CoolFace
Modelpublic

alsoalter/finetuned-gpt-oss-20b-argentina-v02-gguf

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face
0likes20downloads
Model Card

Finetuned GPT OSS 20B Argentina v02 - GGUF

Quantized version optimized for local inference on Mac M1/M2/M3.

Quantizations Available

  • —Q4KM: Best balance of quality and size for 32GB RAM
  • —Q5KM: Higher quality for 64GB RAM (if available)

Usage

bash
# With llama.cpp
./llama-cli -m finetuned-gpt_oss_20b-argentina-v02-q4_k_m.gguf -p "Your prompt here"

Model Details

  • —Base: GPT OSS 20B
  • —Finetuned on: Argentina-specific dataset
  • —Quantization: GGUF format for efficient inference