MiniMind-V User 823ce1e4fc [feat] Implement DPO training for driving safety alignment ... - dataset/driving_preferences.jsonl: 56 (prompt, chosen, rejected) pairs across 8 safety categories: pedestrian, speed, lane change, weather, following distance, traffic signals, emergency vehicles, night driving - scripts/generate_driving_preferences.py: generator script for the dataset - trainer/train_dpo_driving.py: custom DPO trainer * implements DPO loss from scratch (no trl dependency) * frozen reference model + LoRA policy (1.07% trainable params) * logs preference accuracy per step * saves LoRA adapters or full weights * 1 epoch on 56 pairs: loss 0.11, pref_acc 96.4% Usage: uv run python trainer/train_dpo_driving.py --data_path dataset/driving_preferences.jsonl --use_lora --epochs 3		2026-03-06 05:00:42 +00:00
..
eval_images	[feat] update images	2025-10-17 00:03:58 +08:00
__init__.py	250427	2025-04-27 18:13:22 +08:00
driving_preferences.jsonl	[feat] Implement DPO training for driving safety alignment	2026-03-06 05:00:42 +00:00
lm_dataset.py	[update] dataset name	2026-02-05 00:22:58 +08:00