diff --git a/DCGAN-PyTorch/01-DCGAN-PL.ipynb b/DCGAN-PyTorch/01-DCGAN-PL.ipynb
index d3687e449d70fa5ff984345f1da4f321817e8f06..6647aea04e763dcba088694529ff31b14956fe8b 100644
--- a/DCGAN-PyTorch/01-DCGAN-PL.ipynb
+++ b/DCGAN-PyTorch/01-DCGAN-PL.ipynb
@@ -4,15 +4,30 @@
"cell_type": "markdown",
"metadata": {},
"source": [
- "# GAN using PyTorch Lightning \n",
+ "<img width=\"800px\" src=\"../fidle/img/00-Fidle-header-01.svg\"></img>\n",
"\n",
- "See : \n",
- "- https://pytorch-lightning.readthedocs.io/en/stable/notebooks/lightning_examples/basic-gan.html\n",
- "- https://www.assemblyai.com/blog/pytorch-lightning-for-dummies/\n",
+ "# <!-- TITLE --> [SHEEP3] - A DCGAN to Draw a Sheep, with Pytorch Lightning\n",
+ "<!-- DESC --> Episode 1 : Draw me a sheep, revisited with a DCGAN, writing in Pytorch Lightning\n",
+ "<!-- AUTHOR : Jean-Luc Parouty (CNRS/SIMaP) -->\n",
"\n",
+ "## Objectives :\n",
+ " - Build and train a DCGAN model with the Quick Draw dataset\n",
+ " - Understanding DCGAN\n",
"\n",
- "Note : Need \n",
- "```pip install ipywidgets lightning tqdm```"
+ "The [Quick draw dataset](https://quickdraw.withgoogle.com/data) contains about 50.000.000 drawings, made by real people... \n",
+ "We are using a subset of 117.555 of Sheep drawings \n",
+ "To get the dataset : [https://github.com/googlecreativelab/quickdraw-dataset](https://github.com/googlecreativelab/quickdraw-dataset) \n",
+ "Datasets in numpy bitmap file : [https://console.cloud.google.com/storage/quickdraw_dataset/full/numpy_bitmap](https://console.cloud.google.com/storage/quickdraw_dataset/full/numpy_bitmap) \n",
+ "Sheep dataset : [https://storage.googleapis.com/quickdraw_dataset/full/numpy_bitmap/sheep.npy](https://storage.googleapis.com/quickdraw_dataset/full/numpy_bitmap/sheep.npy) (94.3 Mo)\n",
+ "\n",
+ "\n",
+ "## What we're going to do :\n",
+ "\n",
+ " - Have a look to the dataset\n",
+ " - Defining a GAN model\n",
+ " - Build the model\n",
+ " - Train it\n",
+ " - Have a look of the results"
]
},
{
@@ -33,6 +48,7 @@
"source": [
"import os\n",
"import sys\n",
+ "import shutil\n",
"\n",
"import numpy as np\n",
"import torch\n",
@@ -82,7 +98,7 @@
"generator_class = 'Generator_2'\n",
"discriminator_class = 'Discriminator_2' \n",
" \n",
- "scale = .01\n",
+ "scale = .005\n",
"epochs = 5\n",
"batch_size = 32\n",
"num_img = 36\n",
@@ -92,6 +108,25 @@
"data_shape = (28,28,1)"
]
},
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "#### Cleaning"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# You can comment these lines to keep each run...\n",
+ "shutil.rmtree(f'{run_dir}/figs', ignore_errors=True)\n",
+ "shutil.rmtree(f'{run_dir}/models', ignore_errors=True)\n",
+ "shutil.rmtree(f'{run_dir}/tb_logs', ignore_errors=True)"
+ ]
+ },
{
"cell_type": "markdown",
"metadata": {},
@@ -168,9 +203,19 @@
},
"outputs": [],
"source": [
+ "# ---- A little piece of black magic to instantiate Generator and Discriminator from their class names\n",
+ "#\n",
+ "def get_classByName(class_name, **args):\n",
+ " module=sys.modules['__main__']\n",
+ " class_ = getattr(module, class_name)\n",
+ " instance_ = class_(**args)\n",
+ " return instance_\n",
+ "\n",
+ "# ----Get it, and play with them\n",
+ "#\n",
"print('\\nInstantiation :\\n')\n",
- "generator = Generator_2(latent_dim=latent_dim, data_shape=data_shape)\n",
- "discriminator = Discriminator_2(latent_dim=latent_dim, data_shape=data_shape)\n",
+ "generator = get_classByName( generator_class, latent_dim=latent_dim, data_shape=data_shape)\n",
+ "discriminator = get_classByName( discriminator_class, latent_dim=latent_dim, data_shape=data_shape)\n",
"\n",
"print('\\nFew tests :\\n')\n",
"z = torch.randn(batch_size, latent_dim)\n",
@@ -273,7 +318,6 @@
"\n",
"trainer = Trainer(\n",
" accelerator = \"auto\",\n",
- "# devices = 1 if torch.cuda.is_available() else None, # limiting got iPython runs\n",
" max_epochs = epochs,\n",
" callbacks = [callback_progressBar, callback_checkpoints],\n",
" log_every_n_steps = batch_size,\n",
@@ -284,10 +328,12 @@
]
},
{
+ "attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
- "## Step 6 - Reload a checkpoint"
+ "## Step 6 - Reload our best model\n",
+ "Note : "
]
},
{
@@ -296,7 +342,7 @@
"metadata": {},
"outputs": [],
"source": [
- "gan = GAN.load_from_checkpoint('./run/SHEEP3/models/last-v1.ckpt')"
+ "gan = GAN.load_from_checkpoint('./run/SHEEP3/models/bestModel.ckpt')"
]
},
{
@@ -307,8 +353,6 @@
"source": [
"nb_images = 32\n",
"\n",
- "# z = np.random.normal(size=(nb_images,latent_dim))\n",
- "\n",
"z = torch.randn(nb_images, latent_dim)\n",
"print('z size : ',z.size())\n",
"\n",
@@ -325,14 +369,17 @@
"execution_count": null,
"metadata": {},
"outputs": [],
- "source": []
+ "source": [
+ "fidle.end()"
+ ]
},
{
- "cell_type": "code",
- "execution_count": null,
+ "cell_type": "markdown",
"metadata": {},
- "outputs": [],
- "source": []
+ "source": [
+ "---\n",
+ "<img width=\"80px\" src=\"../fidle/img/00-Fidle-logo-01.svg\"></img>"
+ ]
}
],
"metadata": {
diff --git a/DCGAN-PyTorch/modules/Discriminators.py b/DCGAN-PyTorch/modules/Discriminators.py
index 86c2f706115d86ef5ce835d4d6b61ab750107c8e..2458a61dc7bf036f22227634001eafe702de3a9c 100644
--- a/DCGAN-PyTorch/modules/Discriminators.py
+++ b/DCGAN-PyTorch/modules/Discriminators.py
@@ -35,7 +35,6 @@ class Discriminator_1(nn.Module):
)
def forward(self, img):
- # img_flat = img.view(img.size(0), -1)
validity = self.model(img)
return validity
@@ -79,7 +78,7 @@ class Discriminator_2(nn.Module):
)
def forward(self, img):
- img_nchw = img.permute(0, 3, 1, 2) # from NHWC to NCHW
+ img_nchw = img.permute(0, 3, 1, 2) # reformat from NHWC to NCHW
validity = self.model(img_nchw)
return validity
\ No newline at end of file
diff --git a/DCGAN-PyTorch/modules/Generators.py b/DCGAN-PyTorch/modules/Generators.py
index 70e523d747187daf5cd1acf38d9ea32cd2f44cad..9b104d579469f51dfda08b1332c9b100b6fddaa4 100644
--- a/DCGAN-PyTorch/modules/Generators.py
+++ b/DCGAN-PyTorch/modules/Generators.py
@@ -67,14 +67,14 @@ class Generator_2(nn.Module):
nn.Linear(latent_dim, 7*7*64),
nn.Unflatten(1, (64,7,7)),
- nn.UpsamplingNearest2d( scale_factor=2 ),
- # nn.UpsamplingBilinear2d( scale_factor=2 ),
+ # nn.UpsamplingNearest2d( scale_factor=2 ),
+ nn.UpsamplingBilinear2d( scale_factor=2 ),
nn.Conv2d( 64,128, (3,3), stride=(1,1), padding=(1,1) ),
nn.ReLU(),
nn.BatchNorm2d(128),
- nn.UpsamplingNearest2d( scale_factor=2 ),
- # nn.UpsamplingBilinear2d( scale_factor=2 ),
+ # nn.UpsamplingNearest2d( scale_factor=2 ),
+ nn.UpsamplingBilinear2d( scale_factor=2 ),
nn.Conv2d( 128,256, (3,3), stride=(1,1), padding=(1,1)),
nn.ReLU(),
nn.BatchNorm2d(256),
@@ -85,9 +85,10 @@ class Generator_2(nn.Module):
)
def forward(self, z):
- img = self.model(z)
- img = img.view(img.size(0), *self.img_shape) # batch_size x 1 x W x H => batch_size x W x H x 1
- return img
+ img_nchw = self.model(z)
+ img_nhwc = img_nchw.permute(0, 2, 3, 1) # reformat from NCHW to NHWC
+ # img = img.view(img.size(0), *self.img_shape) # reformat from NCHW to NHWC
+ return img_nhwc