topdog/supersonic
1
0
Fork 0
mirror of https://github.com/tencentmusic/supersonic.git synced 2025-12-11 03:58:14 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

(improvement) (common) Update sys parameter init (#400)

Browse Source 
Co-authored-by: jolunoluo
This commit is contained in:
LXW
2023-11-17 18:34:58 +08:00
committed by GitHub
parent d6a386ad03
commit 18211a215d
3 changed files with 41 additions and 40 deletions
Download Patch File Download Diff File Expand all files Collapse all files

79
common/src/main/java/com/tencent/supersonic/common/pojo/SysParameter.java
View File

@@ -8,7 +8,6 @@ import java.util.stream.Collectors;
import lombok.Data;
import org.apache.commons.lang3.StringUtils;
import org.springframework.util.CollectionUtils;
import retrofit2.http.HEAD;
@Data
public class SysParameter {
@@ -31,7 +30,7 @@ public class SysParameter {
return "";
}
Map<String, String> nameToValue = parameters.stream()
.collect(Collectors.toMap(a -> a.getName(), a -> a.getValue(), (k1, k2) -> k1));
.collect(Collectors.toMap(Parameter::getName, Parameter::getValue, (k1, k2) -> k1));
return nameToValue.get(name);
}
@@ -47,52 +46,54 @@ public class SysParameter {
parameters = Lists.newArrayList();
admins = Lists.newArrayList("admin");
//llm config
parameters.add(new Parameter("llm.model.name", "gpt4",
"模型名称(大语言模型相关配置)", "string", "大语言模型相关配置"));
Parameter parameter = new Parameter("llm.model.name", "gpt3.5",
"模型名称", "list", "大语言模型相关配置");
parameter.setCandidateValues(Lists.newArrayList("gpt3.5", "gpt3.5-16k"));
parameters.add(parameter);
parameters.add(new Parameter("llm.api.key", "sk-afdasdasd",
"模型密钥(大语言模型相关配置)", "string", "大语言模型相关配置"));
"模型密钥", "string", "大语言模型相关配置"));
parameters.add(new Parameter("llm.temperature", "0.0",
"温度值", "number", "大语言模型相关配置"));
//detect config
parameters.add(new Parameter("one.detection.size", "8",
"一次探测个数(hanlp相关配置)", "number", "hanlp相关配置"));
parameters.add(new Parameter("one.detection.max.size", "20",
"一次探测最大个数(hanlp相关配置)", "number", "hanlp相关配置"));
//mapper config
parameters.add(new Parameter("metric.dimension.min.threshold", "0.3",
"指标名、维度名最小文本相似度(mapper模糊匹配相关配置)", "number", "mapper模糊匹配相关配置"));
parameters.add(new Parameter("metric.dimension.threshold", "0.3",
"指标名、维度名文本相似度(mapper模糊匹配相关配置)", "number", "mapper模糊匹配相关配置"));
parameters.add(new Parameter("dimension.value.threshold", "0.5",
"维度值最小文本相似度(mapper模糊匹配相关配置)", "number", "mapper模糊匹配相关配置"));
//skip config
parameters.add(new Parameter("query.text.length.threshold", "10",
"文本长短阈值(是否跳过当前parser相关配置)", "number", "是否跳过当前parser相关配置"));
parameters.add(new Parameter("short.text.threshold", "5",
"短文本匹配阈值(是否跳过当前parser相关配置)", "number", "是否跳过当前parser相关配置"));
parameters.add(new Parameter("long.text.threshold", "0.8",
"长文本匹配阈值(是否跳过当前parser相关配置)", "number", "是否跳过当前parser相关配置"));
//embedding mapper config
parameters.add(new Parameter("embedding.mapper.word.min",
"4", "用于向量召回最小的文本长度(向量召回mapper相关配置)", "number", "向量召回mapper相关配置"));
parameters.add(new Parameter("embedding.mapper.word.max", "5",
"用于向量召回最大的文本长度(向量召回mapper相关配置)", "number", "向量召回mapper相关配置"));
parameters.add(new Parameter("embedding.mapper.batch", "50",
"批量向量召回文本请求个数(向量召回mapper相关配置)", "number", "向量召回mapper相关配置"));
parameters.add(new Parameter("embedding.mapper.number", "5",
"批量向量召回文本返回结果个数(向量召回mapper相关配置)", "number", "向量召回mapper相关配置"));
parameters.add(new Parameter("embedding.mapper.distance.threshold",
"0.58", "Mapper阶段向量召回相似度阈值(向量召回mapper相关配置)", "number", "向量召回mapper相关配置"));
//s2SQL config
parameters.add(new Parameter("s2SQL.generation", "2-steps",
"S2SQL生成方式", "string", "S2SQL相关配置"));
parameters.add(new Parameter("s2SQL.linking.value.switch", "true",
"是否将linkingValues提供给大模型", "bool", "S2SQL相关配置"));
//detect config
parameters.add(new Parameter("one.detection.size", "8",
"一次探测个数", "number", "[mapper]hanlp相关配置"));
parameters.add(new Parameter("one.detection.max.size", "20",
"一次探测最大个数", "number", "[mapper]hanlp相关配置"));
//mapper config
parameters.add(new Parameter("metric.dimension.min.threshold", "0.3",
"指标名、维度名最小文本相似度", "number", "[mapper]模糊匹配相关配置"));
parameters.add(new Parameter("metric.dimension.threshold", "0.3",
"指标名、维度名文本相似度", "number", "[mapper]模糊匹配相关配置"));
parameters.add(new Parameter("dimension.value.threshold", "0.5",
"维度值最小文本相似度", "number", "[mapper]模糊匹配相关配置"));
//embedding mapper config
parameters.add(new Parameter("embedding.mapper.word.min",
"4", "用于向量召回最小的文本长度", "number", "[mapper]向量召回相关配置"));
parameters.add(new Parameter("embedding.mapper.word.max", "5",
"用于向量召回最大的文本长度", "number", "[mapper]向量召回相关配置"));
parameters.add(new Parameter("embedding.mapper.batch", "50",
"批量向量召回文本请求个数", "number", "[mapper]向量召回相关配置"));
parameters.add(new Parameter("embedding.mapper.number", "5",
"批量向量召回文本返回结果个数", "number", "[mapper]向量召回相关配置"));
parameters.add(new Parameter("embedding.mapper.distance.threshold",
"0.58", "向量召回相似度阈值", "number", "[mapper]向量召回相关配置"));
//skip config
parameters.add(new Parameter("query.text.length.threshold", "10",
"文本长短阈值", "number", "是否跳过当前parser相关配置"));
parameters.add(new Parameter("short.text.threshold", "5",
"短文本匹配阈值", "number", "是否跳过当前parser相关配置"));
parameters.add(new Parameter("long.text.threshold", "0.8",
"长文本匹配阈值", "number", "是否跳过当前parser相关配置"));
}
}

1
launchers/standalone/src/test/java/com/tencent/supersonic/integration/MetricQueryTest.java
View File

@@ -124,7 +124,6 @@ public class MetricQueryTest extends BaseQueryTest {
expectedResult.setQueryMode(MetricFilterQuery.QUERY_MODE);
expectedParseInfo.setAggType(NONE);
expectedParseInfo.getDimensions().add(DataUtils.getSchemaElement("用户名"));
expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问次数"));
List<String> list = new ArrayList<>();

1
launchers/standalone/src/test/resources/db/schema-h2.sql
View File

@@ -180,6 +180,7 @@ CREATE TABLE IF NOT EXISTS `s2_datasource` (
`depends` varchar(500) DEFAULT NULL ,
`datasource_detail` LONGVARCHAR NOT NULL ,
`status` int(11) DEFAULT NULL ,
`filter_sql` varchar(1000) DEFAULT NULL ,
`created_at` TIMESTAMP NOT NULL ,
`created_by` varchar(100) NOT NULL ,
`updated_at` TIMESTAMP NOT NULL ,