Skip to content

Latest commit

 

History

History
50 lines (30 loc) · 1.26 KB

README.md

File metadata and controls

50 lines (30 loc) · 1.26 KB
Raw

MPT 30B inference code using CPU

Run inference on the latest MPT-30B model using your CPU. This inference code uses a ggml quantized model. To run the model we'll use a library called ctransformers that has bindings to ggml in python.

Turn style with history on latest commit:

Inference Chat

Video of initial demo:

2023-06-25.20-13-24.mp4

Requirements

I recommend you use docker for this model, it will make everything easier for you. Minimum specs system with 32GB of ram. Recommend to use python 3.10.

Tested working on

Will post some numbers for these two later.

  • AMD Epyc 7003 series CPU
  • AMD Ryzen 5950x CPU

Setup

First create a venv.

python -m venv env && source env/bin/activate

Next install dependencies.

pip install -r requirements.txt

Next download the quantized model weights (about 19GB).

python download_model.py

Ready to rock, run inference.

python inference.py

Next modify inference script prompt and generation parameters.