pix2pix.git - Unnamed repository; edit this file 'description' to name the repository.

Age	Commit message (Collapse)	Author
2018-06-07	augment paths	Jules Laplace

2018-06-07	augment paths	Jules Laplace

2018-04-18	k	Jules Laplace

2018-04-18	setting up recursive dataset	Jules Laplace

2018-04-15	fix memory issue in the test_model.py	junyanz

2018-03-27	code refactoring	junyanz

2018-02-18	fix test_model & add timer for data loader & rename identity loss	junyanz

2018-02-10	fix cuda issue	junyanz

2018-02-09	cuda(...) inplace operation	junyanz

2018-02-09	remove unused variable old_lr	junyanz

2018-02-09	code reformatting	junyanz

2018-01-13	fix resize_ issue #170	SsnL

2017-12-15	Update networks.py	guopzhao

2017-12-10	merged conflicts	Taesung Park

2017-12-01	changed the gain of xavier initialization from 1 to 0.02. implemented ↵	Taesung Park
	serial_batches option in unaligned dataset
2017-11-27	fixed a bug in initialization of weights	Taesung Park

2017-11-09	update	SsnL

2017-11-08	working version with handwritten GAN loss. Shift value can be changed	Taesung Park

2017-11-05	fix typo in the cyclegan	junyanz

2017-11-04	fix small issues	junyanz

2017-11-04	add update_html_freq flag	junyanz

2017-11-03	util.tensor2im takes data, not Variable	Robert M Ochshorn

2017-11-02	Merge pull request #130 from LambdaWill/patch-1	Jun-Yan Zhu
	Compatible with the latest version.
2017-11-02	gpu memory leaks	Jean-Philippe Mercier

2017-10-24	Update base_model.py	LambdaWill

2017-10-21	Update networks.py	LambdaWill
	fix typo
2017-10-21	Compatible with the latest version.	LambdaWill
	Since the commit `Change device_id to device in python land #3133`, keyword `device_id` has been changed to `device`
2017-10-19	fix learning rate	junyanz

2017-10-19	update learning rate policy	junyanz

2017-10-19	fix small issue in learning rate policy	junyanz

2017-10-10	fix identity option conflict error, and also not specifying init_type in ↵	Taesung Park
	define_D
2017-10-07	Update cycle_gan_model.py	Jun-Yan Zhu

2017-10-06	fix minor bugs	junyanz

2017-10-06	support nc=1, add new leaerning rate policy and new initialization	junyanz

2017-09-23	Update networks.py	Jun-Yan Zhu

2017-09-22	change the order of concanated channels	LambdaWill
	To be consistent with Cyclegan(Torch), the latter half channels should receive the data from the skip connections.
2017-09-04	Update networks.py	Hungryof

2017-08-26	fix the unaligned dataset	junyanz

2017-08-22	initialized bias of conv to zero,	hazirbas
	deactivated bias for the conv layers which are followed by batch normalization.
2017-07-17	changed the default option to use dropout	Taesung Park

2017-07-05	add reflection padding layer	junyanz

2017-07-03	fix instancenorm & batchnorm	junyanz

2017-06-13	predict single image	junyanz

2017-06-12	update dataset mode	junyanz

2017-05-02	use InstanceNorm2d from the latest pytorch	junyanz

2017-04-27	changed the name of the loaded network from net_G_A to net_G	Taesung Park

2017-04-27	1. Added one_direction_test_model that generates the outputs in only one ↵	Taesung Park
	direction 2. Changed the option naming from ntrain to max_dataset_size
2017-04-23	fix which_direction bug	junyanz

2017-04-22	add comment for python 2/3	junyanz

2017-04-22	add dropout option for G	junyanz