The Simplest DCGAN Implementation

Overview

DCGAN in TensorLayer

This is the TensorLayer implementation of Deep Convolutional Generative Adversarial Networks. Looking for Text to Image Synthesis ? click here

alt tag

  • ๐Ÿ†• ๐Ÿ”ฅ 2019 May: We just update this project to support TF2 and TL2. Enjoy!
  • ๐Ÿ†• ๐Ÿ”ฅ 2019 May: This project is chosen as the default template of TL projects.

Prerequisites

  • Python3.5 3.6
  • TensorFlow==2.0.0a0 pip3 install tensorflow-gpu==2.0.0a0
  • TensorLayer=2.1.0 pip3 install tensorlayer==2.1.0

Usage

First, download the aligned face images from google or baidu to a data folder.

Second, train the GAN:

$ python train.py

Result on celebA

Comments
  • ValueError: Trying to share variable discriminator/d/h4/lin_sigmoid/W

    ValueError: Trying to share variable discriminator/d/h4/lin_sigmoid/W

    hi, I use: Tensorflow V 1.10 CPU version Tensorlayer V 1.10.1 python 3.6 windows 10

    after run : python main.py I get ValueError

    Traceback (most recent call last): File "c:/Users/USER/Desktop/ืคืจื•ื™ื™ืงื˜ ืชื•ืืจ ืฉื ื™/playgroung/dcgan/main.py", line 166, in tf.app.run() File "C:\Python\3.6\lib\site-packages\tensorflow\python\platform\app.py", line 125, in run _sys.exit(main(argv)) File "c:/Users/USER/Desktop/ืคืจื•ื™ื™ืงื˜ ืชื•ืืจ ืฉื ื™/playgroung/dcgan/main.py", line 59, in main _, d2_logits = discriminator(real_images, is_train=True, reuse=True) File "c:\Users\USER\Desktop\ืคืจื•ื™ื™ืงื˜ ืชื•ืืจ ืฉื ื™\playgroung\dcgan\model.py", line 81, in discriminator W_init = w_init, name='d/h4/lin_sigmoid') File "c:\Users\USER\Desktop\ืคืจื•ื™ื™ืงื˜ ืชื•ืืจ ืฉื ื™\playgroung\dcgan\tensorlayer\layers\core.py", line 926, in init W = tf.get_variable(name='W', shape=(n_in, n_units), initializer=W_init, dtype=LayersConfig.tf_dtype, **W_init_args) File "C:\Python\3.6\lib\site-packages\tensorflow\python\ops\variable_scope.py", line 1467, in get_variable aggregation=aggregation) File "C:\Python\3.6\lib\site-packages\tensorflow\python\ops\variable_scope.py", line 1217, in get_variable aggregation=aggregation) File "C:\Python\3.6\lib\site-packages\tensorflow\python\ops\variable_scope.py", line 527, in get_variable aggregation=aggregation) File "C:\Python\3.6\lib\site-packages\tensorflow\python\ops\variable_scope.py", line 481, in _true_getter aggregation=aggregation) File "C:\Python\3.6\lib\site-packages\tensorflow\python\ops\variable_scope.py", line 853, in _get_single_variable found_var.get_shape())) ValueError: Trying to share variable discriminator/d/h4/lin_sigmoid/W, but specified shape (8192, 1) and found shape (2048, 1).

    opened by motkeg 2
  • Error during download

    Error during download

    Tried twice, same error.

    python download.py celebA
    ./data/img_align_celeba.zip: 25.5kB [12:15, 34.7B/s]Traceback (most recent call last):
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/urllib3/response.py", line 302, in _error_catcher
        yield
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/urllib3/response.py", line 601, in read_chunked
        chunk = self._handle_chunk(amt)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/urllib3/response.py", line 557, in _handle_chunk
        value = self._fp._safe_read(amt)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/http/client.py", line 614, in _safe_read
        raise IncompleteRead(b''.join(s), amt)
    http.client.IncompleteRead: IncompleteRead(5887 bytes read, 26881 more expected)
    
    During handling of the above exception, another exception occurred:
    
    Traceback (most recent call last):
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/requests/models.py", line 745, in generate
        for chunk in self.raw.stream(chunk_size, decode_content=True):
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/urllib3/response.py", line 432, in stream
        for line in self.read_chunked(amt, decode_content=decode_content):
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/urllib3/response.py", line 626, in read_chunked
        self._original_response.close()
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/contextlib.py", line 99, in __exit__
        self.gen.throw(type, value, traceback)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/urllib3/response.py", line 320, in _error_catcher
        raise ProtocolError('Connection broken: %r' % e, e)
    urllib3.exceptions.ProtocolError: ('Connection broken: IncompleteRead(5887 bytes read, 26881 more expected)', IncompleteRead(5887 bytes read, 26881 more expected))
    
    During handling of the above exception, another exception occurred:
    
    Traceback (most recent call last):
      File "download.py", line 164, in <module>
        download_celeb_a('./data')
      File "download.py", line 91, in download_celeb_a
        download_file_from_google_drive(drive_id, save_path)
      File "download.py", line 56, in download_file_from_google_drive
        save_response_content(response, destination)
      File "download.py", line 68, in save_response_content
        unit='B', unit_scale=True, desc=destination):
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/tqdm/_tqdm.py", line 962, in __iter__
        for obj in iterable:
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/requests/models.py", line 748, in generate
        raise ChunkedEncodingError(e)
    requests.exceptions.ChunkedEncodingError: ('Connection broken: IncompleteRead(5887 bytes read, 26881 more expected)', IncompleteRead(5887 bytes read, 26881 more expected))
    
    opened by arisliang 2
  • Traceback error - main.py

    Traceback error - main.py

    Trying to run main.py and I get the following error.

    traceback (most recent call last):
      File "main.py", line 23, in <module>
        flags.DEFINE_integer("train_size", np.inf, "The size of train images [np.inf]")
      File "/anaconda2/lib/python2.7/site-packages/tensorflow/python/platform/flags.py", line 58, in wrapper
        return original_function(*args, **kwargs)
      File "/anaconda2/lib/python2.7/site-packages/absl/flags/_defines.py", line 315, in DEFINE_integer
        DEFINE(parser, name, default, help, flag_values, serializer, **args)
      File "/anaconda2/lib/python2.7/site-packages/absl/flags/_defines.py", line 81, in DEFINE
        DEFINE_flag(_flag.Flag(parser, serializer, name, default, help, **args),
      File "/anaconda2/lib/python2.7/site-packages/absl/flags/_flag.py", line 107, in __init__
        self._set_default(default)
      File "/anaconda2/lib/python2.7/site-packages/absl/flags/_flag.py", line 196, in _set_default
        self.default = self._parse(value)
      File "/anaconda2/lib/python2.7/site-packages/absl/flags/_flag.py", line 169, in _parse
        'flag --%s=%s: %s' % (self.name, argument, e))
    absl.flags._exceptions.IllegalFlagValueError: flag --train_size=inf: Expect argument to be a string or int, found <type 'float'>
    

    Has anyone else encountered this?

    opened by 1140lex 2
  • Error running train.py

    Error running train.py

    Running train.py i get the following error:

    Traceback (most recent call last):
      File "train.py", line 60, in <module>
        train()
      File "train.py", line 44, in train
        grad = tape.gradient(g_loss, G.trainable_weights)
    AttributeError: 'Model' object has no attribute 'trainable_weights'
    

    am i running a wrong version of tensorflow/tensorlayer (2.0.0a0/2.0.0)?

    opened by ghost 1
  • The url for CelebA dose not work now

    The url for CelebA dose not work now

    The url for CelebA dataset in download.py dose not work now because of the dropbox problem. So It's better to change it to the url of google-drive or baidu-drive.

    opened by fangpin 1
  • Minor fixes and improvements

    Minor fixes and improvements

    • Added a mini-batch generator function.
    • Renamed generator and discriminator functions.
    • Removed logits from generator as it was redundant.
    • Changed w_init in both models to tf.glorot_normal_initializer since it has shown to improve convergence of model.
    • Using imageio instead of scipy.misc wherever possible since it is deprecated.
    opened by pskrunner14 0
  • TL compatibility and code readabality

    TL compatibility and code readabality

    • Made code compatible with TL1.10.1 and TF1.10.0.
    • Removed out size and batch_size from DeConv2D layers since they're no longer needed.
    • Replaced tl.act.lrelu with tf.nn.leaky_relu since former is deprecated.
    • Replaced tl.layers.initialize_global_variables with tf.global_variables_initializer since former is deprecated.
    • Changed flag train_size dtype from int to float due to an error.
    • Impoved code readability and removed redundant comments and code.
    • Removed pprint in favour of custom logging the flags since pprint was'nt able to access flag values.
    • Fixed wildcard imports and removed unused ones.
    • Updated main module doc.
    • Removed tensorlayer package from root dir since it is installed on system.

    Note: Fixes #9 and #12.

    opened by pskrunner14 0
  • --sample_size option causes ValueError

    --sample_size option causes ValueError

    When I pass --sample_size 256 as a flag to main.py, I get

    ValueError: Cannot feed value of shape (256, 100) for Tensor 'z_noise:0', which has shape '(64, 100)'

    When I pass --sample_size 256 and --batch_size 256, the code went into an endless loop just saying "Sample images updated!" without running each epoch.

    opened by jkraybill 0
  • Performance issue in /data.py (by P3)

    Performance issue in /data.py (by P3)

    Hello! I've found a performance issue in /data.py: ds.batch(batch_size)(here) should be called before ds.map(_map_fn, num_parallel_calls=4)(here), which could make your program more efficient.

    Here is the tensorflow document to support it.

    Besides, you need to check the function _map_fn called in ds.map(_map_fn, num_parallel_calls=4) whether to be affected or not to make the changed code work properly. For example, if _map_fn needs data with shape (x, y, z) as its input before fix, it would require data with shape (batch_size, x, y, z) after fix.

    Looking forward to your reply. Btw, I am very glad to create a PR to fix it if you are too busy.

    opened by DLPerf 1
  • How to train on existing checkpoint

    How to train on existing checkpoint

    I would like to train on an existing saved checkpoint but it seems the GAN learn from scratches.

    Is there some way to load a existing checkpoint ?

    Thanks in advance.

    opened by Fqlox 0
  • local variable 'z' referenced before assignment

    local variable 'z' referenced before assignment

    I have this error! can you help me, please?

    WARNING: Logging before flag parsing goes to stderr. I0718 14:45:17.243686 140030963623808 tl_logging.py:99] [!] checkpoint exists ... I0718 14:45:17.247771 140030963623808 tl_logging.py:99] [!] samples exists ... W0718 14:45:19.318463 140030963623808 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/data/ops/dataset_ops.py:505: py_func (from tensorflow.python.ops.script_ops) is deprecated and will be removed in a future version. Instructions for updating: tf.py_func is deprecated in TF V2. Instead, there are two options available in V2. - tf.py_function takes a python function which manipulates tf eager tensors instead of numpy arrays. It's easy to convert a tf eager tensor to an ndarray (just call tensor.numpy()) but having access to eager tensors means tf.py_functions can use accelerators such as GPUs as well as being differentiable using a gradient tape. - tf.numpy_function maintains the semantics of the deprecated tf.py_func (it is not differentiable, and manipulates numpy arrays). It drops the stateful argument making all functions stateful.

    I0718 14:45:19.464441 140030963623808 tl_logging.py:99] Input _inputlayer_1: [None, 100] I0718 14:45:19.974161 140030963623808 tl_logging.py:99] Dense dense_1: 8192 No Activation I0718 14:45:20.649731 140030963623808 tl_logging.py:99] Reshape reshape_1 I0718 14:45:20.707117 140030963623808 tl_logging.py:99] BatchNorm batchnorm_1: decay: 0.900000 epsilon: 0.000010 act: relu is_train: False I0718 14:45:20.782280 140030963623808 tl_logging.py:99] DeConv2d deconv2d_1: n_filters: 256 strides: (2, 2) padding: SAME act: No Activation dilation: (1, 1) I0718 14:45:22.796400 140030963623808 tl_logging.py:99] BatchNorm batchnorm2d_1: decay: 0.900000 epsilon: 0.000010 act: relu is_train: False I0718 14:45:22.873876 140030963623808 tl_logging.py:99] DeConv2d deconv2d_2: n_filters: 128 strides: (2, 2) padding: SAME act: No Activation dilation: (1, 1) I0718 14:45:23.076492 140030963623808 tl_logging.py:99] BatchNorm batchnorm2d_2: decay: 0.900000 epsilon: 0.000010 act: relu is_train: False I0718 14:45:23.153287 140030963623808 tl_logging.py:99] DeConv2d deconv2d_3: n_filters: 64 strides: (2, 2) padding: SAME act: No Activation dilation: (1, 1) I0718 14:45:23.227763 140030963623808 tl_logging.py:99] BatchNorm batchnorm2d_3: decay: 0.900000 epsilon: 0.000010 act: relu is_train: False I0718 14:45:23.307783 140030963623808 tl_logging.py:99] DeConv2d deconv2d_4: n_filters: 3 strides: (2, 2) padding: SAME act: tanh dilation: (1, 1) I0718 14:45:23.390201 140030963623808 tl_logging.py:99] Input _inputlayer_2: [None, 64, 64, 3] I0718 14:45:23.456669 140030963623808 tl_logging.py:99] Conv2d conv2d_1: n_filter: 64 filter_size: (5, 5) strides: (2, 2) pad: SAME act: I0718 14:45:23.530629 140030963623808 tl_logging.py:99] Conv2d conv2d_2: n_filter: 128 filter_size: (5, 5) strides: (2, 2) pad: SAME act: No Activation I0718 14:45:23.608061 140030963623808 tl_logging.py:99] BatchNorm batchnorm2d_4: decay: 0.900000 epsilon: 0.000010 act: is_train: False I0718 14:45:23.691769 140030963623808 tl_logging.py:99] Conv2d conv2d_3: n_filter: 256 filter_size: (5, 5) strides: (2, 2) pad: SAME act: No Activation I0718 14:45:23.775271 140030963623808 tl_logging.py:99] BatchNorm batchnorm2d_5: decay: 0.900000 epsilon: 0.000010 act: is_train: False I0718 14:45:23.942049 140030963623808 tl_logging.py:99] Conv2d conv2d_4: n_filter: 512 filter_size: (5, 5) strides: (2, 2) pad: SAME act: No Activation I0718 14:45:24.041154 140030963623808 tl_logging.py:99] BatchNorm batchnorm2d_6: decay: 0.900000 epsilon: 0.000010 act: is_train: False I0718 14:45:24.114703 140030963623808 tl_logging.py:99] Flatten flatten_1: I0718 14:45:24.177696 140030963623808 tl_logging.py:99] Dense dense_2: 1 identity I0718 14:45:29.721007 140030963623808 tl_logging.py:99] [*] Saving TL weights into checkpoint/G.npz I0718 14:45:31.268130 140030963623808 tl_logging.py:99] [*] Saved I0718 14:45:31.274845 140030963623808 tl_logging.py:99] [*] Saving TL weights into checkpoint/D.npz I0718 14:45:32.590787 140030963623808 tl_logging.py:99] [*] Saved

    UnboundLocalError Traceback (most recent call last) in () 58 59 if name == 'main': ---> 60 train()

    in train() 53 D.save_weights('{}/D.npz'.format(flags.checkpoint_dir), format='npz') 54 G.eval() ---> 55 result = G(z) 56 G.train() 57 tl.visualize.save_images(result.numpy(), [num_tiles, num_tiles], '{}/train_{:02d}.png'.format(flags.sample_dir, epoch))

    UnboundLocalError: local variable 'z' referenced before assignment

    opened by emnab 4
  • TypeError when run main.py

    TypeError when run main.py

    Tried both the dcgan folder version of tensorlayer (1.4.5) and installed version of tensorlayer (1.8.1). tensorflow version is 1.6.0.

    $ python main.py 
    /home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
      from ._conv import register_converters as _register_converters
    Traceback (most recent call last):
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/absl/flags/_flag.py", line 166, in _parse
        return self.parser.parse(argument)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/absl/flags/_argument_parser.py", line 152, in parse
        val = self.convert(argument)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/absl/flags/_argument_parser.py", line 268, in convert
        type(argument)))
    TypeError: Expect argument to be a string or int, found <class 'float'>
    
    During handling of the above exception, another exception occurred:
    
    Traceback (most recent call last):
      File "main.py", line 23, in <module>
        flags.DEFINE_integer("train_size", np.inf, "The size of train images [np.inf]")
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/tensorflow/python/platform/flags.py", line 58, in wrapper
        return original_function(*args, **kwargs)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/absl/flags/_defines.py", line 315, in DEFINE_integer
        DEFINE(parser, name, default, help, flag_values, serializer, **args)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/absl/flags/_defines.py", line 81, in DEFINE
        DEFINE_flag(_flag.Flag(parser, serializer, name, default, help, **args),
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/absl/flags/_flag.py", line 107, in __init__
        self._set_default(default)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/absl/flags/_flag.py", line 196, in _set_default
        self.default = self._parse(value)
      File "/home/ly/anaconda3/envs/learning/lib/python3.6/site-packages/absl/flags/_flag.py", line 169, in _parse
        'flag --%s=%s: %s' % (self.name, argument, e))
    absl.flags._exceptions.IllegalFlagValueError: flag --train_size=inf: Expect argument to be a string or int, found <class 'float'>
    
    
    opened by arisliang 5
Releases(0.5.0)
Owner
TensorLayer Community
A neutral open community to promote AI technology.
TensorLayer Community
Labelbox is the fastest way to annotate data to build and ship artificial intelligence applications

Labelbox Labelbox is the fastest way to annotate data to build and ship artificial intelligence applications. Use this github repository to help you s

labelbox 1.7k Dec 29, 2022
StarGAN - Official PyTorch Implementation (CVPR 2018)

StarGAN - Official PyTorch Implementation ***** New: StarGAN v2 is available at https://github.com/clovaai/stargan-v2 ***** This repository provides t

Yunjey Choi 5.1k Jan 04, 2023
Artificial intelligence technology inferring issues and logically supporting facts from raw text

๊ฐœ์š” ๋น„์ •ํ˜• ํ…์ŠคํŠธ๋ฅผ ํ•™์Šตํ•˜์—ฌ ์Ÿ์ ๋ณ„ ์‚ฌ์‹ค๊ณผ ๋…ผ๋ฆฌ์  ๊ทผ๊ฑฐ ์ถ”๋ก ์ด ๊ฐ€๋Šฅํ•œ ์ธ๊ณต์ง€๋Šฅ ์›์ฒœ๊ธฐ์ˆ  Artificial intelligence techno

6 Dec 29, 2021
The codes for the work "Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation"

Swin-Unet The codes for the work "Swin-Unet: Unet-like Pure Transformer for Medical Image Segmentation"(https://arxiv.org/abs/2105.05537). A validatio

869 Jan 07, 2023
A dataset for online Arabic calligraphy

Calliar Calliar is a dataset for Arabic calligraphy. The dataset consists of 2500 json files that contain strokes manually annotated for Arabic callig

ARBML 114 Dec 28, 2022
Code for Generating Disentangled Arguments with Prompts: A Simple Event Extraction Framework that Works

GDAP Code for Generating Disentangled Arguments with Prompts: A Simple Event Extraction Framework that Works Environment Python (verified: v3.8) CUDA

45 Oct 29, 2022
Latex code for making neural networks diagrams

PlotNeuralNet Latex code for drawing neural networks for reports and presentation. Have a look into examples to see how they are made. Additionally, l

Haris Iqbal 18.6k Jan 01, 2023
The aim of the game, as in the original one, is to find a specific image from a group of different images of a person's face

GUESS WHO Main Links: [Github] [App] Related Links: [CLIP] [Celeba] The aim of the game, as in the original one, is to find a specific image from a gr

Arnau - DIMAI 3 Jan 04, 2022
Trafffic prediction analysis using hybrid models - Machine Learning

Hybrid Machine learning Model Clone the Repository Create a new Directory as assests and download the model from the below link Model Link To Start th

1 Feb 08, 2022
The repo contains the code to train and evaluate a system which extracts relations and explanations from dialogue.

The repo contains the code to train and evaluate a system which extracts relations and explanations from dialogue. How do I cite D-REX? For now, cite

Alon Albalak 6 Mar 31, 2022
using STGCN to achieve egg classification task

EEG Classification โ€‚โ€‚The task requires us to classify electroencephalography(EEG) into six categories, including human body, human face, animal body,

4 Jun 13, 2022
[NeurIPS 2020] Code for the paper "Balanced Meta-Softmax for Long-Tailed Visual Recognition"

Balanced Meta-Softmax Code for the paper Balanced Meta-Softmax for Long-Tailed Visual Recognition Jiawei Ren, Cunjun Yu, Shunan Sheng, Xiao Ma, Haiyu

Jiawei Ren 65 Dec 21, 2022
Unsupervised Representation Learning via Neural Activation Coding

Neural Activation Coding This repository contains the code for the paper "Unsupervised Representation Learning via Neural Activation Coding" published

yookoon park 5 May 26, 2022
Julia package for multiway (inverse) covariance estimation.

TensorGraphicalModels TensorGraphicalModels.jl is a suite of Julia tools for estimating high-dimensional multiway (tensor-variate) covariance and inve

Wayne Wang 3 Sep 23, 2022
Get the partition that a file belongs and the percentage of space that consumes

tinos_eisai_sy Get the partition that a file belongs and the percentage of space that consumes (works only with OSes that use the df command) tinos_ei

Konstantinos Patronas 6 Jan 24, 2022
CS50's Introduction to Artificial Intelligence Test Scripts

CS50's Introduction to Artificial Intelligence Test Scripts ๐Ÿคทโ€โ™‚๏ธ What's this? ๐Ÿคทโ€โ™€๏ธ This repository contains Python scripts to automate tests for mos

Jet Kan 2 Dec 28, 2022
The Instructed Glacier Model (IGM)

The Instructed Glacier Model (IGM) Overview The Instructed Glacier Model (IGM) simulates the ice dynamics, surface mass balance, and its coupling thro

27 Dec 16, 2022
You Only Hypothesize Once: Point Cloud Registration with Rotation-equivariant Descriptors

You Only Hypothesize Once: Point Cloud Registration with Rotation-equivariant Descriptors In this paper, we propose a novel local descriptor-based fra

Haiping Wang 80 Dec 15, 2022
Open source simulator for autonomous vehicles built on Unreal Engine / Unity, from Microsoft AI & Research

Welcome to AirSim AirSim is a simulator for drones, cars and more, built on Unreal Engine (we now also have an experimental Unity release). It is open

Microsoft 13.8k Jan 03, 2023
This repo contains the pytorch implementation for Dynamic Concept Learner (accepted by ICLR 2021).

DCL-PyTorch Pytorch implementation for the Dynamic Concept Learner (DCL). More details can be found at the project page. Framework Grounding Physical

Zhenfang Chen 31 Jan 06, 2023