Load a quantized model and chat, all in your tab
Pick the GGUF quant that fits your GPU, from the repo itself
Pick the GGUF quantization that fits your GPU
Hugging Face trending models with rank and likes history