Hard_QG_Generator / README.md
goenkalokesh's picture
Updated model card
0d51924 verified
metadata
license: apache-2.0
library_name: transformers
language:
  - en
Tags:
  - MCQ-Generator
  - T5
  - text-generation-inference

Model Card for Hard_QG_Generator

Model Details

Model Name: Hard_QG_Generator

Model Description: This model can generate MCQs from a given passage.

Model Type: Transformer.

License: Acache-2.0

Author: Lokesh Goenka

Intended Use Primary Use Cases: Generating MCQs from a given Passage. Generate MCQ-based tests for kids

Primary Users: Teachers, Students, Researchers.

Limitations and Risks Limitations: Limitation in the number of generated questions.

Training Data Data Sources:

NCERT Books https://epathshala.nic.in//process.php?id=students&type=eTextbooks&ln=en#content

Usage Example Usage:

Python hard_model = T5ForConditionalGeneration.from_pretrained("goenkalokesh/Hard_QG_Generator")

hard_tokenizer = T5Tokenizer.from_pretrained("goenkalokesh/Hard_QG_Generator")


prompt ='''Your Prompt'''

input_text = f"{prompt}: {passage.strip()}"

input_ids = hard_tokenizer.encode(input_text, return_tensors='pt')

outputs = hard_model.generate(

input_ids, 

max_length=max_length, 

num_beams=num_beams, 

early_stopping=early_stopping,

no_repeat_ngram_size=3

)

output_text = hard_tokenizer.decode(outputs[0], skip_special_tokens=True)

print(output_text)