https://github.com/wardbekker/llama_summerize

Example Gong call transcript summarize script
https://github.com/wardbekker/llama_summerize

Last synced: about 4 hours ago
JSON representation

Example Gong call transcript summarize script

• Host: GitHub
• URL: https://github.com/wardbekker/llama_summerize
• Owner: wardbekker
• Created: 2023-08-21T05:46:56.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2023-08-21T05:53:10.000Z (about 1 year ago)
• Last Synced: 2023-08-21T06:36:21.001Z (about 1 year ago)
• Language: Python
• Size: 0 Bytes
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Gong call transcript summarisation example

This example shows how LLAMA can be used to summarize gong call transcripts

## Usage

- Download LLAMA model from https://huggingface.co/TheBloke/Llama-2-13B-chat-GGML/resolve/main/llama-2-13b-chat.ggmlv3.q5_K_S.bin and place it in a ./models folder

- Create a virtual environment: `python -m venv venv`

- Activate the virtual environment: `source venv/bin/activate`

- Install the requirements: `pip install -r requirements.txt`

- On Mac M1, M2 etc make sure to build llama-cpp-python like this: `CMAKE_ARGS="-DLLAMA_METAL=on" FORCE_CMAKE=1 pip install --force llama-cpp-python`. This will build the Metal version of LLAMA which is much faster than the CPU version.

- Run the summarization script: `python summarize.py   `

## Example

Go to any Gong call, click on the transcript tab, and save the HTML file. Then run the following command:

```

python summarize.py gong-transcript.html "Ward" summary.txt

```

(Main) Speaker name is used for semi-intelligent chunking of the text. 

After a coffee or two this will result in a new text file 'summary.txt' with the summary of the call. At the bottom of the file you will have a total summary of the call, and above that you will have a summary of each chunk of the call.