carpedm20 / carpedm20/DCGAN-tensorflow

wrong size of test_arange images when generating with an existing model

Open
#93 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
JavaScript
Stars
7.2k
Forks
2.6k
PR merge metrics
No merged PRs in 30d

Description

I'm working with my own images. Just as a test, I'm using a set of 55 RGB, 256x256 PNG images.
If I run main.py with --is_train, the size of the test_arange samples is around 4.8MB, which is acceptable as the compression is lossless. However, if I run main.py, after the training round, without --is_train, the samples have size 494KB or less. Also, the images are almost unreadable. I wonder if something is wrong.
Here is the command I run:

`python main.py --dataset=mydata --input_height=256 --input_width=256 --output_height=256 --output_width=256 --c_dim=3 --input_fname_pattern="*.png" --batch_size=32 --is_train`

Any help is welcome.
Thank you.

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.