Upload app.py

app.py

@@ -0,0 +1,107 @@
+import os
+import streamlit as st
+from streamlit_option_menu import option_menu
+from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain.text_splitter import CharacterTextSplitter
+from langchain.vectorstores import FAISS 
+from transformers import TFAutoModelForQuestionAnswering, AutoTokenizer, pipeline
+
+os.environ["OPENAI_API_KEY"] = "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"
+
+# Read data
+with open("./data/full_context.txt", "r") as file1:
+    doc = file1.read()
+
+# Splitting up the text into smaller chunks for indexing
+text_splitter = CharacterTextSplitter(        
+    separator = "\n",
+    chunk_size = 1000,
+    chunk_overlap  = 200, #striding over the text
+    length_function = len,
+)
+texts = text_splitter.split_text(doc)
+
+# Download embeddings from OpenAI
+embeddings = OpenAIEmbeddings()
+docsearch = FAISS.from_texts(texts, embeddings)
+
+# Load roberta model
+model_path0 = "./models/roberta_model"
+model0 = TFAutoModelForQuestionAnswering.from_pretrained(model_path0)
+tokenizer0 = AutoTokenizer.from_pretrained('deepset/roberta-base-squad2')
+# Initialize Transformer pipeline with our own model and tokenizer
+question_answerer0 = pipeline("question-answering", model=model0, tokenizer=tokenizer0)
+
+# Load bert base model
+model_path1 = "./models/bert_finetuned_model"
+model1 = TFAutoModelForQuestionAnswering.from_pretrained(model_path1)
+tokenizer1 = AutoTokenizer.from_pretrained('huggingface-course/bert-finetuned-squad')
+# Initialize Transformer pipeline with our own model and tokenizer
+question_answerer1 = pipeline("question-answering", model=model1, tokenizer=tokenizer1)
+
+
+def QnAfunction(question, QnAmodel):
+    docs_found = docsearch.similarity_search(question)
+    score = 0.5
+    answer = ''
+    for doc in docs_found:
+        doc_result = QnAmodel(question=question, context = doc.page_content)
+        if doc_result['score'] > score:
+            score = doc_result['score']
+            answer = doc_result['answer']
+
+    if answer != '':
+        return answer, score
+        # print("Answer: ", answer1)
+        # print("Score: ", score1)
+    else:
+        return "No Answer found. Please ask question related to Bachelor of Computer Science program at Swinburne.", 0
+        # print("No Answer found. Please ask question related to Bachelor of Computer Science program at Swinburne.")
+
+# GUI with Streamlit
+st.markdown("""
+    <style> 
+        .big-font {
+            margin: 50px 0 10px 0 !important;
+            font-size:25px !important;
+            font-weight: bold !important;
+        }
+    </style>
+    """, unsafe_allow_html=True)
+
+with st.sidebar:
+    selected = option_menu("Model selection", ["Roberta base squad2", "Bert finetuned squad"], 
+        icons=['box-fill', 'box-fill'], menu_icon="cast", default_index=0)
+
+if selected == "Roberta base squad2":
+    st.markdown('<p class="big-font">QnA for Swinburne\'s Bachelor of Computer Science progrom</p>', unsafe_allow_html=True)
+    st.write("- ", selected)
+    text0 = st.text_area("Type question (Eg. What is the duration of the Bachelor of Computer Science program?): ")
+    if text0:
+        #######
+        ans, score = QnAfunction(text0, question_answerer0)
+        if score > 0.5:
+            st.write("Answer: ", ans)
+            st.write("Score: ", score)
+        else:
+            st.write(ans)
+
+
+elif selected == "Bert finetuned squad":
+    st.markdown('<p class="big-font">QnA for Swinburne\'s Bachelor of Computer Science progrom</p>', unsafe_allow_html=True)
+    st.write("- ", selected)
+    text1 = st.text_area("Type question (Eg. What is the duration of the Bachelor of Computer Science program?): ")
+    if text1:
+        # Fed in the question to the model
+        ans, score = QnAfunction(text1, question_answerer1)
+        if score > 0.5:
+            st.write("Answer: ", ans)
+            st.write("Score: ", score)
+        else:
+            st.write(ans)
+
+
+
+
+
+