Make the model register as a quantization, instead of a finetune

Update README.md

Delete flux.1-dev-SRPO-fp8.safetensors

Update README.md

Improve model card: Add pipeline tag, update library_name, and fix abstract typo (#1)

Upload flux.1-dev-SRPO-fp8.safetensors with huggingface_hub

Update README.md

Upload flux.1-dev-SRPO-bf16.safetensors with huggingface_hub

initial commit