Assignment 1 - Updated The Base Model and Added Comments

README.md

@@ -0,0 +1,8 @@
+# mgmt590-lec1
+
+Code For Answering Questions
+
+In this version, we have fit a different model (xlm-roberta-large-squad2) and have added comments for better readability
+
+References -
+https://huggingface.co/deepset/xlm-roberta-large-squad2

answer.py

@@ -1,12 +1,45 @@
 import pandas as pd
 from transformers.pipelines import pipeline
+from flask import Flask
+from flask import request
+from flask import jsonify
 
-hg_comp = pipeline('question-answering', model="distilbert-base-uncased-distilled-squad", tokenizer="distilbert-base-uncased-distilled-squad")
+# Create my flask app
+app = Flask(__name__)
 
-data = pd.read_csv('examples.csv')
+# Define a handler for the / path, which
+# returns "Hello World"
+@app.route("/")
+def hello_world():
+    return "<p>Hello, World!</p>"
 
-for idx, row in data.iterrows():
-    context = row['context']
-    question = row['question']
-    answer = hg_comp({'question': question, 'context': context})['answer']
-    print(answer)
+# Define a handler for the /answer path, which
+# processes a JSON payload with a question and
+# context and returns an answer using a Hugging
+# Face model.
+@app.route("/answer", methods=['POST'])
+def answer():
+
+    # Get the request body data
+    data = request.json
+
+    # Import model
+    hg_comp = pipeline('question-answering', model="distilbert-base-uncased-distilled-squad", tokenizer="distilbert-base-uncased-distilled-squad")
+
+    # Answer the answer
+    answer = hg_comp({'question': data['question'], 'context': data['context']})['answer']
+
+    # Create the response body.
+    out = {
+            "question": data['question'],
+            "context": data['context'],
+            "answer": answer
+            }
+
+    return jsonify(out)
+
+# Run if running "python answer.py"
+if __name__ == '__main__':
+
+    # Run our Flask app and start listening for requests!
+    app.run(host='0.0.0.0', port=8000, threaded=True)