Repository navigation
v2.0.8
Release Notes: Hotfixes for Vision Models & Advanced Search## Vision Model Support
- Fixed projector loading bugs. Prevented VLM models (SmolVLM, LFM2-VL) from failing during dynamic multimodal projector fetching. [1, 2]
Advanced Store Integration
- Resolved model visibility issue. Fixed a bug where newly added VLM models from HuggingFace Explorer would not appear in the model list.
File Sizes in Explorer
- Corrected file size displays. Re-routed Explorer search to use non-recursive tree lookups so file sizes no longer incorrectly display as 0MB.
UI Fixes
- Resolved compiler errors. Fixed exhaustive when compiler errors triggered when rendering tags and icons for Vision models.
Full Changelog: v2.0.7...v2.0.8