docs: Updated the `Transformers` example to use intended temp=0.15
#23
by
casinca
- opened
This is a mirror PR for Devstrall-Small, based on: https://huggingface.co/mistralai/Devstral-2-123B-Instruct-2512/discussions/17
This PR adds the required hparam arguments to enable stochastic sampling (temp 0.15) for the transformers snippet rather than greedy decoding.
This reflects the recommended usage pattern intended by Mistral.
Thanks !
juliendenize
changed pull request status to
merged