Not able to serve openai/gpt-oss newer model

Used huggingface-cli to download openai/gpt-oss-20b model and tried max command as “max serve –model-path openai/gpt-oss-20b”, it shows error as follows:

ValueError: The checkpoint you are trying to load has model type gpt_oss but Transformers does not recognize this architecture. This could be because of an issue with the checkpoint, or because your version of Transformers is out of date.

Tried upgrade of transformers, but still the same issue. Do we need to wait for newer version of modular max?

Currently we offer GPT-OSS at the Mammoth level through Large Scale Inference | SF Compute for batched inference workload targeting Enterprises now. Individual local model support is coming soon so stay tuned!

1 Like

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.