first commit

.gitattributes

@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+effnet_b2-20%-10epochs.pth filter=lfs diff=lfs merge=lfs -text

app.py

@@ -0,0 +1,50 @@
+
+import gradio as gr 
+import torch
+import os
+import torch.nn as nn
+import torchvision
+from model import create_model
+
+from timeit import default_timer as timer
+from typing import Tuple, Dict 
+
+#################
+title = 'FoodVisionMini - Diljith'
+description = 'Classifies an image of food item into either one of 3 classes : pizza, steak or sushi'
+
+effnetb2, transforms, optimizer, lossFunc = create_model(num_classes = 3)
+effnetb2.load_state_dict(torch.load(f = 'effnet_b2-20%-10epochs.pth'), map_location = torch.device('cpu'))
+
+
+class_names = ['pizza', 'steak', 'sushi']
+
+def predict(img):
+    img = transforms(img).unsqueeze(0) 
+    effnetb2.eval()
+    probs_dict = {}
+    start = timer()
+    with torch.inference_mode():
+        pred_probs = effnetb2(img).softmax(dim = 1)
+        pred_label = pred_probs.argmax(dim = 1)
+
+        probs_dict = {class_names[i] : float(pred_probs[0][i]) for i in range(len(class_names))} 
+    end = timer 
+
+    return probs_dict, end-start 
+
+
+
+example_list = ['examples/' + example for example in os.listdir(examples_path)]
+
+projectApp = gr.Interface(
+    fn = predict,
+    inputs = gr.Image(type = 'pil'),
+    outputs = [gr.Label(num_top_classes = len(class_names), label = 'Predictions'),
+               gr.Number(label = 'Prediction time(s)')],
+    examples = example_list,
+    title = title,
+    description = description
+)
+
+projectApp.launch()

effnet_b2-20%-10epochs.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c33bb61dc620af399b1377d061f311da9eff22899b992bf90ae726f9a60333dc
+size 31281451

model.py

@@ -0,0 +1,26 @@
+
+
+import torch
+import torch.nn as nn
+import torchvision
+
+
+
+def create_model(num_classes, seed = 42):
+    weights = torchvision.models.EfficientNet_B2_Weights.DEFAULT
+    auto_transforms = weights.transforms()
+
+    model = torchvision.models.efficientnet_b2(weights = weights)
+
+    for param in model.parameters():
+        param.requires_grad = False
+
+    model.classifier = nn.Sequential(
+        nn.Dropout(p = 0.3, inplace = True),
+        nn.Linear(in_features = 1408, out_features = num_classes, bias = True)
+    )
+
+    optimizer = torch.optim.Adam(params = model.parameters(), lr = 0.001)
+    lossFunc = nn.CrossEntropyLoss()
+
+    return model, auto_transforms, optimizer, lossFunc

requirements.txt

@@ -0,0 +1,4 @@
+
+torch == 1.12.0
+torchvision == 0.13.0
+gradio == 3.1.4