exam app

app.py

@@ -0,0 +1,81 @@
+import streamlit as st
+from phi.assistant import Assistant
+from phi.document.reader.pdf import PDFReader
+from phi.utils.log import logger
+from assistant import get_groq_assistant
+import io
+
+st.set_page_config(
+    page_title="Test Corrector Model"
+)
+st.title("Test Corrector Model")
+st.markdown("##### Upload Model Answer and Student Answer PDFs to get the grades")
+
+def restart_assistant():
+    st.session_state["assistant"] = None
+    st.session_state["assistant_run_id"] = None
+    st.rerun()
+
+def main():
+    # Get LLM model
+    llm_model = st.sidebar.selectbox("Select LLM", options=["llama3-70b-8192", "llama3-8b-8192", "mixtral-8x7b-32768"])
+    embeddings_model = st.sidebar.selectbox("Select Embeddings", options=["nomic-embed-text", "text-embedding-3-small"])
+
+    if "llm_model" not in st.session_state:
+        st.session_state["llm_model"] = llm_model
+    elif st.session_state["llm_model"] != llm_model:
+        st.session_state["llm_model"] = llm_model
+        restart_assistant()
+
+    if "embeddings_model" not in st.session_state:
+        st.session_state["embeddings_model"] = embeddings_model
+    elif st.session_state["embeddings_model"] != embeddings_model:
+        st.session_state["embeddings_model"] = embeddings_model
+        restart_assistant()
+
+#type annotation in Python. It indicates that the variable assistant is expected to be an instance of the Assistant class.
+    assistant: Assistant 
+    if "assistant" not in st.session_state or st.session_state["assistant"] is None:
+        logger.info(f"---*--- Creating {llm_model} Assistant ---*---")
+        assistant = get_groq_assistant(llm_model=llm_model, embeddings_model=embeddings_model)
+        st.session_state["assistant"] = assistant
+    else:
+        assistant = st.session_state["assistant"]
+
+    try:
+        st.session_state["assistant_run_id"] = assistant.create_run()
+    except Exception:
+        st.warning("Could not create assistant, is the database running?")
+        return
+
+    # Upload model answer PDF
+    model_answer_pdf = st.file_uploader("Upload Model Answer PDF", type="pdf")
+    model_answers = []
+    if model_answer_pdf:
+        reader = PDFReader()
+        model_documents = reader.read(io.BytesIO(model_answer_pdf.read()))
+        model_answers = [doc.content for doc in model_documents]
+
+    # Upload student answer PDF
+    student_answer_pdf = st.file_uploader("Upload Student Answer PDF", type="pdf")
+    student_answers = []
+    if student_answer_pdf:
+        reader = PDFReader()
+        student_documents = reader.read(io.BytesIO(student_answer_pdf.read()))
+        student_answers = [doc.content for doc in student_documents]
+
+    # Grade answers
+    if st.button("Grade Answers"):
+        if model_answers and student_answers:
+            grades = []
+            # for model_answer, student_answer in zip(model_answers, student_answers):
+            prompt = f"Grade the following student answer based on the model answer:\n\nModel Answer: {[doc.content for doc in model_documents]}\n\nStudent Answer: {[doc.content for doc in student_documents]}"
+            response_generator = assistant.run(prompt)
+            response = ''.join(list(response_generator))
+            grades.append(response)
+            for i, grade in enumerate(grades, 1):
+                st.write(f"{grade}")
+        else:
+            st.warning("Please upload both Model Answer PDF and Student Answer PDF")
+
+main()

assistant.py

@@ -0,0 +1,63 @@
+from typing import Optional
+from phi.assistant import Assistant
+from phi.knowledge import AssistantKnowledge
+from phi.llm.groq import Groq
+from phi.embedder.openai import OpenAIEmbedder
+from phi.embedder.ollama import OllamaEmbedder
+from phi.vectordb.pgvector import PgVector2
+from phi.storage.assistant.postgres import PgAssistantStorage
+
+# db_url = "postgresql+psycopg://ai:ai@localhost:5532/ai"
+db_url = "postgresql://ai_owner:B9iIwFyus4VO@ep-restless-block-a1e1oiah.ap-southeast-1.aws.neon.tech/ai?sslmode=require"
+
+def get_groq_assistant(
+    llm_model: str = "llama3-70b-8192",
+    embeddings_model: str = "text-embedding-3-small",
+    user_id: Optional[str] = None,
+    run_id: Optional[str] = None,
+    debug_mode: bool = True,
+) -> Assistant:
+    """Get a Groq RAG Assistant."""
+
+    embedder = (
+        OllamaEmbedder(model=embeddings_model, dimensions=768)
+        if embeddings_model == "nomic-embed-text"
+        else OpenAIEmbedder(model=embeddings_model, dimensions=1536)
+    )
+    embeddings_table = (
+        "groq_rag_documents_ollama" if embeddings_model == "nomic-embed-text" else "groq_rag_documents_openai"
+    )
+
+    return Assistant(
+        name="groq_rag_assistant",
+        run_id=run_id,
+        user_id=user_id,
+        llm=Groq(model=llm_model),
+        storage=PgAssistantStorage(table_name="groq_rag_assistant", db_url=db_url),
+        knowledge_base=AssistantKnowledge(
+            vector_db=PgVector2(
+                db_url=db_url,
+                collection=embeddings_table,
+                embedder=embedder,
+            ),
+            num_documents=2,
+        ),
+        # Slytle different instructions results in different grading or grading style.
+        description="You are an AI called 'GroqRAG' and your task is to grade student answers based on model answers.",
+        instructions=[
+            "You will always take two PDF files as input: Model Answer (best answers) and Student Answer.",
+            "Don't give marks to the model answers file only use it as a refrance",
+            "You should give a grade to each question on the student answer based on the model answer.",
+            "Use the model answer as the reference for grading.",
+            "A student who provides the meaning of an answer but uses different words and mentions the entire information given in the model answer will receive full marks.",
+            "A student who provides incomplete or irrelevant information will lose marks based on the quality and completeness of their answer.",
+            "Use a consistent marking technique so that The same answers should always receive the same marks."
+            "A question with no answer should receive zero marks."
+        ],
+        add_references_to_prompt=False,
+        markdown=True,
+        add_chat_history_to_messages=True,
+        num_history_messages=4,
+        add_datetime_to_instructions=True,
+        debug_mode=debug_mode,
+    )

requirements.txt

@@ -0,0 +1,116 @@
+aiohttp==3.9.5
+aiosignal==1.3.1
+altair==5.3.0
+annotated-types==0.6.0
+anyio==4.3.0
+asttokens==2.4.1
+attrs==23.2.0
+beautifulsoup4==4.12.3
+blinker==1.7.0
+bs4==0.0.2
+cachetools==5.3.3
+certifi==2024.2.2
+cffi==1.16.0
+charset-normalizer==3.3.2
+click==8.1.7
+colorama==0.4.6
+comm==0.2.2
+cryptography==42.0.8
+curl_cffi==0.6.3
+datasets==2.20.0
+debugpy==1.8.2
+decorator==5.1.1
+dill==0.3.8
+distro==1.9.0
+duckduckgo_search==5.3.0
+exceptiongroup==1.2.1
+executing==2.0.1
+filelock==3.15.4
+frozenlist==1.4.1
+fsspec==2024.5.0
+gitdb==4.0.11
+GitPython==3.1.43
+greenlet==3.0.3
+groq==0.5.0
+h11==0.14.0
+httpcore==1.0.5
+httpx==0.27.0
+huggingface-hub==0.23.4
+idna==3.7
+ipykernel==6.29.5
+ipython==8.26.0
+jedi==0.19.1
+Jinja2==3.1.3
+jsonschema==4.21.1
+jsonschema-specifications==2023.12.1
+jupyter_client==8.6.2
+jupyter_core==5.7.2
+markdown-it-py==3.0.0
+MarkupSafe==2.1.5
+matplotlib-inline==0.1.7
+mdurl==0.1.2
+multidict==6.0.5
+multiprocess==0.70.16
+nest-asyncio==1.6.0
+numpy==1.26.4
+ollama==0.1.8
+openai==1.23.2
+orjson==3.10.1
+packaging==24.0
+pandas==2.2.2
+parso==0.8.4
+pdfminer==20191125
+pdfminer.six==20240706
+pgvector==0.2.5
+phidata==2.4.20
+pillow==10.3.0
+platformdirs==4.2.2
+prompt_toolkit==3.0.47
+protobuf==4.25.3
+psutil==6.0.0
+psycopg==3.1.18
+psycopg-binary==3.1.18
+psycopg2==2.9.9
+pure-eval==0.2.2
+pyarrow==16.0.0
+pyarrow-hotfix==0.6
+pycparser==2.22
+pycryptodome==3.20.0
+pydantic==2.7.0
+pydantic-settings==2.2.1
+pydantic_core==2.18.1
+pydeck==0.8.1b0
+Pygments==2.17.2
+pypdf==4.2.0
+python-dateutil==2.9.0.post0
+python-dotenv==1.0.1
+pytz==2024.1
+PyYAML==6.0.1
+pyzmq==26.0.3
+referencing==0.34.0
+requests==2.32.3
+rich==13.7.1
+rpds-py==0.18.0
+shellingham==1.5.4
+six==1.16.0
+smmap==5.0.1
+sniffio==1.3.1
+soupsieve==2.5
+SQLAlchemy==2.0.29
+stack-data==0.6.3
+streamlit==1.33.0
+tenacity==8.2.3
+toml==0.10.2
+tomli==2.0.1
+toolz==0.12.1
+tornado==6.4
+tqdm==4.66.4
+traitlets==5.14.3
+typer==0.12.3
+typing_extensions==4.11.0
+tzdata==2024.1
+urllib3==1.26.18
+watchdog==4.0.1
+wcwidth==0.2.13
+xxhash==3.4.1
+yarl==1.9.4