jetson-containers build ollama error

[kentsuiGitHub]

I am trying to build ollama but am not able to build it.

Could you guide me how to fix it.

Error:
Step 12/19 : COPY --from=ollama-l4t-build /opt/ollama/dist/linux-arm64/lib/ollama /usr/lib/ollama
COPY failed: stat opt/ollama/dist/linux-arm64/lib/ollama: file does not exist
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 "/opt/jetson-containers/jetson_containers/build.py", line 112, in
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 "/opt/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 ollama:r36.4.3-ollama --file /opt/jetson-containers/packages/llm/ollama/Dockerfile --build-arg BASE_IMAGE=ollama:r36.4.3-python --build-arg OLLAMA_REPO="ollama/ollama" --build-arg OLLAMA_BRANCH="v0.5.7" --build-arg GOLANG_VERSION="1.22.8" --build-arg CMAKE_VERSION="3.22.1" --build-arg JETPACK_VERSION="6.2" --build-arg CUDA_VERSION_MAJOR="12" --build-arg CMAKE_CUDA_ARCHITECTURES="87" /opt/jetson-containers/packages/llm/ollama 2>&1 | tee /opt/jetson-containers/logs/20250119_015021/build/ollama_r36.4.3-ollama.txt; exit ${PIPESTATUS[0]}' returned non-zero exit status 1.

[iueoa]

I've been seeing the same behavior.

[WiktorRembielakA4BEE]

I am getting the same error

[tokk-nv]

Hi @kentsuiGitHub , @iueoa , @WiktorRembielakA4BEE ,

Just so we can understand the whole picture, did you get to pull the dustynv/ollama:0.5.1-r36.4.0 container image?

jetson-containers run --name ollama $(autotag ollama)

It looks like @kentsuiGitHub is on JetPack 6.2, but from what I test, the container image can run on JetPack 6.2 (but you may need to run ollama serve& at the beginning in the container).
Is there any specific reason why you want to build an Ollama container using jetson-containers?

Just so that we have all the alternative paths in our scope, we can now install Ollama natively as well.
https://www.jetson-ai-lab.com/tutorial_ollama.html#1-native-install
I'm just trying not to miss your specific motivation to build a custom Ollama container.

[cyber-nico]

Hi, I got the same error - also on JetPack 6.2. I was going to build a new container with the latest version of Ollama (v0.5.7), since the latest container image available for JetPack is based on v0.5.1. Issuing the command

jetson-containers build --name ollama-latest ollama

led to the exact same error message as reported by @kentsuiGitHub.

[tokk-nv]

We just fixed the issue with this PR #807.
And we just re-build the container with the latest Ollama (0.5.7), and updated the container image.

If you want to just run the container with the latest Ollama, you can do this.

jetson-containers run dustynv/ollama:main-r36.4.0

Or, you can build by yourself with the follwoing.

jetson-containers build --skip-tests=ollama ollama

[tokk-nv]

I just realized that my former PR was not installing Ollama with GPU support, and thus the resulted dustynv/ollama:main-r36.4.0 pushed (9e63cacd3b60) suffers the same issue.

I filed the new PR #808, and this should enable GPU.

[kentsuiGitHub]

Hi, I got the same error - also on JetPack 6.2. I was going to build a new container with the latest version of Ollama (v0.5.7), since the latest container image available for JetPack is based on v0.5.1. Issuing the command

jetson-containers build --name ollama-latest ollama

led to the exact same error message as reported by @kentsuiGitHub.

Thank you for the clarification @cyber-nico

[kentsuiGitHub]

Hi @kentsuiGitHub , @iueoa , @WiktorRembielakA4BEE ,

Just so we can understand the whole picture, did you get to pull the dustynv/ollama:0.5.1-r36.4.0 container image?

jetson-containers run --name ollama $(autotag ollama)

It looks like @kentsuiGitHub is on JetPack 6.2, but from what I test, the container image can run on JetPack 6.2 (but you may need to run ollama serve& at the beginning in the container). Is there any specific reason why you want to build an Ollama container using jetson-containers?

Just so that we have all the alternative paths in our scope, we can now install Ollama natively as well. https://www.jetson-ai-lab.com/tutorial_ollama.html#1-native-install I'm just trying not to miss your specific motivation to build a custom Ollama container.

Hi @tokk-nv

Thank you. I have followed the procedure and installed ollama natively for workaround recently

Will try the latest container image and post it here.

[kentsuiGitHub]

I just realized that my former PR was not installing Ollama with GPU support, and thus the resulted dustynv/ollama:main-r36.4.0 pushed (9e63cacd3b60) suffers the same issue.

I filed the new PR #808, and this should enable GPU.

Understood.
Will try to run the container again when the new image is ready.