Spaces:

---

openGPT-X

/

european-llm-leaderboard

like 92

Running on CPU Upgrade

Hi,

We have noticed and corrected an error in the "chat_template" in "tokenizer_config.json" of EuroLLM-1.7B-instruct.
I don't know if you've used this chat template directly or if you implemented it yourselves. If you've used it directly, could you repeat the evals?

Thanks