Spaces:
Running
Running
File size: 5,658 Bytes
e615195 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 |
# [1] Core Imports (Updated Packages)
import gradio as gr
from langchain_huggingface import HuggingFaceEmbeddings
from langchain_huggingface import HuggingFacePipeline
from langchain_community.document_loaders import UnstructuredURLLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_chroma import Chroma
from langchain.chains import create_retrieval_chain
from langchain.chains.combine_documents.stuff import create_stuff_documents_chain
from langchain_core.prompts import ChatPromptTemplate
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
import nltk
import validators
nltk.download('punkt', quiet=True)
# [2] Initialize Components
text_splitter = RecursiveCharacterTextSplitter(
chunk_size=1000,
chunk_overlap=100,
separators=["\n\n", "\n"]
)
# Updated embeddings initialization
embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
# [3] Model Setup
MODEL_NAME = "google/flan-t5-large"
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForSeq2SeqLM.from_pretrained(MODEL_NAME)
pipe = pipeline(
"text2text-generation",
model=model,
tokenizer=tokenizer,
max_new_tokens=800,
temperature=0.6,
do_sample=True
)
# Updated pipeline wrapper
llm = HuggingFacePipeline(pipeline=pipe)
# [4] Prompt Template
prompt_template = ChatPromptTemplate.from_messages([
("system", "Generate a clear concise most simplest understanding language answer in about 3-5 bullet or more if you need more to explain points, using ONLY the context below.\n\nContext: {context}"),
("human", "{input}")
])
# [5] Processing Function
def process_inputs(urls_str, question):
try:
print("\n=== New Request ===")
# Validate inputs
if not urls_str.strip() or not question.strip():
print("Missing inputs")
return "❌ Please provide both URLs and a question"
urls = [url.strip() for url in urls_str.split(',') if url.strip()]
print(f"Processing {len(urls)} URLs")
# Validate URLs
for url in urls:
if not validators.url(url):
print(f"Invalid URL: {url}")
return f"❌ Invalid URL format: {url}"
# Load documents
try:
loader = UnstructuredURLLoader(urls=urls)
docs = loader.load()
print(f"Loaded {len(docs)} documents")
except Exception as e:
print(f"Document load failed: {str(e)}")
return f"❌ Failed to load documents: {str(e)}"
if not docs:
print("No content found")
return "❌ No content found in the provided URLs"
# Process documents
unique_content = list({doc.page_content.strip(): doc for doc in docs}.values())
split_docs = text_splitter.split_documents(unique_content)
print(f"Split into {len(split_docs)} chunks")
# Create vector store
try:
vectorstore = Chroma.from_documents(
documents=split_docs,
embedding=embeddings
)
retriever = vectorstore.as_retriever(search_kwargs={"k": 5})
print("Vector store created")
except Exception as e:
print(f"Vector store error: {str(e)}")
return f"❌ Vector store error: {str(e)}"
# Create chain
try:
print("Creating RAG chain")
rag_chain = create_retrieval_chain(
retriever,
create_stuff_documents_chain(
llm=llm,
prompt=prompt_template
)
)
print(f"Processing question: {question}")
response = rag_chain.invoke({"input": question})
print("Answer generated successfully")
return response["answer"]
except Exception as e:
print(f"Generation error: {str(e)}")
return f"❌ Generation error: {str(e)}"
except Exception as e:
print(f"Unexpected error: {str(e)}")
return f"❌ Unexpected error: {str(e)}"
# [6] Gradio Interface (Fixed parameters)
with gr.Blocks(theme=gr.themes.Soft()) as demo:
gr.Markdown("# RAG Chat Interface")
with gr.Row():
with gr.Column():
url_input = gr.Textbox(
label="Paste URLs (comma-separated)",
placeholder="https://example.com, https://another-site.org\nSome websites may not work as they won't allow to fetch data from their site.\nTry other websites in that case.",
lines=3
)
question_input = gr.Textbox(
label="Your Question",
placeholder="Type your question here...",
lines=3
)
submit_btn = gr.Button("Get Answer", variant="primary")
answer_output = gr.Textbox(
label="Generated Answer",
interactive=False,
lines=10 # Removed autoscroll=True
)
gr.Examples(
examples=[
[
"https://generativeai.net/, https://www.ibm.com/think/topics/generative-ai",
"What are the key benefits of generative AI?"
]
],
inputs=[url_input, question_input]
)
submit_btn.click(
fn=process_inputs,
inputs=[url_input, question_input],
outputs=answer_output
)
# [7] Launch
if __name__ == "__main__":
demo.launch() |