You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
feat(models): Added Laguna architecture support (LLM_ARCH_LAGUNA) for Laguna XS.2 / M.1 GGUFs, including chat template and HF-to-GGUF conversion registration.
Confirmed Laguna-XS-2.1-MXFP4_MOE loads and the server listens successfully. The previous load failure was missing Laguna arch support; MXFP4 CUDA kernels were already present.
Note: Multi-platform CI package builds may be unavailable while GitHub Actions billing is locked; this release ships notes (same pattern as v0.3.5) until packaging can run again.