initial commit

app.py

@@ -0,0 +1,57 @@
+import gradio as gr
+import os
+import torch
+
+from model import create_vit_model
+from timeit import default_timer as timer
+from typing import Tuple, Dict
+
+# setup class names
+class_names = ['daisy', 'dandelion', 'rose', 'sunflower', 'tulip']
+### model and transforms preparation
+#create vit model
+vit, vit_transforms = create_vit_model(num_classes=5):
+#load saved weights
+vit.load_state_dict(torch.load(f="pretrained_vit_festure_extractor_flower_classification.pth", 
+                               map_location=torch.device('cpu')))
+### predict function
+# create predict function
+def predict(img) -> Tuple[Dict, float]:
+  """transforms and perfroms a prediction on img and returns prediction and time taken"""
+  #start the time
+  start_time = time.time()
+  #transform the target image and add a batch dim
+  img = vit_transforms(img).unsqueeze(0)
+  #put the model in eval mode and turn on inference
+  vit.eval()
+  with torch.inference_mode():
+    # pass the transformed imag thru the model and turn the prediction logits into prediction probabilities
+    pred_probs = torch.softmax(vit(img), dim =1)
+  # create a prediction label and prediction probability
+  pred_labels_and_probs = {class_names[i]: float(pred_probs[0][i]) for i in range(len(class_names))}
+  # calculate prediction time
+  end_time = time.time()
+  pred_time = round(end_time - start_time, 5)
+  # return the prediction dictionary and prediction time
+  return pred_labels_and_probs, pred_time
+### gradio app
+# create title, description and article strings
+title = "flower classification"
+description = "a vit_16 feature extractor computer vision model to classify images of flowers as: daisy, dandelion, rose, sunflower, and tulip"
+article = "created at [flower classification](githhublink to repository)"
+
+# create examples list from examples dir
+example_list = [["examples/" + example] for example in os.listdir("examples")]
+
+# create gradio demo
+demo = gr.Interface(fn=predict, #mapping function from input to output
+                    inputs=gr.Image(type="pil"), # what are the inputs
+                    outputs=[gr.Label(num_top_classes=5, label="predictions"), #what are the outputs
+                             gr.Number(label="prediction time (s)")], #our fn has two outputs, therefore we have 2 outputs
+                    examples = example_list, 
+                    title=title, 
+                    description=description, 
+                    article=article
+                    )
+#launch the demo
+demo.launch()

model.py

@@ -0,0 +1,17 @@
+import torch
+import torchvision
+
+from torch import nn
+def create_vit_model(num_classes: int = 5):
+  # create vit pretrained weights, transforms and model
+  weights = torchvision.models.ViT_B_16_Weights.DEFAULT
+  transforms = weights.transforms()
+  model = torchvision.models.vit_b_16(weights=weights)
+  # freeze all layers in base model
+  for param in model.parameters():
+    param.requires_grad = False
+  # change the classifier head to suit our problem
+  vit.heads = nn.Sequential(nn.Linear(in_features=768,
+                                      out_features=5,
+                                      bias=True))
+  return model, transforms

pretrained_vit_festure_extractor_flower_classification.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d284db415f480342b18b33cb2f7f16c8b54ce588ef0e6b783e7eb47340b0dad0
+size 343276305

requirements.txt

@@ -0,0 +1,3 @@
+torch
+torchvision
+gradio