I’m trying to run the container in my Jetson Orin AGX (Jetpack 6.2+b77):
jetson-containers run $(autotag torch_tensorrt)
and the whole process get stuck here (Loading: 0 packages loaded).
I break the process after 30 minutes of the same message and the container doesn’t have torch_tensorrt:
My settings:
It’s
import tensorrt
print(tensorrt__.version__)
not torch_tensorrt
No. It’s torch_tensorrt. Just like the default test code example included in the same package.
Dear @esteban.gallardo ,
Just checking if you notice issue with other containers as well like pytorch?
XTTS container no longer works:
jetson-containers run dustynv/xtts:r36.3.0
Running the test.py:
Gives the error:
I need both services. Is there any roadmap to fix the containers? I need the XTTS container to work again urgently.
Dear @esteban.gallardo ,
Jetpacj 6.2 has BSP 36.4.3 . Did you try building locally with jetson-container and noticed issue for xtts as well?
Building also fails. I just want that something that it’s been working fine for a year can work again.
The command '/bin/sh -c cd /opt && git clone --depth=1 https://github.com/NVIDIA-AI-IOT/torch2trt && cd torch2trt && cp /tmp/patches/flattener.py torch2trt && pip3 install --verbose . && sed 's|^set(CUDA_ARCHITECTURES.*|#|g' -i CMakeLists.txt && sed 's|Catch2_FOUND|False|g' -i CMakeLists.txt && cmake -B build -DCUDA_ARCHITECTURES=${CUDA_ARCHITECTURES} . && cmake --build build --target install && ldconfig && pip3 install --no-cache-dir --verbose nvidia-pyindex && pip3 install --no-cache-dir --verbose onnx-graphsurgeon' returned a non-zero code: 1
Traceback (most recent call last):
File "/usr/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/usr/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/home/esteban/Workspace/jetson-containers/jetson_containers/build.py", line 122, in <module>
build_container(args.name, args.packages, args.base, args.build_flags, args.build_args, args.simulate, args.skip_tests, args.test_only, args.push, args.no_github_api, args.skip_packages)
File "/home/esteban/Workspace/jetson-containers/jetson_containers/container.py", line 147, in build_container
status = subprocess.run(cmd.replace(_NEWLINE_, ' '), executable='/bin/bash',shell=True, check=True)
File "/usr/lib/python3.10/subprocess.py", line 526, in run
raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command 'DOCKER_BUILDKIT=0 docker build --network=host --tag xtts:r36.4.0-torch2trt --file /home/esteban/Workspace/jetson-containers/packages/pytorch/torch2trt/Dockerfile --build-arg BASE_IMAGE=xtts:r36.4.0-tensorrt /home/esteban/Workspace/jetson-containers/packages/pytorch/torch2trt 2>&1 | tee /home/esteban/Workspace/jetson-containers/logs/20250211_090007/build/xtts_r36.4.0-torch2trt.txt; exit ${PIPESTATUS[0]}' returned non-zero exit status 1.
Dear @esteban.gallardo,
Could you double check if you have enough space(and space related errors not shown on the terminal log) as I see not sufficient space issue with torch_tensort and then I can login to container and notice same behavior like you. I will verify the same with an external SSD and see if it works.
I’ve free space. I’ve an SSD of 500 Gigas. 140 Gigas left.
esteban@esteban-desktop:~/Workspace/StoryBookEditor$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/nvme0n1p1 456G 294G 140G 68% /
tmpfs 31G 84K 31G 1% /dev/shm
tmpfs 13G 35M 13G 1% /run
tmpfs 5,0M 4,0K 5,0M 1% /run/lock
/dev/nvme0n1p10 63M 118K 63M 1% /boot/efi
tmpfs 6,2G 80K 6,2G 1% /run/user/1000
For torch_tensorrt you could try the development container:
I’m in a difficult situation without XTTS container working. Is it possible to know if someone is or will be looking into that?
Thanks.
Hi,
We have built the XTTS container for r36.3.0.
The fastest way is to set up your environment in JetPack 6.0.
TensorRT has been upgraded to v10.3 from v8.6 in the JetPack 6.1/6.2.
So modifications are required to make it compatible with the new library version.
We don’t have a concrete schedule to update that container for JetPack 6.2.
But these are all open sourced so you can try it directly.
Thanks.
XTTS container has worked with JetPack 6.2 for several weeks before stopping working out of the blue.
Anyway, thanks for the sincere reply, so DIY it is.
This is based on
FROM nvcr.io/nvidia/pytorch:25.01-py3
which is jetpack 6.2 cuda 12.8 and has torch_tensorrt, and many of the requirements.txt already on the base image. With an added GitHub - coqui-ai/TTS: 🐸💬 - a deep learning toolkit for Text-to-Speech, battle-tested in research and production and its requirements.
Put Dockerfile.txt (1.8 KB) in a directory and remove .txt extension.
docker build -t pytorch:tts .
or whatever name you want to give the image.
docker run -it --rm --net=host --runtime nvidia --privileged --ipc=host --ulimit memlock=-1 \
–ulimit stack=67108864 -v /home/userid/git/jetson-containers/packages/speech:/workspace \
pytorch:tts bash
It includes pip show spacy, https://spacy.io/ which I believe is a good place to get models.
From within the container run the following to get a list of models.
python /usr/local/lib/python3.12/dist-packages/TTS/server/server.py --list_models
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.