Repository navigation

feat: support GLM-5.3 Flash dynamic FP8 KV cache with regression cove… - #1621

Merged

shihaobai merged 5 commits into

mainfrom

Oct 29, 2026

Merged

feat: support GLM-5.3 Flash dynamic FP8 KV cache with regression cove…#1621
shihaobai merged 5 commits into
mainfrom
glm5.3_fp8kv

Commits on Oct 29, 2026
- feat: support GLM-5.3 Flash dynamic FP8 KV cache with regression coverage
  niushengxiao
  committed
  b1bec38
- perf: optimize FP8 KV attention and cache writes
  niushengxiao
  committed
  09609a0
- feat: move kernel to triton_kernel folder
  niushengxiao
  committed
  beec5e3
- refactor: reuse backend FP8 KV workspaces and derive query padding dimensions
  niushengxiao
  committed
  bf8fcda
- refactor: simplify FP8 KV paths and runtime kernel parameters
  niushengxiao
  committed
  deb58cf