diff --git a/ais_bench/benchmark/configs/datasets/aa_lcr/README.md b/ais_bench/benchmark/configs/datasets/aa_lcr/README.md
index ae1f86f0..b7476aa2 100644
--- a/ais_bench/benchmark/configs/datasets/aa_lcr/README.md
+++ b/ais_bench/benchmark/configs/datasets/aa_lcr/README.md
@@ -11,12 +11,13 @@ AA-LCR（Artificial Analysis Long Context Retrieval）是一个用于评估语
 ## 数据集部署
 
 - 可以从 ModelScope 的数据集链接 🔗 [https://modelscope.cn/datasets/evalscope/AA-LCR](https://modelscope.cn/datasets/evalscope/AA-LCR) 中获取数据集。
-- AA-LCR 数据集为压缩包格式，建议部署在 `{tool_root_path}/ais_bench/datasets/aa_lcr/` 目录下。
+- AA-LCR 数据集为压缩包格式，建议部署在 `{tool_root_path}/ais_bench/datasets/AA-LCR/` 目录下。
 
-- 在 `{tool_root_path}/ais_bench/datasets/aa_lcr/` 目录下执行 `ls -la` 检查目录结构。如果目录结构如下所示，则数据集部署成功：
+- 在 `{tool_root_path}/ais_bench/datasets/` 目录下执行 `tree AA-LCR/` 检查目录结构。如果目录结构如下所示，则数据集部署成功：
     ```
-    {tool_root_path}/ais_bench/datasets/aa_lcr/
-    └── extracted_text/AA-LCR_extracted-text.zip
+    AA-LCR/
+    └── extracted_text/
+        └── AA-LCR_extracted-text.zip
     ```
 
 
diff --git a/ais_bench/benchmark/configs/datasets/aa_lcr/README_en.md b/ais_bench/benchmark/configs/datasets/aa_lcr/README_en.md
index a9b33cd5..70fc45ae 100644
--- a/ais_bench/benchmark/configs/datasets/aa_lcr/README_en.md
+++ b/ais_bench/benchmark/configs/datasets/aa_lcr/README_en.md
@@ -11,12 +11,13 @@ AA-LCR (Artificial Analysis Long Context Retrieval) is a benchmark for evaluatin
 ## Dataset Deployment
 
 - The dataset can be obtained from the ModelScope dataset link: 🔗 [https://modelscope.cn/datasets/evalscope/AA-LCR](https://modelscope.cn/datasets/evalscope/AA-LCR).
-- The AA-LCR dataset is in compressed archive format and is recommended to be deployed in the `{tool_root_path}/ais_bench/datasets/aa_lcr/` directory.
+- The AA-LCR dataset is in compressed archive format and is recommended to be deployed in the `{tool_root_path}/ais_bench/datasets/AA-LCR/` directory.
 
-- Execute `ls -la` in the `{tool_root_path}/ais_bench/datasets/aa_lcr/` directory to check the directory structure. If the directory structure is as shown below, the dataset has been deployed successfully:
+- Execute `tree AA-LCR/` in the `{tool_root_path}/ais_bench/datasets/` directory to check the directory structure. If the directory structure is as shown below, the dataset has been deployed successfully:
     ```
-    {tool_root_path}/ais_bench/datasets/aa_lcr/
-    └── extracted_text/AA-LCR_extracted-text.zip
+    AA-LCR/
+    └── extracted_text/
+        └── AA-LCR_extracted-text.zip
     ```
 
 
diff --git a/ais_bench/benchmark/configs/datasets/ifbench/README.md b/ais_bench/benchmark/configs/datasets/ifbench/README.md
index 92249c0e..abceea97 100644
--- a/ais_bench/benchmark/configs/datasets/ifbench/README.md
+++ b/ais_bench/benchmark/configs/datasets/ifbench/README.md
@@ -13,12 +13,13 @@ IFBench 是一个用于评估 AI 模型在遵循新颖、具有挑战性且多
 ## 数据集部署
 
 - 可以从 Hugging Face 的数据集链接 🔗 [https://huggingface.co/datasets/allenai/IFBench_test](https://huggingface.co/datasets/allenai/IFBench_test) 中获取数据集。
-- IFBench 数据集为 Parquet 格式，建议部署在 `{tool_root_path}/ais_bench/datasets/ifbench/data/` 目录下。
+- IFBench 数据集为 Parquet 格式，建议部署在 `{tool_root_path}/ais_bench/datasets/IFBench_test/data/` 目录下。
 
-- 在 `{tool_root_path}/ais_bench/datasets/ifbench/data/` 目录下执行 `ls -la` 检查目录结构。如果目录结构如下所示，则数据集部署成功：
+- 在 `{tool_root_path}/ais_bench/datasets/` 目录下执行 `tree IFBench_test/` 检查目录结构。如果目录结构如下所示，则数据集部署成功：
     ```
-    {tool_root_path}/ais_bench/datasets/ifbench/data/
-    └── train-00000-of-00001.parquet
+    IFBench_test/
+    └── data/
+        └── train-00000-of-00001.parquet
     ```
 
 
diff --git a/ais_bench/benchmark/configs/datasets/ifbench/README_en.md b/ais_bench/benchmark/configs/datasets/ifbench/README_en.md
index f2df9322..5a660ce8 100644
--- a/ais_bench/benchmark/configs/datasets/ifbench/README_en.md
+++ b/ais_bench/benchmark/configs/datasets/ifbench/README_en.md
@@ -13,12 +13,13 @@ IFBench is a benchmark for evaluating the reliability of AI models in following
 ## Dataset Deployment
 
 - The dataset can be obtained from the Hugging Face dataset link: 🔗 [https://huggingface.co/datasets/allenai/IFBench_test](https://huggingface.co/datasets/allenai/IFBench_test).
-- The IFBench dataset is in Parquet format and is recommended to be deployed in the `{tool_root_path}/ais_bench/datasets/ifbench/data/` directory.
+- The IFBench dataset is in Parquet format and is recommended to be deployed in the `{tool_root_path}/ais_bench/datasets/IFBench_test/data/` directory.
 
-- Execute `ls -la` in the `{tool_root_path}/ais_bench/datasets/ifbench/data/` directory to check the directory structure. If the directory structure is as shown below, the dataset has been deployed successfully:
+- Execute `tree IFBench_test/` in the `{tool_root_path}/ais_bench/datasets/` directory to check the directory structure. If the directory structure is as shown below, the dataset has been deployed successfully:
     ```
-    {tool_root_path}/ais_bench/datasets/ifbench/data/
-    └── train-00000-of-00001.parquet
+    IFBench_test/
+    └── data/
+        └── train-00000-of-00001.parquet
     ```