Open Issues Need Help
View All on GitHubAI Summary: The user reports that model downloads, especially those sourced from GitHub, consistently fail around 800 MB when initiated from within the application. They found a workaround by manually resuming downloads using `wget -c`. The request is to implement an automatic download resumption feature within the UI to handle such mid-download failures.
Ultimate Vocal Remover 5 with Gradio UI. Separate an audio file into various stems, using multiple models
AI Summary: The user is requesting the ability to load custom models whose components, such as the model checkpoint and its configuration file, are hosted at different, arbitrary remote URLs (e.g., HuggingFace for the checkpoint and GitHub for the config). They provide an example JSON structure demonstrating this multi-source setup. The issue implies a desire for automated remote loading, as the current workaround is manual download and local configuration.
Ultimate Vocal Remover 5 with Gradio UI. Separate an audio file into various stems, using multiple models