New
marco-lu2508/fllm — Prefetch added to the Q2_K·Q3_K·IQ4_XS kernels
fLLM, a local LLM inference engine written purely in Rust, has added memory prefetch to the shared kernels for the Q2_K·Q3_K·IQ4_XS quantization formats.
marco-lu2508/fllm
fLLM, a local LLM inference engine written purely in Rust, has added memory prefetch to the shared kernels for the Q2_K·Q3_K·IQ4_XS quantization formats.
View on GitHub ↗- stars
- ★ 0
- Forks
- 0
- Language
- Rust
- License
- Apache-2.0
Posts
1No matching stories.