first commit

EfficientNet_B2_FT.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9b7c644cec4eb5f7a1fbcd1b81c496ee2a6b0cea5af7c1e86b9223d96d92041
+size 31318643

app.py

@@ -0,0 +1,83 @@
+import gradio as gr
+import os
+import torch
+
+from model import create_effnetb2_model
+from timeit import default_timer as timer
+from typing import Tuple, Dict
+
+# Setup class names
+class_names = ['art_nouveau',
+ 'baroque',
+ 'expressionism',
+ 'impressionism',
+ 'post_impressionism',
+ 'realism',
+ 'renaissance',
+ 'romanticism',
+ 'surrealism',
+ 'ukiyo_e']
+
+### 2. Model and transforms preparation ###
+
+# Create EfficientNet_B2 model
+EfficientNetB2_model, EfficientNetB2_transforms = create_effnetb2_model(num_classes=10,is_TrivialAugmentWide=False)
+
+# Load saved weights
+EfficientNetB2_model.load_state_dict(
+    torch.load(
+        f="EfficientNet_B2_FT.pth",
+        map_location=torch.device("cpu"),  # load to CPU
+    )
+)
+
+### 3. Classifier function ###
+
+# Create Classifier function
+def classifier(img) -> Tuple[Dict, float]:
+    """Transforms and performs a prediction on img and returns prediction and time taken.
+    """
+    # Start the timer
+    start_time = timer()
+    
+    # Transform the target image and add a batch dimension
+    img = EfficientNetB2_transforms(img).unsqueeze(0)
+    
+    # Put model into evaluation mode and turn on inference mode
+    EfficientNetB2_model.eval()
+    with torch.inference_mode():
+        # Pass the transformed image through the model and turn the prediction logits into prediction probabilities
+        pred_probs = torch.softmax(EfficientNetB2_model(img), dim=1)
+    
+    # Create a prediction label and prediction probability dictionary for each prediction class (this is the required format for Gradio's output parameter)
+    pred_labels_and_probs = {class_names[i]: float(pred_probs[0][i]) for i in range(len(class_names))}
+    
+    # Calculate the prediction time
+    pred_time = round(timer() - start_time, 5)
+    
+    # Return the prediction dictionary and prediction time 
+    return pred_labels_and_probs, pred_time
+
+### 4. Gradio app ###
+
+# Create title, description and article strings
+title = "Art Classification 🖼️🎨🖌️"
+description = "An EfficientNetB2 computer vision model to classify Artworks."
+article = "Created with Pytorch."
+
+# Create examples list from "examples/" directory
+example_list = [["examples/" + example] for example in os.listdir("examples")]
+
+# Create the Gradio demo
+demo = gr.Interface(fn=classifier, # mapping function from input to output
+                    inputs=gr.Image(type="pil"), 
+                    outputs=[gr.Label(num_top_classes=3, label="Predictions"), #1st output: pred_probs
+                             gr.Number(label="Prediction time (s)")], #2nd output; pred_time
+                    # Create examples list from "examples/" directory
+                    examples=example_list, 
+                    title=title,
+                    description=description,
+                    article=article)
+
+# Launch the demo!
+demo.launch()

model.py

@@ -0,0 +1,47 @@
+import torch
+import torchvision
+
+from torch import nn
+
+def create_effnetb2_model(num_classes:int=10,
+                          seed:int=42,
+                          is_TrivialAugmentWide = True,
+                          freeze_layers=True):
+    """Creates an EfficientNetB2 feature extractor model and transforms.
+
+    Args:
+        num_classes (int, optional): number of classes in the classifier head. Defaults to 10.
+        seed (int, optional): random seed value. Defaults to 42.
+        is_TrivialAugmentWide (boolean): Artificially increase the diversity of a training dataset
+                                      with data augmentation, default = True
+
+    Returns:
+        effnetb2_model (torch.nn.Module): EffNetB2 feature extractor model.
+        effnetb2_transforms (torchvision.transforms): EffNetB2 image transforms.
+    """
+    # 1, 2, 3. Create EffNetB2 pretrained weights, transforms and model
+    weights = torchvision.models.EfficientNet_B2_Weights.DEFAULT
+    effnetb2_transforms = weights.transforms()
+
+    if is_TrivialAugmentWide:
+      effnetb2_transforms = torchvision.transforms.Compose([
+          torchvision.transforms.TrivialAugmentWide(),
+          effnetb2_transforms,
+      ])
+
+
+    effnetb2_model = torchvision.models.efficientnet_b2(weights=weights)
+
+    # 4. Freeze all layers in base model
+    if freeze_layers:
+        for param in effnetb2_model.parameters():
+            param.requires_grad = False
+
+    # 5. Change classifier head with random seed for reproducibility
+    torch.manual_seed(seed)
+    effnetb2_model.classifier = nn.Sequential(
+        nn.Dropout(p=0.3, inplace=True),
+        nn.Linear(in_features=1408, out_features=num_classes),
+    )
+
+    return effnetb2_model, effnetb2_transforms

requirements.txt

@@ -0,0 +1,3 @@
+torch==2.0.0
+torchvision==0.15.1
+gradio==4.10.0