I joined a workflow with 400 prompts.I trained the model on CivitAI, so I don't have performance or loss metrics like I would with local training using tools like Wandb. As a result, the model may have overfitted. I’ve included the halfway training checkpoint in case you think overfitting occurred.Loss monitoring would be a great feature to add to CivitAI training, along with a "resume training" option.

Fortnite Environment FLUX

If you think that the model overfit, this is half training epoch (10)

Files with SHA256: 3ab0b0dde0b1eb686184288e4d857a1669b94025cd5b1f110f082ae735b1d5a6