Verus_Vision_1.0b / README.md
SG161222's picture
Update README.md
a2b33b3 verified
|
raw
history blame
1.4 kB
metadata
license: other
license_name: flux-1-dev-non-commercial-license
license_link: https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md

Verus Vision β€” Model for Realistic and Photorealistic Image Generation

Overview

Verus Vision is a fine-tuned model designed for generating high-quality, realistic and photorealistic images. Built on the Flux Dev (de-distill) model with 12 billion parameters.

Support fine-tuning of this model on my Boosty

This version of the model is not final. Ongoing development is underway to further enhance the quality of outputs and minimize errors.


Verus Vision 1.0b Model Training

Stage 1: Data Preparation

  • [ β˜‘ ] Collect images for the dataset
  • [ β˜‘ ] Dataset editing (crop and resize)
  • [ β˜‘ ] Image captioning

Stage 2: Model Training

  • [ β˜‘ ] Prepare dataset for model input
  • [ β˜‘ ] Model training (Main Dataset: Part 10 / Part 10)

Stage 3: Testing

  • [ β–Ά ] Model testing

Stage 4: Release

  • [ ☐ ] Release Transformer Version
  • [ ☐ ] Release Compact Version