Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fixed brute_force_solver and MNIST example #22

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Fixed brute_force_solver and MNIST example #22

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
162 changes: 130 additions & 32 deletions examples/MNIST Addition - Video.ipynb
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,16 +16,24 @@
},
{
"cell_type": "code",
"execution_count": 2,
"execution_count": 1,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Users/hanliying/opt/anaconda3/envs/pytorch/lib/python3.8/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
" from .autonotebook import tqdm as notebook_tqdm\n"
]
},
{
"data": {
"text/plain": [
"<torch._C.Generator at 0x7fe93808f970>"
"<torch._C.Generator at 0x7fabfcfe9030>"
]
},
"execution_count": 2,
"execution_count": 1,
"metadata": {},
"output_type": "execute_result"
}
Expand All @@ -47,7 +55,7 @@
},
{
"cell_type": "code",
"execution_count": 3,
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
Expand Down Expand Up @@ -86,9 +94,92 @@
},
{
"cell_type": "code",
"execution_count": 4,
"execution_count": 3,
"metadata": {},
"outputs": [],
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz\n",
"Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to ./MNIST/MNIST/raw/train-images-idx3-ubyte.gz\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"9913344it [00:01, 6405934.93it/s] \n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Extracting ./MNIST/MNIST/raw/train-images-idx3-ubyte.gz to ./MNIST/MNIST/raw\n",
"\n",
"Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz\n",
"Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to ./MNIST/MNIST/raw/train-labels-idx1-ubyte.gz\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"29696it [00:00, 7006866.09it/s] "
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Extracting ./MNIST/MNIST/raw/train-labels-idx1-ubyte.gz to ./MNIST/MNIST/raw\n",
"\n",
"Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz\n",
"Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to ./MNIST/MNIST/raw/t10k-images-idx3-ubyte.gz\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\n",
"1649664it [00:00, 18619900.63it/s] \n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Extracting ./MNIST/MNIST/raw/t10k-images-idx3-ubyte.gz to ./MNIST/MNIST/raw\n",
"\n",
"Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz\n",
"Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz to ./MNIST/MNIST/raw/t10k-labels-idx1-ubyte.gz\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"5120it [00:00, 2585772.00it/s] "
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Extracting ./MNIST/MNIST/raw/t10k-labels-idx1-ubyte.gz to ./MNIST/MNIST/raw\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\n"
]
}
],
"source": [
"transform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.5,), (0.5,))])\n",
"mnist_train_data = torchvision.datasets.MNIST(root='./MNIST', train=True, download=True,transform=transform)\n",
Expand All @@ -107,7 +198,7 @@
},
{
"cell_type": "code",
"execution_count": 5,
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
Expand Down Expand Up @@ -145,7 +236,7 @@
},
{
"cell_type": "code",
"execution_count": 6,
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
Expand All @@ -162,7 +253,7 @@
},
{
"cell_type": "code",
"execution_count": 7,
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
Expand All @@ -189,7 +280,7 @@
},
{
"cell_type": "code",
"execution_count": 8,
"execution_count": 7,
"metadata": {},
"outputs": [],
"source": [
Expand All @@ -199,7 +290,7 @@
"\n",
"from pylon.constraint import constraint\n",
"\n",
"def enforce_sum(img1, img2, kwargs):\n",
"def enforce_sum(img1, img2, **kwargs):\n",
" return img1 + img2 == kwargs['summation']\n",
"\n",
"\n",
Expand All @@ -215,7 +306,7 @@
},
{
"cell_type": "code",
"execution_count": 9,
"execution_count": 8,
"metadata": {
"scrolled": false
},
Expand All @@ -224,134 +315,136 @@
"name": "stderr",
"output_type": "stream",
"text": [
" 11%|█ | 1003/9000 [00:54<21:14, 6.27it/s]"
" 0%| | 0/9000 [00:00<?, ?it/s]/Users/hanliying/Documents/UCLA/Research/pylon/pylon/examples/../pylon/brute_force_solver.py:50: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
" else torch.tensor(data=self.cond(*sample, **kwargs), dtype=torch.bool) for sample in samples ])\n",
" 11%|█ | 1005/9000 [00:54<25:23, 5.25it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 6282/10000 (63%)\n",
"Test set: Accuracy: 6328/10000 (63%)\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
" 22%|██▏ | 2004/9000 [01:50<25:17, 4.61it/s]"
" 22%|██▏ | 2010/9000 [01:22<11:46, 9.89it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 9333/10000 (93%)\n",
"Test set: Accuracy: 9310/10000 (93%)\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
" 33%|███▎ | 3003/9000 [02:46<20:46, 4.81it/s]"
" 33%|███▎ | 3009/9000 [02:01<13:27, 7.42it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 9581/10000 (96%)\n",
"Test set: Accuracy: 9530/10000 (95%)\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
" 44%|████▍ | 4003/9000 [03:46<15:12, 5.47it/s]"
" 45%|████▍ | 4009/9000 [02:27<11:41, 7.12it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 9630/10000 (96%)\n",
"Test set: Accuracy: 9650/10000 (96%)\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
" 56%|█████▌ | 5002/9000 [04:46<16:40, 4.00it/s]"
" 56%|█████▌ | 5009/9000 [02:57<11:37, 5.72it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 9492/10000 (95%)\n",
"Test set: Accuracy: 9472/10000 (95%)\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
" 67%|██████▋ | 6002/9000 [05:47<12:38, 3.95it/s]"
" 67%|██████▋ | 6007/9000 [03:30<08:38, 5.77it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 9679/10000 (97%)\n",
"Test set: Accuracy: 9681/10000 (97%)\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
" 78%|███████▊ | 7004/9000 [06:49<06:48, 4.88it/s]"
" 78%|███████▊ | 7008/9000 [04:00<04:04, 8.16it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 9685/10000 (97%)\n",
"Test set: Accuracy: 9659/10000 (97%)\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
" 89%|████████▉ | 8004/9000 [07:51<03:23, 4.88it/s]"
" 89%|████████▉ | 8006/9000 [04:26<01:35, 10.37it/s]"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 9716/10000 (97%)\n",
"Test set: Accuracy: 9593/10000 (96%)\n",
"\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 9000/9000 [08:52<00:00, 16.91it/s]\n"
"100%|██████████| 9000/9000 [04:52<00:00, 30.78it/s]\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test set: Accuracy: 9685/10000 (97%)\n",
"Test set: Accuracy: 9669/10000 (97%)\n",
"\n"
]
}
Expand Down Expand Up @@ -392,7 +485,7 @@
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"display_name": "Python 3.8.12 ('pytorch')",
"language": "python",
"name": "python3"
},
Expand All @@ -406,7 +499,12 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.7"
"version": "3.8.12"
},
"vscode": {
"interpreter": {
"hash": "7a719078278c4492d805b55cadf911a94c7633b48007fee7a7c47218b923b9ef"
}
}
},
"nbformat": 4,
Expand Down
2 changes: 1 addition & 1 deletion pylon/brute_force_solver.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -47,7 +47,7 @@ def loss(self, *logits, **kwargs):
# Get all possible decodings
samples = self.all_samples(log_probs)
indices = torch.stack([torch.tensor(data=self.cond(*sample), dtype=torch.bool) if kwargs == {}
else torch.tensor(data=self.cond(*sample), dtype=torch.bool) for sample in samples ])
else torch.tensor(data=self.cond(*sample, **kwargs), dtype=torch.bool) for sample in samples ])

losses = torch.stack([decoding_loss(sample, log_probs) for sample in samples])

Expand Down