Skip to content

moving trtis to 19.08 in BERT - #283

Merged
nvpstr merged 9 commits into
NVIDIA:masterfrom
swethmandava:master
Nov 25, 2019
Merged

moving trtis to 19.08 in BERT#283
nvpstr merged 9 commits into
NVIDIA:masterfrom
swethmandava:master

Conversation

@swethmandava

Copy link
Copy Markdown
Contributor

No description provided.

@lapolonio

Copy link
Copy Markdown
Contributor

i had to make these changes to get the docker/build to work

diff --git a/TensorFlow/LanguageModeling/BERT/Dockerfile b/TensorFlow/LanguageModeling/BERT/Dockerfile
index a56ef26..552d709 100644
--- a/TensorFlow/LanguageModeling/BERT/Dockerfile
+++ b/TensorFlow/LanguageModeling/BERT/Dockerfile
@@ -17,10 +17,10 @@ RUN git clone https://github.com/titipata/pubmed_parser
 RUN pip3 install /workspace/pubmed_parser

 #Copy the perf_client over
-COPY --from=trt /workspace/install/ /workspace/install/
+COPY --from=trt /workspace/build/dist/dist/ /workspace/install/

 #Install the python wheel with pip
-RUN pip install /workspace/install/python/tensorrtserver*.whl
+RUN pip install /workspace/install/tensorrtserver*.whl

 WORKDIR /workspace/bert
 COPY . .

@lapolonio

lapolonio commented Nov 8, 2019

Copy link
Copy Markdown
Contributor

also need to add apt-get install libcurl3 to bert container
to fix

root@instance-2:/workspace/bert# python
Python 3.6.8 (default, Jan 14 2019, 11:02:34)
[GCC 8.0.1 20180414 (experimental) [trunk revision 259383]] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from tensorrtserver.api import ProtocolType, InferContext, ServerStatusContext, grpc_service_pb2_grpc, grpc_service_pb2, model_config_pb2
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python3.6/dist-packages/tensorrtserver/api/__init__.py", line 50, in <module>
    _crequest = cdll.LoadLibrary(_crequest_path)
  File "/usr/lib/python3.6/ctypes/__init__.py", line 426, in LoadLibrary
    return self._dlltype(name)
  File "/usr/lib/python3.6/ctypes/__init__.py", line 348, in __init__
    self._handle = _dlopen(self._name, mode)
OSError: /usr/lib/x86_64-linux-gnu/libcurl.so.4: version `CURL_OPENSSL_3' not found (required by /usr/local/lib/python3.6/dist-packages/tensorrtserver/api/libcrequest.so)

added in swethmandava#3

default trtis_model_name in run_trtis.sh should be consistent with export_model.sh
add libcurl3 install required by python when importing tensorrtserver.api module
@lapolonio

Copy link
Copy Markdown
Contributor

also is there a reason not to use 19.09 or 19.10? I'm interested in using TensorRT 6.

@swethmandava

Copy link
Copy Markdown
Contributor Author

you might be using an older submodule, please update tensorrt-inference-server to use branch r19.08
No, please go ahead and use the latest containers. 19.08 was the latest container at the time of publishing.

@nvpstr
nvpstr merged commit 437eaba into NVIDIA:master Nov 25, 2019
PeganovAnton pushed a commit to PeganovAnton/DeepLearningExamples that referenced this pull request Sep 8, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants