You can deploy this model and start earning money today!

madlag/
bert-base-uncased-squadv1-x1.84-f88.7-d36-hybrid-filled-v1

Name: bert-base-uncased-squadv1-x1.84-f88.7-d36-hybrid-filled-v1
Availability: InStock

Views

No views yet

BERT-base uncased model fine-tuned on SQuAD v1

This model was created using the nn_pruning python library: the linear layers contains 36.0% of the original weights.

The model contains 50.0% of the original weights overall (the embeddings account for a significant part of the model, and they are not pruned by this method).

With a simple resizing of the linear matrices it ran 1.84x as fast as the dense model on the evaluation. This is possible because the pruning method lead to structured matrices: to visualize them, hover below on the plot to see the non-zero/zero parts of each matrix.

In terms of accuracy, its F1 is 88.72, compared with 88.5 for the dense version, a F1 gain of 0.22.

Fine-Pruning details

This model was fine-tuned from the HuggingFace model checkpoint on SQuAD1.1, and distilled from the model csarron/bert-base-uncased-squad-v1 This model is case-insensitive: it does not make a difference between english and English.

A side-effect of the block pruning is that some of the attention heads are completely removed: 48 heads were removed on a total of 144 (33.3%). Here is a detailed view on how the remaining heads are distributed in the network after pruning.

Details of the SQuAD1.1 dataset

Dataset	Split	# samples
SQuAD1.1	train	90.6K
SQuAD1.1	eval	11.1k

Fine-tuning

Python: 3.8.5
Machine specs:

CPU:

Memory: 64 GiB
GPUs: 1 GeForce GTX 3090, with 24GiB memory
GPU driver: 455.23.05, CUDA: 11.1

Results

Pytorch model file size: 379MB (original BERT: 420MB)

Metric	# Value	# Original (Table 2)	Variation
EM	81.69	80.8	+0.89
F1	88.72	88.5	+0.22

Example Usage

Install nn_pruning: it contains the optimization script, which just pack the linear layers into smaller ones by removing empty rows/columns.

pip install nn_pruning

Then you can use the transformers library almost as usual: you just have to call optimize_model when the pipeline has loaded.

python

from transformers import pipeline
from nn_pruning.inference_model_patcher import optimize_model

qa_pipeline = pipeline(
    "question-answering",
    model="madlag/bert-base-uncased-squadv1-x1.84-f88.7-d36-hybrid-filled-v1",
    tokenizer="madlag/bert-base-uncased-squadv1-x1.84-f88.7-d36-hybrid-filled-v1"
)

print("/home/lagunas/devel/hf/nn_pruning/nn_pruning/analysis/tmp_finetune parameters: 218.0M")
print(f"Parameters count (includes only head pruning, not feed forward pruning)={int(qa_pipeline.model.num_parameters() / 1E6)}M")
qa_pipeline.model = optimize_model(qa_pipeline.model, "dense")

print(f"Parameters count after complete optimization={int(qa_pipeline.model.num_parameters() / 1E6)}M")
predictions = qa_pipeline({
    'context': "Frédéric François Chopin, born Fryderyk Franciszek Chopin (1 March 1810 – 17 October 1849), was a Polish composer and virtuoso pianist of the Romantic era who wrote primarily for solo piano.",
    'question': "Who is Frederic Chopin?",
})
print("Predictions", predictions)