Init push

_Deploy_effB2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a76208f15a11b5f6f2a6fcebe2e28408e00dec0741f5af4d0cfb68fe65d14b77
+size 31276413

app.py

@@ -0,0 +1,45 @@
+import os
+from timeit import default_timer as timer
+from typing import Tuple, Dict
+
+import gradio as gr
+import torch
+from model import create_model
+
+
+title = "Food Vision Mini by Edesak"
+desc = "EffitientNetB2 for recognition of Pizza,Steak,Sushi from [Zero To Mastery Course](https://www.udemy.com/course/pytorch-for-deep-learning/)"
+article = "My Github page [Edesak](https://github.com/Edesak)"
+
+class_names = ["pizza", "steak", "sushi"]
+model,transform = create_model()
+model.load_state_dict(torch.load(f="_Deploy_effB2.pth"))
+model.to("cpu")
+example_list = [["Examples/" + example] for example in os.listdir("Examples")]
+
+def predict(img) -> Tuple[Dict, float]:
+    start_timer = timer()
+
+    img = transform(img).unsqueeze(0)
+
+    model.eval()
+    with torch.inference_mode():
+        y = torch.softmax(model(img), dim=1)
+    pred_labels = {class_names[i]: float(y[0][i]) for i in range(len(class_names))}
+    end_time = timer()
+    pred_time = round(end_time - start_timer, 4)
+
+    return pred_labels, pred_time
+
+
+def greet(name):
+    return "Hello " + name + "!"
+
+
+demo = gr.Interface(fn=predict, inputs=gr.Image(type="pil"),
+                    outputs=[gr.Label(num_top_classes=3, label="Predictions"), gr.Number(label="Prediction time (s)")],
+                    title=title,
+                    description=desc,
+                    article=article)
+
+demo.launch()

model.py

@@ -0,0 +1,16 @@
+from torch.nn import Dropout, Linear
+from torchvision.models import EfficientNet_B2_Weights, efficientnet_b2
+import torch
+def create_model():
+    weights = EfficientNet_B2_Weights.DEFAULT
+    model = efficientnet_b2(weights=weights)
+    transform = weights.transforms()
+
+    classifier = torch.nn.Sequential(
+        Dropout(p=0.3, inplace=True),
+        Linear(in_features=1408, out_features=3)
+    )
+    for layer in model.features:
+        layer.requires_grad_(False)
+    model.classifier = classifier
+    return model,transform

requirements.txt

@@ -0,0 +1,3 @@
+gradio==3.32.0
+torch==1.13.0
+torchvision==0.14.0