ComfyUI Node
VLMImageEditingPrompt
使用VLM模型根据输入图像和指令生成增强的编辑提示词（支持 fp16/int8/int4 量化）
VLMImageEditingPrompt
- image_1
- image_2
- enhanced_prompt
◄task_typeedit►
◄instructionMake the first image have the same style as the second image►
◄quantizationint4►
◄max_new_tokens1024►
CategorySwan
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| image_1 | IMAGE | — | |
| task_type | COMBO | edit | 2 options: edit, generate |
| instruction | STRING | Make the first image have the same style as the second image | — |
| quantization | COMBO | int4 | 3 options: fp16, int8, int4 |
| image_2opt | IMAGE | — | |
| max_new_tokensopt | INT | 102464–2048 | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| enhanced_prompt | STRING | — |