AndroFlux - V8-10k
by ModelsLabNote: This version still sucks. I don't really know what I'm doing wrong.
Training changes from the previously posted run (v5):
Training Steps: Maximum of
6000>> 10000Base Model Precision:
int8-quanto>> fp8-quantoLearning Rate: 1.0
Optimizer: Prodigy
Batch Size:
1>> 2Gradient Accumulation Steps: 1
Rank:
16>> 8Gradient clipping:
default (probably 2)>> 0.1Caption Strategy:
Triple>> Instance prompt
androflux-v8-10kInput
Per image generation will cost 0.0047$
For premium plan image generation will cost 0.00$ i.e Free.
Output
Unknown content type
Related Models
Discover similar models you might be interested in
![PC-98 Style [FLUX] - v2-caption](https://images.stablediffusionapi.com/?Image=https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/7d1ab23f-3c51-4c11-81a9-46b7630de3af/width=832/45149581.jpeg)
![N0Ds [Flux] - Modern Military Helmet with Night Vision Goggles - v1.0](https://images.stablediffusionapi.com/?Image=https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/62cb6dfc-ec03-4d96-a008-7178cb51ed24/width=1440/36710950.mp4)
![Harry Mason - Silent Hill: Shattered Memories [LoRA] - v1.0 FLUX.1-dev (WD14)](https://images.stablediffusionapi.com/?Image=https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/0d020f57-f839-4178-a819-868d9724bf36/width=768/37041785.jpeg)
AndroFlux - V8-10k Readme
Note: This version still sucks. I don't really know what I'm doing wrong.
Training changes from the previously posted run (v5):
Training Steps: Maximum of
6000>> 10000Base Model Precision:
int8-quanto>> fp8-quantoLearning Rate: 1.0
Optimizer: Prodigy
Batch Size:
1>> 2Gradient Accumulation Steps: 1
Rank:
16>> 8Gradient clipping:
default (probably 2)>> 0.1Caption Strategy:
Triple>> Instance prompt