akhaliq HF staff commited on
Commit
0d8a6c3
1 Parent(s): def3c63

Create app.py

Browse files
Files changed (1) hide show
  1. app.py +56 -0
app.py ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import torch
2
+ from PIL import Image
3
+ from torchvision import transforms
4
+ import gradio as gr
5
+
6
+ model = torch.hub.load('PingoLH/Pytorch-HarDNet', 'hardnet68', pretrained=True)
7
+ # or any of these variants
8
+ # model = torch.hub.load('PingoLH/Pytorch-HarDNet', 'hardnet85', pretrained=True)
9
+ # model = torch.hub.load('PingoLH/Pytorch-HarDNet', 'hardnet68ds', pretrained=True)
10
+ # model = torch.hub.load('PingoLH/Pytorch-HarDNet', 'hardnet39ds', pretrained=True)
11
+ model.eval()
12
+ torch.hub.download_url_to_file("https://github.com/pytorch/hub/raw/master/images/dog.jpg", "dog.jpg")
13
+
14
+ # sample execution (requires torchvision)
15
+ def inference(input_image):
16
+ preprocess = transforms.Compose([
17
+ transforms.Resize(256),
18
+ transforms.CenterCrop(224),
19
+ transforms.ToTensor(),
20
+ transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
21
+ ])
22
+ input_tensor = preprocess(input_image)
23
+ input_batch = input_tensor.unsqueeze(0) # create a mini-batch as expected by the model
24
+
25
+ # move the input and model to GPU for speed if available
26
+ if torch.cuda.is_available():
27
+ input_batch = input_batch.to('cuda')
28
+ model.to('cuda')
29
+
30
+ with torch.no_grad():
31
+ output = model(input_batch)
32
+ # The output has unnormalized scores. To get probabilities, you can run a softmax on it.
33
+ probabilities = torch.nn.functional.softmax(output[0], dim=0)
34
+ # Download ImageNet labels
35
+ !wget https://raw.githubusercontent.com/pytorch/hub/master/imagenet_classes.txt
36
+ # Read the categories
37
+ with open("imagenet_classes.txt", "r") as f:
38
+ categories = [s.strip() for s in f.readlines()]
39
+ # Show top categories per image
40
+ top5_prob, top5_catid = torch.topk(probabilities, 5)
41
+ result = {}
42
+ for i in range(top5_prob.size(0)):
43
+ result[categories[top5_catid[i]]] = top5_prob[i].item()
44
+ return result
45
+
46
+ inputs = gr.inputs.Image(type='pil')
47
+ outputs = gr.outputs.Label(type="confidences",num_top_classes=5)
48
+
49
+ title = "HARDNET"
50
+ description = "Gradio demo for HARDNET, Harmonic DenseNet pre-trained on ImageNet. To use it, simply upload your image, or click one of the examples to load them. Read more at the links below."
51
+ article = "<p style='text-align: center'><a href='https://arxiv.org/abs/1909.00948'>HarDNet: A Low Memory Traffic Network</a> | <a href='https://github.com/PingoLH/Pytorch-HarDNet'>Github Repo</a></p>"
52
+
53
+ examples = [
54
+ ['dog.jpg']
55
+ ]
56
+ gr.Interface(inference, inputs, outputs, title=title, description=description, article=article, examples=examples, analytics_enabled=False).launch()