element.chen/ai-g3sb-backman2.0
1
0
Fork 0
Code Pull Requests Activity
Files
85ad31348eabab8a12acd0fb44fb3529dea67755
ai-g3sb-backman2.0/__pycache__/api_server.cpython-312.pyc
T

205 lines
35 KiB
Plaintext
Raw Normal View History

0.1.1 暂存
2026-04-14 10:28:22 +08:00
Ë
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
ðAßiµcãó.
Refactor embedding configuration to remove local model support, transitioning to a unified remote API approach. Update environment variables and documentation accordingly. Enhance error handling in the orchestrator and related modules to reflect these changes. This update simplifies the embedding process and improves overall system reliability.
2026-04-15 09:49:18 +08:00
—dZddlmZddlZddlZddlZddlZddlZddl Z ddl
0.1.1 暂存
2026-04-14 10:28:22 +08:00
m Z ddl m
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
Z
0.1.1 暂存
2026-04-14 10:28:22 +08:00
mZmZmZmZddlmZddlmZmZmZm ZddlmZdd lmZmZdd
lmZm Z dd l!m"Z"e e#«jI«jJZ&e&d z Z'e(e'«ejRvr"ejRjUde(e'««dd
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
l+m,Z,m-Z-m.Z.m/Z/ddl0m1Z1ddl2m3Z3ddl4m5Z5m6Z6ddl7m8Z8m9Z9m:Z:e jve jxdd¬«e jze>«Z?drd„Z@e@«ZAeAj…«re"eA«e?j‡deA›�«ne?j‰deA›�«daEdaFd„ZGGd„de«ZHGd„de«ZIGd„de«ZJGd„d e«ZKGd!„d"e«ZLGd#„d$e«ZMGd%„d&e«ZNGd'„d(e«ZOGd)„d*e«ZPGd+„d,e«ZQGd-„d.e«ZRdsd/„ZSdtd0„ZT dud1„ZUdvd2„ZVdwd3„ZWdxd4„ZX dy dzd5„ZYd{d6„ZZed|d7„«Z[ed8d9d:e[¬;«Z\e\j»ed<gd=d<gd<g¬>«e\j½d?«d@„«Z_e\jÁdA«d}dB„«Zae\jÁdC«d}dD„«Zbe\jÁdE«d~dF„«Zce\j½dE« d d€dG„«Zde\j½dH«edIdJ¬K«dddLdf d�dM„«Zee\jÍdN«d‚dO„«Zge\jÍdP« dƒdQ„«Zhe\jÓdR« d„ d…dS„«Zje\jÁdT«d†dU„«Zke\jÁdV«edI«fd‡dW„«Zle\j½dV« dˆ d‰dX„«Zme\j½dY« d„ dŠdZ„«Zne\jÁdY«d‹d[„«Zoe\jÍd\«eep¬]«ddf dŒd^„«Zqe\jÓd\« d„ d�d_„«Zre\j½d`« dŽ d�da„«Zse\jÁdb«dc„«Zte\j½dd«de„«Zue>dfk(rŒddlvZvddlwZwexejòdgdh««Zzejòdidj«Z{e?j‡dke{›dlez›�«e?j‡dme{›dlez›dn�«ewjø«evjúe\e{ezdodp¬q«yy)�uo
FastAPI �务入� - Text2SQL NL Chat API
包装 Text2SQLOrchestrator 为 REST API �务，供�端调用
Refactor embedding configuration to remove local model support, transitioning to a unified remote API approach. Update environment variables and documentation accordingly. Enhance error handling in the orchestrator and related modules to reflect these changes. This update simplifies the embedding process and improves overall system reliability.
2026-04-15 09:49:18 +08:00
é)Ú annotationsN)ÚPath)ÚOptionalÚListÚDictÚAnyÚ
AsyncIterator)Úasynccontextmanager)ÚFastAPIÚ
HTTPExceptionÚBodyr)ÚCORSMiddleware)ÚStreamingResponseÚ JSONResponse)Ú BaseModelÚField)Ú load_dotenvÚbackend)Úsetup_environmentÚ load_schemaÚcreate_orchestratorÚresolve_sql_dialect)ÚGenerationResult)Ú
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
lite_nl_store)Ú DialogIntentÚclassify_dialog)Úlast_assistant_was_data_queryÚmessages_to_text2sql_contextÚis_likely_follow_upz1%(asctime)s [%(levelname)s] %(name)s: %(message)sz%Y-%m-%d %H:%M:%S)ÚlevelÚformatÚdatefmtcóì—ddl}tj«dz }|j«r|St |dd«r6t |d«r*t|j «dz }|j«r|St |dd«r8t|j«j}|dz }|j«r|S|Stt«j«j}|dz }|j«r|S|S)u-查找 .env 文件，支æŒ�多ç§�è¿�行环境rNz.envÚfrozenFÚ_MEIPASS) ÚsysrÚcwdÚexistsÚgetattrÚhasattrr%Ú
0.1.1 暂存
2026-04-14 10:28:22 +08:00
executableÚparentÚ__file__Úresolve)r&Úcwd_envÚ meipass_envÚexe_dirÚenv_pathÚ
Refactor embedding configuration to remove local model support, transitioning to a unified remote API approach. Update environment variables and documentation accordingly. Enhance error handling in the orchestrator and related modules to reflect these changes. This update simplifies the embedding process and improves overall system reliability.
2026-04-15 09:49:18 +08:00
script_dirs ú
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
api_server.pyÚ_find_env_filer51sÞ€ãô�h‰h‹j˜6Ñ!€G؇~�~Ô؈ôˆs�H˜eÔ$¬°°jÔ)Aܘ3Ÿ<™<Ó(¨6Ñ1ˆ Ø × Ñ Ô ØÐ ôˆs�H˜eÔ$ä�s—~‘~Ó&×-Ñ-ˆØ˜VÑ#ˆØ �?‰?Ô ØˆOð €Nô
Refactor embedding configuration to remove local model support, transitioning to a unified remote API approach. Update environment variables and documentation accordingly. Enhance error handling in the orchestrator and related modules to reflect these changes. This update simplifies the embedding process and improves overall system reliability.
2026-04-15 09:49:18 +08:00
œ(“^×+Ñ+Ó-×4Ñ4ˆ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
Ø Ñ&ˆØ �?‰?Ô ØˆOð €Nóu[OK] 已加载é…�置文件: u[WARN] 未找到 .env 文件: có(—t�tSt«s td«‚tjdd«}tjdd«}t ||«aGd„d«}tt |««atjd«tS) u获å�–或åˆ�始化 orchestratorNuÃ环境检查失败（请查看上方 WARNING/INFO，常è§�原因：未é…�好 OPENAI_* / MODELSCOPE_* / DASHSCOPE_* 远程 Embedding；或 Schema 文件路径ä¸�对ã€�DEEPSEEK_API_KEY æœªè®¾ç½®ï¼‰Ú SCHEMA_PATHz9./data/schemas/G3SB_MCDataDictionary_table_structure.jsonÚSCHEMA_META_PATHcó"—eZdZejd«Zejdd«Zeejdd««Ze ejdd««Z
e ejdd ««Z ejd
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
d «Z d Z
ejd
Refactor embedding configuration to remove local model support, transitioning to a unified remote API approach. Update environment variables and documentation accordingly. Enhance error handling in the orchestrator and related modules to reflect these changes. This update simplifies the embedding process and improves overall system reliability.
2026-04-15 09:49:18 +08:00
d«j«dvZe ejdd««Ze ejdd««Zejdd«j%«j«dvZy)úget_orchestrator.<locals>.ArgsÚDEEPSEEK_API_KEYÚ
MODEL_PRIMARYz
deepseek-chatÚ TEMPERATUREz0.3Ú
MAX_TOKENSÚ4096Ú MAX_RETRYÚ2ÚVECTOR_DB_PATHz./data/embeddings/chromaFÚFEWSHOT_ENABLEDÚtrue)rEÚ1ÚyesÚ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
FEWSHOT_TOP_KÚ3ÚFEWSHOT_MIN_RATINGÚ7ÚTRANSLATE_EN_TO_ZH)Ú0ÚfalseÚnoÚoffN)Ú__name__Ú
0.1.1 暂存
2026-04-14 10:28:22 +08:00
__module__Ú __qualname__ÚosÚgetenvÚapi_keyÚmodelÚfloatÚ temperatureÚintÚ
max_tokensÚ max_retryÚ vector_dbÚno_vector_searchÚlowerÚ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
no_fewshotÚ
fewshot_top_kÚfewshot_min_ratingÚstripÚno_translate_en©r6r4ÚArgsr;os„Ø�"—)‘)Ð.Ó/ˆØ�— ‘ ˜/¨?Ó;ˆÙ˜I˜BŸI™I m°UÓ;Ó<ˆ Ù˜˜Ÿ™ <°Ó8Ó9ˆ
Ù˜ ˜Ÿ ™  +¨sÓ3Ó4ˆ Ø�B—I‘IÐ.Ð0JÓKˆ Ø ÐØ"˜Ÿ™Ð#4°fÓ=×CÑCÓEÐI]Ð]ˆ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
Ù˜I˜BŸI™I o°sÓ;Ó<ˆ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
Ù   §¡Ð+?ÀÓ!EÓFÐØ#˜"Ÿ)™)Ð$8¸&ÓA×GÑGÓI×OÑOÓQðV
ð
‰r6rfu![OK] Orchestrator åˆ�始化完æˆ�)
Ú orchestratorrÚ RuntimeErrorrTrUrÚschema_managerrÚloggerÚinfo)Ú schema_pathÚschema_meta_pathrfs r4Úget_orchestratorrn\sˆ€ôÐÜÐä Ô Üð
0.1.1 暂存
2026-04-14 10:28:22 +08:00
Ló
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
ð
ô —)‘)˜MÐ+fÓg€KÜ—y‘yÐ!3°TÓ:Ðä  Ð.>Ó?€N÷
ñ
ô$'¤~±t³vÓ>€LÜ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
‡K�KÐ3Ô4ä Ðr6cóþ—eZdZUedd¬«Zded<edd¬«Zded <ed
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
d ¬«Zded <edd
¬«Zded<edd¬«Z ded<edd¬«Z
0.1.1 暂存
2026-04-14 10:28:22 +08:00
ded<edd¬«Z ded<edd¬«Z ded<y)Ú
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
NLChatRequest.u!用户输入的自然语言问题©Ú descriptionÚstrÚmessageNu模型路由: openai | deepseekú
Optional[str]Ú service_codeÚautou语言: zh | en | tc | autoÚ lang_codeu任务IDÚtaskIdõ会è¯�IDÚ
session_idu访客业务IDÚvisitor_biz_idu用户IDÚuser_iduæµ�å¼�节æµ�å�‚æ•°ú
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
Optional[int]Ústreaming_throttle)
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
rQrRrSrrtÚ__annotations__rvrxryr{r|r}rrer6r4rprp‡s�…Ù˜Ð*MÔN€GˆSÓNÙ"'¨Ð:[Ô"\€L�-Ó\Ù$ VÐ9VÔW€Iˆ}ÓWÙ! $°JÔ?€FˆMÓ?Ù % d¸
Ô C€J�
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
ÓCÙ$)¨$Ð<LÔ$M€N�MÓMÙ" 4°ZÔ@€Gˆ]Ó@Ù(-¨dÐ@TÔ(UИ
ÔUr6rpcó4—eZdZUded<dZded<dZded<y)Ú
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
IntentPayloadrsÚintentNzOptional[float]Ú
confidenceruÚreason)rQrRrSr€r„r…rer6r4r‚r‚’s…Ø ƒKØ"&€J�Ó&Ø €FˆMÔ r6r‚cóX—eZdZUedd¬«Zded<eed¬«Zded <eed
0.1.1 暂存
2026-04-14 10:28:22 +08:00
¬«Zd ed <ed
d¬«Z ded<edd¬«Z
ded<edd¬«Z ded<edd¬«Z ded<edd¬«Z
ded<edd ¬«Zded!<edd"¬«Zded#<edd$¬«Zded%<y)&ÚDataQueryResult.u生æˆ�çš„SQL语å�¥rqrsÚsqlu查询结果列å��)Údefault_factoryrrz List[str]Úcolumnsu查询结果行数æ�®zList[Dict[str, Any]]Úrowsru 结果行数rZÚ row_countFu 是å�¦æˆªæ–­ÚboolÚ truncatedNuSQL自然语言说明ruÚ sql_explainTu是å�¦å…�许导出zOptional[bool]Ú
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
can_exportu_库探针：1=有数æ�®è¡Œï¼Œ0=无行需追问补充，-1=执行失败(é‡�试)，None=未探针r~Údb_execution_statusu7探针0时的无行说明与追问（å�«é—®é¢˜åˆ†æž�）Údb_empty_feedbackuNTrue 表示探针为0：需用户补充æ�¡ä»¶å�Žé‡�æ–°æ��问以é‡�新生æˆ�SQLÚfollow_up_requiredu4探针为1æ—¶LLM生æˆ�çš„é�¢å�‘用户SQL交付说明Úsql_delivery_message)rQrRrSrrˆr€ÚlistrŠr‹rŒrŽr�r�r‘r’r“r”rer6r4r‡r‡˜så…Ù�SÐ&:Ô;€CˆÓ;Ù¨tÐAUÔV€GˆYÓVÙ!&°tÐI`Ô!a€DÐ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
ÓaÙ˜1¨.Ô9€IˆsÓ9Ù˜E¨~Ô>€IˆtÓ>Ù!& tÐ9PÔ!Q€K�ÓQÙ!& tÐ9MÔ!N€J�ÓNÙ).Ø Øuô*Иóñ(-Ø ÐSô(Ð�}óñ %Ø
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
Ødô Иóñ+0Ø ÐPô+И-ôr6r‡có^—eZdZUded<ded<edd¬«Zded <edd
¬«Zded <y) ÚNLChatSuccessDatar‚rƒr‡Ú
branch_resultNu$æµ�å¼�å�™è¿°ï¼ˆSQLå�—上方说明）rqruÚstream_narrativeu$æµ�å¼�å�™è¿°ï¼ˆSQLå�—下方说明）Ústream_narrative_after)rQrRrSr€rr™ršrer6r4r—r—°s4…Ø ÓØ"Ó"Ù&+¨DÐ>dÔ&eÐ�mÓeÙ,1°$ÐDjÔ,kИMÔkr6r—cóh—eZdZUedd¬«Zded<edd¬«Zded <ed
d ¬«Zd ed
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
<y
0.1.1 暂存
2026-04-14 10:28:22 +08:00
)Ú ApiEnvelopeéÈu状æ€�ç �，200表示æˆ�功rqrZÚcodeÚu消æ�¯rsÚmsgNu å“�应数æ�®zOptional[NLChatSuccessData]Údata)rQrRrSrržr€r r¡rer6r4rœrœ·s7…Ù�cÐ'DÔE€Dˆ#ÓEÙ�R XÔ.€CˆÓ.Ù(-¨dÀÔ(O€DÐ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
%ÔOr6rœcó0—eZdZUded<ded<dZded<y)Ú
0.1.1 暂存
2026-04-14 10:28:22 +08:00
ErrorResponserZržrsr Nz
Optional[Any]r¡)rQrRrSr€r¡rer6r4r£r£½s…Ø
ƒIØ ƒHØ€Dˆ-Ôr6r£có8—eZdZUdZded<dZded<dZded<y)ÚSessionCreateBodyNruÚtitler}r|©rQrRrSr¦r€r}r|rer6r4r¥r¥Ãó …Ø€Eˆ=ÓØ!€Gˆ]Ó!Ø$(€N�MÔ(r6r¥có8—eZdZUdZded<dZded<dZded<y)ÚSessionTitlePatchBodyNrur¦r}r|r§rer6r4rªrªÉr¨r6rªcó4—eZdZUded<dZded<dZded<y)ÚSessionMessagePatchBodyrsÚcontentNrur}r|)rQrRrSr€r}r|rer6r4r¬r¬Ïs…Ø
ƒLØ!€Gˆ]Ó!Ø$(€N�MÔ(r6r¬có^—eZdZUded<dZded<dZded<dZded<dZded <dZded
<y) ÚSqlExecuteBodyrsrˆNr~Úmax_rowsrur}r|Úchat_session_idÚchat_message_id) rQrRrSr€r°r}r|r±r²rer6r4r¯r¯Õs9…Ø ƒHØ"€HˆmÓ"Ø!€Gˆ]Ó!Ø$(€N�MÓ(Ø%)€O�]Ó)Ø%)€O�]Ô)r6r¯cóª—eZdZUedd¬«Zded<dZded<dZd ed
<dZd ed <dZ d ed <dZ
d ed
<dZ d ed<dZ d ed<dZ
d ed<dZd ed<y)ÚFavoriteCreateBody.zsql | function | reportrqrsÚfav_typerŸÚnameNruÚdescr}r|rˆr�ÚpathÚ
reportPathÚparams)rQrRrSrrµr€r¶r·r}r|rˆr�r¸r¹rºrer6r4r´r´Þsm…Ù˜#Ð+DÔE€HˆcÓEØ€Dˆ#ƒNØ€Dˆ-ÓØ!€Gˆ]Ó!Ø$(€N�MÓ(Ø€CˆÓØ!%€K�Ó%Ø€Dˆ-ÓØ $€J�
Ó$Ø €FˆMÔ r6r´cóV—dtj|d¬«›d�jd«S)Nzdata: F©Ú ensure_asciiz
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
zutf-8)ÚjsonÚdumpsÚencode)Úobjs r4Ú _sse_datarÂës(€Ø
”D—J‘J˜s°Ô7Ð8¸Ð =× DÑ DÀWÓ MÐMr6cóH—|xsdj«xsd}ddddœd|idœS) uP寒暄/元问题等：与å‰�端 BUSINESS_MANUAL + branch_result.answer 一致。rŸu 您好。ÚBUSINESS_MANUALçð?Ú conversation©rƒr„r…Úanswer©rƒr˜)rc)ÚreplyÚrs r4Ú_conversation_nl_dictrÌïs4€à Š�"×ÑÓÒ, €Aà.¸cÈ^Ñ\Ø" A˜ñ ðr6cƒóNK—|jxsdj«}|sytj|j|j
|dd¬«ƒd{–—†}|sy|j
d«xsg}|syt|«}t|«rdnd}t||¬ «\}}||fS7ŒM­w)
u}
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
在写入本轮之�读�会�历�，构造 Text2SQL 上文，并判断上一轮助手是�为数�查询。
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
rŸ)rŸFr�r)ÚlimitÚoffsetNÚitemsé)Ú max_pairs)
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
r{rcrÚ get_messagesr}r|Úgetrrr) ÚrequestÚ user_textÚsidr¡rÐÚ last_datarÒÚblockÚ_ns r4Ú_load_session_text2sql_contextrÛøs®èø€ð × Ñ Ò # ×
0.1.1 暂存
2026-04-14 10:28:22 +08:00
*Ñ
*Ó
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
,€CÙ ØÜ×+Ñ+Ø�‰Ø×ÑØ ØØô ÷ €Dñ ØØ �H‰H�WÓ Ò # €EÙ ØÜ-¨eÓ4€Iä(¨Ô3‘¸€IÜ,¨U¸iÔH�I€Eˆ2Ø �)Ð Ðð! ús‚AB%ÁB#ÁAB%cƒó„K—|jr|jj«sy tj|d¬«}t j
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
|j |j|jj«||«ƒd{–—†y7Œ#t$r"}tjd|›�«Yd}~yd}~wwxYw­w)NFr¼u[API] 会è¯�è�½åº“跳过: ) r{rcr¾r¿rÚappend_exchanger}r|Ú ExceptionrjÚwarning)rÕrÖr¡Úassistant_jsonÚes r4Ú_append_session_if_neededrâs èø€Ø × Ò  7×#5Ñ#5×#;Ñ#;Ô#=Øð
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
9ÜŸ™ D°uÔ=ˆÜ×+Ñ+Ø �O‰OØ × "Ñ "Ø × Ñ × $Ñ $Ó &Ø Ø ó 
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
÷
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
ò
ùô ò9Ü�‰Ð3°A°3Ð7×8Ñ8ûð9üsA‚(C«ABÂ
B BÂCÂB B=ÂB8Â3CÂ8B=Â=Ccó"—|jxsdj«}g}|jjd«r'|j t |jd««|j r"|jd„|j D««|jr"|jd„|jD««|rdj|«nd}|ggdddœ}|r||d <|jjd
«}|�||d
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
<|jjd«}|r||d<t|jxr|dk(«|d <|jjd «}|rt |«|d <|jrd
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
nd}|jrdt|j«›d�} n|jr|jdnd} d|| dœ|dœ}
|jjd«} |jjd«} | r
| r| | dœ|
d<|
S)NrŸr’c3ó2K—|]}t|«–—Œy­w©N©rs)Ú.0Úws r4Ú <genexpr>z+_nl_dict_from_generation.<locals>.<genexpr>+sèø€Ð=©_¨œS ŸV©_ùó‚c3ó2K—|]}t|«–—Œy­wråræ©rçrás r4réz+_nl_dict_from_generation.<locals>.<genexpr>-sèø€Ð;©]¨œS ŸV©]ùrêú; rF)rˆrŠr‹rŒrŽr�r‘r“r”rÅgu
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
使用了 u 张表uSQL生æˆ�未通过验è¯�Ú
DATA_QUERYrÇrÉÚquestion_originalÚquestion_zh_normalized)ÚoriginalÚzhÚquery_normalization)rˆrcÚmetadatarÔÚappendrsÚwarningsÚextendÚerrorsÚjoinr�ÚvalidÚlenÚ tables_used)
ÚresultrˆÚ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
explain_partsÚexplainr˜ÚdbsÚdbeÚsdmÚconfr…ÚpayloadÚqoÚqzs
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
r4Ú_nl_dict_from_generationr%sî€Ø �:‰:Ò ˜×
0.1.1 暂存
2026-04-14 10:28:22 +08:00
"Ñ
"Ó
$€CØ!€MØ
‡�×ÑÐ.Ô/Ø×ÑœS §¡Ð1DÑ!EÓFÔGØ
‡‚Ø×ÑÑ=¨V¯_ª_Ó=Ô=Ø
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
‡}‚}Ø×ÑÑ;¨V¯]ª]Ó;Ô;Ù*7ˆd�i‰i˜
Ô&¸T€GàØØØØñ %€MñØ'.ˆ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
�mÑ$Ø
�/‰/×
Ñ
Ð3Ó
4€CØ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
€Ø/2ˆ
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
Ð+Ñ,Ø
�/‰/×
Ñ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
Ð1Ó
2€CÙ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
Ø-0ˆ
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
Ð)Ñ*Ü*.¨v¯|©|Ò/HÀÀqÁÓ*I€MÐ&Ñ'Ø
�/‰/×
Ñ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
Ð4Ó
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
5€CÙ
Ü03°C³ˆ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
Ð,Ñ-Ø—,’,‰3 C€DØ
‡|‚|Øœc &×"4Ñ"4Ó5Ð6°gÐ>‰à%+§]¢]�—‘˜qÒ!Ð8Rˆà)¸ÈÑPØ&ñ€Gð
�‰× Ñ Ð0Ó 1€BØ �‰× Ñ Ð5Ó 6€BÙ ‰bØ68ÀÑ)CˆÐ%Ñ&Ø €Nr6c óÚ—|jxsdj«}tjd|id¬«}d|›d�g}d}|jj d«r>|j
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
dtjt|jd««›d �«|jj d
0.1.1 暂存
2026-04-14 10:28:22 +08:00
«rt|jd
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
«}nV|jrt|jd «}n1|jr%d jd
0.1.1 暂存
2026-04-14 10:28:22 +08:00
„|jddD««}|j«r(|j
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
dtj|«›d �«dj|«S)NrŸrˆFr¼z<data>z</data>r”z<div class="sql-delivery-body">z</div>r’rríc3ó2K—|]}t|«–—Œy­wrårærìs r4réz'_sql_gen_stream_html.<locals>.<genexpr>`sèø€Ð>Ñ,= qœC ŸFÑ,=ùrêéz<div class="sql-explain-body">)
0.1.1 暂存
2026-04-14 10:28:22 +08:00
rˆrcr¾r¿rôrÔrõÚhtml_libÚescapersrörørù)rýrˆÚinnerÚpartsrÿs r4Ú_sql_gen_stream_htmlrRs(€Ø �:‰:Ò ˜×
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
"Ñ
"Ó
$€CÜ �J‰J˜˜s�|°%Ô 8€EØ�e�W˜GÐ
$Ð %€EØ€GØ
‡�×ÑÐ1Ô2Ø
0.1.1 暂存
2026-04-14 10:28:22 +08:00
� ‰ Ø-¬h¯o©o¼cÀ&Ç/Á/ÐRhÑBiÓ>jÓ.kÐ-lÐlrÐ sô
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
ð‡�×ÑÐ.Ô/Ü�f—o‘oÐ&9Ñ:Ó;‰Ø �ŠÜ�f—o‘o aÑ(Ó)‰Ø �ŠØ—)‘)Ñ>¨F¯M©M¸"¸1Ñ,=Ó>Ó>ˆØ‡}�}„Ø
0.1.1 暂存
2026-04-14 10:28:22 +08:00
� ‰ Ð5´h·o±oÀgÓ6NÐ5OÈvÐVÔWØ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
�7‰7�5‹>Ðr6cƒó臇‡‡‡K—t«Šttjdd««Š|xsdj «xsdŠdˆˆˆˆˆfd„ }t j |«ƒd{–—†S7Œ­w)NÚTEXT2SQL_DIALECTÚ sqlserverrŸcóJ•—‰j‰j«‰‰‰¬«S)N)ÚquestionÚdialectÚtop_k_candidatesÚdialog_context)Úgeneraterc)ÚdcrÚorchrÚtop_ks€€€€€r4Ú_callz_run_generate.<locals>._callos,ø€Ø�}‰}Ø—^‘^Ó%ØØ"Øð ó
ð
r6)Úreturnr)rnrrTrUrcÚasyncioÚ to_thread)rrrrrrrs`` @@@r4Ú
_run_generater fsbüèø€ô
Ó €DÜ!¤"§)¡)Ð,>À Ó"LÓM€GØ
Ò
˜B× %Ñ %Ó 'Ò /¨4€B÷
0.1.1 暂存
2026-04-14 10:28:22 +08:00
ñ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
ô×"Ñ" 5Ó)× )Ð)Ð )ús‡A$A2Á+A0Á,A2cóJK—|jr|jj«stddddœ«­–—y|jj«}t«}t ||«ƒd{–—†\}}t j t|||xsd|j¬«ƒd{–—†}|jtjk(r||jxsd}tjd«tddd d
0.1.1 暂存
2026-04-14 10:28:22 +08:00
œ«­–—td d|d
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
œ«­–—t|«}td d
0.1.1 暂存
2026-04-14 10:28:22 +08:00
|dœ«­–—t!|||«ƒd{–—†ytdddd
œ«­–— t#||xsd¬«ƒd{–—†}tddt+|«d
Refactor embedding configuration to remove local model support, transitioning to a unified remote API approach. Update environment variables and documentation accordingly. Enhance error handling in the orchestrator and related modules to reflect these changes. This update simplifies the embedding process and improves overall system reliability.
2026-04-15 09:49:18 +08:00
œ«­–—t-|«}|j.rd
nd}
td |
0.1.1 暂存
2026-04-14 10:28:22 +08:00
|dœ«­–—t!|||«ƒd{–—†y7�Œ_7�Œ+7Œ•7Œj#t$$r=} tj'd| ›�«tdt)| «ddœ«­–—Yd} ~ yd} ~ wwxYw7ŒX­w)Né�õ消æ�¯å†…容ä¸�能为空©ržr r¡©Úlast_turn_was_data_queryrÚ
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
llm_clientrŸuZ[API/stream] 对è¯�æ„�图: conversation（跳过 Text2SQL，与 CLI single_query 一致）rgr­ÚCHAT)ÚstageÚ stream_kindr­Úchatr�Úsuccessrî©ruPI/stream] 生æˆ�异常: éôÚsql_genÚpartial)rtrcrÂrnrÛrrrÚdeepseekrƒrÚ CONVERSATIONÚreply_suggestionrjrkrÌrâr rÞÚerrorrsrrrú) rÕÚtextrÚ dialog_blockrØÚ
0.1.1 暂存
2026-04-14 10:28:22 +08:00
classifiedrÊÚ data_dictrýrár s r4Ú_chat_stream_eventsr9zsõèø€Ø �?Š? '§/¡/×"7Ñ"7Ô"9Ü Ð-GÐQUÑVÓWÓWØØ �?‰?× Ñ Ó "€DÜ Ó €DÜ$BÀ7ÈDÓ$Q×QÑ€L�)Ü×(Ñ(ÜØ Ø!*Ø#Ò+ tØ—=‘=ô ÷€Jð×ÑœL×5Ñ5Ò5Ø×+Ñ+Ò1¨rˆÜ� ‰ ÐpÔqÜ .ÀÐW]Ñ^Ó_Ó_Ü &¸ÈuÑUÓVÓVÜ)¨%Ó0ˆ Ü ¨YÀ ÑJÓKÓKÜ'¨°°yÓA×AÐAØä
˜n¸YÐS_Ñ`Ó
Enhance dialog context handling for Text2SQL queries by integrating session history. Introduce new methods for summarizing previous assistant messages and determining if the last interaction was a data query. Update environment configuration for embedding options and improve error handling in SQL generation. Add user-facing delivery messages for successful SQL execution. This update supports more coherent follow-up questions and improves user experience in conversational interactions.
2026-04-14 18:02:12 +08:00
aÓaðÜ$ T¸,Ò:NÈ$ÔO×Oˆô
0.1.1 暂存
2026-04-14 10:28:22 +08:00
˜i¸ ÔNbÐciÓNjÑkÓ
lÓlÜ(¨Ó0€IØ—|’|‰)¨€CÜ
˜S¨°iÑ@Ó
AÓAÜ
# G¨T°9Ó
=×=Ñ=ð;Rùðùð Bøð
PùÜ òÜ� ‰ Ð2°1°#Ð6Ô7Ü ¬S°«V¸TÑBÓCÓCÜûðúð>ús€‚A-H#Á/GÁ06H#Â&GÂ'BH#Ä>GÄ?H#ÅGÅ+GÅ,GÅ0AH#ÇH!Ç H#ÇH#ÇH#ÇGÇ HÇ!3HÈH#ÈHÈH#cóRK—tjd«tjd«tjd« t«tjd«d­–—tjd«y#t$r}tj d|›�«‚d}~wwxYw­w)Nz<============================================================u Text2SQL API Server å�¯åŠ¨ä¸­...u[OK] æœ�务已就绪uæœ�务åˆ�始化失败: uText2SQL API Server 已关闭)rjrkrnrÞr4)Úapprás r4Úlifespanr< s}èø€ä
‡K�K�ÔÜ
‡K�KÐ2Ô3Ü
‡K�K�ÔðÜÔÜ� ‰ Ð*Ô+ô

ä
‡K�KÐ/Õ0øô
òÜ� ‰ Ð.¨q¨cÐ2Ô3Ø
ûðüs*‚AB'ÁA=Á"B'Á= B$ÂBÂB$Â$B'zText2SQL NL Chat APIu!自然语言转SQL的对è¯�接å�£z1.0.0)r¦rrÚversionr<Ú*T)Ú
allow_originsÚallow_credentialsÚ
allow_methodsÚ
allow_headersz/healthcƒóK—dddœS­w)uå�¥åº·æ£€æŸ¥æŽ¥å�£ÚokzText2SQL API is running)Ústatusrtrerer6r4Ú health_checkrFÂsèø€ðÐ'@Ñ AÐAùs‚z/g3sb/api/nl/chatcƒóŠK— |jr|jj«s
tdd¬«‚|jj«}tj d|dd›d�«t «}t
||«ƒd{–—†\}}tjt|||xsd|j¬«ƒd{–—†}|jtjk(rK|jxsd }tj d
«t|«}t!|||«ƒd{–—†d d |d
œSt#||xsd¬«ƒd{–—†}t%|«}t&j)|«} |j*rd nd}
|j*r+tj d|j,xsd dd›d�«n"tj/d|j0›�«t!|||«ƒd{–—†t3d |
| ¬
«S7�Œi7�Œ57ŒÕ7Œ¸7Œ#t$r‚t4$r3} tj7d| ›�«tdt9| «¬«‚d} ~ wwxYw­w)uš
自然语言对�接�（���），�应结构与��结�包一致。
寒暄/致谢等先� classify_dialog，与 CLI 一致。
r"r#)Ú status_codeÚdetailu[API] 问题: Nédz...r%rŸu:[API/chat] 对è¯�æ„�图: conversation（跳过 Text2SQL）r�r,r$r-r0u[API] 生æˆ�æˆ�功: éPu[API] 生æˆ�失败: u[API] 异常: r.)rtrcr rjrkrnrÛrrrr1rƒrr2r3rÌrâr rr—Úmodel_validaterúrˆrßrørœrÞr4rs) rÕr5rr6rØr7rÊr8rýÚ
Refactor embedding configuration to remove local model support, transitioning to a unified remote API approach. Update environment variables and documentation accordingly. Enhance error handling in the orchestrator and related modules to reflect these changes. This update simplifies the embedding process and improves overall system reliability.
2026-04-15 09:49:18 +08:00
response_datar rás r4Únl_chatrNÈsèø€ð #<Ø�Š g§o¡o×&;Ñ&;Ô&=ܨCÐ8RÔSÐ SØ�‰×$Ñ$Ó&ˆÜ� ‰ �n T¨$¨3 Z L°Ð4Ô5ÜÓ!ˆÜ(FÀwÐPTÓ(U×"Uш �iÜ"×,Ñ,Ü Ø Ø%.Ø'Ò/¨4Ø—}‘}ô 
0.1.1 暂存
2026-04-14 10:28:22 +08:00
÷
Implement follow-up detection in dialog context to prevent context pollution in multi-topic conversations. Introduce is_likely_follow_up function to determine when to inject session history based on user input. Adjust _load_session_text2sql_context to conditionally include context based on follow-up status, enhancing SQL generation accuracy. Update relevant API endpoints to pass user text for improved context handling.
2026-04-15 17:33:07 +08:00
ˆ
Copy Permalink