Linkedin Message Assistant

anchen

Writing assistant for LinkedIn messages using the local LLM. - Install the extension - Click on the extension icon. - Paste your local LLM url (e.g. TGI endpoint `http://localhost:8000/generate` or Llamacpp endpoint `http://localhost:8000/completion`) and click on `Save`. - Open Linkedin message page in your browser. - Type your objective or intention of the message in the text box and click the button at the bottom to generate the message. - Enjoy! Limitations: - Currently, only the dolphin-2.2-yi-34b model prompt format is supported due to it being the best model I found for writing LinkedIn messages assistant. - The backend API only supports the HuggingFace text generation inference because it is the most production-ready and easy to use.