Spaces:
Build error
Build error
NimaBoscarino
commited on
Commit
•
d015ad6
1
Parent(s):
cb42512
Part 1 demo for Gradio guide
Browse files
app.py
CHANGED
@@ -1,11 +1,8 @@
|
|
1 |
-
import gradio as gr
|
2 |
-
import torch
|
3 |
-
from huggingface_hub import hf_hub_download
|
4 |
from torch import nn
|
5 |
-
from torchvision.utils import save_image
|
6 |
-
|
7 |
|
8 |
class Generator(nn.Module):
|
|
|
|
|
9 |
def __init__(self, nc=4, nz=100, ngf=64):
|
10 |
super(Generator, self).__init__()
|
11 |
self.network = nn.Sequential(
|
@@ -26,19 +23,24 @@ class Generator(nn.Module):
|
|
26 |
output = self.network(input)
|
27 |
return output
|
28 |
|
|
|
|
|
29 |
|
30 |
model = Generator()
|
31 |
weights_path = hf_hub_download('nateraw/cryptopunks-gan', 'generator.pth')
|
32 |
-
model.load_state_dict(torch.load(weights_path, map_location=torch.device('cpu')))
|
33 |
|
|
|
34 |
|
35 |
def predict(seed):
|
|
|
36 |
torch.manual_seed(seed)
|
37 |
-
z = torch.randn(
|
38 |
punks = model(z)
|
39 |
save_image(punks, "punks.png", normalize=True)
|
40 |
return 'punks.png'
|
41 |
|
|
|
42 |
|
43 |
gr.Interface(
|
44 |
predict,
|
@@ -46,8 +48,4 @@ gr.Interface(
|
|
46 |
gr.inputs.Slider(label='Seed', minimum=0, maximum=1000, default=42),
|
47 |
],
|
48 |
outputs="image",
|
49 |
-
|
50 |
-
description="These CryptoPunks do not exist. Generate random punks with an initial seed!",
|
51 |
-
article="<p style='text-align: center'><a href='https://arxiv.org/pdf/1511.06434.pdf'>Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks</a> | <a href='https://github.com/teddykoker/cryptopunks-gan'>Github Repo</a></p>",
|
52 |
-
examples=[[123], [42], [456], [1337]],
|
53 |
-
).launch(cache_examples=True)
|
|
|
|
|
|
|
|
|
1 |
from torch import nn
|
|
|
|
|
2 |
|
3 |
class Generator(nn.Module):
|
4 |
+
# Refer to the link below for explanations about nc, nz, and ngf
|
5 |
+
# https://pytorch.org/tutorials/beginner/dcgan_faces_tutorial.html#inputs
|
6 |
def __init__(self, nc=4, nz=100, ngf=64):
|
7 |
super(Generator, self).__init__()
|
8 |
self.network = nn.Sequential(
|
|
|
23 |
output = self.network(input)
|
24 |
return output
|
25 |
|
26 |
+
from huggingface_hub import hf_hub_download
|
27 |
+
import torch
|
28 |
|
29 |
model = Generator()
|
30 |
weights_path = hf_hub_download('nateraw/cryptopunks-gan', 'generator.pth')
|
31 |
+
model.load_state_dict(torch.load(weights_path, map_location=torch.device('cpu'))) # Use 'cuda' if you have a GPU available
|
32 |
|
33 |
+
from torchvision.utils import save_image
|
34 |
|
35 |
def predict(seed):
|
36 |
+
num_punks = 4
|
37 |
torch.manual_seed(seed)
|
38 |
+
z = torch.randn(num_punks, 100, 1, 1)
|
39 |
punks = model(z)
|
40 |
save_image(punks, "punks.png", normalize=True)
|
41 |
return 'punks.png'
|
42 |
|
43 |
+
import gradio as gr
|
44 |
|
45 |
gr.Interface(
|
46 |
predict,
|
|
|
48 |
gr.inputs.Slider(label='Seed', minimum=0, maximum=1000, default=42),
|
49 |
],
|
50 |
outputs="image",
|
51 |
+
).launch()
|
|
|
|
|
|
|
|