Add Flagging

.gitignore

@@ -1,2 +1,3 @@
 .venv/
-.idea/
+.idea/
+.gradio/

app.py

@@ -1,9 +1,11 @@
+from typing import Any, Sequence
 import gradio as gr
+from gradio import CSVLogger, FlaggingCallback
+from gradio.components import Component
 import torch
 import tensorflow as tf
-from transformers import AutoModelForSequenceClassification, DebertaV2Tokenizer,TFAutoModelForSequenceClassification
 
-USE_TENSORFLOW=True
+USE_TENSORFLOW = True
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 CLASSES = {
@@ -19,17 +21,29 @@ if not USE_TENSORFLOW:
         model.half()
 story = open('story.txt').read().replace("\n\n", "\n").replace("\n", " ").strip()
 
+
 def ask(question):
-    input = tokenizer(story, question, truncation=True, padding=True,return_tensors='tf' if USE_TENSORFLOW else 'pt')
+    input = tokenizer(story, question, truncation=True, padding=True, return_tensors='tf' if USE_TENSORFLOW else 'pt')
     if not USE_TENSORFLOW:
         input = {key: value.to(device) for key, value in input.items()}
-        output=model(**input)
+        output = model(**input)
         prediction = torch.softmax(output.logits, 1).squeeze()
         return {c: round(prediction[i].item(), 3) for c, i in CLASSES.items()}
     else:
-      output=model(input, training=False)
-      prediction = tf.nn.softmax(output.logits, axis=-1).numpy().squeeze()
-      return {c: round(prediction[i], 3) for c, i in CLASSES.items()}
+        output = model(input, training=False)
+        prediction = tf.nn.softmax(output.logits, axis=-1).numpy().squeeze()
+        return {c: round(prediction[i], 3) for c, i in CLASSES.items()}
+
+
+class Flagger(FlaggingCallback):
+    def __init__(self):
+        self.base_logger = CSVLogger()
+
+    def setup(self, components: Sequence[Component], flagging_dir: str):
+        self.base_logger.setup(components=components, flagging_dir=flagging_dir)
+
+    def flag(self, flag_data: list[Any], flag_option: str | None = None, username: str | None = None) -> int:
+        return self.base_logger.flag(flag_data=flag_data, flag_option=flag_option, username=username)
 
 
 gradio = gr.Interface(
@@ -37,6 +51,9 @@ gradio = gr.Interface(
     inputs=[gr.Textbox(value="", label="Your question, as an affirmative sentence:")],
     outputs=[gr.Label(label="Answer", num_top_classes=3)],
     title="The Seagull Story",
+    flagging_mode='manual',
+    flagging_callback=Flagger(),
+    flagging_options=['Yes', 'No', 'Irrelevant'],
     description="“ Albert and Dave find themselves on the pier. They go to a nearby restaurant where Albert orders "
                 "seagull meat. The waiter promptly serves Albert the meal. After taking a bite, he realizes "
                 "something. Albert pulls a gun out of his ruined jacket and shoots himself. ”\n\nWhy did Albert shoot "