V4.6.6-2 (#673)

projects/app/src/service/moduleDispatch/agent/classifyQuestion.ts

@@ -9,8 +9,9 @@ import type { ModuleDispatchProps } from '@fastgpt/global/core/module/type.d';
 import { replaceVariable } from '@fastgpt/global/common/string/tools';
 import { Prompt_CQJson } from '@/global/core/prompt/agent';
 import { FunctionModelItemType } from '@fastgpt/global/core/ai/model.d';
-import { getCQModel } from '@/service/core/ai/model';
+import { ModelTypeEnum, getCQModel } from '@/service/core/ai/model';
 import { getHistories } from '../utils';
+import { formatModelPrice2Store } from '@/service/support/wallet/bill/utils';
 
 type Props = ModuleDispatchProps<{
   [ModuleInputKeyEnum.aiModel]: string;
@@ -42,7 +43,7 @@ export const dispatchClassifyQuestion = async (props: Props): Promise<CQResponse
 
   const chatHistories = getHistories(history, histories);
 
-  const { arg, tokens } = await (async () => {
+  const { arg, inputTokens, outputTokens } = await (async () => {
     if (cqModel.toolChoice) {
       return toolChoice({
         ...props,
@@ -59,13 +60,21 @@ export const dispatchClassifyQuestion = async (props: Props): Promise<CQResponse
 
   const result = agents.find((item) => item.key === arg?.type) || agents[agents.length - 1];
 
+  const { total, modelName } = formatModelPrice2Store({
+    model: cqModel.model,
+    inputLen: inputTokens,
+    outputLen: outputTokens,
+    type: ModelTypeEnum.cq
+  });
+
   return {
     [result.key]: result.value,
     [ModuleOutputKeyEnum.responseData]: {
-      price: user.openaiAccount?.key ? 0 : cqModel.price * tokens,
-      model: cqModel.name || '',
+      price: user.openaiAccount?.key ? 0 : total,
+      model: modelName,
       query: userChatInput,
-      tokens,
+      inputTokens,
+      outputTokens,
       cqList: agents,
       cqResult: result.value,
       contextTotalLen: chatHistories.length + 2
@@ -140,7 +149,8 @@ ${systemPrompt}
 
     return {
       arg,
-      tokens: response.usage?.total_tokens || 0
+      inputTokens: response.usage?.prompt_tokens || 0,
+      outputTokens: response.usage?.completion_tokens || 0
     };
   } catch (error) {
     console.log(agentFunction.parameters);
@@ -150,7 +160,8 @@ ${systemPrompt}
 
     return {
       arg: {},
-      tokens: 0
+      inputTokens: 0,
+      outputTokens: 0
     };
   }
 }
@@ -182,12 +193,12 @@ Human:${userChatInput}`
     stream: false
   });
   const answer = data.choices?.[0].message?.content || '';
-  const totalTokens = data.usage?.total_tokens || 0;
 
   const id = agents.find((item) => answer.includes(item.key))?.key || '';
 
   return {
-    tokens: totalTokens,
+    inputTokens: data.usage?.prompt_tokens || 0,
+    outputTokens: data.usage?.completion_tokens || 0,
     arg: { type: id }
   };
 }

projects/app/src/service/moduleDispatch/agent/extract.ts

@@ -10,7 +10,8 @@ import { Prompt_ExtractJson } from '@/global/core/prompt/agent';
 import { replaceVariable } from '@fastgpt/global/common/string/tools';
 import { FunctionModelItemType } from '@fastgpt/global/core/ai/model.d';
 import { getHistories } from '../utils';
-import { getExtractModel } from '@/service/core/ai/model';
+import { ModelTypeEnum, getExtractModel } from '@/service/core/ai/model';
+import { formatModelPrice2Store } from '@/service/support/wallet/bill/utils';
 
 type Props = ModuleDispatchProps<{
   [ModuleInputKeyEnum.history]?: ChatItemType[];
@@ -42,7 +43,7 @@ export async function dispatchContentExtract(props: Props): Promise<Response> {
   const extractModel = getExtractModel(model);
   const chatHistories = getHistories(history, histories);
 
-  const { arg, tokens } = await (async () => {
+  const { arg, inputTokens, outputTokens } = await (async () => {
     if (extractModel.toolChoice) {
       return toolChoice({
         ...props,
@@ -79,16 +80,24 @@ export async function dispatchContentExtract(props: Props): Promise<Response> {
     }
   }
 
+  const { total, modelName } = formatModelPrice2Store({
+    model: extractModel.model,
+    inputLen: inputTokens,
+    outputLen: outputTokens,
+    type: ModelTypeEnum.extract
+  });
+
   return {
     [ModuleOutputKeyEnum.success]: success ? true : undefined,
     [ModuleOutputKeyEnum.failed]: success ? undefined : true,
     [ModuleOutputKeyEnum.contextExtractFields]: JSON.stringify(arg),
     ...arg,
     [ModuleOutputKeyEnum.responseData]: {
-      price: user.openaiAccount?.key ? 0 : extractModel.price * tokens,
-      model: extractModel.name || '',
+      price: user.openaiAccount?.key ? 0 : total,
+      model: modelName,
       query: content,
-      tokens,
+      inputTokens,
+      outputTokens,
       extractDescription: description,
       extractResult: arg,
       contextTotalLen: chatHistories.length + 2
@@ -181,10 +190,10 @@ ${description || '根据用户要求获取适当的 JSON 字符串。'}
     }
   })();
 
-  const tokens = response.usage?.total_tokens || 0;
   return {
     rawResponse: response?.choices?.[0]?.message?.tool_calls?.[0]?.function?.arguments || '',
-    tokens,
+    inputTokens: response.usage?.prompt_tokens || 0,
+    outputTokens: response.usage?.completion_tokens || 0,
     arg
   };
 }
@@ -223,7 +232,8 @@ Human: ${content}`
     stream: false
   });
   const answer = data.choices?.[0].message?.content || '';
-  const totalTokens = data.usage?.total_tokens || 0;
+  const inputTokens = data.usage?.prompt_tokens || 0;
+  const outputTokens = data.usage?.completion_tokens || 0;
 
   // parse response
   const start = answer.indexOf('{');
@@ -232,7 +242,8 @@ Human: ${content}`
   if (start === -1 || end === -1)
     return {
       rawResponse: answer,
-      tokens: totalTokens,
+      inputTokens,
+      outputTokens,
       arg: {}
     };
 
@@ -244,13 +255,15 @@ Human: ${content}`
   try {
     return {
       rawResponse: answer,
-      tokens: totalTokens,
+      inputTokens,
+      outputTokens,
       arg: JSON.parse(jsonStr) as Record<string, any>
     };
   } catch (error) {
     return {
       rawResponse: answer,
-      tokens: totalTokens,
+      inputTokens,
+      outputTokens,
       arg: {}
     };
   }

projects/app/src/service/moduleDispatch/chat/oneapi.ts

@@ -6,7 +6,7 @@ import { sseResponseEventEnum } from '@fastgpt/service/common/response/constant'
 import { textAdaptGptResponse } from '@/utils/adapt';
 import { getAIApi } from '@fastgpt/service/core/ai/config';
 import type { ChatCompletion, StreamChatType } from '@fastgpt/global/core/ai/type.d';
-import { countModelPrice } from '@/service/support/wallet/bill/utils';
+import { formatModelPrice2Store } from '@/service/support/wallet/bill/utils';
 import type { ChatModelItemType } from '@fastgpt/global/core/ai/model.d';
 import { postTextCensor } from '@/service/common/censor';
 import { ChatCompletionRequestMessageRoleEnum } from '@fastgpt/global/core/ai/constant';
@@ -151,7 +151,7 @@ export const dispatchChatCompletion = async (props: ChatProps): Promise<ChatResp
     }
   );
 
-  const { answerText, totalTokens, completeMessages } = await (async () => {
+  const { answerText, inputTokens, outputTokens, completeMessages } = await (async () => {
     if (stream) {
       // sse response
       const { answer } = await streamResponse({
@@ -165,21 +165,26 @@ export const dispatchChatCompletion = async (props: ChatProps): Promise<ChatResp
         value: answer
       });
 
-      const totalTokens = countMessagesTokens({
-        messages: completeMessages
-      });
-
       targetResponse({ res, detail, outputs });
 
       return {
         answerText: answer,
-        totalTokens,
+        inputTokens: countMessagesTokens({
+          messages: filterMessages
+        }),
+        outputTokens: countMessagesTokens({
+          messages: [
+            {
+              obj: ChatRoleEnum.AI,
+              value: answer
+            }
+          ]
+        }),
         completeMessages
       };
     } else {
       const unStreamResponse = response as ChatCompletion;
       const answer = unStreamResponse.choices?.[0]?.message?.content || '';
-      const totalTokens = unStreamResponse.usage?.total_tokens || 0;
 
       const completeMessages = filterMessages.concat({
         obj: ChatRoleEnum.AI,
@@ -188,20 +193,27 @@ export const dispatchChatCompletion = async (props: ChatProps): Promise<ChatResp
 
       return {
         answerText: answer,
-        totalTokens,
+        inputTokens: unStreamResponse.usage?.prompt_tokens || 0,
+        outputTokens: unStreamResponse.usage?.completion_tokens || 0,
         completeMessages
       };
     }
   })();
 
+  const { total, modelName } = formatModelPrice2Store({
+    model,
+    inputLen: inputTokens,
+    outputLen: outputTokens,
+    type: ModelTypeEnum.chat
+  });
+
   return {
     answerText,
     responseData: {
-      price: user.openaiAccount?.key
-        ? 0
-        : countModelPrice({ model, tokens: totalTokens, type: ModelTypeEnum.chat }),
-      model: modelConstantsData.name,
-      tokens: totalTokens,
+      price: user.openaiAccount?.key ? 0 : total,
+      model: modelName,
+      inputTokens,
+      outputTokens,
       query: userChatInput,
       maxToken: max_tokens,
       quoteList: filterQuoteQA,
@@ -227,8 +239,7 @@ function filterQuote({
       a: item.a,
       source: item.sourceName,
       sourceId: String(item.sourceId || 'UnKnow'),
-      index: index + 1,
-      score: item.score?.toFixed(4)
+      index: index + 1
     });
   }
 

projects/app/src/service/moduleDispatch/dataset/search.ts

@@ -1,13 +1,12 @@
 import type { moduleDispatchResType } from '@fastgpt/global/core/chat/type.d';
-import { countModelPrice } from '@/service/support/wallet/bill/utils';
+import { formatModelPrice2Store } from '@/service/support/wallet/bill/utils';
 import type { SelectedDatasetType } from '@fastgpt/global/core/module/api.d';
 import type { SearchDataResponseItemType } from '@fastgpt/global/core/dataset/type';
 import type { ModuleDispatchProps } from '@fastgpt/global/core/module/type.d';
 import { ModelTypeEnum } from '@/service/core/ai/model';
-import { searchDatasetData } from '@/service/core/dataset/data/pg';
+import { searchDatasetData } from '@/service/core/dataset/data/controller';
 import { ModuleInputKeyEnum, ModuleOutputKeyEnum } from '@fastgpt/global/core/module/constants';
 import { DatasetSearchModeEnum } from '@fastgpt/global/core/dataset/constant';
-import { searchQueryExtension } from '@fastgpt/service/core/ai/functions/queryExtension';
 
 type DatasetSearchProps = ModuleDispatchProps<{
   [ModuleInputKeyEnum.datasetSelectList]: SelectedDatasetType;
@@ -15,6 +14,7 @@ type DatasetSearchProps = ModuleDispatchProps<{
   [ModuleInputKeyEnum.datasetLimit]: number;
   [ModuleInputKeyEnum.datasetSearchMode]: `${DatasetSearchModeEnum}`;
   [ModuleInputKeyEnum.userChatInput]: string;
+  [ModuleInputKeyEnum.datasetSearchUsingReRank]: boolean;
 }>;
 export type DatasetSearchResponse = {
   [ModuleOutputKeyEnum.responseData]: moduleDispatchResType;
@@ -27,7 +27,7 @@ export async function dispatchDatasetSearch(
   props: DatasetSearchProps
 ): Promise<DatasetSearchResponse> {
   const {
-    inputs: { datasets = [], similarity = 0.4, limit = 5, searchMode, userChatInput }
+    inputs: { datasets = [], similarity, limit = 1500, usingReRank, searchMode, userChatInput }
   } = props as DatasetSearchProps;
 
   if (!Array.isArray(datasets)) {
@@ -52,14 +52,21 @@ export async function dispatchDatasetSearch(
   const concatQueries = [userChatInput];
 
   // start search
-  const { searchRes, tokenLen } = await searchDatasetData({
+  const { searchRes, tokens, usingSimilarityFilter } = await searchDatasetData({
     rawQuery: userChatInput,
     queries: concatQueries,
     model: vectorModel.model,
     similarity,
     limit,
     datasetIds: datasets.map((item) => item.datasetId),
-    searchMode
+    searchMode,
+    usingReRank
+  });
+
+  const { total, modelName } = formatModelPrice2Store({
+    model: vectorModel.model,
+    inputLen: tokens,
+    type: ModelTypeEnum.vector
   });
 
   return {
@@ -67,17 +74,14 @@ export async function dispatchDatasetSearch(
     unEmpty: searchRes.length > 0 ? true : undefined,
     quoteQA: searchRes,
     responseData: {
-      price: countModelPrice({
-        model: vectorModel.model,
-        tokens: tokenLen,
-        type: ModelTypeEnum.vector
-      }),
+      price: total,
       query: concatQueries.join('\n'),
-      model: vectorModel.name,
-      tokens: tokenLen,
-      similarity,
+      model: modelName,
+      inputTokens: tokens,
+      similarity: usingSimilarityFilter ? similarity : undefined,
       limit,
-      searchMode
+      searchMode,
+      searchUsingReRank: usingReRank
     }
   };
 }

projects/app/src/service/moduleDispatch/tools/cfr.ts

@@ -4,7 +4,8 @@ import { ModuleInputKeyEnum, ModuleOutputKeyEnum } from '@fastgpt/global/core/mo
 import { getHistories } from '../utils';
 import { getAIApi } from '@fastgpt/service/core/ai/config';
 import { replaceVariable } from '@fastgpt/global/common/string/tools';
-import { getExtractModel } from '@/service/core/ai/model';
+import { ModelTypeEnum, getExtractModel } from '@/service/core/ai/model';
+import { formatModelPrice2Store } from '@/service/support/wallet/bill/utils';
 
 type Props = ModuleDispatchProps<{
   [ModuleInputKeyEnum.aiModel]: string;
@@ -75,13 +76,22 @@ A: ${systemPrompt}
   // );
   // console.log(answer);
 
-  const tokens = result.usage?.total_tokens || 0;
+  const inputTokens = result.usage?.prompt_tokens || 0;
+  const outputTokens = result.usage?.completion_tokens || 0;
+
+  const { total, modelName } = formatModelPrice2Store({
+    model: extractModel.model,
+    inputLen: inputTokens,
+    outputLen: outputTokens,
+    type: ModelTypeEnum.extract
+  });
 
   return {
     [ModuleOutputKeyEnum.responseData]: {
-      price: extractModel.price * tokens,
-      model: extractModel.name || '',
-      tokens,
+      price: total,
+      model: modelName,
+      inputTokens,
+      outputTokens,
       query: userChatInput,
       textOutput: answer
     },