1.6.1

• Reduced (flat) VRAM overhead for GDN/KDA/Mamba drafting
• Support recurrent draft models, switch DFlash to sliding-attention (reduced VRAM overhead)
• New hybrid drafting mode (draft model + long-ngram)
• Workaround for Triton issue breaking kernel autotune on mixed-architecture setups
• Lots of ROCm fixes and optimizations
• Turing (sm_75) optimizations
• AVX2 optimizations
• Many bugfixes

Full Changelog: v1.6.0...v1.6.1