SLM_vs_LLM_experiments
/
max_seq_length_128_experiments
/LoRA
/Qwen
/Qwen1.5_7B_LoRA_MAdAiLab
/twitter_disaster
/checkpoint-250
/trainer_state.json
| { | |
| "best_metric": 0.4901912808418274, | |
| "best_model_checkpoint": "../experiments_checkpoints/LoRA/Qwen/Qwen1.5_7B_LoRA_MAdAiLab/twitter_disaster/checkpoint-250", | |
| "epoch": 0.9191176470588235, | |
| "eval_steps": 50, | |
| "global_step": 250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 24.977563858032227, | |
| "learning_rate": 4.9387254901960786e-05, | |
| "loss": 1.7697, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 142.99278259277344, | |
| "learning_rate": 4.877450980392157e-05, | |
| "loss": 1.666, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 87.81151580810547, | |
| "learning_rate": 4.816176470588236e-05, | |
| "loss": 1.0546, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 50.04526901245117, | |
| "learning_rate": 4.7549019607843135e-05, | |
| "loss": 0.813, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 22.704946517944336, | |
| "learning_rate": 4.6936274509803925e-05, | |
| "loss": 0.8422, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "eval_accuracy": 0.7178308823529411, | |
| "eval_f1_macro": 0.6371659017461913, | |
| "eval_f1_micro": 0.7178308823529411, | |
| "eval_loss": 0.6453067660331726, | |
| "eval_runtime": 13.5072, | |
| "eval_samples_per_second": 80.55, | |
| "eval_steps_per_second": 2.517, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 20.923370361328125, | |
| "learning_rate": 4.632352941176471e-05, | |
| "loss": 0.5966, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 66.72351837158203, | |
| "learning_rate": 4.571078431372549e-05, | |
| "loss": 0.7947, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 157.05340576171875, | |
| "learning_rate": 4.5098039215686275e-05, | |
| "loss": 0.8614, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 68.06228637695312, | |
| "learning_rate": 4.448529411764706e-05, | |
| "loss": 0.6034, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 15.797639846801758, | |
| "learning_rate": 4.387254901960784e-05, | |
| "loss": 0.6082, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.7472426470588235, | |
| "eval_f1_macro": 0.7123361820896584, | |
| "eval_f1_micro": 0.7472426470588235, | |
| "eval_loss": 0.5489143133163452, | |
| "eval_runtime": 13.5318, | |
| "eval_samples_per_second": 80.403, | |
| "eval_steps_per_second": 2.513, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 45.799476623535156, | |
| "learning_rate": 4.325980392156863e-05, | |
| "loss": 0.5511, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 49.33269500732422, | |
| "learning_rate": 4.2647058823529415e-05, | |
| "loss": 0.4801, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 36.33636474609375, | |
| "learning_rate": 4.20343137254902e-05, | |
| "loss": 0.4603, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 82.08908081054688, | |
| "learning_rate": 4.142156862745099e-05, | |
| "loss": 0.5596, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 13.81618595123291, | |
| "learning_rate": 4.0808823529411765e-05, | |
| "loss": 0.4305, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_accuracy": 0.7251838235294118, | |
| "eval_f1_macro": 0.5776786815440837, | |
| "eval_f1_micro": 0.7251838235294118, | |
| "eval_loss": 0.5571720004081726, | |
| "eval_runtime": 13.5624, | |
| "eval_samples_per_second": 80.222, | |
| "eval_steps_per_second": 2.507, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 57.01852035522461, | |
| "learning_rate": 4.0196078431372555e-05, | |
| "loss": 0.5056, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 53.905147552490234, | |
| "learning_rate": 3.958333333333333e-05, | |
| "loss": 0.567, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 47.106292724609375, | |
| "learning_rate": 3.897058823529412e-05, | |
| "loss": 0.5069, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 8.995195388793945, | |
| "learning_rate": 3.8357843137254904e-05, | |
| "loss": 0.5449, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 79.07156372070312, | |
| "learning_rate": 3.774509803921569e-05, | |
| "loss": 0.5021, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.7720588235294118, | |
| "eval_f1_macro": 0.7436838605490643, | |
| "eval_f1_micro": 0.7720588235294118, | |
| "eval_loss": 0.49997127056121826, | |
| "eval_runtime": 13.6431, | |
| "eval_samples_per_second": 79.747, | |
| "eval_steps_per_second": 2.492, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 15.758883476257324, | |
| "learning_rate": 3.713235294117647e-05, | |
| "loss": 0.5018, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 183.47061157226562, | |
| "learning_rate": 3.6519607843137254e-05, | |
| "loss": 0.616, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 48.367374420166016, | |
| "learning_rate": 3.5906862745098044e-05, | |
| "loss": 0.4927, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 5.8350114822387695, | |
| "learning_rate": 3.529411764705883e-05, | |
| "loss": 0.4508, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 15.554094314575195, | |
| "learning_rate": 3.468137254901961e-05, | |
| "loss": 0.4715, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_accuracy": 0.7766544117647058, | |
| "eval_f1_macro": 0.7450627015924902, | |
| "eval_f1_micro": 0.7766544117647058, | |
| "eval_loss": 0.4901912808418274, | |
| "eval_runtime": 13.5595, | |
| "eval_samples_per_second": 80.239, | |
| "eval_steps_per_second": 2.507, | |
| "step": 250 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 816, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 50, | |
| "total_flos": 4.17576786067456e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |