This is the README for a VS code extension called "Zero" used to run Local LLM's.. for free.
This VS extension works in conjuction with Ollama and run models such as DeepSeek.
The model being run will be dependendant on your computer hardware, I recommend starting with a lower model e.g. DeepSeek-R1-Distill-Qwen-1.5B if you are unsure.
Rough guide for gpu+ram combo:
- 8gb ram - DeepSeek-R1-Distill-Qwen-1.5B
- 16gb ram - DeepSeek-R1-Distill-Llama-8B
- 32gb ram - DeepSeek-R1-Distill-Qwen-14B
Users appreciate release notes as you update your extension.
Initial release of ...
//### 1.0.1
This project requires the following dependencies,
Implement this VS code extensionm template,
npx --package yo --package generator-code -- yo code
alternate extension build https://www.npmjs.com/package/generator-code
Download and install Ollama from https://ollama.com or https://github.com/ollama/ollama/blob/main/docs/faq.md#how-does-ollama-handle-concurrent-requests
ollama run llama3.2
Within ollama download the model you would like to use, e.g. Deepseek https://ollama.com/library/deepseek-r1:8b
- ensure the model is appropriate for the spec of the machine it is being run on.
additionally add the es6-string-html extension in VS code for Syntax Highlighting.
Once installed replace the extension.ts, package.json and the tsconfig.json in your VS code extension path.
To get it up and running you can run the >debug mode in VS code
Followed by cmd + shift + p
Run the extension and you should now be running a local VS code LLM.
Initial tutorial by Fireship.io, Check out there channel for more tutorials, https://www.youtube.com/watch?v=clJCDHml2cA https://fireship.io/courses/
Enjoy!