Added tiny model for free and fast usage. I created a dataset of ~100 axamples of advanced queries and generated ~4500 synthetic examples. The model was trained on this data based on Qwen2.5-Coder-0.5B-Instruct. The model is available for free and hosted on my vm. I'm planning to add self-hosted version of the model.
- You should add OpenAI API key to plugin settings or set to use special model.
- You can describe what the query should return in logseq block, then access the generation via /
Generate advanced query with ai
- Add more complex prompts for complex queries with visualizations
- Train tiny model for faster and free usage
- Add self-hosted tiny model for free and fast usage
If you find this plugin useful, I'll be glad if you support me by buing me a coffe