search

wchen342 / SketchyGAN

Code for paper "SketchyGAN: Towards Diverse and Realistic Sketch to Image Synthesis"
MIT License
131 stars 32 forks source link

Could you please share the pre-train model? #4

Closed yuqing-liu-dut closed 5 years ago

yuqing-liu-dut commented 6 years ago

Hello, I'm trying to train the model. However it is hard to download the entire dataset and perform training because of the hardware constrain. Meanwhile, when downloading the dataset, there are errors in the zip file. tim 20180522223756 We select all files and download them, unfortunately there are errors from the OneDrive. Could you please share the pre-train model? It will be so thankful.

wchen342 commented 6 years ago

Can you please try downloading them again? I just verified those files and I can download them here with no problems. Maybe you hit the rate limit of OneDrive? About the pre-trained model, I do have old ones but unfortunately they won't work with current version of code. I can still share them if you want, but you need to do some hack to make them work.

yuqing-liu-dut commented 6 years ago

@wchen342 If possible, I want to try the old model. And I will change my tensorflow and other libraries' version to fit it. Meanwhile, I will try to re-download the files from OneDrive. Thank you very much for your kindness and help.

wchen342 commented 6 years ago

I have uploaded an old model here. I also included the code to construct the correct model.

yuqing-liu-dut commented 6 years ago

@wchen342 Thank you very much for your kind help. However, there are problems applying this model. After downloading this model, the name of 2 folders are with not "skgan" but "wgan". I try to rename them to fit the code, there goes an error, however. The error goes as below. qq 20180607225959

When checking the files, I've found that there is no json file in neither ckpt folder nor log folder. So I try to copy the .pys from log folder to src_single. But there is no file with same name. Meanwhile, all the files in log folder are some definitions about layers but not entry.

So could you please tell me how to apply this pre-train model? Thank you very much!

wchen342 commented 6 years ago

I forgot to include the json file: link. The graph and model files are included so you can modify the corresponding files in srcsingle to build the model that fits the checkpoint. The other files, like train, are just a framework for convinience. You can also try to load the graph definition into default graph directly by tf.import_graph_def. inception_score.py has a good example of that.