diff --git a/yudao-module-llm/yudao-module-llm-biz/src/main/java/cn/iocoder/yudao/module/llm/service/dataset/DatasetServiceImpl.java b/yudao-module-llm/yudao-module-llm-biz/src/main/java/cn/iocoder/yudao/module/llm/service/dataset/DatasetServiceImpl.java
index 0842208da..ec4fb0d0a 100644
--- a/yudao-module-llm/yudao-module-llm-biz/src/main/java/cn/iocoder/yudao/module/llm/service/dataset/DatasetServiceImpl.java
+++ b/yudao-module-llm/yudao-module-llm-biz/src/main/java/cn/iocoder/yudao/module/llm/service/dataset/DatasetServiceImpl.java
@@ -3,6 +3,7 @@ package cn.iocoder.yudao.module.llm.service.dataset;
 
 import cn.hutool.core.io.IoUtil;
 import cn.iocoder.yudao.framework.common.exception.ErrorCode;
+import cn.iocoder.yudao.framework.common.exception.util.ServiceExceptionUtil;
 import cn.iocoder.yudao.framework.common.pojo.PageResult;
 import cn.iocoder.yudao.framework.common.util.object.BeanUtils;
 import cn.iocoder.yudao.module.infra.api.file.FileApi;
@@ -29,10 +30,13 @@ import com.opencsv.exceptions.CsvValidationException;
 import lombok.extern.slf4j.Slf4j;
 import org.apache.commons.compress.archivers.ArchiveEntry;
 import org.apache.commons.compress.archivers.tar.TarArchiveInputStream;
+import org.apache.commons.compress.archivers.tar.TarArchiveEntry;
+import org.apache.commons.compress.compressors.gzip.GzipCompressorInputStream;
 import org.apache.poi.ss.usermodel.Cell;
 import org.apache.poi.ss.usermodel.Sheet;
 import org.apache.poi.ss.usermodel.Workbook;
 import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
 import org.springframework.validation.annotation.Validated;
 
 import javax.annotation.Resource;
@@ -70,6 +74,8 @@ public class DatasetServiceImpl implements DatasetService {
     @Resource
     private FileApi fileApi;
 
+
+    @Transactional
     @Override
     public Long createDataset (DatasetSaveReqVO createReqVO) {
         // 校验
@@ -270,7 +276,53 @@ public class DatasetServiceImpl implements DatasetService {
         });
         parseFile(res);
     }
-
+    public void readTarGzFile(List<DatasetFilesDO> tarGzFiles, List<DatasetFilesSaveReqVO> datasetFiles) {
+        DatasetFilesSaveReqVO datasetFilesSaveReqVO = datasetFiles.get(0);
+        List<DatasetFilesSaveReqVO> res = new ArrayList<>();
+        tarGzFiles.forEach(datasetFilesDO -> {
+            HttpURLConnection connection = DataSetReadFileUtils.readFile(datasetFilesDO.getDatasetFileUrl());
+            if (connection != null) {
+                try {
+                    InputStream inputStream = connection.getInputStream();
+                    GzipCompressorInputStream gzipInputStream = new GzipCompressorInputStream(inputStream);
+                    TarArchiveInputStream tarArchiveInputStream = new TarArchiveInputStream(gzipInputStream, Charset.forName("GBK").name());
+                    TarArchiveEntry tarEntry;
+                    try {
+                        while ((tarEntry = tarArchiveInputStream.getNextTarEntry()) != null) {
+                            if (!tarEntry.isDirectory()) {
+                                try {
+                                    final String name = tarEntry.getName();
+                                    byte[] fileBytes = IoUtil.readBytes(tarArchiveInputStream, false);
+                                    Map<String, Object> map = fileApi.llmCreateFile(name, fileBytes);
+                                    String url = map.get("url").toString();
+                                    Long id = Long.parseLong(map.get("id").toString());
+                                    DatasetFilesSaveReqVO vo = new DatasetFilesSaveReqVO();
+                                    vo.setDatasetId(datasetFilesSaveReqVO.getDatasetId());
+                                    vo.setDatasetFile(id);
+                                    vo.setDatasetFileUrl(url);
+                                    vo.setId(datasetFilesSaveReqVO.getId());
+                                    res.add(vo);
+                                } catch (Exception e) {
+                                    // Handle exception for individual file
+                                }
+                            }
+                        }
+                    } catch (IOException e) {
+                        tarArchiveInputStream.close();
+                    } finally {
+                        tarArchiveInputStream.close();
+                    }
+                } catch (Exception e) {
+                    throw ServiceExceptionUtil.exception(new ErrorCode(
+                            11001, "请正确上传tar.gz格式的数据！！！"));
+                } finally {
+                    connection.disconnect();
+                }
+            }
+        });
+        parseFile(res);
+    }
+    // 暂时先不用
     public void readTarFile (List<DatasetFilesDO> tarFiles, List<DatasetFilesSaveReqVO> datasetFiles) {
         DatasetFilesSaveReqVO datasetFilesSaveReqVO = datasetFiles.get(0);
         List<DatasetFilesSaveReqVO> res = new ArrayList<>();
@@ -444,7 +496,7 @@ public class DatasetServiceImpl implements DatasetService {
      *
      * @param xlsxFiles
      */
-    public void readXlsxFile(List<DatasetFilesDO> xlsxFiles) {
+    public void  readXlsxFile(List<DatasetFilesDO> xlsxFiles) {
         xlsxFiles.forEach(datasetFilesDO -> {
             Workbook sheets = DataSetReadFileUtils.readXlsxFromUrl(datasetFilesDO.getDatasetFileUrl());
             if (sheets != null){
@@ -506,10 +558,10 @@ public class DatasetServiceImpl implements DatasetService {
 
         // tar文件
         List<DatasetFilesDO> tarFiles = insertDatasetFiles.stream()
-                .filter(datasetFilesDO -> datasetFilesDO.getDatasetFileUrl().toLowerCase().endsWith(".tar"))
+                .filter(datasetFilesDO -> datasetFilesDO.getDatasetFileUrl().toLowerCase().endsWith(".tar.gz"))
                 .collect(Collectors.toList());
         if (CollectionUtils.isNotEmpty(tarFiles)) {
-            readTarFile(tarFiles,datasetFiles);
+            readTarGzFile(tarFiles,datasetFiles);
         }
 
         // 提取文件