playgroundai

/

playground-v2.5-1024px-aesthetic

like 524

Hey Playground Team,

Thanks for sharing this model with the community -- this means a lot to us 🤗

Is it possible that you can share with us training details such as:

• Total batch size
• Learning rate: scheduler, warmup, min/max values
• Number of training steps
• Total number of images/samples
• Optimizer hyperparams

This would give the community clear understanding of how you archived such results, so we can leverage this knowledge in our fine-tunes. I hope this is not a secret information to share 🙏

cc: @aykamko @ehsanakh @Daiqing @justinlau