Upload app.py

app.py

@@ -1,7 +1,56 @@
-import gradio as gr
-
-def greet(name):
-    return "Hello " + name + "!!" + "This is version 2!!!"
-
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
-iface.launch()
+from fastai import *
+from fastcore.all import *
+from fastai.vision.all import *
+import pandas as pd 
+import re
+import gradio as gr
+
+
+def get_x(data_set):
+  return Path(data_set['image_path'])
+
+def get_y(data_set):
+  return data_set['bf_est']
+
+  
+
+def new_splitter(df):
+  # Get the unique values in the 'id' column
+  unique_ids = df['id'].unique()
+
+  # Shuffle the unique values
+  np.random.seed(42)
+  np.random.shuffle(unique_ids)
+
+  # Calculate the number of unique values to be included in the first dataframe
+  num_unique_in_test = int(np.ceil(len(unique_ids) * 0.8))
+
+  # Get the first 'num_unique_in_df1' unique values
+  test_ids = unique_ids[:num_unique_in_test]
+
+  # Get the rows of the original dataframe that contain the 'df1_ids'
+  test = df.index[df['id'].isin(test_ids)].tolist()
+
+  # Get the rest of the rows from the original dataframe
+  valid = df.index[~df['id'].isin(test_ids)].tolist()
+  
+  return test, valid
+
+
+
+learner = load_learner("bf_model.pkl")
+
+def predict_bf(img):
+  # pil_img = PILImage.create(img)
+  return round(float(learner.predict(img)[1]),2)
+
+image = gr.Image(shape=(192,192))
+# label = gr.float()
+intf = gr.Interface(fn =predict_bf, inputs = image, outputs = "number" )
+intf.launch(inline= False, share=True)
+
+# def greet(name):
+#     return "Hello " + name + "!!" + "This is version 2!!!"
+
+# iface = gr.Interface(fn=greet, inputs="text", outputs="text")
+# iface.launch()