SeeCoder [WIP]

It provides the capability to generate CLIP from an image input, unlike unCLIP, which works in all models. (To use this extension, you need to download the required model file from Install Models)

By BlenderNeko·Created 3 years ago·Updated 2 years ago· 40

BlenderNeko/ComfyUI_SeeCoder

Nodes4

On cloudLocal install

Category_for_testing, image/upscaling

Stars40

Updated2 years ago

Nodes (4)

ConcatConditioning

The conditioning-staple node you might not need at all

_for_testing

Upscale Image (using Model)

Run an ESRGAN upscaler on your image

image/upscaling

SEECoderImageEncode

Hand SD1.5 a reference image instead

conditioning

Load Upscale Model

Load an ESRGAN model for upscaling

loaders

Readme

ComfyUI SeeCoder nodes

This repo contains 2 experimental WIP nodes for ComfyUI that let's you use SeeCoders.

getting SeeCoders

You can find the seecoders here. They have to be placed at models/seecoders

nodes:

SEECoderImageEncode

this node can be used to create an embedding from an image

image: the image to encode
seecoder_name: the name of the seecoder

ConcatConditioning

this node can be used to concat different embeddings together, so you can e.g. create both a text and a visual embedding and concat them together.

conditioning_to: a set of embeddings to concat something to
conditioning_from: the embedding to concat behind those in conditioning_to

TODO:

[ ] support for non safetensor formats
[ ] bring attention layers in line with ones used in comfy