liuyongkang1-tal

the demo shows hot-keys includes "alt+x", but the release v0.2.0 disappear

微调qwen2.5VL 'weight' must be 2-D

2

rank0]: Traceback (most recent call last): [rank0]: File "/mnt/pfs/jh/Users/lyk/Project/open_item_judge/code/Qwen2.5-VL/qwen-vl-finetune/qwenvl/train/train_qwen.py", line 178, in [rank0]: train(attn_implementation="flash_attention_2") [rank0]: File "/mnt/pfs/jh/Users/lyk/Project/open_item_judge/code/Qwen2.5-VL/qwen-vl-finetune/qwenvl/train/train_qwen.py", line 168, in train [rank0]: trainer.train() [rank0]: File "/mnt/pfs/jh/Users/lyk/env/anaconda3/envs/qwen2.5_VL_torch2.6_py3.12/lib/python3.12/site-packages/transformers/trainer.py", line 2245, in train...