roberta-base for QA

Note: this is a clone of roberta-base-squad2 for internal testing.

This is the roberta-base model, fine-tuned using the SQuAD2.0 dataset. It's been trained on question-answer pairs, including unanswerable questions, for the task of Question Answering.

Overview

Language model: roberta-base
Language: English
Downstream-task: Extractive QA
Training data: SQuAD 2.0
Eval data: SQuAD 2.0
Code: See an example QA pipeline on Haystack
Infrastructure: 4x Tesla v100

Hyperparameters

batch_size = 96
n_epochs = 2
base_LM_model = "roberta-base"
max_seq_len = 386
learning_rate = 3e-5
lr_schedule = LinearWarmup
warmup_proportion = 0.2
doc_stride=128
max_query_length=64

Using a distilled model instead

Please note that we have also released a distilled version of this model called deepset/tinyroberta-squad2. The distilled model has a comparable prediction quality and runs at twice the speed of the base model.

Usage

In Haystack

Haystack is an NLP framework by deepset. You can use this model in a Haystack pipeline to do question answering at scale (over many documents). To load the model in Haystack:

reader = FARMReader(model_name_or_path="deepset/roberta-base-squad2")
# or 
reader = TransformersReader(model_name_or_path="deepset/roberta-base-squad2",tokenizer="deepset/roberta-base-squad2")

For a complete example of roberta-base-squad2 being used for Question Answering, check out the Tutorials in Haystack Documentation

In Transformers

from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline

model_name = "deepset/roberta-base-squad2"

# a) Get predictions
nlp = pipeline('question-answering', model=model_name, tokenizer=model_name)
QA_input = {
    'question': 'Why is model conversion important?',
    'context': 'The option to convert models between FARM and transformers gives freedom to the user and let people easily switch between frameworks.'
}
res = nlp(QA_input)

# b) Load model & tokenizer
model = AutoModelForQuestionAnswering.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)

Performance

Evaluated on the SQuAD 2.0 dev set with the official eval script.

"exact": 79.87029394424324,
"f1": 82.91251169582613,

"total": 11873,
"HasAns_exact": 77.93522267206478,
"HasAns_f1": 84.02838248389763,
"HasAns_total": 5928,
"NoAns_exact": 81.79983179142137,
"NoAns_f1": 81.79983179142137,
"NoAns_total": 5945

Using the official question answering notebook from transformers yields:

{'HasAns_exact': 77.93522267206478,
 'HasAns_f1': 83.93715663402219,
 'HasAns_total': 5928,
 'NoAns_exact': 81.90075693860386,
 'NoAns_f1': 81.90075693860386,
 'NoAns_total': 5945,
 'best_exact': 79.92082877116145,
 'best_exact_thresh': 0.0,
 'best_f1': 82.91749890730902,
 'best_f1_thresh': 0.0,
 'exact': 79.92082877116145,
 'f1': 82.91749890730917,
 'total': 11873}

which is consistent with the officially reported results. Using the question answering Evaluator from evaluate gives:

 {'HasAns_exact': 77.91835357624831,
 'HasAns_f1': 84.07820736158186,
 'HasAns_total': 5928,
 'NoAns_exact': 81.91757779646763,
 'NoAns_f1': 81.91757779646763,
 'NoAns_total': 5945,
 'best_exact': 79.92082877116145,
 'best_exact_thresh': 0.996823787689209,
 'best_f1': 82.99634576260925,
 'best_f1_thresh': 0.996823787689209,
 'exact': 79.92082877116145,
 'f1': 82.9963457626089,
 'latency_in_seconds': 0.016523243643392558,
 'samples_per_second': 60.52080460605492,
 'total': 11873,
 'total_time_in_seconds': 196.18047177799986}

which is also consistent with the officially reported results.

Authors

Branden Chan: branden.chan@deepset.ai
Timo Möller: timo.moeller@deepset.ai
Malte Pietsch: malte.pietsch@deepset.ai
Tanay Soni: tanay.soni@deepset.ai

About us

deepset is the company behind the open-source NLP framework Haystack which is designed to help you build production ready NLP systems that use: Question answering, summarization, ranking etc.

Some of our other work: