Upload 9 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+ipc_vector_db/index.faiss filter=lfs diff=lfs merge=lfs -text

Ingest.py

@@ -0,0 +1,18 @@
+from langchain_community.document_loaders import PyPDFLoader,DirectoryLoader
+from langchain.embeddings import HuggingFaceEmbeddings
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.vectorstores import FAISS
+
+loader = DirectoryLoader('data', glob="./*.pdf", loader_cls=PyPDFLoader)
+documents = loader.load()
+
+text_splitter = RecursiveCharacterTextSplitter(chunk_size=1024, chunk_overlap=200)
+texts = text_splitter.split_documents(documents)
+
+embedings = HuggingFaceEmbeddings(model_name="nomic-ai/nomic-embed-text-v1",model_kwargs={"trust_remote_code":True,"revision":"289f532e14dbbbd5a04753fa58739e9ba766f3c7"})
+
+# Creates vector embeddings and saves it in the FAISS DB
+faiss_db = FAISS.from_documents(texts, embedings)
+
+# Saves and export the vector embeddings databse
+faiss_db.save_local("ipc_vector_db")

README.md

@@ -1,12 +1,12 @@
 ---
-title: Aibot
-emoji: 📈
-colorFrom: indigo
-colorTo: green
+title: LawGPT - RAG based AI Attorney Chatbot
+emoji: ⚖️
+colorFrom: red
+colorTo: pink
 sdk: streamlit
-sdk_version: 1.32.0
+sdk_version: 1.31.1
 app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,121 @@
+from langchain_community.vectorstores import FAISS
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain.prompts import PromptTemplate
+from langchain_together import Together
+import os
+from langchain.memory import ConversationBufferWindowMemory
+from langchain.chains import ConversationalRetrievalChain
+import streamlit as st
+import time
+
+st.set_page_config(page_title="AttroneyGPT")
+col1, col2, col3 = st.columns([1,8,1])
+with col2:
+    st.image("logo.png")
+
+st.markdown(
+    """
+    <style>
+    div[data-baseweb="input"] input {
+            border-color: #000000;
+        }
+    margin-top: 0 !important;
+div.stButton > button:first-child {
+    background-color: #808080;
+    color:white;
+}
+div.stButton > button:active {
+    background-color: #808080;
+    color : white;
+}
+
+   div[data-testid="stStatusWidget"] div button {
+        display: none;
+        }
+    
+    .reportview-container {
+            margin-top: -2em;
+        }
+        #MainMenu {visibility: hidden;}
+        .stDeployButton {display:none;}
+        footer {visibility: hidden;}
+        #stDecoration {display:none;}
+    button[title="View fullscreen"]{
+    visibility: hidden;}
+        </style>
+""",
+    unsafe_allow_html=True,
+)
+
+def reset_conversation():
+  st.session_state.messages = []
+  st.session_state.memory.clear()
+
+if "messages" not in st.session_state:
+    st.session_state.messages = []
+
+if "memory" not in st.session_state:
+    st.session_state.memory = ConversationBufferWindowMemory(k=2, memory_key="chat_history",return_messages=True) 
+
+embeddings = HuggingFaceEmbeddings(model_name="nomic-ai/nomic-embed-text-v1",model_kwargs={"trust_remote_code":True,"revision":"289f532e14dbbbd5a04753fa58739e9ba766f3c7"})
+db = FAISS.load_local("ipc_vector_db", embeddings, allow_dangerous_deserialization=True)
+db_retriever = db.as_retriever(search_type="similarity",search_kwargs={"k": 4})
+
+prompt_template = """<s>[INST]This is a chat template and As a legal chat bot specializing in Sericultural related Queries!!.
+CONTEXT: {context}
+CHAT HISTORY: {chat_history}
+QUESTION: {question}
+ANSWER:
+</s>[INST]
+"""
+
+prompt = PromptTemplate(template=prompt_template,
+                        input_variables=['context', 'question', 'chat_history'])
+
+# You can also use other LLMs options from https://python.langchain.com/docs/integrations/llms. Here I have used TogetherAI API
+TOGETHER_AI_API= os.environ['TOGETHER_AI']="2a7c5dcdbb1049a39117ac0865c4d04008d49db31aa85a3258603817af16dbd0"
+llm = Together(
+    model="mistralai/Mistral-7B-Instruct-v0.2",
+    temperature=0.5,
+    max_tokens=1024,
+    together_api_key=f"{TOGETHER_AI_API}"
+)
+
+qa = ConversationalRetrievalChain.from_llm(
+    llm=llm,
+    memory=st.session_state.memory,
+    retriever=db_retriever,
+    combine_docs_chain_kwargs={'prompt': prompt}
+)
+
+for message in st.session_state.messages:
+    role = message.get("role")
+    content = message.get("content")
+    
+    with st.chat_message(role, avatar="user.svg" if role == "human" else "bot"):
+        st.write(content)
+
+input_prompt = st.chat_input("message LAWGpt.....")
+
+if input_prompt:
+    with st.chat_message("human",avatar="user.svg"):
+        st.write(input_prompt)
+
+    st.session_state.messages.append({"role":"human","content":input_prompt})
+    full_response = " "
+    with st.chat_message("bot"):
+        with st.spinner("Thinking..."):
+            result = qa.invoke(input=input_prompt)
+
+            message_placeholder = st.empty()
+
+            full_response = "⚠️ **_Note: This offers basic legal advice and is not a complete substitute for consulting a human attorney_** \n\n\n"
+        for chunk in result["answer"]:
+            full_response+=chunk
+            time.sleep(0.02)
+            
+            message_placeholder.markdown(full_response+" ▌")
+        st.button('Reset All Chat 🗑️', on_click=reset_conversation)
+
+    st.session_state.messages.append({"role": "ai", "content": result["answer"], "avatar": "bot"})
+

ipc_vector_db/index.faiss

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cac55439623c498f6c66b3effc3672a8a06236532be126130c3c14b117a8e92b
+size 1090605

ipc_vector_db/index.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:185276a728ae14de633abfb4299490fd74d4e8eb78df96527500407234c6f5f7
+size 321925

requirements.txt

@@ -0,0 +1,11 @@
+langchain
+pypdf
+transformers
+sentence-transformers
+accelerate
+faiss-cpu
+streamlit
+langchain-fireworks
+einops
+langchain_together
+