Open
Conversation
Introduces support for Metal Performance Shaders (MPS), allowing the demo to run on Apple Silicon. The inference script now dynamically selects the device (MPS, CUDA, or CPU) and conditionally applies FlashAttention. Dependencies are updated to improve cross-platform compatibility, removing `liger_kernel`, `deepspeed`, and using the general `onnxruntime`. A new example video showcasing Apple Silicon execution is added. Gradio certificate files are now ignored.
… apple_silicon
Collaborator
|
Amazing work! so great 👍🐮🍺 will check on a apple silicon mac then merge that! |
Author
|
Two more remarks:
|
…ge git-lfs installation command to pip and remove conda-forge dependencies
Collaborator
|
Hi LeJeko, your PR is great but since github said: This is because I have made some changes in main branch after your PR. Now I have fixed the key part to support mps following your changes Line 37 in 1dee253 But I did not fix the readme. Would you mind to update your PR with a new markdown file about the environment installation? Thank you so much! 🥺 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pull request introduces support for running the LiveCC application on Apple Silicon (tested on M1 Max and M2 Max, 32GB RAM) and includes related updates to the installation guide, demo files, and device handling logic. The most important changes are grouped into installation instructions, demo updates, and device compatibility improvements.
Performance is not fully real-time, but the solution works and demonstrates that LiveCC can run on these architectures.
Installation Instructions for Apple Silicon:
README.md, including steps for setting up dependencies likexcode,miniconda,brew, and specific versions of tools such asffmpeg. It also includes instructions for compilingdecordfor the arm64 architecture.Demo Updates:
demo/sources/LiveCC_on_Apple_Silicon.mp4) to the list of example inputs indemo/app.py.Device Compatibility Improvements:
demo/infer.pyto dynamically select the device (mps,cuda, orcpu) based on availability. This ensures compatibility with Apple Silicon by using thempsbackend when available.