google

/

gemma-3-1b-it

like 680

Follow

Google 34.3k

When using vLLM, a missing chat_template.json results in repeated fetches of the chat template (presumably from tokenizer_config.json), for every single row that the chat template needs to be applied to.