+ Be sure to select a valid embedding model. Chat models are not + embedding models. See{" "} + + this page + {" "} + for more information. +
+
+ 
+
+ AnythingLLM: あなたが探していたオールインワンAIアプリ。
+ ドキュメントとチャットし、AIエージェントを使用し、高度にカスタマイズ可能で、複数ユーザー対応、面倒な設定は不要です。
+
+
+ 
+ |
+
+ 
+👉 デスクトップ用AnythingLLM（Mac、Windows、Linux対応）！今すぐダウンロード +
+ +これは、任意のドキュメント、リソース、またはコンテンツの断片を、チャット中にLLMが参照として使用できるコンテキストに変換できるフルスタックアプリケーションです。このアプリケーションを使用すると、使用するLLMまたはベクトルデータベースを選択し、マルチユーザー管理と権限をサポートできます。 + + + +デモを見る！
+ +[](https://youtu.be/f95rGD9trL0) + +AnythingLLMのテレメトリーとプライバシーについての詳細
+ +### なぜ？ + +この情報を使用して、AnythingLLMの使用方法を理解し、新機能とバグ修正の優先順位を決定し、AnythingLLMのパフォーマンスと安定性を向上させるのに役立てます。 + +### オプトアウト + +サーバーまたはdockerの.env設定で`DISABLE_TELEMETRY`を「true」に設定して、テレメトリーからオプトアウトします。アプリ内でも、サイドバー > `プライバシー`に移動してテレメトリーを無効にすることができます。 + +### 明示的に追跡するもの + +製品およびロードマップの意思決定に役立つ使用詳細のみを追跡します。具体的には： + +- インストールのタイプ（Dockerまたはデスクトップ） +- ドキュメントが追加または削除されたとき。ドキュメントについての情報はありません。イベントが発生したことのみを知ります。これにより、使用状況を把握できます。 +- 使用中のベクトルデータベースのタイプ。どのベクトルデータベースプロバイダーが最も使用されているかを知り、更新があったときに優先して変更を行います。 +- 使用中のLLMのタイプ。最も人気のある選択肢を知り、更新があったときに優先して変更を行います。 +- チャットが送信された。これは最も一般的な「イベント」であり、すべてのインストールでのこのプロジェクトの日常的な「アクティビティ」についてのアイデアを提供します。再び、イベントのみが送信され、チャット自体の性質や内容に関する情報はありません。 + +これらの主張を検証するには、`Telemetry.sendTelemetry`が呼び出されるすべての場所を見つけてください。また、これらのイベントは出力ログに書き込まれるため、送信された具体的なデータも確認できます。IPアドレスやその他の識別情報は収集されません。テレメトリープロバイダーは[PostHog](https://posthog.com/)です。 + +[ソースコード内のすべてのテレメトリーイベントを表示](https://github.com/search?q=repo%3AMintplex-Labs%2Fanything-llm%20.sendTelemetry\(&type=code) + ++このプロジェクトは[MIT](./LICENSE)ライセンスの下でライセンスされています。 + + + +[back-to-top]: https://img.shields.io/badge/-BACK_TO_TOP-222628?style=flat-square +[profile-link]: https://github.com/mintplex-labs +[vector-admin]: https://github.com/mintplex-labs/vector-admin +[assistant-swarm]: https://github.com/Mintplex-Labs/openai-assistant-swarm +[docker-btn]: ./images/deployBtns/docker.png +[docker-deploy]: ./docker/HOW_TO_USE_DOCKER.md +[aws-btn]: ./images/deployBtns/aws.png +[aws-deploy]: ./cloud-deployments/aws/cloudformation/DEPLOY.md +[gcp-btn]: https://deploy.cloud.run/button.svg +[gcp-deploy]: ./cloud-deployments/gcp/deployment/DEPLOY.md +[do-btn]: https://www.deploytodo.com/do-btn-blue.svg +[do-deploy]: ./cloud-deployments/digitalocean/terraform/DEPLOY.md +[render-btn]: https://render.com/images/deploy-to-render-button.svg +[render-deploy]: https://render.com/deploy?repo=https://github.com/Mintplex-Labs/anything-llm&branch=render +[render-btn]: https://render.com/images/deploy-to-render-button.svg +[render-deploy]: https://render.com/deploy?repo=https://github.com/Mintplex-Labs/anything-llm&branch=render +[railway-btn]: https://railway.app/button.svg +[railway-deploy]: https://railway.app/template/HNSCS1?referralCode=WFgJkn diff --git a/README.zh-CN.md b/locales/README.zh-CN.md similarity index 99% rename from README.zh-CN.md rename to locales/README.zh-CN.md index 2385a04c..13505220 100644 --- a/README.zh-CN.md +++ b/locales/README.zh-CN.md @@ -25,7 +25,7 @@
- English · 简体中文 + English · 简体中文 · 简体中文
diff --git a/server/.env.example b/server/.env.example index f51d6177..6148d594 100644 --- a/server/.env.example +++ b/server/.env.example @@ -125,6 +125,12 @@ JWT_SECRET="my-random-string-for-seeding" # Please generate random string at lea # VOYAGEAI_API_KEY= # EMBEDDING_MODEL_PREF='voyage-large-2-instruct' +# EMBEDDING_ENGINE='litellm' +# EMBEDDING_MODEL_PREF='text-embedding-ada-002' +# EMBEDDING_MODEL_MAX_CHUNK_LENGTH=8192 +# LITE_LLM_BASE_PATH='http://127.0.0.1:4000' +# LITE_LLM_API_KEY='sk-123abc' + ########################################### ######## Vector Database Selection ######## ########################################### diff --git a/server/endpoints/workspaceThreads.js b/server/endpoints/workspaceThreads.js index e2aead97..1c207e52 100644 --- a/server/endpoints/workspaceThreads.js +++ b/server/endpoints/workspaceThreads.js @@ -1,4 +1,9 @@ -const { multiUserMode, userFromSession, reqBody } = require("../utils/http"); +const { + multiUserMode, + userFromSession, + reqBody, + safeJsonParse, +} = require("../utils/http"); const { validatedRequest } = require("../utils/middleware/validatedRequest"); const { Telemetry } = require("../models/telemetry"); const { @@ -168,6 +173,77 @@ function workspaceThreadEndpoints(app) { } } ); + + app.delete( + "/workspace/:slug/thread/:threadSlug/delete-edited-chats", + [ + validatedRequest, + flexUserRoleValid([ROLES.all]), + validWorkspaceAndThreadSlug, + ], + async (request, response) => { + try { + const { startingId } = reqBody(request); + const user = await userFromSession(request, response); + const workspace = response.locals.workspace; + const thread = response.locals.thread; + + await WorkspaceChats.delete({ + workspaceId: Number(workspace.id), + thread_id: Number(thread.id), + user_id: user?.id, + id: { gte: Number(startingId) }, + }); + + response.sendStatus(200).end(); + } catch (e) { + console.log(e.message, e); + response.sendStatus(500).end(); + } + } + ); + + app.post( + "/workspace/:slug/thread/:threadSlug/update-chat", + [ + validatedRequest, + flexUserRoleValid([ROLES.all]), + validWorkspaceAndThreadSlug, + ], + async (request, response) => { + try { + const { chatId, newText = null } = reqBody(request); + if (!newText || !String(newText).trim()) + throw new Error("Cannot save empty response"); + + const user = await userFromSession(request, response); + const workspace = response.locals.workspace; + const thread = response.locals.thread; + const existingChat = await WorkspaceChats.get({ + workspaceId: workspace.id, + thread_id: thread.id, + user_id: user?.id, + id: Number(chatId), + }); + if (!existingChat) throw new Error("Invalid chat."); + + const chatResponse = safeJsonParse(existingChat.response, null); + if (!chatResponse) throw new Error("Failed to parse chat response"); + + await WorkspaceChats._update(existingChat.id, { + response: JSON.stringify({ + ...chatResponse, + text: String(newText), + }), + }); + + response.sendStatus(200).end(); + } catch (e) { + console.log(e.message, e); + response.sendStatus(500).end(); + } + } + ); } module.exports = { workspaceThreadEndpoints }; diff --git a/server/endpoints/workspaces.js b/server/endpoints/workspaces.js index 2657eb97..6d6f29bb 100644 --- a/server/endpoints/workspaces.js +++ b/server/endpoints/workspaces.js @@ -380,7 +380,6 @@ function workspaceEndpoints(app) { const history = multiUserMode(response) ? await WorkspaceChats.forWorkspaceByUser(workspace.id, user.id) : await WorkspaceChats.forWorkspace(workspace.id); - response.status(200).json({ history: convertToChatHistory(history) }); } catch (e) { console.log(e.message, e); @@ -420,6 +419,67 @@ function workspaceEndpoints(app) { } ); + app.delete( + "/workspace/:slug/delete-edited-chats", + [validatedRequest, flexUserRoleValid([ROLES.all]), validWorkspaceSlug], + async (request, response) => { + try { + const { startingId } = reqBody(request); + const user = await userFromSession(request, response); + const workspace = response.locals.workspace; + + await WorkspaceChats.delete({ + workspaceId: workspace.id, + thread_id: null, + user_id: user?.id, + id: { gte: Number(startingId) }, + }); + + response.sendStatus(200).end(); + } catch (e) { + console.log(e.message, e); + response.sendStatus(500).end(); + } + } + ); + + app.post( + "/workspace/:slug/update-chat", + [validatedRequest, flexUserRoleValid([ROLES.all]), validWorkspaceSlug], + async (request, response) => { + try { + const { chatId, newText = null } = reqBody(request); + if (!newText || !String(newText).trim()) + throw new Error("Cannot save empty response"); + + const user = await userFromSession(request, response); + const workspace = response.locals.workspace; + const existingChat = await WorkspaceChats.get({ + workspaceId: workspace.id, + thread_id: null, + user_id: user?.id, + id: Number(chatId), + }); + if (!existingChat) throw new Error("Invalid chat."); + + const chatResponse = safeJsonParse(existingChat.response, null); + if (!chatResponse) throw new Error("Failed to parse chat response"); + + await WorkspaceChats._update(existingChat.id, { + response: JSON.stringify({ + ...chatResponse, + text: String(newText), + }), + }); + + response.sendStatus(200).end(); + } catch (e) { + console.log(e.message, e); + response.sendStatus(500).end(); + } + } + ); + app.post( "/workspace/:slug/chat-feedback/:chatId", [validatedRequest, flexUserRoleValid([ROLES.all]), validWorkspaceSlug], diff --git a/server/models/workspaceChats.js b/server/models/workspaceChats.js index c81992ca..bda40064 100644 --- a/server/models/workspaceChats.js +++ b/server/models/workspaceChats.js @@ -220,6 +220,24 @@ const WorkspaceChats = { console.error(error.message); } }, + + // Explicit update of settings + key validations. + // Only use this method when directly setting a key value + // that takes no user input for the keys being modified. + _update: async function (id = null, data = {}) { + if (!id) throw new Error("No workspace chat id provided for update"); + + try { + await prisma.workspace_chats.update({ + where: { id }, + data, + }); + return true; + } catch (error) { + console.error(error.message); + return false; + } + }, }; module.exports = { WorkspaceChats }; diff --git a/server/package.json b/server/package.json index 804f66de..1b0ba280 100644 --- a/server/package.json +++ b/server/package.json @@ -12,7 +12,7 @@ "scripts": { "dev": "NODE_ENV=development nodemon --ignore documents --ignore vector-cache --ignore storage --ignore swagger --trace-warnings index.js", "start": "NODE_ENV=production node index.js", - "lint": "yarn prettier --write ./endpoints ./models ./utils index.js", + "lint": "yarn prettier --ignore-path ../.prettierignore --write ./endpoints ./models ./utils index.js", "swagger": "node ./swagger/init.js", "sqlite:migrate": "cd ./utils/prisma && node migrateFromSqlite.js" }, diff --git a/server/utils/EmbeddingEngines/liteLLM/index.js b/server/utils/EmbeddingEngines/liteLLM/index.js new file mode 100644 index 00000000..cd22480b --- /dev/null +++ b/server/utils/EmbeddingEngines/liteLLM/index.js @@ -0,0 +1,93 @@ +const { toChunks, maximumChunkLength } = require("../../helpers"); + +class LiteLLMEmbedder { + constructor() { + const { OpenAI: OpenAIApi } = require("openai"); + if (!process.env.LITE_LLM_BASE_PATH) + throw new Error( + "LiteLLM must have a valid base path to use for the api." + ); + this.basePath = process.env.LITE_LLM_BASE_PATH; + this.openai = new OpenAIApi({ + baseURL: this.basePath, + apiKey: process.env.LITE_LLM_API_KEY ?? null, + }); + this.model = process.env.EMBEDDING_MODEL_PREF || "text-embedding-ada-002"; + + // Limit of how many strings we can process in a single pass to stay with resource or network limits + this.maxConcurrentChunks = 500; + this.embeddingMaxChunkLength = maximumChunkLength(); + } + + async embedTextInput(textInput) { + const result = await this.embedChunks( + Array.isArray(textInput) ? textInput : [textInput] + ); + return result?.[0] || []; + } + + async embedChunks(textChunks = []) { + // Because there is a hard POST limit on how many chunks can be sent at once to LiteLLM (~8mb) + // we concurrently execute each max batch of text chunks possible. + // Refer to constructor maxConcurrentChunks for more info. + const embeddingRequests = []; + for (const chunk of toChunks(textChunks, this.maxConcurrentChunks)) { + embeddingRequests.push( + new Promise((resolve) => { + this.openai.embeddings + .create({ + model: this.model, + input: chunk, + }) + .then((result) => { + resolve({ data: result?.data, error: null }); + }) + .catch((e) => { + e.type = + e?.response?.data?.error?.code || + e?.response?.status || + "failed_to_embed"; + e.message = e?.response?.data?.error?.message || e.message; + resolve({ data: [], error: e }); + }); + }) + ); + } + + const { data = [], error = null } = await Promise.all( + embeddingRequests + ).then((results) => { + // If any errors were returned from LiteLLM abort the entire sequence because the embeddings + // will be incomplete. + const errors = results + .filter((res) => !!res.error) + .map((res) => res.error) + .flat(); + if (errors.length > 0) { + let uniqueErrors = new Set(); + errors.map((error) => + uniqueErrors.add(`[${error.type}]: ${error.message}`) + ); + + return { + data: [], + error: Array.from(uniqueErrors).join(", "), + }; + } + return { + data: results.map((res) => res?.data || []).flat(), + error: null, + }; + }); + + if (!!error) throw new Error(`LiteLLM Failed to embed: ${error}`); + return data.length > 0 && + data.every((embd) => embd.hasOwnProperty("embedding")) + ? data.map((embd) => embd.embedding) + : null; + } +} + +module.exports = { + LiteLLMEmbedder, +}; diff --git a/server/utils/helpers/chat/responses.js b/server/utils/helpers/chat/responses.js index d07eae30..609b1819 100644 --- a/server/utils/helpers/chat/responses.js +++ b/server/utils/helpers/chat/responses.js @@ -174,6 +174,7 @@ function convertToChatHistory(history = []) { role: "user", content: prompt, sentAt: moment(createdAt).unix(), + chatId: id, }, { type: data?.type || "chart", diff --git a/server/utils/helpers/index.js b/server/utils/helpers/index.js index e60202a6..8f0df126 100644 --- a/server/utils/helpers/index.js +++ b/server/utils/helpers/index.js @@ -128,6 +128,9 @@ function getEmbeddingEngineSelection() { case "voyageai": const { VoyageAiEmbedder } = require("../EmbeddingEngines/voyageAi"); return new VoyageAiEmbedder(); + case "litellm": + const { LiteLLMEmbedder } = require("../EmbeddingEngines/liteLLM"); + return new LiteLLMEmbedder(); default: return new NativeEmbedder(); } diff --git a/server/utils/helpers/updateENV.js b/server/utils/helpers/updateENV.js index d5cdc68f..1a0e710a 100644 --- a/server/utils/helpers/updateENV.js +++ b/server/utils/helpers/updateENV.js @@ -577,6 +577,7 @@ function supportedEmbeddingModel(input = "") { "lmstudio", "cohere", "voyageai", + "litellm", ]; return supported.includes(input) ? null