add network tinybert scipy version info

official/nlp/tinybert/README.md

@@ -65,6 +65,7 @@ The backbone structure of TinyBERT is transformer, the transformer contains four
     - Prepare hardware environment with Ascend or GPU processor.
 - Framework
     - [MindSpore](https://gitee.com/mindspore/mindspore)
+    - scipy>=1.7
 - For more information, please check the resources below：
     - [MindSpore Tutorials](https://www.mindspore.cn/tutorials/en/master/index.html)
     - [MindSpore Python API](https://www.mindspore.cn/docs/api/en/master/index.html)
@@ -100,7 +101,7 @@ The backbone structure of TinyBERT is transformer, the transformer contains four
     Please follow the instructions in the link below:
     https:gitee.com/mindspore/models/tree/master/utils/hccl_tools.
 
-    For dataset, if you want to set the format and parameters, a schema configuration file with JSON format needs to be created, please refer to [tfrecord](https://www.mindspore.cn/docs/programming_guide/en/master/dataset_loading.html#tfrecord) format.
+    For dataset, if you want to set the format and parameters, a schema configuration file with JSON format needs to be created, please refer to [tfrecord](https://www.mindspore.cn/docs/en/master/api_python/dataset/mindspore.dataset.TFRecordDataset.html) format.
 
     ```text
     For general task, schema file contains ["input_ids", "input_mask", "segment_ids"].
@@ -627,6 +628,8 @@ In gd_config.yaml and td_config/*.yaml, we set the hidden_dropout_prob and atten
 
 In run_general_distill.py, we set the random seed to make sure distribute training has the same init weight.
 
+If accuracy < standard, may be scipy version < 1.7.
+
 # [ModelZoo Homepage](#contents)
 
 Please check the official [homepage](https://gitee.com/mindspore/models).

official/nlp/tinybert/README_CN.md

@@ -70,6 +70,7 @@ TinyBERT模型的主干结构是转换器，转换器包含四个编码器模块
     - 使用Ascend或GPU处理器准备硬件环境。
 - 框架
     - [MindSpore](https://gitee.com/mindspore/mindspore)
+    - scipy>=1.7
 - 更多关于Mindspore的信息，请查看以下资源：
     - [MindSpore教程](https://www.mindspore.cn/tutorials/zh-CN/master/index.html)
     - [MindSpore Python API](https://www.mindspore.cn/docs/api/zh-CN/master/index.html)
@@ -105,7 +106,7 @@ TinyBERT模型的主干结构是转换器，转换器包含四个编码器模块
     详情参见如下链接：
     https:gitee.com/mindspore/models/tree/master/utils/hccl_tools.
 
-    如需设置数据集格式和参数，请创建JSON格式的视图配置文件，详见[TFRecord](https://www.mindspore.cn/docs/programming_guide/zh-CN/master/dataset_loading.html#tfrecord) 格式。
+    如需设置数据集格式和参数，请创建JSON格式的视图配置文件，详见[TFRecord](https://www.mindspore.cn/docs/zh-CN/master/api_python/dataset/mindspore.dataset.TFRecordDataset.html) 格式。
 
     ```text
     For general task, schema file contains ["input_ids", "input_mask", "segment_ids"].
@@ -623,6 +624,8 @@ gd_config.yaml和td_config/*.yaml文件中设置了hidden_dropout_prob和attenti
 
 run_general_distill.py文件中设置了随机种子，确保分布式训练初始权重相同。
 
+若结果精度不达标，可能原因为使用的scipy版本低于1.7
+
 # ModelZoo主页
 
 请浏览官网[主页](https://gitee.com/mindspore/models)。

official/nlp/tinybert/requirements.txt

 easydict
+scipy>=1.7.1
\ No newline at end of file

official/nlp/tinybert/run_task_distill.py

@@ -410,8 +410,7 @@ def run_main():
         if lists:
             lists.sort(key=lambda fn: os.path.getmtime(td_phase1_save_ckpt_dir + '/' + fn))
             name_ext = os.path.splitext(lists[-1])
-            if name_ext[-1] != ".ckpt":
-                raise ValueError("Invalid file, checkpoint file should be .ckpt file")
+            assert name_ext[-1] == ".ckpt", "Invalid file, checkpoint file should be .ckpt file"
             newest_ckpt_file = os.path.join(td_phase1_save_ckpt_dir, lists[-1])
             # run task distill
             run_task_distill(newest_ckpt_file)