# Low VRAM Consumption

This is the result of GPTQ quantization on Qwen/Qwen3-Reranker-4B, significantly reducing VRAM usage.

The Flux model employs a 4-bit Transformer and T5 encoder for text-to-image generation tasks, supporting non-commercial use.

A multilingual instruction-following language model based on LLaMA 7B, supporting English, Chinese, and Japanese, suitable for chatbots and instruction-following tasks.