| { | |
| "model_id": "deepseek-ai/DeepSeek-V4-Flash-0731", | |
| "max_seq_len": 65536, | |
| "data_dir": "/workspace/data_merged", | |
| "output_dir": "/checkpoints/dsv4-flash-grug-fft-partial", | |
| "epochs": 1, | |
| "micro_batch": 1, | |
| "grad_accum": 8, | |
| "lr": 1e-5, | |
| "warmup_steps": 50, | |
| "weight_decay": 0.01, | |
| "max_grad_norm": 1.0, | |
| "packing": true, | |
| "save_steps": 500, | |
| "save_total_limit": 4 | |
| } |