ivylilili comments

Repositories
Issues
Comments

Results 2 comments of


                                            ivylilili

can you provide the specific hyper params of train.sh for causalvae training ?

> We have updated the training script and added training parameter descriptions in the document. PR: #201 Hello~ I noticed that your provided training hyper params for causalvae is 256\*256\*17....

The density_for_timestep_sampling and loss_weighting for SD3 Training！！！

> yes, if you look at the timestep selection distribution using the SD3 style training, it effectively does not ever train the 900-1000 or 0-100 range of timesteps. they are...