LeoXing1996

Oh, my fault! I miss this function! But I still confuse about only train `disA2` and `disB2` via random attribute code and only train `disA1` and `disB1` via attribute from...

I will attempt to reproduce your error on windows. You can add 'mmgen' to `PYTHON_PATH` for a temporary solution. Then you can use it like it has already been installed.

Seems that you use a self-defined config is modified from `stylegan3_r_ada_fp16_gamma3.3_metfaces_1024_b4x8 `. In the original config, `dataset` is defined in `https://github.com/open-mmlab/mmgeneration/blob/master/configs/_base_/datasets/ffhq_flip.py` Maybe you remove `RepeatDataset` from `data['train']` but still remain...

This [url](https://github.com/TommyZihao/MMGeneration_Tutorials) is a Google Colab tutorial for MMGeneration and we will merge these to our repo later. To train model with python API, you can use the following code:...

You should add `CheckpointHook` and `work_dirs` in cfg at the start of your code. ```python checkpoint_config = dict(interval=5000, by_epoch=False, max_keep_ckpts=10) cfg.checkpoint_config = checkpoint_config cfg.work_dir = 'my-work-space' ```

@oschan77 Seems that your system Cuda version (11.1) and PyTorch Cuda version (11.3) do not match. @asafberreby Please run python `mmgen/utils/collect_env.py` to collect necessary environment information.

You can use this [script](https://github.com/NVlabs/stylegan3/blob/main/dataset_tool.py) to generate image files. We will add this script to our repo later as well.

This is caused by `gcc --version` output being different between Windows and Ubuntu. We will fix this later. A temporary solution is to remove the following lines: https://github.com/open-mmlab/mmgeneration/blob/6bce6ed0315438b3f7955019a2c6a64078cfd0e8/mmgen/utils/collect_env.py#L50-L52

Please provide the command you execute and the full error info.

Can you show how you instantiate `SAGAN_generator` and `SAGAN_discriminator` and the console output? We prefer to build models by `build_model(MODEL_CONFIG)`. Configs for SAGAN can be found in this [URL](https://github.com/open-mmlab/mmgeneration/tree/master/configs/sagan).