BERT (Bidirectional Encoder Representations from Transformers)

Contents

Pre-trained Models

Access to Pretrained Checkpoints

Restoring from Checkpoints

Access to Pretrained hub modules.

Set Up

Process Datasets

Pre-training

Fine-tuning

Fine-tuning with BERT

Cloud GPUs and TPUs

Sentence and Sentence-pair Classification Tasks

SQuAD 1.1