[ihales]

I had Claude add it as an edit-prediction provider (running locally on llama.cpp on my Macbook Pro). It's been working well so far (including next-edit prediction!), though it could use more testing and tuning. If you want to try it out you can build my branch: https://github.com/ihales/zed/tree/sweep-local-edit-predicti...

If you have llama.cpp installed, you can start the model with `llama-server -hf sweepai/sweep-next-edit-1.5B --port 11434`

Add the following to your settings.json:

```

  "features": {
    "edit_prediction_provider": { "experimental": "sweep-local" },
  },
  "edit_predictions": {
    "sweep_local": {
      "api_url": "http://localhost:11434/v1/completions",
    },
  }

```

Other settings you can add in `edit_predictions.sweep_local` include:

- `model` - defaults to "sweepai/sweep-next-edit-1.5B"

- `max_tokens` - defaults to 2048

- `max_editable_tokens` - defaults to 600

- `max_context_tokens` - defaults to 1200

I haven't had time to dive into Zed edit predictions and do a thorough review of Claude's code (it's not much, but my rust is... rusty, and I'm short on free time right now), and there hasn't been much discussion of the feature, so I don't feel comfortable submitting a PR yet, but if someone else wants to take it from here, feel free!

[oakesm9]

This is great and similar to what I was thinking of doing at some point. I just wasn't sure if it needed to be specific to Sweep Local or if it could be a generic llama.cpp provider.

[ihales]

I was thinking about this too. Zed officially supports self-hosting Zeta, and so one option would be to create a proxy that uses the Zeta wire format, but is packed by llama.cpp (or any model backend). In the proxy you could configure prompts, context, templates, etc., while still using a production build of Zed. I'll give it a shot if I have time.