import gradio as gr
from transformers import TextIteratorStreamer
from threading import Thread
from transformers import StoppingCriteria, StoppingCriteriaList
import torch
import spaces
import os 
model_name = "microsoft/Phi-3-medium-128k-instruct"
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained(model_name, device_map='cuda', torch_dtype=torch.float16, trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(model_name)

class StopOnTokens(StoppingCriteria):
    def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs) -> bool:
        stop_ids = [29, 0]
        for stop_id in stop_ids:
            if input_ids[0][-1] == stop_id:
                return True
        return False
@spaces.GPU()
def predict(message, history):
    history_transformer_format = history + [[message, ""]]
    stop = StopOnTokens()
    messages = "".join(["".join(["\n<|end|>\n<|user|>\n"+item[0], "\n<|end|>\n<|assistant|>\n"+item[1]]) for item in history_transformer_format])
    #messages = "".join(["".join(["<user>"+item[0], "<output>"+item[1]]) for item in history_transformer_format])
    model_inputs = tokenizer([messages], return_tensors="pt").to("cuda")
    streamer = TextIteratorStreamer(tokenizer, timeout=10., skip_prompt=True, skip_special_tokens=True)
    generate_kwargs = dict(
        model_inputs,
        streamer=streamer,
        max_new_tokens=8192,
        do_sample=True,
        top_p=0.8,
        top_k=40,
        temperature=0.9,
        stopping_criteria=StoppingCriteriaList([stop])
    )
    t = Thread(target=model.generate, kwargs=generate_kwargs)
    t.start()
    partial_message = ""
    for new_token in streamer:
        if new_token != '<':
            partial_message += new_token
        yield partial_message

demo = gr.ChatInterface(fn=predict, examples=["Write me a python snake game code", "Write me a ping pong game code"], title="Phi-3-medium-128k-instruct")
demo.launch(share=True)