fjxmlzn / pacgan Goto Github PK
View Code? Open in Web Editor NEW[NeurIPS 2018] [JSAIT] PacGAN: The power of two samples in generative adversarial networks
Home Page: https://arxiv.org/abs/1712.04086
License: MIT License
[NeurIPS 2018] [JSAIT] PacGAN: The power of two samples in generative adversarial networks
Home Page: https://arxiv.org/abs/1712.04086
License: MIT License
I'm trying to replicate this work for a research I am working on, but there´s been multiple problems that I am trying to fix, but there seems to be no end to it.
Most of them seem to be caused by environment incompatibilities.
here are some of the latest I've found:
Problems downloading datasets
PacGAN\stacked_MNIST_experiments\VEEGAN_experiment\model.py", line 156, in train
writer.writeheader()
...
TypeError: a bytes-like object is required, not 'str'
PacGAN\stacked_MNIST_experiments\VEEGAN_experiment\model.py", line 200, in train
}, **feed_dict_z))
TypeError: keywords must be strings
PacGAN\stacked_MNIST_experiments\VEEGAN_experiment\model.py", line 281, in log_metrics
if len(results) + len(new_results) > self.num_test_sample:
TypeError: object of type 'zip' has no len()
PacGAN\stacked_MNIST_experiments\VEEGAN_experiment\utils.py", line 67, in imsave
return scipy.misc.imsave(path, image)
AttributeError: module 'scipy.misc' has no attribute 'imsave'
PacGAN\stacked_MNIST_experiments\VEEGAN_experiment\model.py", line 297, in log_metrics
p[0:num_mode] = map.values()
TypeError: float() argument must be a string or a number, not 'dict_values'
Is there any conda environment.yml or requirements.txt file one can use to replicate the original environment?
Or probably a newer/alternative repository for this work?
Just in case, attached you´ll find the last version of my requirements.txt file for the stacked MNIST experiments.
requirements.txt
No errors, no warning. When I tried to reproduce the celebA results, the program exits in two seconds. Does anyone have the same issue?
Here is the log.
(pacgan) ➜ celeba_experiments git:(master) ✗ python main.py --config="./configs/CelebA_128x128_N2M2S64.yaml"
2019-11-23 21:02:41,303 INFO gpu0 receives task packing_num-1,run-0,
2019-11-23 21:02:41,304 INFO gpu0 skips task packing_num-1,run-0,
2019-11-23 21:02:41,305 INFO gpu0 receives task packing_num-2,run-0,
2019-11-23 21:02:41,306 INFO gpu0 skips task packing_num-2,run-0,
2019-11-23 21:02:41,307 INFO gpu0 receives task packing_num-3,run-0,
2019-11-23 21:02:41,307 INFO gpu0 skips task packing_num-3,run-0,
2019-11-23 21:02:41,308 INFO gpu0 receives task packing_num-4,run-0,
2019-11-23 21:02:41,309 INFO gpu0 skips task packing_num-4,run-0,
2019-11-23 21:02:41,310 INFO gpu0 receives task packing_num-1,run-1,
2019-11-23 21:02:41,310 INFO gpu0 skips task packing_num-1,run-1,
2019-11-23 21:02:41,311 INFO gpu0 receives task packing_num-2,run-1,
2019-11-23 21:02:41,311 INFO gpu0 skips task packing_num-2,run-1,
2019-11-23 21:02:41,312 INFO gpu0 receives task packing_num-3,run-1,
2019-11-23 21:02:41,312 INFO gpu0 skips task packing_num-3,run-1,
2019-11-23 21:02:41,313 INFO gpu0 receives task packing_num-4,run-1,
2019-11-23 21:02:41,313 INFO gpu0 skips task packing_num-4,run-1,
2019-11-23 21:02:41,314 INFO gpu0 finished
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.