Update flashinfer to 0.6.8 (#39959)

Signed-off-by: bai <v@gor.io>
This commit is contained in:
bai
2026-04-20 10:37:23 -07:00
committed by GitHub
parent b9cf629bd0
commit 191e3fdaa1
9 changed files with 31 additions and 61 deletions
+3 -24
View File
@@ -580,33 +580,12 @@ RUN --mount=type=cache,target=/root/.cache/uv \
# Install FlashInfer JIT cache (requires CUDA-version-specific index URL)
# https://docs.flashinfer.ai/installation.html
# From versions.json: .flashinfer.version
# 0.6.7: CUTLASS 4.4.2 bump, fixes TMA grouped GEMM on SM12x (flashinfer#2798)
# TODO: bump to 0.6.8 when released for NVFP4/MXFP4 group GEMMs on
# SM120/SM121 (RTX 50 / DGX Spark) via flashinfer#2738
ARG FLASHINFER_VERSION=0.6.7
ARG FLASHINFER_VERSION=0.6.8.post1
RUN --mount=type=cache,target=/root/.cache/uv \
uv pip install --system flashinfer-jit-cache==${FLASHINFER_VERSION} \
--extra-index-url https://flashinfer.ai/whl/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.') \
&& flashinfer show-config
# Pre-download FlashInfer TRTLLM BMM headers for air-gapped environments.
# At runtime, MoE JIT compilation downloads these from edge.urm.nvidia.com
# which fails without internet. This step caches them at build time.
RUN python3 <<'PYEOF'
from flashinfer.jit import env as jit_env
from flashinfer.jit.cubin_loader import download_trtllm_headers, get_cubin
from flashinfer.artifacts import ArtifactPath, CheckSumHash
download_trtllm_headers(
'bmm',
jit_env.FLASHINFER_CUBIN_DIR / 'flashinfer' / 'trtllm' / 'batched_gemm' / 'trtllmGen_bmm_export',
f'{ArtifactPath.TRTLLM_GEN_BMM}/include/trtllmGen_bmm_export',
ArtifactPath.TRTLLM_GEN_BMM,
get_cubin(f'{ArtifactPath.TRTLLM_GEN_BMM}/checksums.txt', CheckSumHash.TRTLLM_GEN_BMM),
)
print('FlashInfer TRTLLM BMM headers downloaded successfully')
PYEOF
&& flashinfer show-config \
&& flashinfer download-cubin
# ============================================================
# OPENAI API SERVER DEPENDENCIES
+2 -5
View File
@@ -217,16 +217,13 @@ RUN pip install setuptools==75.6.0 packaging==23.2 ninja==1.11.1.3 build==1.2.2.
# build flashinfer for torch nightly from source around 10 mins
# release version: v0.6.7
# 0.6.7: CUTLASS 4.4.2 bump, fixes TMA grouped GEMM on SM12x (flashinfer#2798)
# TODO: bump to 0.6.8 when released for NVFP4/MXFP4 group GEMMs on
# SM120/SM121 (RTX 50 / DGX Spark) via flashinfer#2738
# release version: v0.6.8.post1
# todo(elainewy): cache flashinfer build result for faster build
ENV CCACHE_DIR=/root/.cache/ccache
RUN --mount=type=cache,target=/root/.cache/ccache \
--mount=type=cache,target=/root/.cache/uv \
echo "git clone flashinfer..." \
&& git clone --depth 1 --branch v0.6.7 --recursive https://github.com/flashinfer-ai/flashinfer.git \
&& git clone --depth 1 --branch v0.6.8.post1 --recursive https://github.com/flashinfer-ai/flashinfer.git \
&& cd flashinfer \
&& git submodule update --init --recursive \
&& echo "finish git clone flashinfer..." \
+1 -1
View File
@@ -65,7 +65,7 @@
"default": "true"
},
"FLASHINFER_VERSION": {
"default": "0.6.7"
"default": "0.6.8.post1"
},
"GDRCOPY_CUDA_VERSION": {
"default": "12.8"