Created by: NextGuido

我更改了rec_chinese_lite_train.yml配置文件的内容，如下：

Global:
  algorithm: CRNN
  use_gpu: true
  epoch_num: 300
  log_smooth_window: 20
  print_batch_step: 10
  save_model_dir: ./output/rec_CRNN
  save_epoch_step: 3
  eval_batch_step: 2000
  train_batch_size_per_card: 160
  test_batch_size_per_card: 160
  image_shape: [3, 32, 320]
  max_text_length: 25
  character_type: ch
  character_dict_path: ./ppocr/utils/18548.txt
  loss_type: ctc
  reader_yml: ./configs/rec/rec_chinese_reader_csv.yml
  pretrain_weights: ./pretrain_models/rec_mv3_none_bilstm_ctc/best_accuracy
  checkpoints:
  save_inference_dir:
Architecture:
  function: ppocr.modeling.architectures.rec_model,RecModel

Backbone:
  function: ppocr.modeling.backbones.rec_mobilenet_v3,MobileNetV3
  scale: 0.5
  model_name: small

Head:
  function: ppocr.modeling.heads.rec_ctc_head,CTCPredict
  encoder_type: rnn
  SeqRNN:
    hidden_size: 48
    
Loss:
  function: ppocr.modeling.losses.rec_ctc_loss,CTCLoss

Optimizer:
  function: ppocr.optimizer,AdamDecay
  base_lr: 0.0005
  beta1: 0.9
  beta2: 0.999

其实只更改了epoch_num，batch_size，character_dict_path，reader_yml，pretrain_weights这几处，但是迭代了100轮左右，现在的train acc只有0.2左右。我的字符集有18000多个，请问这个轻量级的网络目前来看是不是欠拟合了？因为我发现如果我换成resnet50作为backbone，网络可以训练的不错