| import requests | |
| import pandas as pd | |
| from bs4 import BeautifulSoup | |
| import gradio as gr | |
| from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline | |
| model_id = "meta-llama/Meta-Llama-3-8B-Instruct" | |
| tokenizer = AutoTokenizer.from_pretrained(model_id) | |
| model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", torch_dtype="auto") | |
| llm_pipeline = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512) | |
| def run_llm(prompt: str): | |
| result = llm_pipeline(prompt)[0]["generated_text"] | |
| return result.replace(prompt, "") | |
| def fetch_financial_news(query="markets", max_articles=3): | |
| url = f"https://www.google.com/search?q={query}+site:reuters.com&tbm=nws" | |
| headers = {"User-Agent": "Mozilla/5.0"} | |
| response = requests.get(url, headers=headers) | |
| soup = BeautifulSoup(response.text, "html.parser") | |
| links = [] | |
| for g in soup.find_all('a', href=True): | |
| href = g['href'] | |
| if "reuters.com" in href and len(links) < max_articles: | |
| links.append(href.split("&")[0].replace("/url?q=", "")) | |
| return links | |
| def summarize_news_article(url): | |
| try: | |
| r = requests.get(url, headers={"User-Agent": "Mozilla/5.0"}) | |
| soup = BeautifulSoup(r.text, "html.parser") | |
| paragraphs = soup.find_all('p') | |
| text = "\n".join(p.get_text() for p in paragraphs[:10]) | |
| prompt = f"You are a financial analyst. Summarize the key points from this article:\n\n{text}\n\nReturn a concise summary suitable for investors." | |
| return run_llm(prompt) | |
| except Exception as e: | |
| return f"Failed to summarize article: {e}" | |
| def analyze_stock_data(symbol="AAPL"): | |
| try: | |
| url = f"https://query1.finance.yahoo.com/v7/finance/download/{symbol}?period1=1682899200&period2=1685577600&interval=1d&events=history" | |
| df = pd.read_csv(url) | |
| df["Date"] = pd.to_datetime(df["Date"]) | |
| closing_prices = df[["Date", "Close"]].tail(10) | |
| data_str = closing_prices.to_string(index=False) | |
| prompt = f"You're a financial analyst. Given the following recent closing prices of {symbol}, analyze the trend and summarize in plain English:\n\n{data_str}" | |
| return run_llm(prompt) | |
| except Exception as e: | |
| return f"Failed to fetch stock data: {e}" | |
| def analyze(query, stock_symbol): | |
| output = "" | |
| output += "📈 Fetching Financial News...\n" | |
| urls = fetch_financial_news(query) | |
| for url in urls: | |
| output += f"\n📰 {url}\n" | |
| output += summarize_news_article(url) + "\n" | |
| output += "\n📊 Analyzing Stock Trends...\n" | |
| output += analyze_stock_data(stock_symbol) | |
| return output | |
| gr.Interface( | |
| fn=analyze, | |
| inputs=[ | |
| gr.Textbox(label="Financial News Topic", value="tech stocks"), | |
| gr.Textbox(label="Stock Symbol", value="AAPL") | |
| ], | |
| outputs=gr.Textbox(label="Financial Summary", lines=20), | |
| title="🧠 Financial Analyst Agent (LLaMA 3.1 + LangChain Style)", | |
| description="Summarizes financial news and stock data using LLaMA 3.1 + LangChain-style prompts." | |
| ).launch() |