taming-transformers

taming-transformers copied to clipboard

small question about the vq-vae paper

1

hello! thank you for your great work! I have a question about loss function in paper. L = log p(x|z(x)) + ||sg[z(x)] − e|| + β||z(x) − sg[e]|| the author...

ghost

Openimages version.

2

I want know How much images in Openimages used at VQGAN, 1.7M or 9.2M? And at stage2 still use Openimages or changing to ImageNet-1k? Thank you. @asanakoy @rom1504 @pesser @tgisaturday...

ghost

Implementation question - Resize then crop - Model trained on low resolution image?

2

Hi, from the config file `sflckr_cond_stage.yaml`, the image is resized to SmallestMaxSize=256 and then crop. So the model was trained on smaller (resized) images. Is the model ckpt also trained...

HieuPhan33

AttributeError: module 'signal' has no attribute 'SIGUSR1'

1

hi, I have a problem: ```configs/faceshq_vqgan.yaml -t True --gpus 0, Running on GPUs 0, Working with z of shape (1, 256, 16, 16) = 65536 dimensions. loaded pretrained LPIPS loss...

nhw649

pytorch still raised "out of memory" but my PYTORCH_CUDA_ALLOC_CONF = "max_split_size_mb:128"

1

Do I see properly that my Nvidia free memory is 451MiB? If true than why pytorch still raise Exception "out of memory"? `torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate...

uranity

According to the compatible list released by torch-lightning official web site, the following environment yaml describe the incompatible package version : pytorch=1.7.0 pytorch-lightning==1.0.8 ======================= Official version :  Basically the...

HuangChiEn

Hi! The [vqgan_imagenet_f16_16384](https://heibox.uni-heidelberg.de/d/a7530b09fed84f80a887/) model in particular has been used by other interesting projects. This repository has an open source license (the MIT license), but the model is hosted separately and...

Aspie96

Hi, How should I calculate the negative log-likelihood metric (or something like bits/dim) which is usually reported in generative modeling papers? Should I just use the reconstruction loss as NLL?...

khanhvu207

How to train vqgan on coco dataset?

1

Can't find coco_vqgan.yaml in “config” dir

kevinchow1993

Hello! I was very impressed by your paper. I am interested in trying out the training myself. Do you happen to have weights for a pretrained VQGAN model on faces...

scj0709