Published Two Quantized GGUF Models

With permission from DATAtab - published two quantized Open Weights Models on Hugging Face - in GGUF format - for easy use with LMSTUDIO or eventually Ollama.

Their Serbian is good and they do provide high quality answers - smaller Q4_0 has the output tensor copied to improve response quality.

Detailed comparison and study of quality loss due to quantization is planned.

Models are available at:

https://huggingface.co/MarkoRadojcic/YugoGPT-Florida_Q8_0-GGUF

https://huggingface.co/MarkoRadojcic/YugoGPT-Florida_Q4_0.GGUF