Upload 6 files

.gitattributes

@@ -33,3 +33,5 @@ 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
+LawGPT/data/ipc_law.pdf filter=lfs diff=lfs merge=lfs -text
+LawGPT/Lawbot511.png filter=lfs diff=lfs merge=lfs -text

LawGPT/LawGPT.py

@@ -0,0 +1,69 @@
+from langchain.vectorstores import Chroma
+from transformers import AutoTokenizer, AutoModelForSeq2SeqLM 
+from transformers import pipeline
+import torch
+from langchain.llms import HuggingFacePipeline
+from langchain.embeddings import SentenceTransformerEmbeddings
+from langchain.chains import RetrievalQA
+import gradio as gr 
+
+def chat(chat_history, user_input):
+
+    bot_response = qa_chain({"query": user_input})
+    bot_response = bot_response['result']
+    response = ""
+    for letter in ''.join(bot_response):
+        response += letter + ""
+        yield chat_history + [(user_input, response)]
+
+checkpoint = "MBZUAI/LaMini-Flan-T5-783M"
+tokenizer = AutoTokenizer.from_pretrained(checkpoint)
+base_model = AutoModelForSeq2SeqLM.from_pretrained(
+    checkpoint,
+    device_map="auto",
+    torch_dtype = torch.float32)
+
+embeddings = SentenceTransformerEmbeddings(model_name="sentence-transformers/multi-qa-mpnet-base-dot-v1")
+
+db = Chroma(persist_directory="ipc_vector_data", embedding_function=embeddings)
+
+pipe = pipeline(
+    'text2text-generation',
+    model = base_model,
+    tokenizer = tokenizer,
+    max_length = 512,
+    do_sample = True,
+    temperature = 0.3,
+    top_p= 0.95
+)
+local_llm = HuggingFacePipeline(pipeline=pipe)
+
+qa_chain = RetrievalQA.from_chain_type(llm=local_llm,
+        chain_type='stuff',
+        retriever=db.as_retriever(search_type="similarity", search_kwargs={"k":2}),
+        return_source_documents=True,
+        )
+
+with gr.Blocks() as gradioUI:
+    
+    gr.Image('lawbot511.png')
+    
+    with gr.Row():
+        chatbot = gr.Chatbot()
+    with gr.Row():
+        input_query = gr.TextArea(label='Input',show_copy_button=True)
+
+    with gr.Row():
+        with gr.Column():
+            submit_btn = gr.Button("Submit", variant="primary")
+        with gr.Column():
+            clear_input_btn = gr.Button("Clear Input")
+        with gr.Column():
+            clear_chat_btn = gr.Button("Clear Chat")
+
+    submit_btn.click(chat, [chatbot, input_query], chatbot)
+    submit_btn.click(lambda: gr.update(value=""), None, input_query, queue=False)
+    clear_input_btn.click(lambda: None, None, input_query, queue=False)
+    clear_chat_btn.click(lambda: None, None, chatbot, queue=False)
+
+gradioUI.queue().launch()

LawGPT/VectorEmbeddings.py

@@ -0,0 +1,15 @@
+from langchain.document_loaders import PyPDFLoader, DirectoryLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.embeddings import SentenceTransformerEmbeddings
+from langchain.vectorstores import Chroma
+   
+loader = DirectoryLoader('data', glob="./*.pdf", loader_cls=PyPDFLoader)
+documents = loader.load()
+   
+text_splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=200)
+texts = text_splitter.split_documents(documents)
+
+embeddings = SentenceTransformerEmbeddings(model_name="multi-qa-mpnet-base-dot-v1")
+persist_directory = "ipc_vector_data"
+db = Chroma.from_documents(texts, embeddings, persist_directory=persist_directory)
+

LawGPT/data/ipc_law.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e67161633a056f77848221ab30c49b26199c66cc844ee559ac47d2ca5dea9256
+size 20102169

LawGPT/requirements.txt

@@ -0,0 +1,8 @@
+langchain
+transformers
+torch
+gradio
+sentence-transformers
+accelerate
+chromadb
+pypdf