Nie-Yingying

[QUESTION] OOM when load XCOMET-XXL in A100 with 40G memory for prediction

4

## ❓ Questions and Help ### Before asking: 1. Search for similar [issues](https://github.com/Unbabel/COMET/issues). 3. Search the [docs](https://unbabel.github.io/COMET/html/index.html). #### What is your question? I can predict scores with only cpu successfully....

chat template test failed when I chose DeepSeek-R1-Distill-Qwen in model family

3

 xinference 1.2.1 vllm 0.6.1.post2 vllm-flash-attn 2.6.1 transformers 4.48.3 torch 2.4.0