'Re-Dream' Combining prompts generated with Florence2 from 2 images and combined Latents. V.2
5.0
0 reviewsDescription
A better working version of this workflow:
This workflow takes in 2 images and then Florence2 generates prompts from those images.
These prompts are then averaged by 'ConditioningAvarge' node.
(set the 'ConditioningAvarge' node between 0.45 and .55 to make the result more like one or the other image.
This version is different from the previous one because it doesn't work on an 'Empty Latent' but instead also combines the Latents encoded from the input images witch gives a lot better results then the previous version.
The images used for the previews for this workflow are attached on the right side of this workflow page as 'Assets'.
I like feedback and comments ! So please leave them below ! :)
ENJOY !
#NeuraLunk
Discussion
(No comments yet)
Loading...
Resources (2)
Reviews
No reviews yet
Versions (1)
- latest (a month ago)
Node Details
Primitive Nodes (9)
DownloadAndLoadFlorence2Model (1)
Florence2Run (2)
FluxGuidance (1)
ModelSamplingFlux (1)
Note (2)
Reroute (2)
Custom Nodes (23)
- CM_NearestSDXLResolution (1)
- CR Image Output (1)
ComfyUI
- KSamplerSelect (1)
- VAEDecode (1)
- VAEEncode (2)
- LatentBlend (1)
- ImageScale (2)
- SamplerCustomAdvanced (1)
- BasicScheduler (1)
- RandomNoise (1)
- BasicGuider (1)
- UNETLoader (1)
- DualCLIPLoader (1)
- VAELoader (1)
- CLIPTextEncode (2)
- ConditioningAverage (1)
- LoadImage (2)
- ShowText|pysssss (2)
Model Details
Checkpoints (0)
LoRAs (0)