Did Dynamic Prompt or Prompt Magic - interrupt Generation Batch Count?

[cxbignekoc]

version: f2.0.1v1.10.1-previous-561-g82eb7566  •  python: 3.10.11  •  torch: 2.3.1+cu121  •  xformers: 0.0.27  •  gradio: 4.40.0

Steps: 30, Sampler: DPM++ 2M SDE, Schedule type: Karras, CFG scale: 5, Seed: 310178740, Size: 1048x792, Model hash: 8805a1afb1, Model: SDXL_juggernautXL_juggXIByRundiffusion, Clip skip: 2, ControlNet 0: "Module: None, Model: diffusion_pytorch_model_promax [9460e4db], Weight: 0.4, Resize Mode: Crop and Resize, Processor Res: 512, Threshold A: 0.5, Threshold B: 0.5, Guidance Start: 0.0, Guidance End: 1.0, Pixel Perfect: True, Control Mode: Balanced, Hr Option: Both", ControlNet 1: "Module: None, Model: None, Weight: 1, Resize Mode: Crop and Resize, Processor Res: 512, Threshold A: 0.5, Threshold B: 0.5, Guidance Start: 0.0, Guidance End: 1.0, Pixel Perfect: True, Control Mode: Balanced, Hr Option: Both", Version: f2.0.1v1.10.1-previous-561-g82eb7566

Weird Behaviour? Buggy Stuff.

i got:

• no 8 Picture Output but i set it up
• twice of an Single Picture from Control Net in the Output
• twice of an single generated Picture always if one is generated
• variations of the Output using Combinatorial batches but its "ignored" by Max generations =0
• Combinatorial batches 4 generated 16 picture by set up Batch Count of 8 (+ 2 Control Net Image)

If i Setup Batch Count to 8, in a random situation, i generate just one Picture twice.
i expected 8 variations. The Second Picture has another prompt variation and weight then of a random sentence like

• "misty eerie cityscape" turns into "(misty eerie cityscape:0.63).

Before all 8 Picture has no change at all in the Picture Prompt but are different.
The Prompt contains NO WEIGHTS.

Before this "1 Picture Output" happens i got no Problems getting 8 Pictures.
Combinatorial batches of Dynamic Prompts is set up to "1" anyways it is ignored by max gernarion is "0".
But Combinatorial batches seems interrupting the Batch Count of the Outputs.
i expected that should not happen.

Now i also got Control Net running and i got no issues until was switching from Depth to Canny.
Mostly i used at ALL because its seens working fine.

• I always need to reload the Control Net Model also, its another issue.

So now the Batch Count seems interrupted to generated just one Picture.
I figured out that an higher Combinatorial batches from 1 to 2 generating 4 Pictures and Doubles the Control Net Single Image in the Output. But its ignored by max generations = 0.

If Magic Prompt is on alles is working as expected.

So what the hell is going on here?