app.py

import gradio as gr
from huggingface_hub import login
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
from threading import Thread
import torch

MODEL = "m-a-p/OpenCodeInterpreter-DS-33B"
CHAT_TEMPLATE = "{%- set found_item = false -%}\n{%- for message in messages -%}\n    {%- if message['role'] == 'system' -%}\n        {%- set found_item = true -%}\n    {%- endif -%}\n{%- endfor -%}\n{%- if not found_item -%}\n{{'You are an AI programming assistant, utilizing the Deepseek Coder model, developed by Deepseek Company, and you only answer questions related to computer science. For politically sensitive questions, security and privacy issues, and other non-computer science questions, you will refuse to answer.\\n'}}\n{%- endif %}\n{%- for message in messages %}\n    {%- if message['role'] == 'system' %}\n{{ message['content'] }}\n    {%- else %}\n        {%- if message['role'] == 'user' %}\n{{'### Instruction:\\n' + message['content'] + '\\n'}}\n        {%- else %}\n{{'### Response:\\n' + message['content'] + '\\n<|EOT|>\\n'}}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{{'### Response:\\n'}}\n"

system_message = "You are a computer programmer that can translate python code to C++ in order to improve performance"

def user_prompt_for(python):
    return f"Rewrite this python code to C++. You must search for the maximum performance. \
    Format your response in Markdown. This is the python Code: \
    \n\n\
    {python}"

def messages_for(python):
    return [
        {"role": "system", "content": system_message},
        {"role": "user", "content": user_prompt_for(python)}
    ]

tokenizer = AutoTokenizer.from_pretrained(MODEL)
tokenizer.chat_template = CHAT_TEMPLATE
model = AutoModelForCausalLM.from_pretrained(MODEL, torch_dtype=torch.bfloat16, device_map="auto")
model.eval()

decode_kwargs = dict(skip_special_tokens=True)
streamer = TextIteratorStreamer(tokenizer, decode_kwargs=decode_kwargs)

cplusplus = None
def translate(python, progress=gr.Progress()):
    progress(0, desc="Starting")
    formatted_prompt = tokenizer.apply_chat_template(
                        messages_for(python),
                        tokenize=False,
                        add_generation_prompt=True,
                        return_tensors="pt")
    inputs = tokenizer(formatted_prompt, return_tensors="pt", padding=True).to(model.device)
    attention_mask = inputs["attention_mask"]
    input_ids = inputs['input_ids']

    outputs = model.generate(
        input_ids,
        attention_mask=attention_mask,
        max_new_tokens=1024,
        do_sample=False,
        pad_token_id=tokenizer.eos_token_id,
        eos_token_id=tokenizer.eos_token_id,
    )
    progress(1, desc="Finished")
    return tokenizer.decode(outputs[0][len(input_ids[0]):], skip_special_tokens=True)
    '''
    generation_kwargs = dict(
                            inputs,
                            streamer=streamer,
                            max_new_tokens=1024,
                            do_sample=False,
                            pad_token_id=tokenizer.eos_token_id,
                            eos_token_id=tokenizer.eos_token_id
                        )
    thread = Thread(target=model.generate, kwargs=generation_kwargs)
    thread.start()
    cplusplus = ""
    for chunk in streamer:
        cplusplus += chunk
        yield cplusplus
    '''

demo = gr.Interface(fn=translate, inputs="code", outputs="markdown")
demo.launch()