Closed MartinEthier closed 6 months ago
carlinds
commented
6 months ago Hi! Thanks for reporting this issue. I see that we missed a line in the dockerfile, which should set the default shell to bash and solve your first issue. Adding the following before the torch installation should work:
SHELL ["/bin/bash", "-c"]
The second error can be resolved by changing back the project name in pyproject.toml to "nerfstudio".
I will put up a fix with both of these changes.
TurtleZhong
commented
6 months ago Hi! Thanks for reporting this issue. I see that we missed a line in the dockerfile, which should set the default shell to bash and solve your first issue. Adding the following before the torch installation should work:
SHELL ["/bin/bash", "-c"]The second error can be resolved by changing back the project name in pyproject.toml to "nerfstudio".
I will put up a fix with both of these changes.
Hi, Following your solution, I have fixed in the tiny-cuda-nn install step, I also have changed back the project name in pyproject.toml to "nerfstudio", but I still failed in this command RUN ns-install-cli --mode install, I got error like this:
=> [15/17] RUN cd /nerfstudio && python3.10 -m pip install --no-cache-dir -e . 190.4s
=> [16/17] WORKDIR /workspace 0.1s => ERROR [17/17] RUN ns-install-cli --mode install 10.2s ------
> [17/17] RUN ns-install-cli --mode install: #0 0.345 [09:17:56] 🤷 .zshrc not found, skipping. install.py:369
#0 0.346 🔍 Found .bashrc! install.py:371
#0 0.609 ✔ Wrote new completion to /nerfstudio/nerfstudio/scripts/completions/bash/_ns-dev-test! install.py:133
#0 0.636 ✔ Wrote new completion to /nerfstudio/nerfstudio/scripts/completions/bash/_ns-install-cli! install.py:133
#0 2.826 [09:17:58] ✔ Wrote new completion to install.py:133
#0 2.826 /nerfstudio/nerfstudio/scripts/completions/bash/_ns-dev-sync-viser-message-defs!
#0 3.454 [09:17:59] ✔ Wrote new completion to /nerfstudio/nerfstudio/scripts/completions/bash/_ns-download-data! install.py:133
#0 3.809 ✔ Wrote new completion to /nerfstudio/nerfstudio/scripts/completions/bash/_ns-process-data! install.py:133
#0 6.006 [09:18:01] ✔ Wrote new completion to /nerfstudio/nerfstudio/scripts/completions/bash/_ns-render! install.py:133
#0 6.043 ✔ Wrote new completion to /nerfstudio/nerfstudio/scripts/completions/bash/_ns-viewer! install.py:133
#0 6.127 ✔ Wrote new completion to /nerfstudio/nerfstudio/scripts/completions/bash/_ns-eval! install.py:133
#0 6.616 [09:18:02] ❌ Completion script generation failed: ['ns-export', '--tyro-print-completion', 'bash'] install.py:123
#0 6.619 Traceback (most recent call last): install.py:127
#0 6.619 File "/usr/local/bin/ns-export", line 5, in <module>
#0 6.619 from nerfstudio.scripts.exporter import entrypoint
#0 6.619 File "/nerfstudio/nerfstudio/scripts/exporter.py", line 43, in <module>
#0 6.619 from nerfstudio.fields.sdf_field import SDFField # noqa
#0 6.619 ModuleNotFoundError: No module named 'nerfstudio.fields.sdf_field'
#0 6.619
#0 6.620 Traceback (most recent call last):
#0 6.620 File "/usr/local/bin/ns-install-cli", line 8, in <module>
#0 6.620 sys.exit(entrypoint())
#0 6.620 File "/nerfstudio/nerfstudio/scripts/completions/install.py", line 411, in entrypoint
#0 6.621 tyro.cli(main, description=__doc__)
#0 6.621 File "/usr/local/lib/python3.10/dist-packages/tyro/_cli.py", line 184, in cli
#0 6.621 output = _cli_impl(
#0 6.621 File "/usr/local/lib/python3.10/dist-packages/tyro/_cli.py", line 446, in _cli_impl
#0 6.621 out, consumed_keywords = _calling.call_from_args(
#0 6.621 File "/usr/local/lib/python3.10/dist-packages/tyro/_calling.py", line 241, in call_from_args
#0 6.621 return unwrapped_f(*positional_args, **kwargs), consumed_keywords # type: ignore
#0 6.621 File "/nerfstudio/nerfstudio/scripts/completions/install.py", line 392, in main
#0 6.621 _generate_completions_files(completions_dir, scripts_dir, shells_supported, shells_found)
#0 6.621 File "/nerfstudio/nerfstudio/scripts/completions/install.py", line 323, in _generate_completions_files
#0 6.622 completion_paths = list(
#0 6.622 File "/usr/lib/python3.10/concurrent/futures/_base.py", line 621, in result_iterator
#0 6.622 yield _result_or_cancel(fs.pop())
#0 6.622 File "/usr/lib/python3.10/concurrent/futures/_base.py", line 319, in _result_or_cancel
#0 6.622 return fut.result(timeout)
#0 6.622 File "/usr/lib/python3.10/concurrent/futures/_base.py", line 458, in result
#0 6.622 return self.__get_result()
#0 6.622 File "/usr/lib/python3.10/concurrent/futures/_base.py", line 403, in __get_result
#0 6.623 raise self._exception
#0 6.623 File "/usr/lib/python3.10/concurrent/futures/thread.py", line 58, in run
#0 6.623 result = self.fn(*self.args, **self.kwargs)
#0 6.623 File "/nerfstudio/nerfstudio/scripts/completions/install.py", line 325, in <lambda>
#0 6.623 lambda path_or_entrypoint_and_shell: _generate_completion(
#0 6.623 File "/nerfstudio/nerfstudio/scripts/completions/install.py", line 128, in _generate_completion
#0 6.623 raise e
#0 6.623 File "/nerfstudio/nerfstudio/scripts/completions/install.py", line 115, in _generate_completion
#0 6.623 new = subprocess.run(
#0 6.623 File "/usr/lib/python3.10/subprocess.py", line 526, in run
#0 6.624 raise CalledProcessError(retcode, process.args,
#0 6.624 subprocess.CalledProcessError: Command '['ns-export', '--tyro-print-completion', 'bash']' returned non-zero exit status 1.
#0 10.16 [09:18:05] ✔ Wrote new completion to /nerfstudio/nerfstudio/scripts/completions/bash/_ns-train! install.py:133
------
Dockerfile:154
--------------------
152 | SHELL ["/bin/sh", "-c"]
153 | # Install nerfstudio cli auto completion
154 | >>> RUN ns-install-cli --mode install
155 |
156 | # Bash as default entrypoint.
--------------------
ERROR: failed to solve: process "/bin/sh -c ns-install-cli --mode install" did not complete successfully: exit code: 1I haven't been able to fix this issue yet. I'm wondering if you have any ideas on what is going wrong.
carlinds
commented
6 months ago Hi! This has already been fixed in the latest version. Please pull the latest changes and try again :)
HulkM
commented
6 months ago Hi! I meet some new wrong that shows in 79th row of Dockerfile.
georghess
commented
6 months ago Hi @HulkM
The Connection timed out indicates that there might be an issue with your internet connection.
As a side note, you could remove any ceres and colmap installation as they are not needed for any of the AD datasets.
Crescent-Saturn
commented
6 months ago Hi! I meet some new wrong that shows in 79th row of Dockerfile.
@HulkM Hi, check the connection to https://ceres-solver.googlesource.com/ceres-solver.git. It is not easy when you are in China mainland. Most google services seem to be blocked. Good luck:
HulkM
commented
6 months ago Hi @HulkM
The
Connection timed outindicates that there might be an issue with your internet connection.As a side note, you could remove any ceres and colmap installation as they are not needed for any of the AD datasets.
Hi! I meet some new wrong that shows in 79th row of Dockerfile.
@HulkM Hi, check the connection to https://ceres-solver.googlesource.com/ceres-solver.git. It is not easy when you are in China mainland. Most google services seem to be blocked. Good luck:
Thanks! I will try it! :)
TurtleZhong
commented
6 months ago Hi! This has already been fixed in the latest version. Please pull the latest changes and try again :)
I have tried the latest version, and it's OK.
When trying to build the docker image, I run into a few issues. First, I get the following error:
This seems to be due to the command using bash specific syntax. I am using Ubuntu which means the docker commands are ran using /bin/sh instead of /bin/bash. This can be fixed by replacing the CUDA_VER assignment to something that does not use bash specific syntax:
CUDA_VER=$(echo $CUDA_VERSION | awk 'BEGIN{FS=OFS="."}NF--' | sed -e 's/\.//g')After fixing the previous error I then I get another error:
I haven't been able to fix this issue yet. I'm wondering if you have any ideas on what is going wrong.