Upload 8 files

ConvNeXt_Tiny_101classes_20_10epochs.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:930ce0c2c48294e4cf8215c27a1ec1799593fc2c42363441a7063adf499462df
+size 111662445

app.py

@@ -0,0 +1,51 @@
+
+import gradio as gr
+import os
+import torch
+
+from model import create_model
+from timeit import default_timer as timer
+
+
+with open("class_names.txt", "r") as f: 
+    class_names = [class_name.strip() for class_name in f.readlines()]
+    
+
+model, model_transforms = create_model(num_classes=len(class_names))
+model.load_state_dict(torch.load("ConvNeXt_Tiny_101classes_20_10epochs.pth")) # load -> load_state_dict
+model = model.to('cpu') # Place the model on the CPU
+
+
+def predict(img):
+    
+    time_start = timer()
+        
+    weights = torchvision.models.ConvNeXt_Tiny_Weights.DEFAULT 
+    transform_convnext_tiny = weights.transforms()
+    img_tensor = transform_convnext_tiny(img).unsqueeze(dim=0) # [Channels, Height, Width] -> [Batch_size, Channels, Height, Width]
+
+    model.eval()
+    with torch.inference_mode():
+        predicted_probs = model(img_tensor).softmax(dim=1)
+    
+    # Class name & predicted probability for each class (required by Gradio)
+    pred_labels_probs = {}
+    for i in range(len(CLASS_NAMES)):
+        pred_labels_probs[CLASS_NAMES[i]] = float(predicted_probs[0][i])
+
+    return pred_labels_probs, round(timer() - time_start, 5)
+
+
+app = gr.Interface(fn=predict, # mapping function for [ input -> output ]
+                   
+                   inputs=gr.Image(type="pil"), # Input data
+                   outputs=[gr.Label(num_top_classes=3, label="Predictions"), # Output data (fn function's return values)
+                            gr.Number(label="Inference time (s)")], 
+                   
+                   examples=[["examples/" + example] for example in os.listdir("examples")]
+                   
+                   title='ConvNeXt_Food101',
+                   description='A ConvNext CV model to classify 101 foods',
+                   article='Model trained on 150 images per class')
+
+app.launch()

class_names.txt

@@ -0,0 +1,101 @@
+apple_pie
+baby_back_ribs
+baklava
+beef_carpaccio
+beef_tartare
+beet_salad
+beignets
+bibimbap
+bread_pudding
+breakfast_burrito
+bruschetta
+caesar_salad
+cannoli
+caprese_salad
+carrot_cake
+ceviche
+cheese_plate
+cheesecake
+chicken_curry
+chicken_quesadilla
+chicken_wings
+chocolate_cake
+chocolate_mousse
+churros
+clam_chowder
+club_sandwich
+crab_cakes
+creme_brulee
+croque_madame
+cup_cakes
+deviled_eggs
+donuts
+dumplings
+edamame
+eggs_benedict
+escargots
+falafel
+filet_mignon
+fish_and_chips
+foie_gras
+french_fries
+french_onion_soup
+french_toast
+fried_calamari
+fried_rice
+frozen_yogurt
+garlic_bread
+gnocchi
+greek_salad
+grilled_cheese_sandwich
+grilled_salmon
+guacamole
+gyoza
+hamburger
+hot_and_sour_soup
+hot_dog
+huevos_rancheros
+hummus
+ice_cream
+lasagna
+lobster_bisque
+lobster_roll_sandwich
+macaroni_and_cheese
+macarons
+miso_soup
+mussels
+nachos
+omelette
+onion_rings
+oysters
+pad_thai
+paella
+pancakes
+panna_cotta
+peking_duck
+pho
+pizza
+pork_chop
+poutine
+prime_rib
+pulled_pork_sandwich
+ramen
+ravioli
+red_velvet_cake
+risotto
+samosa
+sashimi
+scallops
+seaweed_salad
+shrimp_and_grits
+spaghetti_bolognese
+spaghetti_carbonara
+spring_rolls
+steak
+strawberry_shortcake
+sushi
+tacos
+takoyaki
+tiramisu
+tuna_tartare
+waffles

model.py

@@ -0,0 +1,23 @@
+
+import torch
+import torchvision
+
+from torch import nn
+
+
+def create_model(num_classes=101, seed=42):
+    
+    weights = torchvision.models.ConvNeXt_Tiny_Weights.DEFAULT # .DEFAULT = best available weights on ImageNet
+    transforms = weights.transforms()
+    model = torchvision.models.convnext_tiny(weights=weights)
+
+    # Sequential (features)
+    for param in model.features.parameters(): 
+        param.requires_grad = False # "requires" "grad"ient-descent
+
+    # Sequential (classifier)
+    torch.manual_seed(seed)
+    model.classifier[-1] = nn.Linear(in_features=model.classifier[-1].in_features, 
+                                        out_features=num_classes)
+    
+    return model, transforms

requirements.txt

@@ -0,0 +1,4 @@
+
+torch==1.13.0
+torchvision==0.14.0
+gradio==3.12.0