README.md

metadata

license: gpl-3.0
tags:
  - DocVQA
  - Document Question Answering
  - Document Visual Question Answering
datasets:
  - rubentito/mp-docvqa
language:
  - en

BigBird-BASE-ITC fine-tuned on MP-DocVQA

This is BigBird-base trained on TriviaQA from Google hub and fine-tuned on Multipage DocVQA (MP-DocVQA) dataset.

• Due to Huggingface implementation, the global tokens are defined according to the Internal Transformer Construction (ITC) strategy.

This model was used as a baseline in Hierarchical multimodal transformers for Multi-Page DocVQA.

• Results on the MP-DocVQA dataset are reported in Table 2.
• Training hyperparameters can be found in Table 8 of Appendix D.

How to use

Here is how to use this model to get the features of a given text in PyTorch:

from transformers import BigBirdForQuestionAnswering

# by default its in `block_sparse` mode with num_random_blocks=3, block_size=64
model = BigBirdForQuestionAnswering.from_pretrained("rubentito/bigbird-base-itc-mpdocvqa")

# you can change `attention_type` to full attention like this:
model = BigBirdForQuestionAnswering.from_pretrained("rubentito/bigbird-base-itc-mpdocvqa", attention_type="original_full")

# you can change `block_size` & `num_random_blocks` like this:
model = BigBirdForQuestionAnswering.from_pretrained("rubentito/bigbird-base-itc-mpdocvqa", block_size=16, num_random_blocks=2)

question = "Replace me by any text you'd like."
context = "Put some context for answering"
encoded_input = tokenizer(question, context, return_tensors='pt')
output = model(**encoded_input)

Model results

Extended experimentation can be found in Table 2 of Hierarchical multimodal transformers for Multi-Page DocVQA. You can also check the live leaderboard at the RRC Portal.

Model	HF name	ANLS	APPA
Bert large	rubentito/bert-large-mpdocvqa	0.4183	51.6177
Longformer base	rubentito/longformer-base-mpdocvqa	0.5287	71.1696
BigBird ITC base	rubentito/bigbird-base-itc-mpdocvqa	0.4929	67.5433
LayoutLMv3 base	rubentito/layoutlmv3-base-mpdocvqa	0.4538	51.9426
T5 base	rubentito/t5-base-mpdocvqa	0.5050	0.0000
Hi-VT5	TBA	0.6201	79.23

Citation Information

@article{tito2022hierarchical,
  title={Hierarchical multimodal transformers for Multi-Page DocVQA},
  author={Tito, Rub{\`e}n and Karatzas, Dimosthenis and Valveny, Ernest},
  journal={arXiv preprint arXiv:2212.05935},
  year={2022}
}