438597e-05, | |
| "loss": 0.452, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 20.546531677246094, | |
| "learning_rate": 3.6842105263157895e-05, | |
| "loss": 0.4047, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "eval_accuracy": 0.8649538866930171, | |
| "eval_f1_macro": 0.8467453346863307, | |
| "eval_f1_micro": 0.8649538866930171, | |
| "eval_loss": 0.4545969069004059, | |
| "eval_runtime": 19.151, | |
| "eval_samples_per_second": 79.265, | |
| "eval_steps_per_second": 2.506, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 46.09516525268555, | |
| "learning_rate": 3.640350877192983e-05, | |
| "loss": 0.4348, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 35.31401443481445, | |
| "learning_rate": 3.5964912280701756e-05, | |
| "loss": 0.5397, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 22.361942291259766, | |
| "learning_rate": 3.5526315789473684e-05, | |
| "loss": 0.5323, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 42.461910247802734, | |
| "learning_rate": 3.508771929824561e-05, | |
| "loss": 0.5028, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 35.93864059448242, | |
| "learning_rate": 3.4649122807017546e-05, | |
| "loss": 0.568, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_accuracy": 0.8708827404479579, | |
| "eval_f1_macro": 0.8540867659285116, | |
| "eval_f1_micro": 0.8708827404479579, | |
| "eval_loss": 0.42495009303092957, | |
| "eval_runtime": 19.1783, | |
| "eval_samples_per_second": 79.152, | |
| "eval_steps_per_second": 2.503, | |
| "step": 350 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "total_flos": 5.846222510227456e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |