lerobot FFmpeg version issue in Ubuntu

OS: Ubuntu 22.0.4 python: 3.10 virtual env: miniconda ffmpeg: 7.1.1 open CV: 4.11 av == 14.2.0 torchvision == 0.21.0 av and torchvision were default. After update to above version, they gave the same result.

Issue: In order to record video, newest FFmpeg is a must since certain encoder is used. When I do the following command:

pip install --no-binary=av -e ".[feetech]"

or

pip install --no-binary=av -e .

Terminal gave the same output as follow, which indicate: " error: ‘AV_OPT_TYPE_CHANNEL_LAYOUT’ undeclared (first use in this function); did you mean ‘AV_OPT_TYPE_CHLAYOUT’? " " Possible reason is: FFmpeg changed AV_OPT_TYPE_CHANNEL_LAYOUT into AV_OPT_TYPE_CHLAYOUT in new version. It is not compatible in PyAV.

Suggetion: can we have more control of versions by writing a requirement.txt indicating wanted package versions in order to avoid conflicts？A complete package list can be exported from a successful environment using a json file.

The following paragraphs are the terminal outputs:

  building 'av.format' extension
  gcc -pthread -B /home/guyu/miniconda3/envs/lerobot/compiler_compat -Wno-unused-result -Wsign-compare -DNDEBUG -fwrapv -O2 -Wall -fPIC -O2 -isystem /home/guyu/miniconda3/envs/lerobot/include -fPIC -O2 -isystem /home/guyu/miniconda3/envs/lerobot/include -fPIC -I/usr/include/x86_64-linux-gnu -I/home/guyu/miniconda3/envs/lerobot/include/python3.10 -c src/av/format.c -o build/temp.linux-x86_64-cpython-310/src/av/format.o
  src/av/format.c: In function ‘__pyx_pf_2av_6format_15ContainerFormat___cinit__’:
  src/av/format.c:3373:24: warning: assignment discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
   3373 |     __pyx_v_self->iptr = av_find_input_format(__pyx_t_4);
        |                        ^
  src/av/format.c:3410:24: warning: assignment discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers]
   3410 |     __pyx_v_self->optr = av_guess_format(__pyx_t_5, NULL, NULL);
        |                        ^
  gcc -pthread -B /home/guyu/miniconda3/envs/lerobot/compiler_compat -shared -Wl,--allow-shlib-undefined -Wl,-rpath,/home/guyu/miniconda3/envs/lerobot/lib -Wl,-rpath-link,/home/guyu/miniconda3/envs/lerobot/lib -L/home/guyu/miniconda3/envs/lerobot/lib -Wl,--allow-shlib-undefined -Wl,-rpath,/home/guyu/miniconda3/envs/lerobot/lib -Wl,-rpath-link,/home/guyu/miniconda3/envs/lerobot/lib -L/home/guyu/miniconda3/envs/lerobot/lib build/temp.linux-x86_64-cpython-310/src/av/format.o -lavformat -lavcodec -lavdevice -lavutil -lavfilter -lswscale -lswresample -o build/lib.linux-x86_64-cpython-310/av/format.cpython-310-x86_64-linux-gnu.so
  building 'av.bytesource' extension
  gcc -pthread -B /home/guyu/miniconda3/envs/lerobot/compiler_compat -Wno-unused-result -Wsign-compare -DNDEBUG -fwrapv -O2 -Wall -fPIC -O2 -isystem /home/guyu/miniconda3/envs/lerobot/include -fPIC -O2 -isystem /home/guyu/miniconda3/envs/lerobot/include -fPIC -I/usr/include/x86_64-linux-gnu -I/home/guyu/miniconda3/envs/lerobot/include/python3.10 -c src/av/bytesource.c -o build/temp.linux-x86_64-cpython-310/src/av/bytesource.o
  gcc -pthread -B /home/guyu/miniconda3/envs/lerobot/compiler_compat -shared -Wl,--allow-shlib-undefined -Wl,-rpath,/home/guyu/miniconda3/envs/lerobot/lib -Wl,-rpath-link,/home/guyu/miniconda3/envs/lerobot/lib -L/home/guyu/miniconda3/envs/lerobot/lib -Wl,--allow-shlib-undefined -Wl,-rpath,/home/guyu/miniconda3/envs/lerobot/lib -Wl,-rpath-link,/home/guyu/miniconda3/envs/lerobot/lib -L/home/guyu/miniconda3/envs/lerobot/lib build/temp.linux-x86_64-cpython-310/src/av/bytesource.o -lavformat -lavcodec -lavdevice -lavutil -lavfilter -lswscale -lswresample -o build/lib.linux-x86_64-cpython-310/av/bytesource.cpython-310-x86_64-linux-gnu.so
  building 'av.dictionary' extension
  gcc -pthread -B /home/guyu/miniconda3/envs/lerobot/compiler_compat -Wno-unused-result -Wsign-compare -DNDEBUG -fwrapv -O2 -Wall -fPIC -O2 -isystem /home/guyu/miniconda3/envs/lerobot/include -fPIC -O2 -isystem /home/guyu/miniconda3/envs/lerobot/include -fPIC -I/usr/include/x86_64-linux-gnu -I/home/guyu/miniconda3/envs/lerobot/include/python3.10 -c src/av/dictionary.c -o build/temp.linux-x86_64-cpython-310/src/av/dictionary.o
  gcc -pthread -B /home/guyu/miniconda3/envs/lerobot/compiler_compat -shared -Wl,--allow-shlib-undefined -Wl,-rpath,/home/guyu/miniconda3/envs/lerobot/lib -Wl,-rpath-link,/home/guyu/miniconda3/envs/lerobot/lib -L/home/guyu/miniconda3/envs/lerobot/lib -Wl,--allow-shlib-undefined -Wl,-rpath,/home/guyu/miniconda3/envs/lerobot/lib -Wl,-rpath-link,/home/guyu/miniconda3/envs/lerobot/lib -L/home/guyu/miniconda3/envs/lerobot/lib build/temp.linux-x86_64-cpython-310/src/av/dictionary.o -lavformat -lavcodec -lavdevice -lavutil -lavfilter -lswscale -lswresample -o build/lib.linux-x86_64-cpython-310/av/dictionary.cpython-310-x86_64-linux-gnu.so
  building 'av.option' extension
  gcc -pthread -B /home/guyu/miniconda3/envs/lerobot/compiler_compat -Wno-unused-result -Wsign-compare -DNDEBUG -fwrapv -O2 -Wall -fPIC -O2 -isystem /home/guyu/miniconda3/envs/lerobot/include -fPIC -O2 -isystem /home/guyu/miniconda3/envs/lerobot/include -fPIC -I/usr/include/x86_64-linux-gnu -I/home/guyu/miniconda3/envs/lerobot/include/python3.10 -c src/av/option.c -o build/temp.linux-x86_64-cpython-310/src/av/option.o
  src/av/option.c: In function ‘__pyx_pymod_exec_option’:
  src/av/option.c:7416:51: error: ‘AV_OPT_TYPE_CHANNEL_LAYOUT’ undeclared (first use in this function); did you mean ‘AV_OPT_TYPE_CHLAYOUT’?
   7416 |   __pyx_t_3 = __Pyx_PyInt_From_enum__AVOptionType(AV_OPT_TYPE_CHANNEL_LAYOUT); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 36, __pyx_L1_error)
        |                                                   ^~~~~~~~~~~~~~~~~~~~~~~~~~
        |                                                   AV_OPT_TYPE_CHLAYOUT
  src/av/option.c:7416:51: note: each undeclared identifier is reported only once for each function it appears in
  error: command '/usr/bin/gcc' failed with exit code 1
  [end of output]

note: This error originates from a subprocess, and is likely not a problem with pip. ERROR: Failed building wheel for av Successfully built lerobot Failed to build av ERROR: Failed to build installable wheels for some pyproject.toml based projects (av) (lerobot) guyu@GuLinux:~/Documents/5_lerobot/lerobot$

What you should do according to the new update is the following:
Fetch the latest changes (if you haven't already done that).
Create a new virtual environment using conda:
conda create -n lerobot python=3.10 -y
conda activate lerobot
Install ffmpeg using conda. This is in accordance with the README, just a few additions to ensures ffmpeg version ≥ 7:
conda install -c conda-forge 'ffmpeg>=7.0' -y
Install lerobot (change the extras based on the robot you're using; this one is for Aloha):
pip install --no-binary=av -e ".[dynamixel, intelrealsense]"
These steps should be enough to avoid any issues with encoders or jpeg12_write_raw_data.

Originally posted by @shantanuparab-tr in #679

Apr 06 '25 07:04 gu021

follow my suggestions in issue #727 :)

Apr 07 '25 13:04 Tavish9

I don't understand the meaning of this issue. Did you follow the recommended instructions for installation?

Download our source code:

git clone https://github.com/huggingface/lerobot.git
cd lerobot

Create a virtual environment with Python 3.10 and activate it, e.g. with miniconda:

conda create -y -n lerobot python=3.10
conda activate lerobot

When using miniconda, if you don't have fffmpeg in your environment:

conda install ffmpeg

Install 🤗 LeRobot:

pip install --no-binary=av -e .

NOTE: If you encounter build errors, you may need to install additional dependencies (cmake, build-essential, and ffmpeg libs). On Linux, run: sudo apt-get install cmake build-essential python-dev pkg-config libavformat-dev libavcodec-dev libavdevice-dev libavutil-dev libswscale-dev libswresample-dev libavfilter-dev pkg-config. For other systems, see: Compiling PyAV

This should install ffmpeg 6.1.1. The issue of having ffmpeg 7 is that it is not compatible with the version packaged by opencv

Apr 07 '25 15:04 imstevenpmwork

Here's what I have noticed. Just running

conda install ffmpeg

installs ffmpeg version 6.1.1. This version does not support libsvtav1 encoder

ffmpeg version 6.1.1 Copyright (c) 2000-2023 the FFmpeg developers
  built with gcc 11.2.0 (Anaconda gcc)
  configuration: --prefix=/home/interbotix/miniconda3/envs/ffmpeg_test --cc=/croot/ffmpeg_1743153284778/_build_env/bin/x86_64-conda-linux-gnu-cc --ar=/croot/ffmpeg_1743153284778/_build_env/bin/x86_64-conda-linux-gnu-ar --nm=/croot/ffmpeg_1743153284778/_build_env/bin/x86_64-conda-linux-gnu-nm --ranlib=/croot/ffmpeg_1743153284778/_build_env/bin/x86_64-conda-linux-gnu-ranlib --strip=/croot/ffmpeg_1743153284778/_build_env/bin/x86_64-conda-linux-gnu-strip --disable-doc --enable-swresample --enable-swscale --enable-openssl --enable-libxml2 --enable-libtheora --enable-demuxer=dash --enable-postproc --enable-hardcoded-tables --enable-libfreetype --enable-libharfbuzz --enable-libfontconfig --enable-libdav1d --enable-zlib --enable-libaom --enable-pic --enable-shared --disable-static --disable-gpl --enable-version3 --disable-sdl2 --enable-libopenh264 --enable-libopus --enable-libmp3lame --enable-libopenjpeg --enable-libvorbis --enable-pthreads --enable-libtesseract --enable-libvpx
  libavutil      58. 29.100 / 58. 29.100
  libavcodec     60. 31.102 / 60. 31.102
  libavformat    60. 16.100 / 60. 16.100
  libavdevice    60.  3.100 / 60.  3.100
  libavfilter     9. 12.100 /  9. 12.100
  libswscale      7.  5.100 /  7.  5.100
  libswresample   4. 12.100 /  4. 12.100
 V....D libaom-av1           libaom AV1 (codec av1)
 V....D libopenh264          OpenH264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 (codec h264)
 VF.... libopenjpeg          OpenJPEG JPEG 2000 (codec jpeg2000)
 V....D libtheora            libtheora Theora (codec theora)
 V....D libvpx               libvpx VP8 (codec vp8)
 V....D libvpx-vp9           libvpx VP9 (codec vp9)
 VF...D zlib                 LCL (LossLess Codec Library) ZLIB
 A....D libmp3lame           libmp3lame MP3 (MPEG audio layer 3) (codec mp3)
 A....D libopus              libopus Opus (codec opus)
 A....D libvorbis            libvorbis (codec vorbis)

Installing ffmpeg using conda install -c conda-forge 'ffmpeg>=7.0' -y gives you all the required encoders.

ffmpeg version 7.1.1 Copyright (c) 2000-2025 the FFmpeg developers
  built with gcc 13.3.0 (conda-forge gcc 13.3.0-2)
  configuration: --prefix=/home/interbotix/miniconda3/envs/lerobot_trossen_env --cc=/home/conda/feedstock_root/build_artifacts/ffmpeg_1743376049581/_build_env/bin/x86_64-conda-linux-gnu-cc --cxx=/home/conda/feedstock_root/build_artifacts/ffmpeg_1743376049581/_build_env/bin/x86_64-conda-linux-gnu-c++ --nm=/home/conda/feedstock_root/build_artifacts/ffmpeg_1743376049581/_build_env/bin/x86_64-conda-linux-gnu-nm --ar=/home/conda/feedstock_root/build_artifacts/ffmpeg_1743376049581/_build_env/bin/x86_64-conda-linux-gnu-ar --disable-doc --enable-openssl --enable-demuxer=dash --enable-hardcoded-tables --enable-libfreetype --enable-libharfbuzz --enable-libfontconfig --enable-libopenh264 --enable-libdav1d --disable-gnutls --enable-libvpx --enable-libass --enable-pthreads --enable-alsa --enable-libpulse --enable-vaapi --enable-libopenvino --enable-gpl --enable-libx264 --enable-libx265 --enable-libmp3lame --enable-libaom --enable-libsvtav1 --enable-libxml2 --enable-pic --enable-shared --disable-static --enable-version3 --enable-zlib --enable-libvorbis --enable-libopus --enable-librsvg --enable-ffplay --pkg-config=/home/conda/feedstock_root/build_artifacts/ffmpeg_1743376049581/_build_env/bin/pkg-config
  libavutil      59. 39.100 / 59. 39.100
  libavcodec     61. 19.101 / 61. 19.101
  libavformat    61.  7.100 / 61.  7.100
  libavdevice    61.  3.100 / 61.  3.100
  libavfilter    10.  4.100 / 10.  4.100
  libswscale      8.  3.100 /  8.  3.100
  libswresample   5.  3.100 /  5.  3.100
  libpostproc    58.  3.100 / 58.  3.100
 V....D libaom-av1           libaom AV1 (codec av1)
 V..... libsvtav1            SVT-AV1(Scalable Video Technology for AV1) encoder (codec av1)
 V....D libx264              libx264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 (codec h264)
 V....D libx264rgb           libx264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 RGB (codec h264)
 V....D libopenh264          OpenH264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 (codec h264)
 V....D libx265              libx265 H.265 / HEVC (codec hevc)
 V....D libvpx               libvpx VP8 (codec vp8)
 V....D libvpx-vp9           libvpx VP9 (codec vp9)
 VF...D zlib                 LCL (LossLess Codec Library) ZLIB
 A....D libmp3lame           libmp3lame MP3 (MPEG audio layer 3) (codec mp3)
 A....D libopus              libopus Opus (codec opus)
 A....D libvorbis            libvorbis (codec vorbis)

Using ffmpeg 6.1.1 results in encoder not found issue as mentioned in

#822
#705

I'm using Linux 22.04 , and I don't face any issues with pyav installations as I have downloaded all the required packages as mentioned in the Readme.

I hope this information helps. @imstevenpmwork

Apr 07 '25 15:04 shantanuparab-tr

Hi, I am using Ubuntu 22.04, and I also faced the same av build issue of using ffmpeg==7.1.1. Downgrading to 6.1.1 helps the installation.

I am new to this repository, so I haven't run anything to face the encoder issue yet.

Apr 07 '25 16:04 BruceGeLi

Hi @shantanuparab-tr,

Thank you for the detailed explanation. It looks like this is an additional challenge on top of the previous dependency version conflicts and the imshow() issue we encountered.

After checking, I confirm that the conda-installed FFmpeg doesn’t include libsvtav1, but the Brew-installed version (on macOS) does. While upgrading to FFmpeg 7 would include libsvtav1, it would reintroduce conflicts with OpenCV, leading to the same imshow() problem as before.

Interestingly, on Ubuntu 24, installing FFmpeg via sudo apt install ffmpeg provides version 6.1.1—which is compatible with our codebase—and includes libsvtav1. This suggests we could update the installation instructions to recommend installing FFmpeg through Brew (macOS) or APT (Ubuntu 24) instead of Conda, ensuring compatibility while keeping encoder support. (https://github.com/huggingface/lerobot/pull/950)

One caveat: Ubuntu 22’s default APT repository installs FFmpeg 4.X, so we’d need separate instructions for that version. Another solution would be to build manually FFmpeg with all the blows and whistles and then point all the packages to the built version. But I'm afraid that would complicate the installation for an entry user.

Apr 07 '25 20:04 imstevenpmwork

This is really a nice suggestion:

This suggests we could update the installation instructions to recommend installing FFmpeg through Brew (macOS) or APT (Ubuntu 24) instead of Conda, ensuring compatibility while keeping encoder support. (#950)

libsvtav1 was absent 1.5 weeks ago when I tried to record data set episode. We need at least one working Ubuntu solution since Pi0 project is based on Ubuntu only. May I ask who will do the verification of above solution and do the update of installation instruction? A nice and reliable start is essential to a good result.

Apr 08 '25 05:04 gu021

Here's an update, I tried a few things and noticed that the order of installation really matters.

Here's the workflow that works for me on Linux 22.04.

Create a conda virtual environment.

conda create -n lerobot python=3.10 -y
conda activate lerobot

Installing LeRobot

pip install --no-binary=av -e ".[dynamixel, intelrealsense]"

Installing ffmpeg

conda install -c conda-forge ffmpeg -y

How this help, this allows you to build pyav from source without getting this error

  src/av/option.c: In function ‘__pyx_pymod_exec_option’:
  src/av/option.c:7416:51: error: ‘AV_OPT_TYPE_CHANNEL_LAYOUT’ undeclared (first use in this function); did you mean ‘AV_OPT_TYPE_CHLAYOUT’?
   7416 |   __pyx_t_3 = __Pyx_PyInt_From_enum__AVOptionType(AV_OPT_TYPE_CHANNEL_LAYOUT); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 36, __pyx_L1_error)
        |                                                   ^~~~~~~~~~~~~~~~~~~~~~~~~~
        |                                                   AV_OPT_TYPE_CHLAYOUT
  src/av/option.c:7416:51: note: each undeclared identifier is reported only once for each function it appears in
  error: command '/usr/bin/gcc' failed with exit code 1
  [end of output]

Installing ffmpeg later allows you to get the correct version of ffmpeg that supports libsvtav1

Note: Here's why the previous solution worked Isuccessfully build pyav from source in one of my environments, whihc was then cached, this cached version was used in the other environments as well, therefore, I didn't run into the pyav not built error.

When I used pip install --no-binary=av -e . --no-cache-dir --force-reinstall I was able to recreate the issue. This led me to the above solution.

@gu021

Apr 08 '25 15:04 shantanuparab-tr

Nice solution! I believe this time it will work. You found the root cause. I will try it out soon. Thanks a lot, Shantanuparab!

Apr 08 '25 17:04 gu021

We're working on a more elaborated solution, I recommend you to give first this PR a try (from a fresh environment and following the new installation instructions in the main readme): https://github.com/huggingface/lerobot/pull/903

Apr 09 '25 07:04 imstevenpmwork

Hello! I created an issue #964 as I have having some related issues. Using the fixes provided by @shantanuparab-tr I had done conda install -c conda-forge 'ffmpeg>=7.0' -y and since then my installation worked fine, I was able to run the control_robot.py script perfectly and collect a dataset. Now when I tried to train I had an issue where I get :

NFO 2025-04-09 16:58:03 ts/train.py:202 Start offline training on a fixed dataset
Traceback (most recent call last):
  File "/home/moonshot/lerobot/lerobot/scripts/train.py", line 288, in <module>
    train()
  File "/home/moonshot/lerobot/lerobot/configs/parser.py", line 227, in wrapper_inner
    response = fn(cfg, *args, **kwargs)
  File "/home/moonshot/lerobot/lerobot/scripts/train.py", line 205, in train
    batch = next(dl_iter)
  File "/home/moonshot/lerobot/lerobot/common/datasets/utils.py", line 588, in cycle
    yield next(iterator)
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 701, in __next__
    data = self._next_data()
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 1465, in _next_data
    return self._process_data(data)
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 1491, in _process_data
    data.reraise()
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torch/_utils.py", line 715, in reraise
    raise exception
RuntimeError: Caught RuntimeError in DataLoader worker process 0.
Original Traceback (most recent call last):
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torch/utils/data/_utils/worker.py", line 351, in _worker_loop
    data = fetcher.fetch(index)  # type: ignore[possibly-undefined]
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 52, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 52, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/home/moonshot/lerobot/lerobot/common/datasets/lerobot_dataset.py", line 739, in __getitem__
    video_frames = self._query_videos(query_timestamps, ep_idx)
  File "/home/moonshot/lerobot/lerobot/common/datasets/lerobot_dataset.py", line 711, in _query_videos
    frames = decode_video_frames(video_path, query_ts, self.tolerance_s, self.video_backend)
  File "/home/moonshot/lerobot/lerobot/common/datasets/video_utils.py", line 66, in decode_video_frames
    return decode_video_frames_torchcodec(video_path, timestamps, tolerance_s)
  File "/home/moonshot/lerobot/lerobot/common/datasets/video_utils.py", line 190, in decode_video_frames_torchcodec
    from torchcodec.decoders import VideoDecoder
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torchcodec/__init__.py", line 10, in <module>
    from . import decoders, samplers  # noqa
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torchcodec/decoders/__init__.py", line 7, in <module>
    from ._core import VideoStreamMetadata
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torchcodec/decoders/_core/__init__.py", line 8, in <module>
    from ._metadata import (
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torchcodec/decoders/_core/_metadata.py", line 15, in <module>
    from torchcodec.decoders._core.video_decoder_ops import (
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torchcodec/decoders/_core/video_decoder_ops.py", line 59, in <module>
    load_torchcodec_extension()
  File "/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torchcodec/decoders/_core/video_decoder_ops.py", line 44, in load_torchcodec_extension
    raise RuntimeError(
RuntimeError: Could not load libtorchcodec. Likely causes:
          1. FFmpeg is not properly installed in your environment. We support
             versions 4, 5, 6 and 7.
          2. The PyTorch version (2.5.1+cu124) is not compatible with
             this version of TorchCodec. Refer to the version compatibility
             table:
             https://github.com/pytorch/torchcodec?tab=readme-ov-file#installing-torchcodec.
          3. Another runtime dependency; see exceptions below.
        The following exceptions were raised as we tried to load libtorchcodec:
        
[start of libtorchcodec loading traceback]
libavutil.so.59: cannot open shared object file: No such file or directory
/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torchcodec/libtorchcodec6.so: undefined symbol: _ZNK3c1011StorageImpl27throw_data_ptr_access_errorEv
libavutil.so.57: cannot open shared object file: No such file or directory
/home/moonshot/miniconda3/envs/lerobot/lib/python3.10/site-packages/torchcodec/libtorchcodec4.so: undefined symbol: _ZNK3c1011StorageImpl27throw_data_ptr_access_errorEv
[end of libtorchcodec loading traceback].

I have mentioned the versions for ffmpeg and libtorchcodec on #964, I then came across this issue and tried to downgrade ffmpeg to 6.1.1 and retry, but was not successful. After donwgrading to 6.1.1, I even have issues running control_robot script originating from opencv again. Any suggestions?

thanks

Apr 09 '25 15:04 shrutichakraborty

I was testing against /feat/ branch. Then I noticed that it is brought to main. Then I tested main branch again. At least so far, I do not have any issue with FFmpeg version any more. I would like to close this issue. Thanks a lot, all!

Apr 09 '25 17:04 gu021

lerobot lerobot copied to clipboard

FFmpeg version issue in Ubuntu

lerobot
lerobot copied to clipboard