01yingxuan

Results 1 issues of 01yingxuan

When performing SFT on the Qwen2.5 32B model, I encountered an OOM (Out of Memory) issue. I would like to ask if the Qwen series of models supports model parallelism...