app.py

import spaces
import re
import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig
import torch
import json

LEAN4_DEFAULT_HEADER = (
    "import Mathlib\n"
    "import Aesop\n\n"
    "set_option maxHeartbeats 0\n\n"
    "open BigOperators Real Nat Topology Rat\n"
)

title = "# 🙋🏻‍♂️Welcome to 🌟Tonic's 🌕💉👨🏻‍🔬Moonshot Math"

description = """
     **Kimina-Prover-72B** is a state-of-the-art large formal reasoning model for Lean 4, achieving **80%+ pass rate** on the miniF2F benchmark, outperforming all prior works.\
Trained with Reinforcement Learning, 72B parameters, and a 32K token context window.\
- [Kimina-Prover-Preview GitHub](https://github.com/MoonshotAI/Kimina-Prover-Preview)\
- [Hugging Face: AI-MO/Kimina-Prover-72B](https://huggingface.co/AI-MO/Kimina-Prover-72B)\
- [Kimina Prover blog](https://huggingface.co/blog/AI-MO/kimina-prover)\
- [unimath dataset](https://huggingface.co/datasets/introspector/unimath)\
"""

citation = """> **Citation:**
> ```
> @article{kimina_prover_2025,
>   title = {Kimina-Prover Preview: Towards Large Formal Reasoning Models with Reinforcement Learning},
>   author = {Wang, Haiming and Unsal, Mert and ...},
>   year = {2025},
>   url = {http://arxiv.org/abs/2504.11354},
> }
> ```
"""


joinus ="""
### Join us:
🌟TeamTonic🌟 is always making cool demos! Join our active builder's 🛠️community 👻  
[![Join us on Discord](https://img.shields.io/discord/1109943800132010065?label=Discord&logo=discord&style=flat-square)](https://discord.gg/qdfnvSPcqP)  
On 🤗Huggingface: [MultiTransformer](https://huggingface.co/MultiTransformer)  
On 🌐Github: [Tonic-AI](https://github.com/tonic-ai) & contribute to🌟 [Build Tonic](https://git.tonic-ai.com/contribute)  
🤗Big thanks to Yuvi Sharma and all the folks at huggingface for the community grant 🤗
"""

# Build the initial system prompt
SYSTEM_PROMPT = "You are an expert in mathematics and Lean 4."

# Helper to build a Lean4 code block
def build_formal_block(formal_statement, informal_prefix=""):
    return (
        f"{LEAN4_DEFAULT_HEADER}\n"
        f"{informal_prefix}\n"
        f"{formal_statement}"
    )

# Helper to extract the first Lean4 code block from text
def extract_lean4_code(text):
    code_block = re.search(r"```lean4(.*?)(```|$)", text, re.DOTALL)
    if code_block:
        code = code_block.group(1)
        lines = [line for line in code.split('\n') if line.strip()]
        return '\n'.join(lines)
    return text.strip()

# Example problems
unimath1 = """Goal:
  X : UU
  Y : UU
  P : UU
  xp : (X → P) → P
  yp : (Y → P) → P
  X0 : X × Y → P
  x : X
  ============================
   (Y → P)"""

unimath2 = """Goal:
    R : ring  M : module R
  ============================
   (islinear (idfun M))"""

unimath3 = """Goal:
    X : UU  i : nat  b : hProptoType (i < S i)  x : Vector X (S i)  r : i = i
  ============================
   (pr1 lastelement = pr1 (i,, b))"""

unimath4 = """Goal:
    X : dcpo  CX : continuous_dcpo_struct X  x : pr1hSet X  y : pr1hSet X
  ============================
   (x ⊑ y ≃ (∀ i : approximating_family CX x, approximating_family CX x i ⊑ y))"""

additional_info_prompt = "/-Explain using mathematics-/\n"

examples = [
    [unimath1, additional_info_prompt, 2500],
    [unimath2, additional_info_prompt, 2500],
    [unimath3, additional_info_prompt, 2500],
    [unimath4, additional_info_prompt, 2500]
]

model_name = "AI-MO/Kimina-Prover-72B"
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, device_map="auto", trust_remote_code=True)

# Set generation config
model.generation_config = GenerationConfig.from_pretrained(model_name)
model.generation_config.pad_token_id = model.generation_config.eos_token_id
model.generation_config.do_sample = True
model.generation_config.temperature = 0.6
model.generation_config.top_p = 0.95

# Initialize chat history with system prompt
def init_chat(formal_statement, informal_prefix):
    user_prompt = (
        "Think about and solve the following problem step by step in Lean 4.\n"
        "# Problem: Provide a formal proof for the following statement.\n"
        f"# Formal statement:\n```lean4\n{build_formal_block(formal_statement, informal_prefix)}\n```\n"
    )
    return [
        {"role": "system", "content": SYSTEM_PROMPT},
        {"role": "user", "content": user_prompt}
    ]

# Gradio chat handler
@spaces.GPU
def chat_handler(user_message, informal_prefix, max_tokens, chat_history):
    # If chat_history is empty, initialize with system and first user message
    if not chat_history or len(chat_history) < 2:
        chat_history = init_chat(user_message, informal_prefix)
        display_history = [("user", user_message)]
    else:
        # Append new user message
        chat_history.append({"role": "user", "content": user_message})
        display_history = []
        for msg in chat_history:
            if msg["role"] == "user":
                display_history.append(("user", msg["content"]))
            elif msg["role"] == "assistant":
                display_history.append(("assistant", msg["content"]))
    # Format prompt using chat template
    prompt = tokenizer.apply_chat_template(chat_history, tokenize=False, add_generation_prompt=True)
    input_ids = tokenizer(prompt, return_tensors="pt").input_ids.to(model.device)
    attention_mask = torch.ones_like(input_ids)
    outputs = model.generate(
        input_ids,
        attention_mask=attention_mask,
        max_length=max_tokens + input_ids.shape[1],
        pad_token_id=model.generation_config.pad_token_id,
        temperature=model.generation_config.temperature,
        top_p=model.generation_config.top_p,
    )
    result = tokenizer.decode(outputs[0], skip_special_tokens=True)
    # Extract only the new assistant message (after the prompt)
    new_response = result[len(prompt):].strip()
    # Add assistant message to chat history
    chat_history.append({"role": "assistant", "content": new_response})
    display_history.append(("assistant", new_response))
    # Extract Lean4 code
    code = extract_lean4_code(new_response)
    # Prepare output
    output_data = {
        "model_input": prompt,
        "model_output": result,
        "lean4_code": code,
        "chat_history": chat_history
    }
    return display_history, json.dumps(output_data, indent=2), code, chat_history

def main():
    with gr.Blocks() as demo:
        # Title and Model Description
        gr.Markdown("""# 🙋🏻‍♂️Welcome to 🌟Tonic's 🌕💉👨🏻‍🔬Moonshot Math""")
        gr.Markdown(description)
        gr.Markdown(joinus)        
        with gr.Row():
            with gr.Column():
                chat = gr.Chatbot(label="Chat History")
                user_input = gr.Textbox(label="Your message or formal statement", lines=4)
                informal = gr.Textbox(value=additional_info_prompt, label="Optional informal prefix")
                max_tokens = gr.Slider(minimum=150, maximum=4096, value=2500, label="Max Tokens")
                submit = gr.Button("Send")
            with gr.Column():
                json_out = gr.JSON(label="Full Output")
                code_out = gr.Code(label="Extracted Lean4 Code", language="lean4")
        state = gr.State([])
        # On submit, call chat_handler
        submit.click(chat_handler, [user_input, informal, max_tokens, state], [chat, json_out, code_out, state])
        gr.Markdown(citation)
    demo.launch()

if __name__ == "__main__":
    main()