def create_and_check_openai_gpt_model(self, config, input_ids, head_mask, token_type_ids, *args):
model = OpenAIGPTModel(config=config)
model.to(torch_device)
model.eval()
result = model(input_ids, token_type_ids=token_type_ids, head_mask=head_mask)
result = model(input_ids, token_type_ids=token_type_ids)
result = model(input_ids)
self.parent.assertEqual(result.last_hidden_state.shape, (self.batch_size, self.seq_length, self.hidden_size))
def __init__(self, config):
super().__init__(config)
config.num_labels = 1
self.transformer = OpenAIGPTModel(config)
self.lm_head = nn.Linear(config.n_embd, config.vocab_size, bias=False)
self.multiple_choice_head = SequenceSummary(config)
self.persona_head = SequenceSummary(config)
self.init_weights()
def create_and_check_openai_gpt_model(self, config, input_ids, head_mask, token_type_ids, *args):
model = OpenAIGPTModel(config=config)
model.to(torch_device)
model.eval()
model(input_ids, token_type_ids=token_type_ids, head_mask=head_mask)
model(input_ids, token_type_ids=token_type_ids)
(sequence_output,) = model(input_ids)
result = {"sequence_output": sequence_output}
self.parent.assertListEqual(
list(result["sequence_output"].size()), [self.batch_size, self.seq_length, self.hidden_size],
)
def convert_openai_checkpoint_to_pytorch(openai_checkpoint_folder_path, openai_config_file, pytorch_dump_folder_path):
# Construct model
if openai_config_file == "":
config = OpenAIGPTConfig()
else:
config = OpenAIGPTConfig.from_json_file(openai_config_file)
model = OpenAIGPTModel(config)
# Load weights from numpy
load_tf_weights_in_openai_gpt(model, config, openai_checkpoint_folder_path)
# Save pytorch-model
pytorch_weights_dump_path = pytorch_dump_folder_path + '/' + WEIGHTS_NAME
pytorch_config_dump_path = pytorch_dump_folder_path + '/' + CONFIG_NAME
print("Save PyTorch model to {}".format(pytorch_weights_dump_path))
torch.save(model.state_dict(), pytorch_weights_dump_path)
print("Save configuration file to {}".format(pytorch_config_dump_path))
with open(pytorch_config_dump_path, "w", encoding="utf-8") as f:
f.write(config.to_json_string())