!31983 add has_trained_epoch in history

Merge pull request !31983 from liutongtong9/fix_hisoty

docs/api/api_python/train/mindspore.train.callback.History.rst

@@ -7,6 +7,14 @@
     .. note::
         通常使用在 `mindspore.Model.train` 中。
 
+    **参数：**
+
+    - **has_trained_epoch** (int) - 表示已经训练了多少个epoch，如果设置了该参数，History将监控该数值之后epoch的网络输出信息。默认值：0。
+
+    **异常：**
+
+    - **ValueError** - 当 `has_trained_epoch` 不是整数或小于零。
+
     .. py:method:: begin(run_context)
 
         训练开始时初始化History对象的epoch属性。

mindspore/python/mindspore/train/callback/_history.py

@@ -16,6 +16,7 @@
 
 import numpy as np
 from mindspore.common.tensor import Tensor
+from mindspore._checkparam import Validator
 from ._callback import Callback
 
 
@@ -31,6 +32,13 @@ class History(Callback):
     Note:
         Normally used in `mindspore.Model.train`.
 
+    Args:
+        has_trained_epoch (int): How many epochs has trained. If this parameter is set, History will record the
+                network output information after has_trained_epoch's epoch. Default: 0.
+
+    Raises:
+        ValueError: If has_trained_epoch is not an integer or less than zero.
+
     Examples:
         >>> import numpy as np
         >>> import mindspore.dataset as ds
@@ -49,9 +57,11 @@ class History(Callback):
         {'epoch': [1, 2]}
         {'net_output': [1.607877, 1.6033841]}
     """
-    def __init__(self):
+    def __init__(self, has_trained_epoch=0):
         super(History, self).__init__()
+        Validator.check_non_negative_int(has_trained_epoch)
         self.history = {}
+        self._has_trained_epoch = has_trained_epoch
 
     def begin(self, run_context):
         """
@@ -70,7 +80,10 @@ class History(Callback):
             run_context (RunContext): Context of the `mindspore.Model.{train | eval}`.
         """
         cb_params = run_context.original_args()
-        epoch = cb_params.get("cur_epoch_num", 1)
+        if "cur_epoch_num"  in cb_params:
+            epoch = cb_params.get("cur_epoch_num") + self._has_trained_epoch
+        else:
+            epoch = 1
         self.epoch.get("epoch").append(epoch)
         net_output = cb_params.net_outputs
         if isinstance(net_output, (tuple, list)):