- Minor: add a default value for the "index" and "text" fields of the completion stream's chunk.
1.3.6
- Support Phi3 FP8 conversion. - Hotfix for safetensor checkpoint saver.
1.3.5
- Optimize CPU RAM usage during quantization with offloading - Support FP8 conversion for DBRX, Mixtral, and Command R+
1.3.4
- Hotfix for LoRA checkpoint saving error.
1.3.3
New Features
- **FP8 Checkpoint Conversion**: We've introduced a new feature for FP8 checkpoint conversion. - **Sharded Safetensors Checkpoint Saving**: Added the ability to save sharded safetensors checkpoints. - **LoRA Support on Mistral Model**: We have added support for LoRA (Low-Rank Adaptation) on the Mistral model.
Bug Fixes
- **BF16 Hotfix**: Addressed an urgent issue with bf16 processing. - **BFloat Safetensors Conversion**: Fixed an issue related to bfloat conversion for safetensors. - **Automatic Token Refresh**: Resolved a bug affecting automatic token refresh.