I just changed the /Long-RL-main/examples/new_supports/qwen2_5_vl_3b_video_grpo.sh file for Qwen2.5-VL-7B-Instruct, however the model can not be trained. Qwen2.5-VL-3B-Instruct can be trained. It's strange and hard to debug.
the code just stop here and can not forward.
Can you provide a sh file for Qwen2.5-VL-7B training
I just changed the /Long-RL-main/examples/new_supports/qwen2_5_vl_3b_video_grpo.sh file for Qwen2.5-VL-7B-Instruct, however the model can not be trained. Qwen2.5-VL-3B-Instruct can be trained. It's strange and hard to debug.
Can you provide a sh file for Qwen2.5-VL-7B training