args.preprocess_path, args.glove_path)
embedding_table = np.loadtxt(
os.path.join(args.preprocess_path, "weight.txt")).astype(np.float32)
# DynamicRNN in this network on Ascend platform only support the condition that the shape of input_size
# and hiddle_size is multiples of 16, this problem will be solved later.
if args.device_target == 'Ascend':
pad_num = int(np.ceil(cfg.embed_size / 16) * 16 - cfg.embed_size)
if pad_num > 0:
embedding_table = np.pad(embedding_table, [(0, 0), (0, pad_num)],
'constant')
cfg.embed_size = int(np.ceil(cfg.embed_size / 16) * 16)
network = SentimentNet(vocab_size=embedding_table.shape[0],
embed_size=cfg.embed_size,
num_hiddens=cfg.num_hiddens,
num_layers=cfg.num_layers,
bidirectional=cfg.bidirectional,
num_classes=cfg.num_classes,
weight=Tensor(embedding_table),
batch_size=cfg.batch_size)
# pre_trained
if args.pre_trained:
load_param_into_net(network, load_checkpoint(args.pre_trained))
ds_train = lstm_create_dataset(args.preprocess_path, cfg.batch_size, 1)
loss = nn.SoftmaxCrossEntropyWithLogits(sparse=True, reduction='mean')
if cfg.dynamic_lr:
lr = Tensor(
get_lr(global_step=cfg.global_step,
lr_init=cfg.lr_init,
lr_end=cfg.lr_end,
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='MindSpore LSTM Exporter')
parser.add_argument('--preprocess_path',
type=str,
default='./preprocess',
help='path where the pre-process data is stored.')
parser.add_argument('--ckpt_file',
type=str,
required=True,
help='lstm ckpt file.')
args = parser.parse_args()
embedding_table = np.loadtxt(
os.path.join(args.preprocess_path, "weight.txt")).astype(np.float32)
network = SentimentNet(vocab_size=embedding_table.shape[0],
embed_size=cfg.embed_size,
num_hiddens=cfg.num_hiddens,
num_layers=cfg.num_layers,
bidirectional=cfg.bidirectional,
num_classes=cfg.num_classes,
weight=Tensor(embedding_table),
batch_size=cfg.batch_size)
param_dict = load_checkpoint(args.ckpt_file)
load_param_into_net(network, param_dict)
input_arr = Tensor(
np.random.uniform(0.0, 1e5, size=[64, 500]).astype(np.int32))
export(network, input_arr, file_name="lstm", file_format="MINDIR")