Hugging Face က Pre-Trained Model ကိုသုံးသည့် chatbot

zawhtutwin

Zaw Htut Win

Posted on April 20, 2023

Hugging Face က Pre-Trained Model ကိုသုံးသည့် chatbot

Python Version က 3.8.5 ကိုသုံးတာပါ။
Torch ပဲ install လုပ်ဖို့ လိုပါတယ်။

pip install torch

ပြီးရင် ကိုယ်သုံးချင်တဲ့ model ကို Hugging Face မှာရှာရမှာပါ။

ssmadha/gpt2-finetuned-scientific-articles ဆိုတဲ့ Model ပါ။

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

# Load the tokenizer and model
tokenizer = AutoTokenizer.from_pretrained("ssmadha/gpt2-finetuned-scientific-articles")
model = AutoModelForCausalLM.from_pretrained("ssmadha/gpt2-finetuned-scientific-articles")

# Start the chat loop
while True:
    # Get user input
    user_input = input("You: ")

    # Tokenize user input and generate response
    input_ids = tokenizer.encode(user_input, return_tensors='pt')
    output = model.generate(
     input_ids,
     max_length=50,
     pad_token_id=tokenizer.eos_token_id,
     do_sample=False,
     top_p=0.9,
     top_k=0
    )
    response_text = tokenizer.decode(output[0], skip_special_tokens=True)

    print("ChatBot:", response_text)

Enter fullscreen mode Exit fullscreen mode
💖 💪 🙅 🚩
zawhtutwin
Zaw Htut Win

Posted on April 20, 2023

Join Our Newsletter. No Spam, Only the good stuff.

Sign up to receive the latest update from our blog.

Related