ComfyUI Node
Youtu-VL (GGUF Advanced)
A ComfyUI node in 🧪AILab/YoutuVL with 11 inputs and 1 output.
Youtu-VL (GGUF Advanced)
- image
- text
◄modelYoutu-VL-4B-Instruct-GGUF-Q8►
◄n_gpu_layers-1►
◄n_ctx4096►
◄preset_prompt🖼️ Describe Image►
◄custom_prompt►
◄max_tokens512►
◄temperature0.10►
◄top_p0.001►
◄keep_model_loadedtrue►
◄seed1►
Category🧪AILab/YoutuVL
Inputs (11)
| Name | Type | Default | Description |
|---|---|---|---|
| model | COMBO | Youtu-VL-4B-Instruct-GGUF-Q8 | Select the GGUF quantized model. |
| n_gpu_layers | INT | -1-1–100 | Number of layers to offload to GPU. -1 = all layers. |
| n_ctx | INT | 4096512–32768 | Context window size. |
| preset_prompt | COMBO | 🖼️ Describe Image | 6 options: 🖼️ Describe Image, 📝 Detailed Description, 🔍 Analyze Elements, 🏷️ Generate Tags, 📄 OCR Text, 🎨 Art Style Analysis |
| custom_prompt | STRING | — | |
| max_tokens | INT | 51264–32768 | Maximum number of new tokens to generate. |
| temperature | FLOAT | 0.100.01–2 | Sampling temperature. Lower = more focused. |
| top_p | FLOAT | 0.0010.001–1 | Nucleus sampling cutoff. |
| keep_model_loaded | BOOLEAN | true | Keep model loaded for faster subsequent inference. |
| seed | INT | 11–4294967295 | — |
| imageopt | IMAGE | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| text | STRING | — |