A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 08:15 UTC
Add hf_transfer to testing image (#10096)
Michael Goin committed
Nov 8, 2024 at 03:35 UTC
aea6ad629ff92f072a11b21dcdb1105677744007
1 file changed
+5
Dockerfile
+5
@@ -191,6 +191,11 @@ ADD . /vllm-workspace/
191
RUN --mount=type=cache,target=/root/.cache/pip \
192
python3 -m pip install -r requirements-dev.txt
193
194
+# enable fast downloads from hf (for testing)
195
+RUN --mount=type=cache,target=/root/.cache/pip \
196
+ python3 -m pip install hf_transfer
197
+ENV HF_HUB_ENABLE_HF_TRANSFER 1
198
+
199
# Copy in the v1 package for testing (it isn't distributed yet)
200
COPY vllm/v1 /usr/local/lib/python3.12/dist-packages/vllm/v1
201