app.py

# -*- coding: utf-8 -*-
"""translation practice.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1KrnodZGBZrUFdaJ9FIn8IhtWtCL7peoE
"""
import requests
import gradio as gr
from dotenv import load_dotenv
import os
#from openai import OpenAI
from langchain_openai import OpenAI
import spacy
#from langchain.chat_models import ChatOpenAI
from langchain_openai import ChatOpenAI
from langchain.schema import AIMessage, HumanMessage

# Load environment variables from .env file
load_dotenv()

# Access the env
HF_TOKEN = os.getenv('HUGGING_FACE_TOKEN')

# openai setup
client = OpenAI(
  api_key=os.getenv('OPENAI_API_KEY')
)

# hugging face setup
#model_name = "mmnga/ELYZA-japanese-Llama-2-7b-instruct-gguf"
API_URL = f"https://api-inference.huggingface.co/models/"
#API_URL = f"https://api-inference.huggingface.co/models/{model_name}"
headers = {"Authorization": f"Bearer {HF_TOKEN}"}

# Global variable to control debug printing
DEBUG_MODE = True

def debug_print(*args, **kwargs):
    if DEBUG_MODE:
        print(*args, **kwargs)

def split_sentences_ginza(input_text):
    nlp = spacy.load("ja_core_news_sm")
    doc = nlp(input_text)
    sentences = [sent.text for sent in doc.sents]
    return sentences

def query_hf(payload, model_name):
    # HTTP POST Request
    response = requests.post(API_URL+model_name, headers=headers, json=payload)
    return response.json()

def translate_hf(input_text):
    debug_print("Translating... ", input_text)

    sentences = split_sentences_ginza(input_text)  # split into sentences
    translated_sentences = []

    debug_print("Split sentences... ", sentences)

    for sentence in sentences:
        if sentence.strip():  # Ensure sentence is not empty
            # API Request for each sentence:
            response = query_hf({
                "inputs": sentence.strip(),
                "options": {"wait_for_model": True}
            }, "Helsinki-NLP/opus-mt-ja-en")

            debug_print("response: ", response)
            translated_sentence = response[0]["translation_text"]
            translated_sentences.append(translated_sentence)

    # Join the translated sentences
    translation = ' '.join(translated_sentences)

    return translation


def translate_openai(input_text):

    prompt = "Translate the following text into Japanese language: " + input_text

    response = client.chat.completions.create( # get translation from GPT
        messages=[
            {
                "role": "user",
                "content": prompt,
            }
        ],
        model="gpt-3.5-turbo",
        temperature=0 # should be the same translation every time
      )
    translation = response.choices[0].message.content
    debug_print("GPT translation:", translation)

    return translation


llm = ChatOpenAI(temperature=0.7, model='gpt-3.5-turbo')

def predict(message, history):
    history_langchain_format = []
    for human, ai in history:
        history_langchain_format.append(HumanMessage(content=human)) # converted to str to avoid error; not compatible with multimodal
        history_langchain_format.append(AIMessage(content=ai))
    history_langchain_format.append(HumanMessage(content=message))
    gpt_response = llm(history_langchain_format)
    return gpt_response.content

app = gr.ChatInterface(fn=predict, title="Translation Chatbot")#, multimodal=True)

app.launch()