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 15, 12:15 UTC
[TPU][V1][CI] Replace `python3 setup.py develop` with standard `pip install --e` on TPU (#17374)
Signed-off-by: NickLucche <nlucches@redhat.com>
Nicolò Lucchesi committed
Apr 29, 2025 at 19:36 UTC
792595b59d01eb7060b6cbc6dec20c129ea917b2
2 files changed
+2
-2
docker/Dockerfile.tpu
+1
-1
@@ -23,7 +23,7 @@ RUN --mount=type=cache,target=/root/.cache/pip \
23
--mount=type=bind,source=.git,target=.git \
24
python3 -m pip install \
25
-r requirements/tpu.txt
26
-RUN python3 setup.py develop
26
+RUN python3 -m pip install -e .
27
28
# install development dependencies (for testing)
29
RUN python3 -m pip install -e tests/vllm_test_utils
docs/source/getting_started/installation/ai_accelerator/tpu.inc.md
+1
-1
@@ -158,7 +158,7 @@ sudo apt-get install libopenblas-base libopenmpi-dev libomp-dev
158
Run the setup script:
159
160
```bash
161
-VLLM_TARGET_DEVICE="tpu" python setup.py develop
161
+VLLM_TARGET_DEVICE="tpu" python -m pip install -e .
162
```
163
164
## Set up using Docker