{ "model_base": "Qwen/Qwen2.5-32B-Instruct", "model_name": "RegTech-32B-Instruct", "dataset": "./train.jsonl", "env_file": "/home/ubuntu/sophia-core-server/.tuning/.env.32B", "train_samples": 2330, "eval_samples": 258, "params": { "rank": 16, "alpha": 16, "dropout": 0.05, "lr": 2e-05, "scheduler": "cosine", "epochs": 1, "effective_batch": 4, "max_seq_length": 4096, "neftune_alpha": 0.0, "target_modules": [ "q_proj", "k_proj", "v_proj", "o_proj", "gate_proj", "up_proj", "down_proj" ] }, "results": { "total_steps": 583, "final_train_loss": 0.931, "best_eval_loss": 0.9896443486213684, "best_eval_step": 560, "best_token_accuracy": 0.7659, "elapsed_minutes": 37.7 }, "loss_history": { "train": [ [ 5, 1.7845 ], [ 10, 1.8161 ], [ 15, 1.8696 ], [ 20, 2.0645 ], [ 25, 1.6354 ], [ 30, 1.5951 ], [ 35, 1.7529 ], [ 40, 1.9265 ], [ 45, 1.7091 ], [ 50, 1.4318 ], [ 55, 1.4539 ], [ 60, 1.4346 ], [ 65, 1.4506 ], [ 70, 1.254 ], [ 75, 1.3812 ], [ 80, 1.2598 ], [ 85, 1.4447 ], [ 90, 1.1975 ], [ 95, 1.1675 ], [ 100, 1.394 ], [ 105, 1.347 ], [ 110, 1.2377 ], [ 115, 1.1502 ], [ 120, 1.2652 ], [ 125, 1.2239 ], [ 130, 1.1843 ], [ 135, 1.2547 ], [ 140, 1.2079 ], [ 145, 1.1555 ], [ 150, 1.1216 ], [ 155, 1.101 ], [ 160, 1.1316 ], [ 165, 1.1601 ], [ 170, 1.2139 ], [ 175, 1.0971 ], [ 180, 0.9276 ], [ 185, 1.2373 ], [ 190, 1.1523 ], [ 195, 1.1123 ], [ 200, 1.1195 ], [ 205, 1.1249 ], [ 210, 1.106 ], [ 215, 1.1916 ], [ 220, 1.0094 ], [ 225, 0.9779 ], [ 230, 1.1532 ], [ 235, 1.1116 ], [ 240, 1.0537 ], [ 245, 1.12 ], [ 250, 1.065 ], [ 255, 0.9508 ], [ 260, 0.9716 ], [ 265, 0.9991 ], [ 270, 1.0961 ], [ 275, 1.04 ], [ 280, 1.0153 ], [ 285, 0.9728 ], [ 290, 1.029 ], [ 295, 0.9904 ], [ 300, 0.9582 ], [ 305, 1.0926 ], [ 310, 1.0106 ], [ 315, 0.98 ], [ 320, 0.8666 ], [ 325, 1.0373 ], [ 330, 0.9106 ], [ 335, 1.0044 ], [ 340, 0.9259 ], [ 345, 1.017 ], [ 350, 1.0305 ], [ 355, 0.9852 ], [ 360, 1.0174 ], [ 365, 0.8547 ], [ 370, 0.9254 ], [ 375, 0.8651 ], [ 380, 0.9753 ], [ 385, 0.9514 ], [ 390, 1.0401 ], [ 395, 0.9505 ], [ 400, 1.0557 ], [ 405, 0.9294 ], [ 410, 0.9874 ], [ 415, 0.998 ], [ 420, 1.0099 ], [ 425, 0.9127 ], [ 430, 0.9525 ], [ 435, 0.9343 ], [ 440, 0.9187 ], [ 445, 0.9685 ], [ 450, 0.8536 ], [ 455, 0.9055 ], [ 460, 0.9866 ], [ 465, 0.9282 ], [ 470, 0.8396 ], [ 475, 0.9417 ], [ 480, 1.011 ], [ 485, 0.9357 ], [ 490, 0.8992 ], [ 495, 1.0237 ], [ 500, 0.8904 ], [ 505, 0.9034 ], [ 510, 0.9129 ], [ 515, 0.9439 ], [ 520, 0.8588 ], [ 525, 1.0242 ], [ 530, 0.8042 ], [ 535, 0.8632 ], [ 540, 0.8258 ], [ 545, 0.9191 ], [ 550, 0.965 ], [ 555, 0.911 ], [ 560, 1.0103 ], [ 565, 0.9916 ], [ 570, 0.9791 ], [ 575, 0.9193 ], [ 580, 0.931 ] ], "eval": [ [ 40, 1.946990728378296 ], [ 80, 1.5292835235595703 ], [ 120, 1.3830845355987549 ], [ 160, 1.2611178159713745 ], [ 200, 1.186440110206604 ], [ 240, 1.1333062648773193 ], [ 280, 1.089249610900879 ], [ 320, 1.0572314262390137 ], [ 360, 1.0289398431777954 ], [ 400, 1.0110489130020142 ], [ 440, 1.000611662864685 ], [ 480, 0.993800699710846 ], [ 520, 0.9907287955284119 ], [ 560, 0.9896443486213684 ] ], "token_accuracy": [ [ 40, 0.6795 ], [ 80, 0.7006 ], [ 120, 0.7146 ], [ 160, 0.7261 ], [ 200, 0.734 ], [ 240, 0.7425 ], [ 280, 0.7489 ], [ 320, 0.7535 ], [ 360, 0.7587 ], [ 400, 0.7615 ], [ 440, 0.7639 ], [ 480, 0.7651 ], [ 520, 0.7657 ], [ 560, 0.7659 ] ] } }