FR: Integrated Local Download of Ollama Models

[LeonelRFF]

Is your feature request related to a problem? Please describe.

Hello, The drawback I find is that to use the models locally, you need to use another tool such as Termux. From there, you can install Ollama and then run it in Swift-Chat.

Describe the solution you'd like

would it be possible to download models from the app to run them locally? I mean, in the Ollama section, you could write the name of the model to download it from here.

Describe alternatives you've considered

Something similar to what I'm referring to is in the project here github.com/sunshine0523/OllamaServer, where it is not necessary to use Termux, you can simply do all this from the application.

Additional context

No response

[zhu-xiaowei]

Thank you for suggesting this feature. Having the ability to start and run the Ollama server on the device, as well as download and manage models directly within the app, would certainly enhance the user experience. However, implementing this well requires significant additional effort, such as designing a more user-friendly interface, ensuring stable operation, and improving overall usability.

Currently, the main goal of SwiftChat is to efficiently invoke large models and display their markdown results. Moreover, there are already tools like Termux that can run the Ollama server locally, so this functionality is not within the current development scope of the SwiftChat app.

If you are interested in implementing this feature, you are very welcome to submit a PR.