| { | |
| "base_model": "Qwen/Qwen3-8B", | |
| "quantization": "4bit", | |
| "max_length": 4096, | |
| "softcap": null, | |
| "steps": 2000, | |
| "train_rows": 119858, | |
| "dev_zero_shot": { | |
| "accuracy": 0.7218543046357616, | |
| "nll": 3.117936041840196, | |
| "count": 151 | |
| }, | |
| "dev_final": { | |
| "accuracy": 0.8543046357615894, | |
| "nll": 0.41393711996655635, | |
| "count": 151 | |
| }, | |
| "recipe": { | |
| "base_model": "Qwen/Qwen3-8B", | |
| "include_teacher": true, | |
| "mix": "news:type=0.22,news:certainty=0.148,news:outlook=0.12,news:main=0.08,news:role=0.03,filing:risk=0.12,filing:outlook=0.09,filing:legal=0.08,filing:going_concern=0.012,filing:customer=0.02,filing:cyber=0.03,human=0.04", | |
| "mix_by": "question", | |
| "quantization": "4bit", | |
| "minimum_train": 100, | |
| "minimum_dev": 20, | |
| "lora_rank": 32, | |
| "lora_alpha": 64, | |
| "lora_lr": 2e-05, | |
| "head_lr": 0.0001, | |
| "micro_batch": 1, | |
| "grad_accum": 16, | |
| "epochs": 3.0, | |
| "steps": 2000, | |
| "max_length": 4096, | |
| "max_grad_norm": 1.0, | |
| "stats_rows": 512, | |
| "log_every": 5, | |
| "eval_every": 250, | |
| "seed": 1 | |
| }, | |
| "training_data": "100,000 news + 11,752 SEC 10-K + 8,000 news-outlook decisions labelled by a local Qwen3.8-27B teacher, plus 106 human-reviewed 10-K decisions", | |
| "layout": { | |
| "full_model": "root (LoRA merged into Qwen3-8B, bf16 safetensors)", | |
| "adapter": "adapter/ (LoRA for Qwen/Qwen3-8B, trained with 4-bit QLoRA)" | |
| } | |
| } |