Model Summary

Purpose

A coding assistant optimized for Apple Silicon.

Key Features

• Advanced reasoning and coding capabilities.

Use Cases

• Designed for use with VSCode extension Continue or other Ollama-supported front ends.

Performance

No performance metrics provided yet.

Dependencies

• Requires Ollama or llama.cpp using the GGUF format.
• Recommended front end VSCode extension Continue

README

Installation Instructions

To run the latest version of the model, use the following command:

ollama run ih0dl/octopus:latest

Note: The vision model (ih0dl/octopus:vision) will be available soon.

Usage Examples

You can prompt the model directly. For example, this octopus was prompted using the VS code extension and a context file:

Prompt: Can we change 'run_cli' in @sd3c.py to accept external variables?

Response:

Yes, you can modify the `run_cli()` function to accept external variables as arguments. Here's how you can do it:

```python
def run(frames, use_cyberpunk_palette, save_as_gif, output_path, blend_mode, save_faces, custom_prompt, negative_prompt, font_size, gif_frame_speed, use_existing_image, existing_image_path):
    print("Welcome to the Cyberpunk Image Generator!")

    main(
        frames=frames,
        use_cyberpunk_palette=use_cyberpunk_palette,
        save_as_gif=save_as_gif,
        output_path=output_path,
        blend_mode=blend_mode,
        save_simulated_vision=save_faces,
        custom_prompt=custom_prompt,
        font_size=font_size,
        gif_frame_speed=gif_frame_speed,
        use_existing_image=use_existing_image,
        negative_prompt=negative_prompt,
        existing_image_path=existing_image_path if use_existing_image else None
    )

Now you can call the run() function with your variables directly:

run(frames=4,
    use_cyberpunk_palette=True,
    save_as_gif=True,
    output_path="./cypher_subjects/",
    blend_mode="multiply",
    save_faces=True,
    custom_prompt="A close up of a woman's face in a photorealistic manner with high contrast and vivid colors against a black background",
    negative_prompt="frame, border",
    font_size=32,
    gif_frame_speed=100,
    use_existing_image=False,
    existing_image_path=None)

Configuration Options

You should edit the system message and add in your device hardware. It will help the model immensely when suggesting commands.

Troubleshooting

There are no specific troubleshooting steps provided.

Contributors

Currently, the main contributor to this model is ih0dl.

License

This model is licensed under the Mistral AI Non-Production License.