dev

GGitHubperf: use ggml operators to optimize cpu ffn forwarding (#94)

bb431d95创建于 2025年8月19日历史提交

文件	最后提交记录	最后更新时间
mod.rs	perf: bench ffn w/ weight transfer (#91)	10 个月前
xpu_ffn.rs	perf: use ggml operators to optimize cpu ffn forwarding (#94) * perf: use ggml operators to optimize cpu ffn forwarding * perf: supports bf16 on ggml backend * chore: make clippy happy * chore: align the types * chore: tuning & fix serialization * fix: fix padding and context size * feat: allow dropping cache after loading expert backend * chore: statically link ggml * feat: allocating tensor data from rust side * feat: allow specifying computation backend * chore: clippy & format * chore: tuning * chore: delete unused feature flags * chore: remove ggml-cuda * fix: ggml-cpu.h includes ggml.h * perf: single thread for better throughput	10 个月前
xpu_ffn_activate.rs	perf: bench ffn using Criterion (#80) * perf: bench ffn using Criterion * chore: clippy & format * better ffn benchmarking * chore: explain why we use cuda to initialize weights	10 个月前
xpu_ffn_queue.rs	perf: remove useless locks & replace 3rd OnceCell with std OnceLock (#85) * perf: remove useless locks & replace 3rd OnceCell with std OnceLock * chore: clippy & format	10 个月前
xpu_ffn_with_weight_transfer.rs	perf: use ggml operators to optimize cpu ffn forwarding (#94) * perf: use ggml operators to optimize cpu ffn forwarding * perf: supports bf16 on ggml backend * chore: make clippy happy * chore: align the types * chore: tuning & fix serialization * fix: fix padding and context size * feat: allow dropping cache after loading expert backend * chore: statically link ggml * feat: allocating tensor data from rust side * feat: allow specifying computation backend * chore: clippy & format * chore: tuning * chore: delete unused feature flags * chore: remove ggml-cuda * fix: ggml-cpu.h includes ggml.h * perf: single thread for better throughput	10 个月前
xpu_transfer.rs	perf: bench ffn using Criterion (#80) * perf: bench ffn using Criterion * chore: clippy & format * better ffn benchmarking * chore: explain why we use cuda to initialize weights	10 个月前