Quality Realtime Interpolation
5.0
2 reviewsDescription
(This template is used for Workflow Contest)
What this workflow does
Interpolates between two prompts. Benefit from the prompting capabilities and the speed of SDXL Turbo combined with the SD1.5 LCM LoRA and a 1.5 model to get decent quality images in 0.6 seconds on an RTX 3090.
How to use this workflow
Enter your prompts, Enable "Extra Options" and then "Auto Queue" for "realtime" interpolations. Modify the strength to see your changes - with quality output!
Multiple "Blending Modes" are available, so play with them all! Want to turn a cat into a rodent (and let's face it, who doesn't)? Enter "cat" in Prompt A, "rodent" in Prompt B, select "bislerp" blending mode and then change the Blending Strength. ("Exclusion" is the only one that doesn't really work).
Tips about this workflow
Generates lots of images :) Watching how things change will really help you up your prompting game too.
🎥 Video demo link (optional)
👉 [Please add here]
Discussion
(No comments yet)
Loading...
Reviews
No reviews yet
Versions (1)
- latest (a year ago)
Node Details
Primitive Nodes (12)
Anything Everywhere (1)
Anything Everywhere3 (1)
Anything Everywhere? (3)
Note (1)
PrimitiveNode (5)
Prompts Everywhere (1)
Custom Nodes (21)
ComfyUI
- CheckpointLoaderSimple (2)
- LoraLoaderModelOnly (1)
- PreviewImage (2)
- VAEDecode (2)
- CLIPTextEncode (6)
- VAEEncode (1)
- EmptyLatentImage (1)
- KSamplerSelect (1)
- SDTurboScheduler (1)
- SamplerCustom (1)
- KSampler (1)
- ConditioningBlend (2)
Model Details
Checkpoints (2)
SD1.5/epicrealism_naturalSin.safetensors
SDXL/sd_xl_turbo_1.0_fp16.safetensors
LoRAs (1)
SD15/sd15_LCM_pytorch_lora_weights.safetensors