diff --git a/official/cv/resnet/README.md b/official/cv/resnet/README.md
index a6fb7febed601e82c9490992d23e07d0aee22396..50837ab5e2f1b2d8685339f1353211ef1a376796 100644
--- a/official/cv/resnet/README.md
+++ b/official/cv/resnet/README.md
@@ -482,15 +482,19 @@ bash run_parameter_server_train_gpu.sh [DATASET_PATH] [CONFIG_PATH] [PRETRAINED_
 
 ```bash
 # evaluation with distributed training Ascend example:
+cd scripts/
 bash run_distribute_train.sh [RANK_TABLE_FILE] [DATASET_PATH] [CONFIG_PATH] [RUN_EVAL](optional) [EVAL_DATASET_PATH](optional)
 
 # evaluation with standalone training Ascend example:
-bash run_standalone_train.sh [RANK_TABLE_FILE] [DATASET_PATH] [CONFIG_PATH] [RUN_EVAL](optional) [EVAL_DATASET_PATH](optional)
+cd scripts/
+bash run_standalone_train.sh [DATASET_PATH] [CONFIG_PATH] [RUN_EVAL](optional) [EVAL_DATASET_PATH](optional)
 
 # evaluation with distributed training GPU example:
+cd scripts/
 bash run_distribute_train_gpu.sh [DATASET_PATH] [CONFIG_PATH] [RUN_EVAL](optional) [EVAL_DATASET_PATH](optional)
 
 # evaluation with standalone training GPU example:
+cd scripts/
 bash run_standalone_train_gpu.sh [DATASET_PATH] [CONFIG_PATH] [RUN_EVAL](optional) [EVAL_DATASET_PATH](optional)
 ```
 
diff --git a/official/cv/resnet/README_CN.md b/official/cv/resnet/README_CN.md
index dcaa4ea345900c0717f773d85e6316c6b3d928e6..51f874592606162cb6c9d951807545fd7b5b42f0 100644
--- a/official/cv/resnet/README_CN.md
+++ b/official/cv/resnet/README_CN.md
@@ -840,10 +840,10 @@ bash run_distribute_train_gpu.sh ../pruner/scop/ ../pruner/scop/resnet50_cifar10
 # 分布式训练
 cd ./golden_stick/scripts/
 # PYTHON_PATH 表示需要应用的算法的'train.py'脚本所在的目录。
-bash run_distribute_train.sh [RANK_TABLE_FILE] [PYTHON_PATH] [DATASET_PATH] [CONFIG_PATH] [PRETRAINED_CKPT_PATH](optional)
+bash run_distribute_train.sh [RANK_TABLE_FILE] [PYTHON_PATH] [CONFIG_PATH] [DATASET_PATH] [CKPT_TYPE](optional) [CKPT_PATH](optional)
 
 # 分布式训练示例(SCOP算法使用多卡训练)
-bash run_standalone_train.sh /path/to/rank_table_file ../pruner/scop/ ../pruner/scop/resnet50_cifar10_config.yaml /path/to/dataset
+bash run_distribute_train.sh /path/to/rank_table_file ../pruner/scop/ ../pruner/scop/resnet50_cifar10_config.yaml /path/to/dataset
 
 # 单机训练
 cd ./golden_stick/scripts/
diff --git a/research/nlp/gpt2/README_CN.md b/research/nlp/gpt2/README_CN.md
index 1822062b707d05d8b870453b1d6489de36fdecb0..c7f48086f0fe99460a1a81e3d6822c6ecbb408fc 100644
--- a/research/nlp/gpt2/README_CN.md
+++ b/research/nlp/gpt2/README_CN.md
@@ -151,6 +151,8 @@ GPT-2脚本及代码结构如下：
 
 `python read_weight_tf.py --ckpt_file_path=/{path}/model.ckpt`
 
+> tf模型以model.ckpt开头的有３个文件，分别为.data，.index，.meta文件，此处传参为model.ckpt，而不是任何一个其中的单独文件
+
 - 在mindspore的环境下，运行`save_weight_ms.py`，示例代码如下：
 
 `python save_weight_ms.py --output_file_name="mindspore_gpt2_small.ckpt"`
diff --git a/research/nlp/gpt2/convert_tf_ckpt/read_weight_tf.py b/research/nlp/gpt2/convert_tf_ckpt/read_weight_tf.py
index 577b656c229f72f9a7d26212f2d362dd2cf8c4be..e37fdb57aada6802be7e379564edcb95a562c35b 100644
--- a/research/nlp/gpt2/convert_tf_ckpt/read_weight_tf.py
+++ b/research/nlp/gpt2/convert_tf_ckpt/read_weight_tf.py
@@ -24,7 +24,7 @@ import argparse
 import tensorflow as tf
 import numpy as np
 
-from .trans_dict import trans_dict_tf
+from trans_dict import trans_dict_tf
 
 
 def read_weight(ckpt_path):