训练时的epoch少于设置的epoch

[wsy-yjys]

问题确认 Search before asking

• [X] 我已经查询历史issue，没有发现相似的bug。I have searched the issues and found no similar bug report.

Bug组件 Bug Component

No response

Bug描述 Describe the Bug

我参照ppyoloe_plus_crn_t_auxhead_300e_cocoyml，自建了一个ppyoloe_plus_crn_t_auxhead_300e_gdut.yml，具体内容如下：

_BASE_: [
  '../datasets/GDUT.yml',
  '../runtime.yml',
  './_base_/optimizer_300e.yml',
  './_base_/ppyoloe_plus_crn_tiny_auxhead.yml',
  './_base_/ppyoloe_plus_reader.yml',
]

epoch: 300

LearningRate:
  base_lr: 0.0025     # base_lr=0.01， 原来8gpu*every_gpu_8,now all_batch=16,
  schedulers:
    - name: CosineDecay
      max_epochs: 360
    - name: LinearWarmup
      start_factor: 0.
      epochs: 5

log_iter: 100
snapshot_epoch: 10     # every epoch save weight
weights: output/ppyoloe_plus_crn_t_auxhead_300e_gdut/model_final

TrainReader:
  batch_size: 4         # 4gpu,per 4 imgs
EvalReader:
  batch_size: 2

# pretrain_weights: https://paddledet.bj.bcebos.com/models/pretrained/CSPResNetb_t_pretrained.pdparams
depth_mult: 0.33
width_mult: 0.375

其中设置的epoch是300，但是训练时打印的日志信息如下：

[04/03 17:07:42] ppdet.engine INFO: Epoch: [0] [ 0/99] learning_rate: 0.000000 loss: 8.639099 loss_cls: 3.119581 loss_iou: 1.255225 loss_dfl: 4.762912 loss_l1: 11.159260 eta: 1 day, 12:28:33 batch_cost: 4.4213 data_cost: 0.0005 ips: 0.9047 images/s

Epoch: [0] [ 0/99]显示epoch只有100轮，如何解决这个问题，求助

复现环境 Environment

• OS: Linux
• paddlepaddle-gpu: 2.4.2.post112
• python: 3.9.16
• CUDA: 11.3
• CUDNN: 8..2.1
• GCC: 9.4.0

Bug描述确认 Bug description confirmation

• [X] 我确认已经提供了Bug复现步骤、代码改动说明、以及环境信息，确认问题是可以复现的。I confirm that the bug replication steps, code change instructions, and environment information have been provided, and the problem can be reproduced.

是否愿意提交PR？ Are you willing to submit a PR?

• [ ] 我愿意提交PR！I'd like to help by submitting a PR!

[nemonameless]

Epoch: [0] [ 0/99] 是第0个epoch，还远远没训到你设置的300epoch。99表示每个epoch的iters数。