Continuing to try and refine this model, just barely stepping foot trying to understand how this all works.
3261412_training_data.zip
seraph.safetensors