Training Details about diffused heads

MStypulkowski / diffused-heads

Official repository for Diffused Heads: Diffusion Models Beat GANs on Talking-Face Generation

Other

463 stars 31 forks source link

Training Details about diffused heads #10

Closed xiapengchng closed 1 year ago

xiapengchng commented 1 year ago

Thank you for your awsome work! Could you tell me the training details about diffuse heads such as learning rate, batch size, optimizer since i want to reimplement the training pipeline.

MStypulkowski commented 1 year ago

Thank you and sorry for late reply!

Learning rate: 5e-5 decreasing below 1e-6 with 0.8 factor. Batch size: 10 per GPU. Optimizer: Adam.