minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 13 Packages Projects Releases Wiki Activity

fix: llama 4 + trtllm gen + fp8 kv cache incompatibility (#12347)

Browse Source
This commit is contained in:
b8zhong
2025-10-29 11:31:02 -07:00
committed by GitHub
parent b53d9e11c6
commit bacb3825fe
1 changed files with 7 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/sglang/srt/server_args.py
+7
View File
@@ -971,6 +971,13 @@ class ServerArgs:
logger.warning( logger.warning(
"Use trtllm_mha as attention backend on sm100 for Llama4 model" "Use trtllm_mha as attention backend on sm100 for Llama4 model"
) )
if is_sm100_supported() and self.attention_backend == "trtllm_mha":
# TODO(brayden): remove this once TRTLLM MHA kernel for FP8 w/ tileSizeKv=128 is available.
# This is a Llama 4 specific issue only.
self.kv_cache_dtype = "bfloat16"
logger.warning(
"Setting kv_cache_dtype to bfloat16 for Llama4 with trtllm_mha backend, due to a missing FlashInfer TRTLLM MHA kernel for FP8 KV Cache"
)
if is_sm100_supported() and self.moe_runner_backend == "auto": if is_sm100_supported() and self.moe_runner_backend == "auto":
if self.quantization in {"fp8", "modelopt_fp8"}: if self.quantization in {"fp8", "modelopt_fp8"}:
self.moe_runner_backend = "flashinfer_trtllm" self.moe_runner_backend = "flashinfer_trtllm"