[DSv4] Adding TRTLLM gen attention kernel (#43827)

This commit is contained in:
Yongye Zhu
2026-06-04 07:35:09 -07:00
committed by GitHub
parent 3e77036768
commit b5235fca2e
20 changed files with 2659 additions and 86 deletions
+2 -1
View File
@@ -55,7 +55,8 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, ops) {
// Horizontally-fused DeepseekV4-MLA: per-head RMSNorm + GPT-J RoPE for Q, and
// GPT-J RoPE + UE8M0 FP8 quant + paged cache insert for KV, all in one
// kernel launch. Registered in _C_stable_libtorch.
// kernel launch. Registered in _C_stable_libtorch (incl. the FlashInfer V4
// full-cache bf16/fp8 variants).
// Quantization ops
#ifndef USE_ROCM