mBart-large-50-KQA

How to use the model

Import model and tokenizer from transformer libray

# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("tykea/mBart-large-50-KQA")
model = AutoModelForSeq2SeqLM.from_pretrained("tykea/mBart-large-50-KQA")

Define function to take question and pass to the model

import torch

#ask function for easier asking 
def ask(custom_question):
# Tokenize the input
    inputs = tokenizer(
        f"qestion: {custom_question}",
        return_tensors="pt",
        truncation=True,
        max_length=512,
        padding="max_length"
    )

    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
    inputs = {key: value.to(device) for key, value in inputs.items()}

    model.eval()
    with torch.no_grad():
        outputs = model.generate(
        input_ids=inputs["input_ids"],
        max_length=50,
        num_beams=4, 
        repetition_penalty=2.0,
        early_stopping=True,
        do_sample=True,     
        top_k=50,           
        top_p=0.95,         
        temperature=0.7,
    )
    answer = tokenizer.decode(outputs[0], skip_special_tokens=True)
    print(f"Question: {custom_question}")
    print(f"Answer: {answer}")

Then call the function #ask function

question = "តើប្អូនកើតនៅប្រទេសណា?"
ask(question)
#output
Question: តើប្អូនកើតនៅប្រទេសណា?
Answer: ប្អូនកើតនៅប្រទេសចិន