diff --git "a/docs/source/LLM/\346\224\257\346\214\201\347\232\204\346\250\241\345\236\213\345\222\214\346\225\260\346\215\256\351\233\206.md" "b/docs/source/LLM/\346\224\257\346\214\201\347\232\204\346\250\241\345\236\213\345\222\214\346\225\260\346\215\256\351\233\206.md"
index e8ab9de00..b82839ff3 100644
--- "a/docs/source/LLM/\346\224\257\346\214\201\347\232\204\346\250\241\345\236\213\345\222\214\346\225\260\346\215\256\351\233\206.md"
+++ "b/docs/source/LLM/\346\224\257\346\214\201\347\232\204\346\250\241\345\236\213\345\222\214\346\225\260\346\215\256\351\233\206.md"
@@ -484,6 +484,7 @@
|chart-qa|[swift/ChartQA](https://modelscope.cn/datasets/swift/ChartQA/summary)||28299|43.1±5.5, min=29, max=77|en, vqa, quality|[HuggingFaceM4/ChartQA](https://huggingface.co/datasets/HuggingFaceM4/ChartQA)|
|chinese-c4|[None](https://modelscope.cn/datasets/None/summary)||-|Dataset is too huge, please click the original link to view the dataset stat.|pretrain, zh, quality|[shjwudp/chinese-c4](https://huggingface.co/datasets/shjwudp/chinese-c4)|
|cinepile|[swift/cinepile](https://modelscope.cn/datasets/swift/cinepile/summary)||-|Dataset is too huge, please click the original link to view the dataset stat.|vqa, en, youtube, video|[tomg-group-umd/cinepile](https://huggingface.co/datasets/tomg-group-umd/cinepile)|
+|classical-chinese-translate|[swift/classical_chinese_translate](https://modelscope.cn/datasets/swift/classical_chinese_translate/summary)||6655|344.0±76.4, min=61, max=815|chat, play-ground|-|
|codealpaca-20k|[AI-ModelScope/CodeAlpaca-20k](https://modelscope.cn/datasets/AI-ModelScope/CodeAlpaca-20k/summary)||20016|100.2±60.1, min=29, max=1776|code, en|[HuggingFaceH4/CodeAlpaca_20K](https://huggingface.co/datasets/HuggingFaceH4/CodeAlpaca_20K)|
|cosmopedia|[None](https://modelscope.cn/datasets/None/summary)|auto_math_text
khanacademy
openstax
stanford
stories
web_samples_v1
web_samples_v2
wikihow|-|Dataset is too huge, please click the original link to view the dataset stat.|multi-domain, en, qa|[HuggingFaceTB/cosmopedia](https://huggingface.co/datasets/HuggingFaceTB/cosmopedia)|
|cosmopedia-100k|[swift/cosmopedia-100k](https://modelscope.cn/datasets/swift/cosmopedia-100k/summary)||100000|1024.5±243.1, min=239, max=2981|multi-domain, en, qa|[HuggingFaceTB/cosmopedia-100k](https://huggingface.co/datasets/HuggingFaceTB/cosmopedia-100k)|
diff --git a/docs/source_en/LLM/Supported-models-datasets.md b/docs/source_en/LLM/Supported-models-datasets.md
index d2e6f57f6..fe1d71105 100644
--- a/docs/source_en/LLM/Supported-models-datasets.md
+++ b/docs/source_en/LLM/Supported-models-datasets.md
@@ -484,6 +484,7 @@ The table below introduces the datasets supported by SWIFT:
|chart-qa|[swift/ChartQA](https://modelscope.cn/datasets/swift/ChartQA/summary)||28299|43.1±5.5, min=29, max=77|en, vqa, quality|[HuggingFaceM4/ChartQA](https://huggingface.co/datasets/HuggingFaceM4/ChartQA)|
|chinese-c4|[None](https://modelscope.cn/datasets/None/summary)||-|Dataset is too huge, please click the original link to view the dataset stat.|pretrain, zh, quality|[shjwudp/chinese-c4](https://huggingface.co/datasets/shjwudp/chinese-c4)|
|cinepile|[swift/cinepile](https://modelscope.cn/datasets/swift/cinepile/summary)||-|Dataset is too huge, please click the original link to view the dataset stat.|vqa, en, youtube, video|[tomg-group-umd/cinepile](https://huggingface.co/datasets/tomg-group-umd/cinepile)|
+|classical-chinese-translate|[swift/classical_chinese_translate](https://modelscope.cn/datasets/swift/classical_chinese_translate/summary)||6655|344.0±76.4, min=61, max=815|chat, play-ground|-|
|codealpaca-20k|[AI-ModelScope/CodeAlpaca-20k](https://modelscope.cn/datasets/AI-ModelScope/CodeAlpaca-20k/summary)||20016|100.2±60.1, min=29, max=1776|code, en|[HuggingFaceH4/CodeAlpaca_20K](https://huggingface.co/datasets/HuggingFaceH4/CodeAlpaca_20K)|
|cosmopedia|[None](https://modelscope.cn/datasets/None/summary)|auto_math_text
khanacademy
openstax
stanford
stories
web_samples_v1
web_samples_v2
wikihow|-|Dataset is too huge, please click the original link to view the dataset stat.|multi-domain, en, qa|[HuggingFaceTB/cosmopedia](https://huggingface.co/datasets/HuggingFaceTB/cosmopedia)|
|cosmopedia-100k|[swift/cosmopedia-100k](https://modelscope.cn/datasets/swift/cosmopedia-100k/summary)||100000|1024.5±243.1, min=239, max=2981|multi-domain, en, qa|[HuggingFaceTB/cosmopedia-100k](https://huggingface.co/datasets/HuggingFaceTB/cosmopedia-100k)|
diff --git a/swift/llm/data/dataset_info.json b/swift/llm/data/dataset_info.json
index df33f3f52..6375824c5 100644
--- a/swift/llm/data/dataset_info.json
+++ b/swift/llm/data/dataset_info.json
@@ -149,6 +149,18 @@
"tags": ["vqa", "en", "youtube", "video"],
"huge_dataset": true
},
+ "classical-chinese-translate": {
+ "dataset_id": "swift/classical_chinese_translate",
+ "conversations": {
+ "user_role": "user",
+ "assistant_role": "assistant",
+ "conversations_key": "conversations",
+ "from_key": "from",
+ "value_key": "value",
+ "error_strategy": "delete"
+ },
+ "tags": ["chat", "play-ground"]
+ },
"tagengo-gpt4": {
"dataset_id": "swift/tagengo-gpt4",
"hf_dataset_id": "lightblue/tagengo-gpt4",
diff --git a/swift/llm/utils/preprocess.py b/swift/llm/utils/preprocess.py
index 21a8582d4..a0df4b4c9 100644
--- a/swift/llm/utils/preprocess.py
+++ b/swift/llm/utils/preprocess.py
@@ -84,8 +84,8 @@ def parse_medias(self, d: Dict[str, Any]):
@property
def empty_row(self):
empty_row = {
- 'query': '',
- 'response': '',
+ 'query': None,
+ 'response': None,
'tools': None,
'system': None,
'history': None,