FastGPT/packages/service/core/ai/rerank/index.ts
heheer 5231f4281f
image compatibility for various content-types (#6119)
* image compatibility for various content-types

* perf: image type detect

* perf: gethistory

* update test

* update rerank log

* perf: login

* fix: query extension use

---------

Co-authored-by: archer <545436317@qq.com>
2025-12-18 23:25:48 +08:00

89 lines
2.2 KiB
TypeScript

import { addLog } from '../../../common/system/log';
import { POST } from '../../../common/api/serverRequest';
import { getDefaultRerankModel } from '../model';
import { getAxiosConfig } from '../config';
import { type RerankModelItemType } from '@fastgpt/global/core/ai/model.d';
import { countPromptTokens } from '../../../common/string/tiktoken';
type PostReRankResponse = {
id: string;
results: {
index: number;
relevance_score: number;
}[];
meta?: {
tokens: {
input_tokens: number;
output_tokens: number;
};
};
};
type ReRankCallResult = {
results: { id: string; score?: number }[];
inputTokens: number;
};
export function reRankRecall({
model = getDefaultRerankModel(),
query,
documents,
headers
}: {
model?: RerankModelItemType;
query: string;
documents: { id: string; text: string }[];
headers?: Record<string, string>;
}): Promise<ReRankCallResult> {
if (!model) {
return Promise.reject('No rerank model');
}
if (documents.length === 0) {
return Promise.resolve({
results: [],
inputTokens: 0
});
}
const { baseUrl, authorization } = getAxiosConfig();
let start = Date.now();
const documentsTextArray = documents.map((doc) => doc.text);
return POST<PostReRankResponse>(
model.requestUrl ? model.requestUrl : `${baseUrl}/rerank`,
{
model: model.model,
query,
documents: documentsTextArray
},
{
headers: {
Authorization: model.requestAuth ? `Bearer ${model.requestAuth}` : authorization,
...headers
},
timeout: 30000
}
)
.then(async (data) => {
addLog.info('ReRank finish:', { time: Date.now() - start });
if (!data?.results || data?.results?.length === 0) {
addLog.error('[Rerank Error]', { message: 'Empty result', data });
}
return {
results: data?.results?.map((item) => ({
id: documents[item.index].id,
score: item.relevance_score
})),
inputTokens:
data?.meta?.tokens?.input_tokens ||
(await countPromptTokens(documentsTextArray.join('\n') + query, ''))
};
})
.catch((err) => {
addLog.error('[Rerank Error]', err);
return Promise.reject(err);
});
}