Décri une images, génère un prompt et une image Flux.

Le décrypteur Florence2 VLM associé au nœud Miaoshouai-Tagger génère de très bonne description d'images.

https://github.com/miaoshouai/ComfyUI-Miaoshouai-Tagger

Vous aurez besoin de ces MODELS :

flux_dev_fp8_scaled_diffusion_model.safetensors --> ComfyUI_windows_portable\ComfyUI\models\unet

https://huggingface.co/comfyanonymous/flux_dev_scaled_fp8_test/blob/main/flux_dev_fp8_scaled_diffusion_model.safetensors

t5xxl_fp8_e4m3fn.safetensors --> ComfyUI_windows_portable\ComfyUI\models\clip

https://huggingface.co/comfyanonymous/flux_text_encoders/blob/main/t5xxl_fp8_e4m3fn.safetensors

clip_l.safetensors --> ComfyUI_windows_portable\ComfyUI\models\clip

https://huggingface.co/comfyanonymous/flux_text_encoders/blob/main/clip_l.safetensors

clip_g.safetensors --> ComfyUI_windows_portable\ComfyUI\models\clip

https://huggingface.co/comfyanonymous/clip_vision_g/blob/main/clip_vision_g.safetensors

Hyper-FLUX.1-dev-8steps-lora.safetensors --> ComfyUI_windows_portable\ComfyUI\models\loras

https://huggingface.co/ByteDance/Hyper-SD/blob/main/Hyper-FLUX.1-dev-8steps-lora.safetensors

Flux_Ultimator.safetensors --> ComfyUI_windows_portable\ComfyUI\models\loras

https://civitai.com/models/792297/flux-ultimator-or-another-flux-detailer

lora-000016.TA_trained.safetensors --> ComfyUI_windows_portable\ComfyUI\models\loras

https://civitai.com/models/805324/fluxhands-final-bonus

-------------------------------------

Nodes to use Florence2 VLM for image tagging and captioning

ComfyUI-Miaoshouai-Tagger --> Via Manager or : https://github.com/miaoshouai/ComfyUI-Miaoshouai-Tagger

-------------------------------------