You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
"Installation Error: jukebox package not found on conda channels (Windows 64-bit, Python 3.9.18)" "Unable to Install Jukebox from Conda (conda 23.11.0, Python 3.9.18)"
#292
Open
KillshotELOx opened this issue
Jan 5, 2024
· 1 comment
Created a new conda environment with Python 3.9: conda create -n jukebox_env python=3.9
Activated the environment: conda activate jukebox_env
Attempted to install the jukebox package: conda install jukebox
Additional Context:
I've already addressed the Python version issue by creating a new environment with Python 3.9.
I've refreshed the channel metadata and updated conda.
I've searched for alternative channels on Anaconda Cloud but haven't found any with the jukebox package.
Request:
Please advise on how to install the jukebox package successfully. I'm open to alternative installation methods or suggestions for different packages with similar functionality.
Thank you for your assistance!
re-attempt @ 1326 PM 1/5/2024
Summary:
Environment: Windows 64-bit
Conda version: 23.11.0
Python version: 3.9.18
Error: "UnavailableInvalidChannel: HTTP 404 NOT FOUND for channel hcc https://anaconda.org/hcc"
Attempted command: conda install jukebox
Cause: The package's configuration is missing specific packages (tensorboardX.beholder and tensorboardX.proto).
Solutions:
Update setuptools: pip install --upgrade setuptools
If using a version control system, ensure all files are committed and up-to-date.
Try installing tensorboardX directly: pip install tensorboardX
2. PackagesNotFoundError for torchvision and pytorch:
Cause: The specified versions (0.3* and 1.1*) aren't available in the current channels.
Solutions:
Use compatible versions: Check available versions with conda search torchvision and conda search pytorch.
Add a channel with the desired versions: Use conda config --add channels <channel_name>.
3. Error Building Wheel for apex:
Cause: Likely related to missing GPU resources or configuration issues.
Potential Solutions:
Install on a system with a GPU if required.
Check apex installation instructions for dependencies and configuration.
Retry installation with --no-cuda option if GPU is not needed.
re-attempting with:
# Required: Sampling
conda create --name jukebox python=3.9.18
conda activate jukebox
conda install mpi4py=3.0.3 # or try: pip install mpi4py==3.0.3
conda install pytorch torchvision cpuonly -c pytorch # <-- Modified for CPU
git clone https://github.com/openai/jukebox.git
cd jukebox
pip install -r requirements.txt
pip install -e .
# Required: Training
conda install av=7.0.01 -c conda-forge
pip install tensorboardX # <-- Simplified installation
# Optional: Apex (not recommended without GPU)
# If you still want to try Apex without GPU:
pip install -v --no-cache-dir --global-option="--cpp_ext" ./apex # <-- Removed --cuda_ext
The text was updated successfully, but these errors were encountered:
well, this is simple, jukebox package doesn't exist in the system. and in fact, you can just do "pip install -e ." to get pip to install jukebox. that would be really helpful if you are getting that ModuleNotFoundError with jukebox.
Here's the updated error report, incorporating the latest details:
Operating System: Windows 64-bit
Conda Version: 23.11.0
Python Version: 3.9.18
Jukebox Version: (If applicable, specify the version you're trying to install)
Channels: pytorch, conda-forge, defaults
Error Message:
PackagesNotFoundError: The following packages are not available from current channels: - jukebox Current channels: - https://conda.anaconda.org/pytorch - https://conda.anaconda.org/conda-forge - defaults
Steps to Reproduce:
Created a new conda environment with Python 3.9: conda create -n jukebox_env python=3.9
Activated the environment: conda activate jukebox_env
Attempted to install the jukebox package: conda install jukebox
Additional Context:
I've already addressed the Python version issue by creating a new environment with Python 3.9.
I've refreshed the channel metadata and updated conda.
I've searched for alternative channels on Anaconda Cloud but haven't found any with the jukebox package.
Request:
Please advise on how to install the jukebox package successfully. I'm open to alternative installation methods or suggestions for different packages with similar functionality.
Thank you for your assistance!
re-attempt @ 1326 PM 1/5/2024
Summary:
Environment: Windows 64-bit
Conda version: 23.11.0
Python version: 3.9.18
Error: "UnavailableInvalidChannel: HTTP 404 NOT FOUND for channel hcc https://anaconda.org/hcc"
Attempted command: conda install jukebox
@ 1330 PM 1/5/2024
re-attempt after removal of channel https://anaconda.org/hcc:
(base) PS C:\Windows\system32> conda config --show channels
channels:
(base) PS C:\Windows\system32> conda install jukebox
Channels:
Platform: win-64
Collecting package metadata (repodata.json): done
Solving environment: failed
PackagesNotFoundError: The following packages are not available from current channels:
Current channels:
To search for alternate channels that may provide the conda package you're
looking for, navigate to
and use the search bar at the top of the page.
(base) PS C:\Windows\system32>
@ 1340
analysis of:
results from analysis:
Cause: The package's configuration is missing specific packages (tensorboardX.beholder and tensorboardX.proto).
Solutions:
Update setuptools: pip install --upgrade setuptools
If using a version control system, ensure all files are committed and up-to-date.
Try installing tensorboardX directly: pip install tensorboardX
2. PackagesNotFoundError for torchvision and pytorch:
Cause: The specified versions (0.3* and 1.1*) aren't available in the current channels.
Solutions:
Use compatible versions: Check available versions with conda search torchvision and conda search pytorch.
Add a channel with the desired versions: Use conda config --add channels <channel_name>.
3. Error Building Wheel for apex:
Cause: Likely related to missing GPU resources or configuration issues.
Potential Solutions:
Install on a system with a GPU if required.
Check apex installation instructions for dependencies and configuration.
Retry installation with --no-cuda option if GPU is not needed.
re-attempting with:
The text was updated successfully, but these errors were encountered: