# 4-bit quantization fine-tuning

This is a fine-tuned Phi-4 model that achieves 2x training acceleration through the Unsloth and TRL libraries and focuses on text generation tasks.