https://github.com/mayooear/free-ai-youtube-summarizer

Free AI Youtube Summarizer on your computer using mistral-instruct-v0.2, langchain and llama_index
https://github.com/mayooear/free-ai-youtube-summarizer

ai langchain llama-index mistral youtube

Last synced: about 1 month ago
JSON representation

Free AI Youtube Summarizer on your computer using mistral-instruct-v0.2, langchain and llama_index

• Host: GitHub
• URL: https://github.com/mayooear/free-ai-youtube-summarizer
• Owner: mayooear
• Created: 2023-12-19T16:29:09.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-12-19T17:10:45.000Z (over 1 year ago)
• Last Synced: 2025-05-06T18:15:20.471Z (about 1 month ago)
• Topics: ai, langchain, llama-index, mistral, youtube
• Language: Jupyter Notebook
• Homepage:
• Size: 2.93 KB
• Stars: 18
• Watchers: 2
• Forks: 4
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
## Free AI Youtube Summarizer on your computer

Free AI Youtube Summarizer built with opensource LLM (mistral-instruct-v0.2), langchain and llama_index.

Max RAM required: 7 GB

### Setup - Download the open source model

If you're using mac or linux, download [ollama](https://ollama.ai/download) and follow these [Readme.md instructions](https://github.com/jmorganca/ollama/blob/main/README.md) to get up and running with your chosen open source large language models locally.

If you're using Windows, download the model via this account [here](https://huggingface.co/TheBloke/Mistral-7B-Instruct-v0.2-GGUF), then adapt the code to use `llamacpp` instead of `ChatOllama` to reference the models path on your local computer. Follow these instructions [here](https://python.langchain.com/docs/integrations/llms/llamacpp)

### Usage

After you've downloaded your chosen LLM, you can change the default model name:

```

chat_model = ChatOllama(

    model="mistral",

)

```

To adapt the model's instructions, you can change the `system_prompt` variable.

Run the notebook cells to download the youtube transcript and call the model to make a prediction.