How can I troubleshoot the usage of context in TabbyML when repositories are not being used?
I'm having trouble with TabbyML not using repositories as context for completions. I've checked data/events/ but the JSON doesn't include any information about the context from repositories. Tabby is running inside Docker, and the scheduler seems to run without issues. I'm looking for ways to troubleshoot this and wondering if there are any logs I might have missed.
Matti Frind
Asked on Mar 19, 2024
To troubleshoot the usage of context in TabbyML, you can review the event log located at ~/.tabby/events. Modifications added by the repository context will be included as part of the prompt. Additionally, if you're using the --webserver option, you can check the Code Browser from the avatar menu in the UI. It's important to note that the options for Git providers and repository configurations in config.toml are mutually exclusive. If you define repositories in config.toml, they will not take effect if you're also using --webserver. Instead, repositories should be defined in Git Providers in the UI. This behavior will be documented soon on the official TabbyML documentation site.