product · inferred from evidence
mistralai/Mixtral-8x7B-Instruct-v0.1
A specific model identifier used in the context of serving with LoRAX.
- Not to able to serve mistralai/Mixtral-8x7B-Instruct-v0.1 on loraX with multiple GPU's
Pull the latest LoRAX image and retry: docker pull ghcr.io/predibase/lorax:latest The maintainer suspected a stale image, and the reporter confirmed multi-GPU serving worked after updating. For NVLink setups, also set `NCCL_P2P_LEVEL=NVL`
- Problem when using pipeline serialization
Haystack: Maintainer anakin87 confirmed this is a real bug: incorrect serialization in telemetry breaks pipeline deserialization with HuggingFaceLocalGenerator kwargs (e.g. bnb_4bit_compute_dtype: torch.bfloat16 needs quotes).
- LoRAX: TensorParallelHead object has no attribute base_layer when serving Mixtral with an adapter
This crash is a LoRAX bug in adapter loading for Mixtral's tensor-parallel head, fixed in PR #131. Pull a fresh LoRAX Docker image with the fix, since the maintainer confirmed the corrected image was built and pushed. If you pinned an image