Update app.py

app.py

@@ -8,6 +8,12 @@ os.system("unzip blobzip.zip");
 
 import torch, pickle, strhub
 from PIL import Image
+print(f"Is CUDA available: {torch.cuda.is_available()}")
+
+# True
+print(f"CUDA device: {torch.cuda.get_device_name(torch.cuda.current_device())}")
+
+
 # from strhub.data.module import SceneTextDataModule
 # from strhub.models.utils import load_from_checkpoint, parse_model_args
 
@@ -39,21 +45,37 @@ parseq = torch.load('tensor.pt', map_location=torch.device('cpu')).eval()
 
 img_transform = get_transform(parseq.hparams.img_size, augment=True)
 
-img = Image.open('oscqt.jpeg').convert('RGB')
+# img = Image.open('oscqt.jpeg').convert('RGB')
+
+# img = img_transform(img).unsqueeze(0)
+# logits = parseq(img)
+# logits.shape
+
+# # # Greedy decoding
+# pred = logits.softmax(-1)
+# label, confidence = parseq.tokenizer.decode(pred)
+# print('Decoded label = {}'.format(label[0]))
+
+
 
-img = img_transform(img).unsqueeze(0)
-logits = parseq(img)
-logits.shape
+# def greet(name):
+#     return "Hello " + name + "!!"
 
-# # Greedy decoding
-pred = logits.softmax(-1)
-label, confidence = parseq.tokenizer.decode(pred)
-print('Decoded label = {}'.format(label[0]))
+# iface = gr.Interface(fn=greet, inputs="text", outputs="text")
+# iface.launch()
 
 
+def captcha_solver(img):
+  img = img.convert('RGB')
+  img = img_transform(img).unsqueeze(0)
 
-def greet(name):
-    return "Hello " + name + "!!"
+  logits = parseq(img)
+  logits.shape
+  
+  # # Greedy decoding
+  pred = logits.softmax(-1)
+  label, confidence = parseq.tokenizer.decode(pred)
+  return label[0]
 
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
-iface.launch()
+demo = gr.Interface(fn=captcha_solver, inputs=gr.inputs.Image(type="pil"), outputs=gr.outputs.Textbox())
+demo.launch()