{ "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": 923, "eval_samples": 102, "params": { "rank": 16, "alpha": 32, "dropout": 0.1, "lr": 1e-05, "scheduler": "cosine", "epochs": 3, "effective_batch": 4, "max_seq_length": 4096, "neftune_alpha": 5.0, "target_modules": [ "q_proj", "k_proj", "v_proj", "o_proj", "gate_proj", "up_proj", "down_proj" ] }, "results": { "total_steps": 693, "final_train_loss": 0.8432, "best_eval_loss": 0.8133148550987244, "best_eval_step": 640, "elapsed_minutes": 40.0 }, "loss_history": { "train": [ [ 5, 1.7606 ], [ 10, 1.8257 ], [ 15, 2.0366 ], [ 20, 1.8063 ], [ 25, 1.9763 ], [ 30, 1.9014 ], [ 35, 1.8005 ], [ 40, 1.6563 ], [ 45, 1.6997 ], [ 50, 1.5501 ], [ 55, 1.273 ], [ 60, 1.5863 ], [ 65, 1.3741 ], [ 70, 1.5504 ], [ 75, 1.5004 ], [ 80, 1.3517 ], [ 85, 1.3193 ], [ 90, 1.2519 ], [ 95, 1.4066 ], [ 100, 1.3787 ], [ 105, 1.317 ], [ 110, 1.273 ], [ 115, 1.366 ], [ 120, 1.1712 ], [ 125, 1.1309 ], [ 130, 1.5067 ], [ 135, 1.1212 ], [ 140, 1.4059 ], [ 145, 1.2398 ], [ 150, 1.2904 ], [ 155, 1.3055 ], [ 160, 1.1672 ], [ 165, 1.2585 ], [ 170, 1.252 ], [ 175, 0.9654 ], [ 180, 1.1089 ], [ 185, 1.1518 ], [ 190, 1.1216 ], [ 195, 1.1329 ], [ 200, 0.9045 ], [ 205, 1.0882 ], [ 210, 1.0312 ], [ 215, 0.9879 ], [ 220, 0.9716 ], [ 225, 1.0379 ], [ 230, 1.0032 ], [ 235, 1.1349 ], [ 240, 0.9745 ], [ 245, 0.9717 ], [ 250, 0.8808 ], [ 255, 0.9055 ], [ 260, 1.0285 ], [ 265, 1.0887 ], [ 270, 0.8846 ], [ 275, 0.9121 ], [ 280, 1.032 ], [ 285, 1.043 ], [ 290, 0.9623 ], [ 295, 0.8949 ], [ 300, 0.8456 ], [ 305, 0.8745 ], [ 310, 0.94 ], [ 315, 0.8265 ], [ 320, 0.9407 ], [ 325, 0.9309 ], [ 330, 0.8971 ], [ 335, 0.8191 ], [ 340, 0.7918 ], [ 345, 0.9414 ], [ 350, 0.783 ], [ 355, 0.8103 ], [ 360, 0.8443 ], [ 365, 0.9374 ], [ 370, 0.9564 ], [ 375, 0.8063 ], [ 380, 0.7323 ], [ 385, 0.826 ], [ 390, 0.8682 ], [ 395, 0.8559 ], [ 400, 0.9216 ], [ 405, 0.8368 ], [ 410, 0.9346 ], [ 415, 0.8632 ], [ 420, 0.8698 ], [ 425, 0.7973 ], [ 430, 0.8807 ], [ 435, 0.7736 ], [ 440, 0.7795 ], [ 445, 0.768 ], [ 450, 0.8426 ], [ 455, 0.8065 ], [ 460, 0.7986 ], [ 465, 1.0072 ], [ 470, 0.7926 ], [ 475, 0.7541 ], [ 480, 0.7158 ], [ 485, 0.7268 ], [ 490, 0.793 ], [ 495, 0.9075 ], [ 500, 0.8299 ], [ 505, 0.8684 ], [ 510, 0.8688 ], [ 515, 0.6847 ], [ 520, 0.7814 ], [ 525, 0.7811 ], [ 530, 0.7919 ], [ 535, 0.8829 ], [ 540, 0.7775 ], [ 545, 0.7112 ], [ 550, 0.7917 ], [ 555, 0.769 ], [ 560, 0.7328 ], [ 565, 0.7695 ], [ 570, 0.5831 ], [ 575, 0.7399 ], [ 580, 0.8321 ], [ 585, 0.8167 ], [ 590, 0.832 ], [ 595, 0.7446 ], [ 600, 0.698 ], [ 605, 0.5891 ], [ 610, 0.8106 ], [ 615, 0.7455 ], [ 620, 0.7738 ], [ 625, 0.7939 ], [ 630, 0.8003 ], [ 635, 0.7349 ], [ 640, 0.7863 ], [ 645, 0.7968 ], [ 650, 0.8229 ], [ 655, 0.7529 ], [ 660, 0.6354 ], [ 665, 0.8064 ], [ 670, 0.9099 ], [ 675, 0.8156 ], [ 680, 0.7462 ], [ 685, 0.907 ], [ 690, 0.8432 ] ], "eval": [ [ 40, 2.072058916091919 ], [ 80, 1.625858187675476 ], [ 120, 1.4893486499786377 ], [ 160, 1.359459400177002 ], [ 200, 1.2217525243759155 ], [ 240, 1.113280177116394 ], [ 280, 1.0186439752578735 ], [ 320, 0.937186598777771 ], [ 360, 0.8894771337509155 ], [ 400, 0.862065315246582 ], [ 440, 0.8410201072692871 ], [ 480, 0.8306860327720642 ], [ 520, 0.8223150372505188 ], [ 560, 0.8172081112861633 ], [ 600, 0.8150315880775452 ], [ 640, 0.8133148550987244 ], [ 680, 0.8135352730751038 ] ] } }