def create_and_check_for_pretraining(
self, config, input_ids, token_type_ids, input_mask, sequence_labels, token_labels, choice_labels
):
model = BigBirdForPreTraining(config=config)
model.to(torch_device)
model.eval()
result = model(
input_ids,
attention_mask=input_mask,
token_type_ids=token_type_ids,
labels=token_labels,
next_sentence_label=sequence_labels,
)
self.parent.assertEqual(result.prediction_logits.shape, (self.batch_size, self.seq_length, self.vocab_size))
self.parent.assertEqual(result.seq_relationship_logits.shape, (self.batch_size, config.num_labels))
def convert_tf_checkpoint_to_pytorch(tf_checkpoint_path, big_bird_config_file, pytorch_dump_path, is_trivia_qa):
# Initialise PyTorch model
config = BigBirdConfig.from_json_file(big_bird_config_file)
print("Building PyTorch model from configuration: {}".format(str(config)))
if is_trivia_qa:
model = BigBirdForQuestionAnswering(config)
else:
model = BigBirdForPreTraining(config)
# Load weights from tf checkpoint
load_tf_weights_in_big_bird(model, tf_checkpoint_path, is_trivia_qa=is_trivia_qa)
# Save pytorch-model
print(f"Save PyTorch model to {pytorch_dump_path}")
model.save_pretrained(pytorch_dump_path)