# Copyright 2020 Huawei Technologies Co., Ltd # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # ============================================================================ """crnn training""" import os import argparse import ast import mindspore.nn as nn from mindspore import context from mindspore.common import set_seed from mindspore.train.model import Model from mindspore.context import ParallelMode from mindspore.nn.wrap import WithLossCell from mindspore.train.callback import TimeMonitor, LossMonitor, CheckpointConfig, ModelCheckpoint from mindspore.communication.management import init, get_group_size, get_rank from src.loss import CTCLoss from src.dataset import create_dataset from src.crnn import crnn from src.crnn_for_train import TrainOneStepCellWithGradClip from src.metric import CRNNAccuracy from src.eval_callback import EvalCallBack set_seed(1) parser = argparse.ArgumentParser(description="crnn training") parser.add_argument("--run_distribute", action='store_true', help="Run distribute, default is false.") parser.add_argument('--dataset_path', type=str, default=None, help='Dataset path, default is None') parser.add_argument('--platform', type=str, default='Ascend', choices=['Ascend'], help='Running platform, only support Ascend now. Default is Ascend.') parser.add_argument('--model', type=str, default='lowercase', help="Model type, default is lowercase") parser.add_argument('--dataset', type=str, default='synth', choices=['synth', 'ic03', 'ic13', 'svt', 'iiit5k']) parser.add_argument('--eval_dataset', type=str, default='svt', choices=['synth', 'ic03', 'ic13', 'svt', 'iiit5k']) parser.add_argument('--eval_dataset_path', type=str, default=None, help='Dataset path, default is None') parser.add_argument("--run_eval", type=ast.literal_eval, default=False, help="Run evaluation when training, default is False.") parser.add_argument("--save_best_ckpt", type=ast.literal_eval, default=True, help="Save best checkpoint when run_eval is True, default is True.") parser.add_argument("--eval_start_epoch", type=int, default=5, help="Evaluation start epoch when run_eval is True, default is 5.") parser.add_argument("--eval_interval", type=int, default=5, help="Evaluation interval when run_eval is True, default is 5.") parser.set_defaults(run_distribute=False) args_opt = parser.parse_args() if args_opt.model == 'lowercase': from src.config import config1 as config else: from src.config import config2 as config context.set_context(mode=context.GRAPH_MODE, device_target=args_opt.platform, save_graphs=False) if args_opt.platform == 'Ascend': device_id = int(os.getenv('DEVICE_ID')) context.set_context(device_id=device_id) def apply_eval(eval_param): evaluation_model = eval_param["model"] eval_ds = eval_param["dataset"] metrics_name = eval_param["metrics_name"] res = evaluation_model.eval(eval_ds) return res[metrics_name] if __name__ == '__main__': lr_scale = 1 if args_opt.run_distribute: if args_opt.platform == 'Ascend': init() lr_scale = 1 device_num = int(os.environ.get("RANK_SIZE")) rank = int(os.environ.get("RANK_ID")) else: init() lr_scale = 1 device_num = get_group_size() rank = get_rank() context.reset_auto_parallel_context() context.set_auto_parallel_context(device_num=device_num, parallel_mode=ParallelMode.DATA_PARALLEL, gradients_mean=True) else: device_num = 1 rank = 0 max_text_length = config.max_text_length # create dataset dataset = create_dataset(name=args_opt.dataset, dataset_path=args_opt.dataset_path, batch_size=config.batch_size, num_shards=device_num, shard_id=rank, config=config) step_size = dataset.get_dataset_size() # define lr lr_init = config.learning_rate lr = nn.dynamic_lr.cosine_decay_lr(0.0, lr_init, config.epoch_size * step_size, step_size, config.epoch_size) loss = CTCLoss(max_sequence_length=config.num_step, max_label_length=max_text_length, batch_size=config.batch_size) net = crnn(config) opt = nn.SGD(params=net.trainable_params(), learning_rate=lr, momentum=config.momentum, nesterov=config.nesterov) net_with_loss = WithLossCell(net, loss) net_with_grads = TrainOneStepCellWithGradClip(net_with_loss, opt).set_train() # define model model = Model(net_with_grads) # define callbacks callbacks = [LossMonitor(), TimeMonitor(data_size=step_size)] save_ckpt_path = os.path.join(config.save_checkpoint_path, 'ckpt_' + str(rank) + '/') if args_opt.run_eval: if args_opt.eval_dataset_path is None or (not os.path.isdir(args_opt.eval_dataset_path)): raise ValueError("{} is not a existing path.".format(args_opt.eval_dataset_path)) eval_dataset = create_dataset(name=args_opt.eval_dataset, dataset_path=args_opt.eval_dataset_path, batch_size=config.batch_size, is_training=False, config=config) eval_model = Model(net, loss, metrics={'CRNNAccuracy': CRNNAccuracy(config)}) eval_param_dict = {"model": eval_model, "dataset": eval_dataset, "metrics_name": "CRNNAccuracy"} eval_cb = EvalCallBack(apply_eval, eval_param_dict, interval=args_opt.eval_interval, eval_start_epoch=args_opt.eval_start_epoch, save_best_ckpt=True, ckpt_directory=save_ckpt_path, besk_ckpt_name="best_acc.ckpt", metrics_name="acc") callbacks += [eval_cb] if config.save_checkpoint and rank == 0: config_ck = CheckpointConfig(save_checkpoint_steps=config.save_checkpoint_steps, keep_checkpoint_max=config.keep_checkpoint_max) ckpt_cb = ModelCheckpoint(prefix="crnn", directory=save_ckpt_path, config=config_ck) callbacks.append(ckpt_cb) model.train(config.epoch_size, dataset, callbacks=callbacks)