mirror of https://github.com/leafspark/AutoGGUF
27 lines
910 B
Markdown
27 lines
910 B
Markdown
# AutoGGUF - automated GGUF model quantizer
|
|
|
|
This application provides a graphical user interface for quantizing GGUF models
|
|
using the llama.cpp library. It allows users to download different versions of
|
|
llama.cpp, manage multiple backends, and perform quantization tasks with various
|
|
options.
|
|
|
|
**Main features**:
|
|
1. Download and manage llama.cpp backends
|
|
2. Select and quantize GGUF models
|
|
3. Configure quantization parameters
|
|
4. Monitor system resources during quantization
|
|
|
|
**Usage**:
|
|
1. Install dependencies, either using the `requirements.txt` file or `pip install PyQt6 requests psutil`.
|
|
2. Run the `run.bat` script to start the application, or run the command `python src/main.py`.
|
|
|
|
**Dependencies**:
|
|
- PyQt6
|
|
- requests
|
|
- psutil
|
|
|
|
**To be implemented:**
|
|
- Actual progress bar tracking
|
|
- Download safetensors from HF and convert to unquanted GGUF
|
|
- Specify multiple KV overrides
|