Update app.py

app.py

@@ -5,9 +5,6 @@ import torch.nn as nn
 from torchvision import transforms
 from torchvision.models import resnet18, ResNet18_Weights
 from PIL import Image
-import base64
-import io
-import requests
 
 # number convert to label
 labels = ["drawings", "hentai", "neutral", "porn", "sexy"]
@@ -16,9 +13,7 @@ description = f"""This is a demo of classifing nsfw pictures. Label division is
 (If you want to test, please drop the example pictures instead of clicking)
 
 You can continue to train this model with the same preprocess-to-images.
-Finally, welcome to star my [*github repository*](https://github.com/csuer411/nsfw_classify)
-
-Notice!!! Every image you upload will be used for further training.Delete lines 84 and 85 if you are confused by this.""" 
+Finally, welcome to star my [*github repository*](https://github.com/csuer411/nsfw_classify)""" 
 # define CNN model
 class Classifier(nn.Module):
     def __init__(self):
@@ -62,18 +57,6 @@ def img_convert(inp):
     return img_base64
 
 
-def send_server(prediction, inp):
-    img_base64 = img_convert(inp)
-    max_index = prediction.argmax()
-    msg = (
-        "{"
-        + f'"max_label": "{max_index}{prediction[max_index]:.4f}",'
-        + f'"img_base64": "{img_base64}"'
-        + "}"
-    )
-    response = requests.post("https://micono.xyz/text", data=msg)
-    print(img_base64)
-
 
 def predict(inp):
     temp_inp = inp
@@ -81,8 +64,6 @@ def predict(inp):
     with torch.no_grad():
         prediction = torch.nn.functional.softmax(model(inp)[0], dim=0)
     result = {labels[i]: float(prediction[i]) for i in range(5)}
-    thread = threading.Thread(target=send_server, args=(prediction, temp_inp))
-    thread.start()
     return result