ComfyUI Node
Timm Backbone Image Encode
A ComfyUI node in image with 4 inputs and 1 output.
Timm Backbone Image Encode
- model
- image
- TENSOR
◄feature_typepooler_output►
◄hidden_state_index-1►
Categoryimage
Inputs (4)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | — | |
| image | IMAGE | — | |
| feature_type | COMBO | pooler_output | 2 options: pooler_output, hidden_state |
| hidden_state_index | INT | -1-128–128 | Index of the hidden state to extract. Only used when feature_type is 'hidden_state'. -1 means the last, -2 means the penultimate. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| TENSOR | TENSOR | — |