Request for at least 2 input image files and text prompt about cog-comfyui

[fofr]

Prompt is a good idea, I'll see if I can add that.

[bertagknowles]

I think I'll try to pass the prompts as variable in json. Thanks.

[fofr]

SAM models are being added here:
#22

[fofr]

You can zip up your inputs and pass in as many as you want. Or you can use URLs in the JSON instead of file paths.

[bertagknowles]

Thanks for the quick response. A variable positive prompt as api input will be very helpful. I am working on an application and i send two image files (source file and reference file) in base64 format through replicate api. I am aware about zip and url inputs. Can I add variables to the url inputs inside the json file? Will the input accept base64 format? Thanks.

[fofr]

The issue I'm seeing when looking into this is that both positive and negative prompts use CLIPTextEncode and look like this:

    "inputs": {
      "text": "some text",
      "clip": [
        "4",
        1
      ]
    },
    "class_type": "CLIPTextEncode",
    "_meta": {
      "title": "CLIP Text Encode (Prompt)"
    }
  },
  "7": {
    "inputs": {
      "text": "some other text",
      "clip": [
        "4",
        1
      ]
    },
    "class_type": "CLIPTextEncode",
    "_meta": {
      "title": "CLIP Text Encode (Prompt)"
    }
  },

I wouldn't know which of these to update automatically with a given prompt field.

[bertagknowles]

Also can you please add SAM segmentation models. The base model will suffice. Thanks.