Extremely new about this. Don’t know where to post this minor issue, whether it’s worth a PR or not.
When running a curl request to the server, when requesting output tokens greater than max_model_len. There’s a space missing in the printed message between the “max_tokens={value}” and “cannot”.
On vllm/vllm/renders/params.py , line 147-148
f"{self.max_output_tokens_param}={max_output_tokens}"
f"cannot be greater than "
This is a valid minor issue: the error message in vllm/renderers/params.py is missing a space between "max_tokens={value}" and "cannot be greater than", due to adjacent f-strings being concatenated without a space. This can be fixed by adding a space at the end of the first f-string or the start of the second. Such formatting bugs are commonly addressed via small PRs in open source projects, and vLLM welcomes even minor contributions (contributing guide).
Would you like more detail on how to submit a PR for this fix?
Sources:
Hint: Mention @RunLLM in the post for followups.