Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix: Fix version for setuptools and grpcio-tools. Remove cudnn 8 installation #7331

Merged
merged 6 commits into from
Jun 18, 2024
Merged

fix: Fix version for setuptools and grpcio-tools. Remove cudnn 8 installation #7331

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
439f866
Fix version for setuptools and grpcio-tools
krishung5 Jun 6, 2024
6f85431
Remove grpcio-tools fixed version
krishung5 Jun 11, 2024
b9be792
Merge remote-tracking branch 'origin/main' into krish-fix-trtllm-test
krishung5 Jun 12, 2024
8cb38a8
Remove cudnn 8 installation
krishung5 Jun 17, 2024
0cd0bde
Specify grpcio-tools version
krishung5 Jun 18, 2024
81cda6b
Add comments
krishung5 Jun 18, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 7 additions & 3 deletions build.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1082,9 +1082,9 @@ def create_dockerfile_linux(
"""
if "tensorrtllm" in backends:
df += """
# Remove TRT contents that are not needed in runtime
RUN apt-get update && apt-get install -y libcudnn8-dev && ldconfig

RUN ldconfig
# Remove contents that are not needed in runtime
RUN ARCH="$(uname -i)" \\
&& rm -fr ${TRT_ROOT}/bin ${TRT_ROOT}/targets/${ARCH}-linux-gnu/bin ${TRT_ROOT}/data \\
&& rm -fr ${TRT_ROOT}/doc ${TRT_ROOT}/onnx_graphsurgeon ${TRT_ROOT}/python \\
Expand All @@ -1094,10 +1094,14 @@ def create_dockerfile_linux(
RUN python3 -m pip install --upgrade pip \\
&& pip3 install transformers

# Install TensorRT-LLM
# ldconfig for TRT-LLM
RUN find /usr -name libtensorrt_llm.so -exec dirname {} \; > /etc/ld.so.conf.d/tensorrt-llm.conf
RUN find /opt/tritonserver -name libtritonserver.so -exec dirname {} \; > /etc/ld.so.conf.d/triton-tensorrtllm-worker.conf

# Setuptools has breaking changes in version 70.0.0, so fix it to 69.5.1
# The generated code in grpc_service_pb2_grpc.py depends on grpcio>=1.64.0, so fix it to 1.64.0
RUN pip3 install setuptools==69.5.1 grpcio-tools==1.64.0
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Any particular reason for pinning the versions here?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Setuptools has some breaking changes of pkg_resources in the latest 70.0.0 version. See here for more info.
The fixed version for grpcio-tools is to resolve the below error:

python3 /opt/tritonserver/tensorrtllm_backend/inflight_batcher_llm/client/end_to_end_grpc_client.py '--prompt=My name is'
/usr/local/lib/python3.10/dist-packages/tritonclient/grpc/service_pb2_grpc.py:21: RuntimeWarning: The grpc package installed is at version 1.62.1, but the generated code in grpc_service_pb2_grpc.py depends on grpcio>=1.64.0. Please upgrade your grpc module to grpcio>=1.64.0 or downgrade your generated code using grpcio-tools<=1.62.1. This warning will become an error in 1.65.0, scheduled for release on June 25, 2024.
  warnings.warn(
client creation failed: Channel.unary_unary() got an unexpected keyword argument '_registered_method'
Exception ignored in: <function InferenceServerClient.__del__ at 0xfffeef103c70>
Traceback (most recent call last):
  File "/usr/local/lib/python3.10/dist-packages/tritonclient/grpc/_client.py", line 257, in __del__
    self.close()
  File "/usr/local/lib/python3.10/dist-packages/tritonclient/grpc/_client.py", line 264, in close
    self.stop_stream()
  File "/usr/local/lib/python3.10/dist-packages/tritonclient/grpc/_client.py", line 1811, in stop_stream
    if self._stream is not None:
AttributeError: 'InferenceServerClient' object has no attribute '_stream'

I have added the reason to the script.


ENV LD_LIBRARY_PATH=/usr/local/tensorrt/lib/:/opt/tritonserver/backends/tensorrtllm:$LD_LIBRARY_PATH
"""
with open(os.path.join(ddir, dockerfile_name), "w") as dfile:
Expand Down
Loading