app.py

import streamlit as st
from transformers import pipeline, AutoTokenizer, AutoModelForSequenceClassification
import json
import os
import requests
import torch
from gensim import summarize, corpora, models, dictionary
import re
from pygments import highlight
from pygments.lexers import PythonLexer
from pygments.formatters import HtmlFormatter
import sys
import time
from threading import Thread
import subprocess
import collections.abc as collections

client = InferenceClient(
    "mistralai/Mixtral-8x7B-Instruct-v0.1"
)

def format_prompt(message, history):
  prompt = "<s>"
  for user_prompt, bot_response in history:
    prompt += f"[INST] {user_prompt} [/INST]"
    prompt += f" {bot_response}</s> "
  prompt += f"[INST] {message} [/INST]"
  return prompt

# --- Constants ---
MODEL_URL = "https://huggingface.co/models"
TASKS_FILE = "tasks.json"
CODE_EXECUTION_ENV = {}
PIPELINE_RUNNING = False

# --- Model Initialization ---
generator = pipeline('text-generation', model='EleutherAI/gpt-neo-2.7B')
sentiment_model_name = "distilbert-base-uncased-finetuned-sst-2-english"
sentiment_tokenizer = AutoTokenizer.from_pretrained(sentiment_model_name)
sentiment_model = AutoModelForSequenceClassification.from_pretrained(sentiment_model_name)

# --- Helper Functions ---

def generate_code(prompt):
    """Generates code based on the given prompt."""
    generated = generator(prompt, max_length=200, do_sample=True, temperature=0.9)
    return generated[0]['generated_text']

def add_task(task_description):
    """Adds a new task to the task list."""
    try:
        with open(TASKS_FILE, "r") as outfile:
            tasks = json.load(outfile)
    except FileNotFoundError:
        tasks = []
    tasks.append({"task": task_description["task"], "description": task_description["description"], "status": "Pending"})
    with open(TASKS_FILE, "w") as outfile:
        json.dump(tasks, outfile)

def display_code(code):
    """Displays the code in a formatted manner."""
    formatter = HtmlFormatter(style='default')
    lexer = PythonLexer()
    html = highlight(code, lexer, formatter)
    st.markdown(html, unsafe_allow_html=True)

def summarize_text(text):
    """Summarizes the given text."""
    return summarize(text)

def analyze_sentiment(text):
    """Analyzes the sentiment of the given text."""
    inputs = sentiment_tokenizer(text, return_tensors='pt')
    outputs = sentiment_model(**inputs)
    probs = torch.nn.functional.softmax(outputs.logits, dim=1)
    return probs.tolist()[0][1]

def run_tests(code):
    """Runs tests on the given code."""
    # Placeholder for testing logic
    return "Tests passed."

def load_model(model_name):
    """Loads a pre-trained model."""
    tokenizer = AutoTokenizer.from_pretrained(model_name)
    model = AutoModelForSequenceClassification.from_pretrained(model_name)
    return model, tokenizer

def save_model(model, tokenizer, file_name):
    """Saves the model and tokenizer."""
    model.save_pretrained(file_name)
    tokenizer.save_pretrained(file_name)

def load_dataset(file_name):
    """Loads a dataset from a file."""
    data = []
    with open(file_name, "r") as infile:
        for line in infile:
            data.append(line.strip())
    return data

def save_dataset(data, file_name):
    """Saves a dataset to a file."""
    with open(file_name, "w") as outfile:
        for item in data:
            outfile.write("%s\n" % item)

def download_file(url, file_name):
    """Downloads a file from a URL."""
    response = requests.get(url)
    if response.status_code == 200:
        with open(file_name, "wb") as outfile:
            outfile.write(response.content)

def get_model_list():
    """Gets a list of available models."""
    response = requests.get(MODEL_URL)
    models = []
    for match in re.finditer("<a href='/models/(\w+/\w+)'", response.text):
        models.append(match.group(1))
    return models

def predict_text(model, tokenizer, text):
    """Predicts the text using the given model and tokenizer."""
    inputs = tokenizer(text, return_tensors='pt')
    outputs = model(**inputs)
    probs = torch.nn.functional.softmax(outputs.logits, dim=1)
    return probs.tolist()[0]

def get_user_input():
    """Gets user input."""
    input_type = st.selectbox("Select an input type", ["Text", "File", "Model"])
    if input_type == "Text":
        prompt = st.text_input("Enter text:")
        return prompt
    elif input_type == "File":
        uploaded_file = st.file_uploader("Choose a file")
        if uploaded_file:
            return uploaded_file.read().decode("utf-8")
        else:
            return ""
    elif input_type == "Model":
        model_name = st.selectbox("Select a model", get_model_list())
        model, tokenizer = load_model(model_name)
        text = st.text_area("Enter text:")
        return text

def get_tasks():
    """Loads tasks from tasks.json."""
    try:
        with open(TASKS_FILE, "r") as outfile:
            tasks = json.load(outfile)
        return tasks
    except FileNotFoundError:
        return []

def complete_task(task_id):
    """Completes a task."""
    tasks = get_tasks()
    if 0 <= task_id < len(tasks):
        tasks[task_id]["status"] = "Completed"
        with open(TASKS_FILE, "w") as outfile:
            json.dump(tasks, outfile)
        st.write(f"Task {task_id} completed.")
    else:
        st.write(f"Invalid task ID: {task_id}")

def delete_task(task_id):
    """Deletes a task."""
    tasks = get_tasks()
    if 0 <= task_id < len(tasks):
        del tasks[task_id]
        with open(TASKS_FILE, "w") as outfile:
            json.dump(tasks, outfile)
        st.write(f"Task {task_id} deleted.")
    else:
        st.write(f"Invalid task ID: {task_id}")

def run_pipeline():
    """Runs the pipeline."""
    global PIPELINE_RUNNING
    PIPELINE_RUNNING = True
    while PIPELINE_RUNNING:
        tasks = get_tasks()
        for i, task in enumerate(tasks):
            if task["status"] == "Pending":
                st.write(f"Processing task {i}: {task['task']}")
                try:
                    code = generate_code(task['description'])
                    st.write(f"Generated code:\n{code}")
                    # Execute code in a separate process
                    process = subprocess.Popen(["python", "-c", code], stdout=subprocess.PIPE, stderr=subprocess.PIPE)
                    output, error = process.communicate()
                    st.write(f"Code output:\n{output.decode('utf-8')}")
                    st.write(f"Code error:\n{error.decode('utf-8')}")
                    # Run tests (replace with actual logic)
                    test_result = run_tests(code)
                    st.write(f"Test result: {test_result}")
                    # Update task status
                    tasks[i]["status"] = "Completed"
                    with open(TASKS_FILE, "w") as outfile:
                        json.dump(tasks, outfile)
                except Exception as e:
                    st.write(f"Error processing task {i}: {e}")
                    tasks[i]["status"] = "Failed"
                    with open(TASKS_FILE, "w") as outfile:
                        json.dump(tasks, outfile)
                time.sleep(1)  # Adjust delay as needed

def stop_pipeline():
    """Stops the pipeline."""
    global PIPELINE_RUNNING
    PIPELINE_RUNNING = False
    st.write("Pipeline stopped.")

def load_model(file_name):
    """Loads a saved model."""
    try:
        with open(file_name, "rb") as f:
            model = pickle.load(f)
        with open(file_name.replace(".sav", "_tokenizer.pkl"), "rb") as f:
            tokenizer = pickle.load(f)
        return model, tokenizer
    except FileNotFoundError:
        st.write(f"Model not found: {file_name}")
        return None, None

def delete_model(file_name):
    """Deletes a saved model."""
    try:
        os.remove(file_name)
        os.remove(file_name.replace(".sav", "_tokenizer.pkl"))
        st.write(f"Model deleted: {file_name}")
    except FileNotFoundError:
        st.write(f"Model not found: {file_name}")

# --- Streamlit App ---

def main():
    """Main function."""
    st.title("AI-Powered Code Interpreter")

    # --- Code Generation and Analysis ---
    st.subheader("Code Generation and Analysis")
    text = get_user_input()

    if text:
        prompt = "Generate a python function that:\n\n" + text
        code = generate_code(prompt)

        summarized_text = ""
        if len(text) > 100:
            summarized_text = summarize_text(text)

        sentiment = ""
        if text:
            sentiment = "Positive" if analyze_sentiment(text) > 0.5 else "Negative"

        tests_passed = ""
        if code:
            tests_passed = run_tests(code)

        st.subheader("Summary:")
        st.write(summarized_text)

        st.subheader("Sentiment:")
        st.write(sentiment)

        st.subheader("Code:")
        display_code(code)

        st.subheader("Tests:")
        st.write(tests_passed)

        if st.button("Save code"):
            file_name = st.text_input("Enter file name:")
            with open(file_name, "w") as outfile:
                outfile.write(code)

    # --- Dataset Management ---
    st.subheader("Dataset Management")
    if st.button("Load dataset"):
        file_name = st.text_input("Enter file name:")
        data = load_dataset(file_name)
        st.write(data)

    if st.button("Save dataset"):
        data = st.text_area("Enter data:")
        file_name = st.text_input("Enter file name:")
        save_dataset(data, file_name)

    # --- Model Management ---
    st.subheader("Model Management")
    if st.button("Download model"):
        model_name = st.selectbox("Select a model", get_model_list())
        url = f"{MODEL_URL}/models/{model_name}/download"
        file_name = model_name.replace("/", "-") + ".tar.gz"
        download_file(url, file_name)

    if st.button("Load model"):
        model_name = st.selectbox("Select a model", get_model_list())
        model, tokenizer = load_model(model_name)

    if st.button("Predict text"):
        text = st.text_area("Enter text:")
        probs = predict_text(model, tokenizer, text)
        st.write(probs)

    if st.button("Save model"):
        file_name = st.text_input("Enter file name:")
        save_model(model, tokenizer, file_name)

    # --- Saved Model Management ---
    st.subheader("Saved Model Management")
    file_name = st.text_input("Enter file name:")
    model, tokenizer = load_model(file_name)

    if st.button("Delete model"):
        delete_model(file_name)

    # --- Task Management ---
    st.subheader("Task Management")
    if st.button("Add task"):
        task = st.text_input("Enter task:")
        description = st.text_area("Enter description:")
        add_task({"task": task, "description": description})

    if st.button("Show tasks"):
        tasks = get_tasks()
        st.write(tasks)

    if st.button("Complete task"):
        task_id = st.number_input("Enter task ID:")
        complete_task(task_id)

    if st.button("Delete task"):
        task_id = st.number_input("Enter task ID:")
        delete_task(task_id)

    # --- Pipeline Management ---
    st.subheader("Pipeline Management")
    if st.button("Run pipeline") and not PIPELINE_RUNNING:
        Thread(target=run_pipeline).start()
    if st.button("Stop pipeline") and PIPELINE_RUNNING:
        stop_pipeline()

    # --- Console Management ---
    st.subheader("Console Management")
    if st.button("Clear console"):
        st.write("")

    if st.button("Quit"):
        sys.exit()

if __name__ == "__main__":
    main()