primasr's picture
Update README.md
68a014f
metadata
language:
  - ms
datasets:
  - squad_v2
metrics:
  - exact_match
  - f1

Overview

This model is an experiment I and my friend did as a researcher internship at the National University of Singapore (NUS). We finetuned the model to our datasets in Finance and Healthcare domain, in the Malay Language.

Details

  • Finetuned from the base model by zhufy
  • The base datasets from SQuAD2.0
  • Our datasets in Finance and Healthcare domain

Finetuned Detail

from transformers import TrainingArguments

training_args = TrainingArguments(
    output_dir='test_trainer', 
    evaluation_strategy='epoch', 
    num_train_epochs=20, 
    optim='adamw_torch',
    report_to='all',
    logging_steps=1,
)

How to use the Model

from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline

model_name = "primasr/malaybert-for-eqa-finetuned"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForQuestionAnswering.from_pretrained(model_name)
nlp = pipeline("question-answering", model=model, tokenizer=tokenizer)