chore: remove comment

.changeset/clever-doors-invite.md

@@ -0,0 +1,5 @@
+---
+"llamaindex": patch
+---
+
+feat: experimental package + json query engine

.changeset/config.json

@@ -1,7 +1,7 @@
 {
   "$schema": "https://unpkg.com/@changesets/config@2.3.1/schema.json",
   "changelog": "@changesets/cli/changelog",
-  "commit": false,
+  "commit": true,
   "fixed": [],
   "linked": [],
   "access": "public",

.changeset/eight-dogs-burn.md

@@ -0,0 +1,12 @@
+---
+"llamaindex": patch
+"@llamaindex/core-test": patch
+---
+
+- Add missing exports:
+  - `IndexStructType`,
+  - `IndexDict`,
+  - `jsonToIndexStruct`,
+  - `IndexList`,
+  - `IndexStruct`
+- Fix `IndexDict.toJson()` method

.changeset/few-chicken-exercise.md

@@ -0,0 +1,5 @@
+---
+"create-llama": patch
+---
+
+Add "Start in VSCode" option to postInstallAction

.changeset/khaki-rivers-unite.md

@@ -0,0 +1,5 @@
+---
+"llamaindex": patch
+---
+
+Add streaming to agents

.changeset/plenty-onions-relate.md

@@ -0,0 +1,5 @@
+---
+"create-llama": patch
+---
+
+Add devcontainers to generated code

.changeset/strange-waves-sit.md

@@ -0,0 +1,5 @@
+---
+"llamaindex": minor
+---
+
+Use parameter object for retrieve function of Retriever (to align usage with query function of QueryEngine)

.cjs.swcrc

@@ -1,13 +1,9 @@
 {
   "jsc": {
     "parser": {
-      "syntax": "typescript",
-      "decorators": true
+      "syntax": "typescript"
     },
-    "target": "esnext",
-    "transform": {
-      "decoratorVersion": "2022-03"
-    }
+    "target": "esnext"
   },
   "module": {
     "type": "commonjs",

.eslintrc.cjs

@@ -1,76 +0,0 @@
-module.exports = {
-  root: true,
-  extends: [
-    "turbo",
-    "prettier",
-    "plugin:@typescript-eslint/recommended-type-checked-only",
-  ],
-  parserOptions: {
-    project: true,
-    __tsconfigRootDir: __dirname,
-  },
-  settings: {
-    react: {
-      version: "999.999.999",
-    },
-  },
-  rules: {
-    "max-params": ["error", 4],
-    "prefer-const": "error",
-    "@typescript-eslint/no-floating-promises": [
-      "error",
-      {
-        ignoreIIFE: true,
-      },
-    ],
-    "@typescript-eslint/await-thenable": "off",
-    "@typescript-eslint/ban-ts-comment": "off",
-    "@typescript-eslint/ban-types": "off",
-    "no-array-constructor": "off",
-    "@typescript-eslint/no-array-constructor": "off",
-    "@typescript-eslint/no-base-to-string": "off",
-    "@typescript-eslint/no-duplicate-enum-values": "off",
-    "@typescript-eslint/no-duplicate-type-constituents": "off",
-    "@typescript-eslint/no-explicit-any": "off",
-    "@typescript-eslint/no-extra-non-null-assertion": "off",
-    "@typescript-eslint/no-for-in-array": "off",
-    "no-implied-eval": "off",
-    "@typescript-eslint/no-implied-eval": "off",
-    "no-loss-of-precision": "off",
-    "@typescript-eslint/no-loss-of-precision": "off",
-    "@typescript-eslint/no-misused-new": "off",
-    "@typescript-eslint/no-misused-promises": "off",
-    "@typescript-eslint/no-namespace": "off",
-    "@typescript-eslint/no-non-null-asserted-optional-chain": "off",
-    "@typescript-eslint/no-redundant-type-constituents": "off",
-    "@typescript-eslint/no-this-alias": "off",
-    "@typescript-eslint/no-unnecessary-type-assertion": "off",
-    "@typescript-eslint/no-unnecessary-type-constraint": "off",
-    "@typescript-eslint/no-unsafe-argument": "off",
-    "@typescript-eslint/no-unsafe-assignment": "off",
-    "@typescript-eslint/no-unsafe-call": "off",
-    "@typescript-eslint/no-unsafe-declaration-merging": "off",
-    "@typescript-eslint/no-unsafe-enum-comparison": "off",
-    "@typescript-eslint/no-unsafe-member-access": "off",
-    "@typescript-eslint/no-unsafe-return": "off",
-    "no-unused-vars": "off",
-    "@typescript-eslint/no-unused-vars": "off",
-    "@typescript-eslint/no-var-requires": "off",
-    "@typescript-eslint/prefer-as-const": "off",
-    "require-await": "off",
-    "@typescript-eslint/require-await": "off",
-    "@typescript-eslint/restrict-plus-operands": "off",
-    "@typescript-eslint/restrict-template-expressions": "off",
-    "@typescript-eslint/triple-slash-reference": "off",
-    "@typescript-eslint/unbound-method": "off",
-  },
-  overrides: [
-    {
-      files: ["examples/**/*.ts"],
-      rules: {
-        "turbo/no-undeclared-env-vars": "off",
-      },
-    },
-  ],
-  ignorePatterns: ["dist/", "lib/", "deps/"],
-};

.eslintrc.js

@@ -0,0 +1,23 @@
+module.exports = {
+  root: true,
+  // This tells ESLint to load the config from the package `eslint-config-custom`
+  extends: ["custom"],
+  settings: {
+    next: {
+      rootDir: ["apps/*/"],
+    },
+  },
+  rules: {
+    "max-params": ["error", 4],
+    "prefer-const": "error",
+  },
+  overrides: [
+    {
+      files: ["examples/**/*.ts"],
+      rules: {
+        "turbo/no-undeclared-env-vars": "off",
+      },
+    },
+  ],
+  ignorePatterns: ["dist/", "lib/"],
+};

.github/workflows/e2e.yml

@@ -0,0 +1,68 @@
+name: E2E Tests
+on:
+  push:
+    branches: [main]
+  pull_request:
+    paths:
+      - "packages/create-llama/**"
+      - ".github/workflows/e2e.yml"
+    branches: [main]
+
+env:
+  POETRY_VERSION: "1.6.1"
+
+jobs:
+  e2e:
+    name: create-llama
+    timeout-minutes: 60
+    strategy:
+      fail-fast: true
+      matrix:
+        node-version: [18, 20]
+        python-version: ["3.11"]
+        os: [macos-latest, windows-latest]
+    defaults:
+      run:
+        shell: bash
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@v4
+      - name: Set up python ${{ matrix.python-version }}
+        uses: actions/setup-python@v4
+        with:
+          python-version: ${{ matrix.python-version }}
+      - name: Install Poetry
+        uses: snok/install-poetry@v1
+        with:
+          version: ${{ env.POETRY_VERSION }}
+      - uses: pnpm/action-setup@v2
+      - name: Setup Node.js ${{ matrix.node-version }}
+        uses: actions/setup-node@v4
+        with:
+          node-version: ${{ matrix.node-version }}
+          cache: "pnpm"
+      - name: Install dependencies
+        run: pnpm install
+      - name: Install Playwright Browsers
+        run: pnpm exec playwright install --with-deps
+        working-directory: ./packages/create-llama
+      - name: Build create-llama
+        run: pnpm run build
+        working-directory: ./packages/create-llama
+      - name: Pack
+        run: pnpm pack --pack-destination ./output
+        working-directory: ./packages/create-llama
+      - name: Extract Pack
+        run: tar -xvzf ./output/*.tgz -C ./output
+        working-directory: ./packages/create-llama
+      - name: Run Playwright tests
+        run: pnpm exec playwright test
+        env:
+          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
+        working-directory: ./packages/create-llama
+      - uses: actions/upload-artifact@v3
+        if: always()
+        with:
+          name: playwright-report
+          path: ./packages/create-llama/playwright-report/
+          retention-days: 30

.github/workflows/lint_on_push_or_pull.yml

@@ -13,7 +13,9 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v4
-      - uses: pnpm/action-setup@v3
+      - uses: pnpm/action-setup@v2
+        with:
+          version: latest
       - name: Setup Node.js
         uses: actions/setup-node@v4
         with:

.github/workflows/publish.yml

@@ -14,14 +14,6 @@ jobs:
 
     steps:
       - uses: actions/checkout@v4
-      - uses: pnpm/action-setup@v3
-      - name: Setup Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version-file: ".nvmrc"
-          cache: "pnpm"
-      - name: Install dependencies
-        run: pnpm install
 
       - name: Publish @llamaindex/env
         run: npx jsr publish

.github/workflows/release.yml

@@ -1,57 +0,0 @@
-name: Release
-
-on:
-  push:
-    branches:
-      - main
-
-concurrency: ${{ github.workflow }}-${{ github.ref }}
-
-jobs:
-  release:
-    name: Release
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout Repo
-        uses: actions/checkout@v4
-
-      - uses: pnpm/action-setup@v3
-
-      - name: Setup Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version-file: ".nvmrc"
-          cache: "pnpm"
-
-      - name: Install dependencies
-        run: pnpm install
-
-      - name: Add auth token to .npmrc file
-        run: |
-          cat << EOF >> ".npmrc"
-            //registry.npmjs.org/:_authToken=$NPM_TOKEN
-          EOF
-        env:
-          NPM_TOKEN: ${{ secrets.NPM_TOKEN }}
-
-      - name: Get changeset status
-        id: get-changeset-status
-        run: |
-          pnpm changeset status --output .changeset/status.json
-          new_version=$(jq -r '.releases[] | select(.name == "llamaindex") | .newVersion' < .changeset/status.json)
-          rm -v .changeset/status.json
-          echo "new-version=${new_version}" >> "$GITHUB_OUTPUT"
-
-      - name: Create Release Pull Request or Publish to npm
-        id: changesets
-        uses: changesets/action@v1
-        with:
-          commit: Release ${{ steps.get-changeset-status.outputs.new-version }}
-          title: Release ${{ steps.get-changeset-status.outputs.new-version }}
-          # update version PR with the latest changesets
-          version: pnpm new-version
-          # build package and call changeset publish
-          publish: pnpm release
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          NPM_TOKEN: ${{ secrets.NPM_TOKEN }}

.github/workflows/test.yml

@@ -1,55 +1,18 @@
 name: Run Tests
 
-on:
-  push:
-    branches:
-      - main
-  pull_request:
-    branches:
-      - main
-
-concurrency:
-  group: ${{ github.workflow }}-${{ github.ref }}
-  cancel-in-progress: true
+on: [push, pull_request]
 
 jobs:
-  e2e:
-    strategy:
-      fail-fast: false
-      matrix:
-        node-version: [18.x, 20.x, 22.x]
-    name: E2E on Node.js ${{ matrix.node-version }}
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-
-      - uses: pnpm/action-setup@v3
-
-      - name: Setup Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version: ${{ matrix.node-version }}
-          cache: "pnpm"
-      - name: Install dependencies
-        run: pnpm install
-      - name: Run E2E Tests
-        run: pnpm run e2e
-
   test:
-    strategy:
-      fail-fast: false
-      matrix:
-        node-version: [18.x, 20.x, 22.x]
-    name: Test on Node.js ${{ matrix.node-version }}
     runs-on: ubuntu-latest
 
     steps:
       - uses: actions/checkout@v4
-      - uses: pnpm/action-setup@v3
+      - uses: pnpm/action-setup@v2
       - name: Setup Node.js
         uses: actions/setup-node@v4
         with:
-          node-version: ${{ matrix.node-version }}
+          node-version-file: ".nvmrc"
           cache: "pnpm"
       - name: Install dependencies
         run: pnpm install
@@ -60,7 +23,7 @@ jobs:
 
     steps:
       - uses: actions/checkout@v4
-      - uses: pnpm/action-setup@v3
+      - uses: pnpm/action-setup@v2
       - name: Setup Node.js
         uses: actions/setup-node@v4
         with:
@@ -70,9 +33,6 @@ jobs:
         run: pnpm install
       - name: Build
         run: pnpm run build --filter llamaindex
-      - name: Use Build For Examples
-        run: pnpm link ../packages/core/
-        working-directory: ./examples
       - name: Run Type Check
         run: pnpm run type-check
       - name: Run Circular Dependency Check
@@ -84,30 +44,12 @@ jobs:
           name: typecheck-build-dist
           path: ./packages/core/dist
           if-no-files-found: error
-  core-edge-runtime:
-    runs-on: ubuntu-latest
-
-    steps:
-      - uses: actions/checkout@v4
-      - uses: pnpm/action-setup@v3
-      - name: Setup Node.js
-        uses: actions/setup-node@v4
-        with:
-          node-version-file: ".nvmrc"
-          cache: "pnpm"
-      - name: Install dependencies
-        run: pnpm install
-      - name: Build
-        run: pnpm run build --filter @llamaindex/edge
-      - name: Build Edge Runtime
-        run: pnpm run build
-        working-directory: ./packages/edge/e2e/test-edge-runtime
   typecheck-examples:
     runs-on: ubuntu-latest
 
     steps:
       - uses: actions/checkout@v4
-      - uses: pnpm/action-setup@v3
+      - uses: pnpm/action-setup@v2
       - name: Setup Node.js
         uses: actions/setup-node@v4
         with:

.npmrc

@@ -1,5 +1,2 @@
 auto-install-peers = true
 enable-pre-post-scripts = true
-prefer-workspace-packages = true
-save-workspace-protocol = true
-link-workspace-packages = true

.swcrc

@@ -1,12 +1,8 @@
 {
   "jsc": {
     "parser": {
-      "syntax": "typescript",
-      "decorators": true
+      "syntax": "typescript"
     },
-    "target": "esnext",
-    "transform": {
-      "decoratorVersion": "2022-03"
-    }
+    "target": "esnext"
   }
 }

.vscode/launch.json

@@ -10,9 +10,8 @@
       "name": "Debug Example",
       "skipFiles": ["<node_internals>/**"],
       "runtimeExecutable": "pnpm",
-      "console": "integratedTerminal",
       "cwd": "${workspaceFolder}/examples",
-      "runtimeArgs": ["npx", "tsx", "${file}"]
+      "runtimeArgs": ["ts-node", "${fileBasename}"]
     }
   ]
 }

CONTRIBUTING.md

@@ -79,22 +79,14 @@ That should start a webserver which will serve the docs on https://localhost:300
 
 Any changes you make should be reflected in the browser. If you need to regenerate the API docs and find that your TSDoc isn't getting the updates, feel free to remove apps/docs/api. It will automatically regenerate itself when you run pnpm start again.
 
-## Changeset
+## Publishing
 
-We use [changesets](https://github.com/changesets/changesets) for managing versions and changelogs. To create a new changeset, run:
+To publish a new version of the library, run
 
+```shell
+pnpm new-llamaindex
+pnpm new-create-llama
+pnpm release
+git push # push to the main branch
+git push --tags
 ```
-pnpm changeset
-```
-
-Please send a descriptive changeset for each PR.
-
-## Publishing (maintainers only)
-
-The [Release Github Action](.github/workflows/release.yml) is automatically generating and updating a
-PR called "Release {version}".
-
-This PR will update the `package.json` and `CHANGELOG.md` files of each package according to
-the current changesets in the [.changeset](.changeset/) folder.
-
-If this PR is merged it will automatically add version tags to the repository and publish the updated packages to NPM.

README-turborepo.md

@@ -0,0 +1,81 @@
+# Turborepo starter
+
+This is an official starter Turborepo.
+
+## Using this example
+
+Run the following command:
+
+```sh
+npx create-turbo@latest
+```
+
+## What's inside?
+
+This Turborepo includes the following packages/apps:
+
+### Apps and Packages
+
+- `docs`: a [Next.js](https://nextjs.org/) app
+- `web`: another [Next.js](https://nextjs.org/) app
+- `ui`: a stub React component library shared by both `web` and `docs` applications
+- `eslint-config-custom`: `eslint` configurations (includes `eslint-config-next` and `eslint-config-prettier`)
+- `tsconfig`: `tsconfig.json`s used throughout the monorepo
+
+Each package/app is 100% [TypeScript](https://www.typescriptlang.org/).
+
+### Utilities
+
+This Turborepo has some additional tools already setup for you:
+
+- [TypeScript](https://www.typescriptlang.org/) for static type checking
+- [ESLint](https://eslint.org/) for code linting
+- [Prettier](https://prettier.io) for code formatting
+
+### Build
+
+To build all apps and packages, run the following command:
+
+```
+cd my-turborepo
+pnpm build
+```
+
+### Develop
+
+To develop all apps and packages, run the following command:
+
+```
+cd my-turborepo
+pnpm dev
+```
+
+### Remote Caching
+
+Turborepo can use a technique known as [Remote Caching](https://turbo.build/repo/docs/core-concepts/remote-caching) to share cache artifacts across machines, enabling you to share build caches with your team and CI/CD pipelines.
+
+By default, Turborepo will cache locally. To enable Remote Caching you will need an account with Vercel. If you don't have an account you can [create one](https://vercel.com/signup), then enter the following commands:
+
+```
+cd my-turborepo
+npx turbo login
+```
+
+This will authenticate the Turborepo CLI with your [Vercel account](https://vercel.com/docs/concepts/personal-accounts/overview).
+
+Next, you can link your Turborepo to your Remote Cache by running the following command from the root of your Turborepo:
+
+```
+npx turbo link
+```
+
+## Useful Links
+
+Learn more about the power of Turborepo:
+
+- [Tasks](https://turbo.build/repo/docs/core-concepts/monorepos/running-tasks)
+- [Caching](https://turbo.build/repo/docs/core-concepts/caching)
+- [Remote Caching](https://turbo.build/repo/docs/core-concepts/remote-caching)
+- [Filtering](https://turbo.build/repo/docs/core-concepts/monorepos/filtering)
+- [Configuration Options](https://turbo.build/repo/docs/reference/configuration)
+- [CLI Usage](https://turbo.build/repo/docs/reference/command-line-reference)

README.md

@@ -83,52 +83,37 @@ Check out our NextJS playground at https://llama-playground.vercel.app/. The sou
 
 - [Node](/packages/core/src/Node.ts): The basic data building block. Most commonly, these are parts of the document split into manageable pieces that are small enough to be fed into an embedding model and LLM.
 
-- [Embedding](/packages/core/src/embeddings/OpenAIEmbedding.ts): Embeddings are sets of floating point numbers which represent the data in a Node. By comparing the similarity of embeddings, we can derive an understanding of the similarity of two pieces of data. One use case is to compare the embedding of a question with the embeddings of our Nodes to see which Nodes may contain the data needed to answer that quesiton. Because the default service context is OpenAI, the default embedding is `OpenAIEmbedding`. If using different models, say through Ollama, use this [Embedding](/packages/core/src/embeddings/OllamaEmbedding.ts) (see all [here](/packages/core/src/embeddings)).
+- [Embedding](/packages/core/src/Embedding.ts): Embeddings are sets of floating point numbers which represent the data in a Node. By comparing the similarity of embeddings, we can derive an understanding of the similarity of two pieces of data. One use case is to compare the embedding of a question with the embeddings of our Nodes to see which Nodes may contain the data needed to answer that quesiton.
 
 - [Indices](/packages/core/src/indices/): Indices store the Nodes and the embeddings of those nodes. QueryEngines retrieve Nodes from these Indices using embedding similarity.
 
-- [QueryEngine](/packages/core/src/engines/query/RetrieverQueryEngine.ts): Query engines are what generate the query you put in and give you back the result. Query engines generally combine a pre-built prompt with selected Nodes from your Index to give the LLM the context it needs to answer your query. To build a query engine from your Index (recommended), use the [`asQueryEngine`](/packages/core/src/indices/BaseIndex.ts) method on your Index. See all query engines [here](/packages/core/src/engines/query).
+- [QueryEngine](/packages/core/src/QueryEngine.ts): Query engines are what generate the query you put in and give you back the result. Query engines generally combine a pre-built prompt with selected Nodes from your Index to give the LLM the context it needs to answer your query.
 
-- [ChatEngine](/packages/core/src/engines/chat/SimpleChatEngine.ts): A ChatEngine helps you build a chatbot that will interact with your Indices. See all chat engines [here](/packages/core/src/engines/chat).
+- [ChatEngine](/packages/core/src/ChatEngine.ts): A ChatEngine helps you build a chatbot that will interact with your Indices.
 
 - [SimplePrompt](/packages/core/src/Prompt.ts): A simple standardized function call definition that takes in inputs and formats them in a template literal. SimplePrompts can be specialized using currying and combined using other SimplePrompt functions.
 
-## Using NextJS
+## Note: NextJS:
 
-If you're using the NextJS App Router, you can choose between the Node.js and the [Edge runtime](https://nextjs.org/docs/app/building-your-application/rendering/edge-and-nodejs-runtimes#edge-runtime).
+If you're using NextJS App Router, you'll need to use the NodeJS runtime (default) and add the following config to your next.config.js to have it use imports/exports in the same way Node does.
 
-With NextJS 13 and 14, using the Node.js runtime is the default. You can explicitly set the Edge runtime in your [router handler](https://nextjs.org/docs/app/building-your-application/routing/route-handlers) by adding this line:
-
-```typescript
-export const runtime = "edge";
+```js
+export const runtime = "nodejs"; // default
 ```
 
-The following sections explain further differences in using the Node.js or Edge runtime.
-
-### Using the Node.js runtime
-
-Add the following config to your `next.config.js` to ignore specific packages in the server-side bundling:
-
 ```js
 // next.config.js
 /** @type {import('next').NextConfig} */
 const nextConfig = {
   experimental: {
-    serverComponentsExternalPackages: [
-      "pdf2json",
-      "@zilliz/milvus2-sdk-node",
-      "sharp",
-      "onnxruntime-node",
-    ],
+    serverComponentsExternalPackages: ["pdf2json"],
   },
   webpack: (config) => {
-    config.externals.push({
-      pdf2json: "commonjs pdf2json",
-      "@zilliz/milvus2-sdk-node": "commonjs @zilliz/milvus2-sdk-node",
-      sharp: "commonjs sharp",
-      "onnxruntime-node": "commonjs onnxruntime-node",
-    });
-
+    config.resolve.alias = {
+      ...config.resolve.alias,
+      sharp$: false,
+      "onnxruntime-node$": false,
+    };
     return config;
   },
 };
@@ -136,61 +121,12 @@ const nextConfig = {
 module.exports = nextConfig;
 ```
 
-### Using the Edge runtime
-
-We publish a dedicated package (`@llamaindex/edge` instead of `llamaindex`) for using the Edge runtime. To use it, first install the package:
-
-```shell
-pnpm install @llamaindex/edge
-```
-
-> _Note_: Ensure that your `package.json` doesn't include the `llamaindex` package if you're using `@llamaindex/edge`.
-
-Then make sure to use the correct import statement in your code:
-
-```typescript
-// replace 'llamaindex' with '@llamaindex/edge'
-import {} from "@llamaindex/edge";
-```
-
-A further difference is that the `@llamaindex/edge` package doesn't export classes from the `readers` or `storage` folders. The reason is that most of these classes are not compatible with the Edge runtime.
-
-If you need any of those classes, you have to import them instead directly. Here's an example for importing the `PineconeVectorStore` class:
-
-```typescript
-import { PineconeVectorStore } from "@llamaindex/edge/storage/vectorStore/PineconeVectorStore";
-```
-
-As the `PDFReader` is not working with the Edge runtime, here's how to use the `SimpleDirectoryReader` with the `LlamaParseReader` to load PDFs:
-
-```typescript
-import { SimpleDirectoryReader } from "@llamaindex/edge/readers/SimpleDirectoryReader";
-import { LlamaParseReader } from "@llamaindex/edge/readers/LlamaParseReader";
-
-export const DATA_DIR = "./data";
-
-export async function getDocuments() {
-  const reader = new SimpleDirectoryReader();
-  // Load PDFs using LlamaParseReader
-  return await reader.loadData({
-    directoryPath: DATA_DIR,
-    fileExtToReader: {
-      pdf: new LlamaParseReader({ resultType: "markdown" }),
-    },
-  });
-}
-```
-
-> _Note_: Reader classes have to be added explictly to the `fileExtToReader` map in the Edge version of the `SimpleDirectoryReader`.
-
-You'll find a complete example of using the Edge runtime with LlamaIndexTS here: https://github.com/run-llama/create_llama_projects/tree/main/nextjs-edge-llamaparse
-
 ## Supported LLMs:
 
 - OpenAI GPT-3.5-turbo and GPT-4
-- Anthropic Claude 3 (Opus, Sonnet, and Haiku) and the legacy models (Claude 2 and Instant)
+- Anthropic Claude Instant and Claude 2
 - Groq LLMs
-- Llama2/3 Chat LLMs (70B, 13B, and 7B parameters)
+- Llama2 Chat LLMs (70B, 13B, and 7B parameters)
 - MistralAI Chat LLMs
 - Fireworks Chat LLMs
 

apps/docs/CHANGELOG.md

@@ -1,14 +1,5 @@
 # docs
 
-## 0.0.5
-
-### Patch Changes
-
-- Updated dependencies [87142b2]
-- Updated dependencies [5a6cc0e]
-- Updated dependencies [87142b2]
-  - llamaindex@0.2.11
-
 ## 0.0.4
 
 ### Patch Changes

apps/docs/docs/examples/agent.mdx

@@ -4,7 +4,82 @@ A built-in agent that can take decisions and reasoning based on the tools provid
 
 ## OpenAI Agent
 
-import CodeBlock from "@theme/CodeBlock";
-import CodeSource from "!raw-loader!../../../../examples/agent/openai";
+```ts
+import { FunctionTool, OpenAIAgent } from "llamaindex";
 
-<CodeBlock language="ts">{CodeSource}</CodeBlock>
+// Define a function to sum two numbers
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
+}
+
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
+}
+
+// Define the parameters of the sum function as a JSON schema
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
+    },
+  },
+  required: ["a", "b"],
+};
+
+// Define the parameters of the divide function as a JSON schema
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The dividend to divide",
+    },
+    b: {
+      type: "number",
+      description: "The divisor to divide by",
+    },
+  },
+  required: ["a", "b"],
+};
+
+async function main() {
+  // Create a function tool from the sum function
+  const sumFunctionTool = new FunctionTool(sumNumbers, {
+    name: "sumNumbers",
+    description: "Use this function to sum two numbers",
+    parameters: sumJSON,
+  });
+
+  // Create a function tool from the divide function
+  const divideFunctionTool = new FunctionTool(divideNumbers, {
+    name: "divideNumbers",
+    description: "Use this function to divide two numbers"
+    parameters: divideJSON,
+  });
+
+  // Create an OpenAIAgent with the function tools
+  const agent = new OpenAIAgent({
+    tools: [sumFunctionTool, divideFunctionTool],
+    verbose: true,
+  });
+
+  // Chat with the agent
+  const response = await agent.chat({
+    message: "How much is 5 + 5? then divide by 2",
+  });
+
+  // Print the response
+  console.log(String(response));
+}
+
+main().then(() => {
+  console.log("Done");
+});
+```

apps/docs/docs/modules/agent/index.md

@@ -11,10 +11,4 @@ An “agent” is an automated reasoning and decision engine. It takes in a user
 
 LlamaIndex.TS comes with a few built-in agents, but you can also create your own. The built-in agents include:
 
-- OpenAI Agent
-- Anthropic Agent
-- ReACT Agent
-
-## Examples
-
-- [OpenAI Agent](../../examples/agent.mdx)
+- [OpenAI Agent](./openai.mdx)

apps/docs/docs/modules/agent/multi_document_agent.mdx

@@ -0,0 +1,314 @@
+# Multi-Document Agent
+
+In this guide, you learn towards setting up an agent that can effectively answer different types of questions over a larger set of documents.
+
+These questions include the following
+
+- QA over a specific doc
+- QA comparing different docs
+- Summaries over a specific doc
+- Comparing summaries between different docs
+
+We do this with the following architecture:
+
+- setup a “document agent” over each Document: each doc agent can do QA/summarization within its doc
+- setup a top-level agent over this set of document agents. Do tool retrieval and then do CoT over the set of tools to answer a question.
+
+## Setup and Download Data
+
+We first start by installing the necessary libraries and downloading the data.
+
+```bash
+pnpm i llamaindex
+```
+
+```ts
+import {
+  Document,
+  ObjectIndex,
+  OpenAI,
+  OpenAIAgent,
+  QueryEngineTool,
+  SimpleNodeParser,
+  SimpleToolNodeMapping,
+  SummaryIndex,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+  storageContextFromDefaults,
+} from "llamaindex";
+```
+
+And then for the data we will run through a list of countries and download the wikipedia page for each country.
+
+```ts
+import fs from "fs";
+import path from "path";
+
+const dataPath = path.join(__dirname, "tmp_data");
+
+const extractWikipediaTitle = async (title: string) => {
+  const fileExists = fs.existsSync(path.join(dataPath, `${title}.txt`));
+
+  if (fileExists) {
+    console.log(`File already exists for the title: ${title}`);
+    return;
+  }
+
+  const queryParams = new URLSearchParams({
+    action: "query",
+    format: "json",
+    titles: title,
+    prop: "extracts",
+    explaintext: "true",
+  });
+
+  const url = `https://en.wikipedia.org/w/api.php?${queryParams}`;
+
+  const response = await fetch(url);
+  const data: any = await response.json();
+
+  const pages = data.query.pages;
+  const page = pages[Object.keys(pages)[0]];
+  const wikiText = page.extract;
+
+  await new Promise((resolve) => {
+    fs.writeFile(path.join(dataPath, `${title}.txt`), wikiText, (err: any) => {
+      if (err) {
+        console.error(err);
+        resolve(title);
+        return;
+      }
+      console.log(`${title} stored in file!`);
+
+      resolve(title);
+    });
+  });
+};
+```
+
+```ts
+export const extractWikipedia = async (titles: string[]) => {
+  if (!fs.existsSync(dataPath)) {
+    fs.mkdirSync(dataPath);
+  }
+
+  for await (const title of titles) {
+    await extractWikipediaTitle(title);
+  }
+
+  console.log("Extration finished!");
+```
+
+These files will be saved in the `tmp_data` folder.
+
+Now we can call the function to download the data for each country.
+
+```ts
+await extractWikipedia([
+  "Brazil",
+  "United States",
+  "Canada",
+  "Mexico",
+  "Argentina",
+  "Chile",
+  "Colombia",
+  "Peru",
+  "Venezuela",
+  "Ecuador",
+  "Bolivia",
+  "Paraguay",
+  "Uruguay",
+  "Guyana",
+  "Suriname",
+  "French Guiana",
+  "Falkland Islands",
+]);
+```
+
+## Load the data
+
+Now that we have the data, we can load it into the LlamaIndex and store as a document.
+
+```ts
+import { Document } from "llamaindex";
+
+const countryDocs: Record<string, Document> = {};
+
+for (const title of wikiTitles) {
+  const path = `./agent/helpers/tmp_data/${title}.txt`;
+  const text = await fs.readFile(path, "utf-8");
+  const document = new Document({ text: text, id_: path });
+  countryDocs[title] = document;
+}
+```
+
+## Setup LLM and StorageContext
+
+We will be using gpt-4 for this example and we will use the `StorageContext` to store the documents in-memory.
+
+```ts
+const llm = new OpenAI({
+  model: "gpt-4",
+});
+
+const ctx = serviceContextFromDefaults({ llm });
+
+const storageContext = await storageContextFromDefaults({
+  persistDir: "./storage",
+});
+```
+
+## Building Multi-Document Agents
+
+In this section we show you how to construct the multi-document agent. We first build a document agent for each document, and then define the top-level parent agent with an object index.
+
+```ts
+const documentAgents: Record<string, any> = {};
+const queryEngines: Record<string, any> = {};
+```
+
+Now we iterate over each country and create a document agent for each one.
+
+### Build Agent for each Document
+
+In this section we define “document agents” for each document.
+
+We define both a vector index (for semantic search) and summary index (for summarization) for each document. The two query engines are then converted into tools that are passed to an OpenAI function calling agent.
+
+This document agent can dynamically choose to perform semantic search or summarization within a given document.
+
+We create a separate document agent for each coutnry.
+
+```ts
+for (const title of wikiTitles) {
+  // parse the document into nodes
+  const nodes = new SimpleNodeParser({
+    chunkSize: 200,
+    chunkOverlap: 20,
+  }).getNodesFromDocuments([countryDocs[title]]);
+
+  // create the vector index for specific search
+  const vectorIndex = await VectorStoreIndex.init({
+    serviceContext: serviceContext,
+    storageContext: storageContext,
+    nodes,
+  });
+
+  // create the summary index for broader search
+  const summaryIndex = await SummaryIndex.init({
+    serviceContext: serviceContext,
+    nodes,
+  });
+
+  const vectorQueryEngine = summaryIndex.asQueryEngine();
+  const summaryQueryEngine = summaryIndex.asQueryEngine();
+
+  // create the query engines for each task
+  const queryEngineTools = [
+    new QueryEngineTool({
+      queryEngine: vectorQueryEngine,
+      metadata: {
+        name: "vector_tool",
+        description: `Useful for questions related to specific aspects of ${title} (e.g. the history, arts and culture, sports, demographics, or more).`,
+      },
+    }),
+    new QueryEngineTool({
+      queryEngine: summaryQueryEngine,
+      metadata: {
+        name: "summary_tool",
+        description: `Useful for any requests that require a holistic summary of EVERYTHING about ${title}. For questions about more specific sections, please use the vector_tool.`,
+      },
+    }),
+  ];
+
+  // create the document agent
+  const agent = new OpenAIAgent({
+    tools: queryEngineTools,
+    llm,
+    verbose: true,
+  });
+
+  documentAgents[title] = agent;
+  queryEngines[title] = vectorIndex.asQueryEngine();
+}
+```
+
+## Build Top-Level Agent
+
+Now we define the top-level agent that can answer questions over the set of document agents.
+
+This agent takes in all document agents as tools. This specific agent RetrieverOpenAIAgent performs tool retrieval before tool use (unlike a default agent that tries to put all tools in the prompt).
+
+Here we use a top-k retriever, but we encourage you to customize the tool retriever method!
+
+Firstly, we create a tool for each document agent
+
+```ts
+const allTools: QueryEngineTool[] = [];
+```
+
+```ts
+for (const title of wikiTitles) {
+  const wikiSummary = `
+    This content contains Wikipedia articles about ${title}.
+    Use this tool if you want to answer any questions about ${title}
+  `;
+
+  const docTool = new QueryEngineTool({
+    queryEngine: documentAgents[title],
+    metadata: {
+      name: `tool_${title}`,
+      description: wikiSummary,
+    },
+  });
+
+  allTools.push(docTool);
+}
+```
+
+Our top level agent will use this document agents as tools and use toolRetriever to retrieve the best tool to answer a question.
+
+```ts
+// map the tools to nodes
+const toolMapping = SimpleToolNodeMapping.fromObjects(allTools);
+
+// create the object index
+const objectIndex = await ObjectIndex.fromObjects(
+  allTools,
+  toolMapping,
+  VectorStoreIndex,
+  {
+    serviceContext,
+    storageContext,
+  },
+);
+
+// create the top agent
+const topAgent = new OpenAIAgent({
+  toolRetriever: await objectIndex.asRetriever({}),
+  llm,
+  verbose: true,
+  prefixMessages: [
+    {
+      content:
+        "You are an agent designed to answer queries about a set of given countries. Please always use the tools provided to answer a question. Do not rely on prior knowledge.",
+      role: "system",
+    },
+  ],
+});
+```
+
+## Use the Agent
+
+Now we can use the agent to answer questions.
+
+```ts
+const response = await topAgent.chat({
+  message: "Tell me the differences between Brazil and Canada economics?",
+});
+
+// print output
+console.log(response);
+```
+
+You can find the full code for this example [here](https://github.com/run-llama/LlamaIndexTS/tree/main/examples/agent/multi-document-agent.ts)

apps/docs/docs/modules/agent/openai.mdx

@@ -0,0 +1,187 @@
+---
+sidebar_position: 0
+---
+
+# OpenAI Agent
+
+OpenAI API that supports function calling, it’s never been easier to build your own agent!
+
+In this notebook tutorial, we showcase how to write your own OpenAI agent
+
+## Setup
+
+First, you need to install the `llamaindex` package. You can do this by running the following command in your terminal:
+
+```bash
+pnpm i llamaindex
+```
+
+Then we can define a function to sum two numbers and another function to divide two numbers.
+
+```ts
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
+}
+
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
+}
+```
+
+## Create a function tool
+
+Now we can create a function tool from the sum function and another function tool from the divide function.
+
+For the parameters of the sum function, we can define a JSON schema.
+
+### JSON Schema
+
+```ts
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
+    },
+  },
+  required: ["a", "b"],
+};
+
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The dividend a to divide",
+    },
+    b: {
+      type: "number",
+      description: "The divisor b to divide by",
+    },
+  },
+  required: ["a", "b"],
+};
+
+const sumFunctionTool = new FunctionTool(sumNumbers, {
+  name: "sumNumbers",
+  description: "Use this function to sum two numbers",
+  parameters: sumJSON,
+});
+
+const divideFunctionTool = new FunctionTool(divideNumbers, {
+  name: "divideNumbers",
+  description: "Use this function to divide two numbers",
+  parameters: divideJSON,
+});
+```
+
+## Create an OpenAIAgent
+
+Now we can create an OpenAIAgent with the function tools.
+
+```ts
+const agent = new OpenAIAgent({
+  tools: [sumFunctionTool, divideFunctionTool],
+  verbose: true,
+});
+```
+
+## Chat with the agent
+
+Now we can chat with the agent.
+
+```ts
+const response = await agent.chat({
+  message: "How much is 5 + 5? then divide by 2",
+});
+
+console.log(String(response));
+```
+
+## Full code
+
+```ts
+import { FunctionTool, OpenAIAgent } from "llamaindex";
+
+// Define a function to sum two numbers
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
+}
+
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
+}
+
+// Define the parameters of the sum function as a JSON schema
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
+    },
+  },
+  required: ["a", "b"],
+};
+
+// Define the parameters of the divide function as a JSON schema
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The argument a to divide",
+    },
+    b: {
+      type: "number",
+      description: "The argument b to divide",
+    },
+  },
+  required: ["a", "b"],
+};
+
+async function main() {
+  // Create a function tool from the sum function
+  const sumFunctionTool = new FunctionTool(sumNumbers, {
+    name: "sumNumbers",
+    description: "Use this function to sum two numbers",
+    parameters: sumJSON,
+  });
+
+  // Create a function tool from the divide function
+  const divideFunctionTool = new FunctionTool(divideNumbers, {
+    name: "divideNumbers",
+    description: "Use this function to divide two numbers",
+    parameters: divideJSON,
+  });
+
+  // Create an OpenAIAgent with the function tools
+  const agent = new OpenAIAgent({
+    tools: [sumFunctionTool, divideFunctionTool],
+    verbose: true,
+  });
+
+  // Chat with the agent
+  const response = await agent.chat({
+    message: "How much is 5 + 5? then divide by 2",
+  });
+
+  // Print the response
+  console.log(String(response));
+}
+
+main().then(() => {
+  console.log("Done");
+});
+```

apps/docs/docs/modules/agent/query_engine_tool.mdx

@@ -0,0 +1,132 @@
+---
+sidebar_position: 1
+---
+
+# OpenAI Agent + QueryEngineTool
+
+QueryEngineTool is a tool that allows you to query a vector index. In this example, we will create a vector index from a set of documents and then create a QueryEngineTool from the vector index. We will then create an OpenAIAgent with the QueryEngineTool and chat with the agent.
+
+## Setup
+
+First, you need to install the `llamaindex` package. You can do this by running the following command in your terminal:
+
+```bash
+pnpm i llamaindex
+```
+
+Then you can import the necessary classes and functions.
+
+```ts
+import {
+  OpenAIAgent,
+  SimpleDirectoryReader,
+  VectorStoreIndex,
+  QueryEngineTool,
+} from "llamaindex";
+```
+
+## Create a vector index
+
+Now we can create a vector index from a set of documents.
+
+```ts
+// Load the documents
+const documents = await new SimpleDirectoryReader().loadData({
+  directoryPath: "node_modules/llamaindex/examples/",
+});
+
+// Create a vector index from the documents
+const vectorIndex = await VectorStoreIndex.fromDocuments(documents);
+```
+
+## Create a QueryEngineTool
+
+Now we can create a QueryEngineTool from the vector index.
+
+```ts
+// Create a query engine from the vector index
+const abramovQueryEngine = vectorIndex.asQueryEngine();
+
+// Create a QueryEngineTool with the query engine
+const queryEngineTool = new QueryEngineTool({
+  queryEngine: abramovQueryEngine,
+  metadata: {
+    name: "abramov_query_engine",
+    description: "A query engine for the Abramov documents",
+  },
+});
+```
+
+## Create an OpenAIAgent
+
+```ts
+// Create an OpenAIAgent with the query engine tool tools
+
+const agent = new OpenAIAgent({
+  tools: [queryEngineTool],
+  verbose: true,
+});
+```
+
+## Chat with the agent
+
+Now we can chat with the agent.
+
+```ts
+const response = await agent.chat({
+  message: "What was his salary?",
+});
+
+console.log(String(response));
+```
+
+## Full code
+
+```ts
+import {
+  OpenAIAgent,
+  SimpleDirectoryReader,
+  VectorStoreIndex,
+  QueryEngineTool,
+} from "llamaindex";
+
+async function main() {
+  // Load the documents
+  const documents = await new SimpleDirectoryReader().loadData({
+    directoryPath: "node_modules/llamaindex/examples/",
+  });
+
+  // Create a vector index from the documents
+  const vectorIndex = await VectorStoreIndex.fromDocuments(documents);
+
+  // Create a query engine from the vector index
+  const abramovQueryEngine = vectorIndex.asQueryEngine();
+
+  // Create a QueryEngineTool with the query engine
+  const queryEngineTool = new QueryEngineTool({
+    queryEngine: abramovQueryEngine,
+    metadata: {
+      name: "abramov_query_engine",
+      description: "A query engine for the Abramov documents",
+    },
+  });
+
+  // Create an OpenAIAgent with the function tools
+  const agent = new OpenAIAgent({
+    tools: [queryEngineTool],
+    verbose: true,
+  });
+
+  // Chat with the agent
+  const response = await agent.chat({
+    message: "What was his salary?",
+  });
+
+  // Print the response
+  console.log(String(response));
+}
+
+main().then(() => {
+  console.log("Done");
+});
+```

apps/docs/docs/modules/agent/react_agent.mdx

@@ -0,0 +1,203 @@
+# ReAct Agent
+
+The ReAct agent is an AI agent that can reason over the next action, construct an action command, execute the action, and repeat these steps in an iterative loop until the task is complete.
+
+In this notebook tutorial, we showcase how to write your ReAct agent using the `llamaindex` package.
+
+## Setup
+
+First, you need to install the `llamaindex` package. You can do this by running the following command in your terminal:
+
+```bash
+pnpm i llamaindex
+```
+
+And then you can import the `OpenAIAgent` and `FunctionTool` from the `llamaindex` package.
+
+```ts
+import { FunctionTool, OpenAIAgent } from "llamaindex";
+```
+
+Then we can define a function to sum two numbers and another function to divide two numbers.
+
+```ts
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
+}
+
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
+}
+```
+
+## Create a function tool
+
+Now we can create a function tool from the sum function and another function tool from the divide function.
+
+For the parameters of the sum function, we can define a JSON schema.
+
+### JSON Schema
+
+```ts
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
+    },
+  },
+  required: ["a", "b"],
+};
+
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The dividend a to divide",
+    },
+    b: {
+      type: "number",
+      description: "The divisor b to divide by",
+    },
+  },
+  required: ["a", "b"],
+};
+
+const sumFunctionTool = new FunctionTool(sumNumbers, {
+  name: "sumNumbers",
+  description: "Use this function to sum two numbers",
+  parameters: sumJSON,
+});
+
+const divideFunctionTool = new FunctionTool(divideNumbers, {
+  name: "divideNumbers",
+  description: "Use this function to divide two numbers",
+  parameters: divideJSON,
+});
+```
+
+## Create an ReAct
+
+Now we can create an OpenAIAgent with the function tools.
+
+```ts
+const agent = new ReActAgent({
+  tools: [sumFunctionTool, divideFunctionTool],
+  verbose: true,
+});
+```
+
+## Chat with the agent
+
+Now we can chat with the agent.
+
+```ts
+const response = await agent.chat({
+  message: "How much is 5 + 5? then divide by 2",
+});
+
+console.log(String(response));
+```
+
+The output will be:
+
+```bash
+Thought: I need to use a tool to help me answer the question.
+Action: sumNumbers
+Action Input: {"a":5,"b":5}
+
+Observation: 10
+Thought: I can answer without using any more tools.
+Answer: The sum of 5 and 5 is 10, and when divided by 2, the result is 5.
+
+The sum of 5 and 5 is 10, and when divided by 2, the result is 5.
+```
+
+## Full code
+
+```ts
+import { FunctionTool, ReActAgent } from "llamaindex";
+
+// Define a function to sum two numbers
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
+}
+
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
+}
+
+// Define the parameters of the sum function as a JSON schema
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
+    },
+  },
+  required: ["a", "b"],
+};
+
+// Define the parameters of the divide function as a JSON schema
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The argument a to divide",
+    },
+    b: {
+      type: "number",
+      description: "The argument b to divide",
+    },
+  },
+  required: ["a", "b"],
+};
+
+async function main() {
+  // Create a function tool from the sum function
+  const sumFunctionTool = new FunctionTool(sumNumbers, {
+    name: "sumNumbers",
+    description: "Use this function to sum two numbers",
+    parameters: sumJSON,
+  });
+
+  // Create a function tool from the divide function
+  const divideFunctionTool = new FunctionTool(divideNumbers, {
+    name: "divideNumbers",
+    description: "Use this function to divide two numbers",
+    parameters: divideJSON,
+  });
+
+  // Create an OpenAIAgent with the function tools
+  const agent = new OpenAIAgent({
+    tools: [sumFunctionTool, divideFunctionTool],
+    verbose: true,
+  });
+
+  // Chat with the agent
+  const response = await agent.chat({
+    message: "I want to sum 5 and 5 and then divide by 2",
+  });
+
+  // Print the response
+  console.log(String(response));
+}
+
+main().then(() => {
+  console.log("Done");
+});
+```

apps/docs/docs/modules/embeddings/available_embeddings/huggingface.md

@@ -3,14 +3,17 @@
 To use HuggingFace embeddings, you need to import `HuggingFaceEmbedding` from `llamaindex`.
 
 ```ts
-import { HuggingFaceEmbedding, Settings } from "llamaindex";
+import { HuggingFaceEmbedding, serviceContextFromDefaults } from "llamaindex";
 
-// Update Embed Model
-Settings.embedModel = new HuggingFaceEmbedding();
+const huggingFaceEmbeds = new HuggingFaceEmbedding();
+
+const serviceContext = serviceContextFromDefaults({ embedModel: openaiEmbeds });
 
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 
 const queryEngine = index.asQueryEngine();
 
@@ -26,8 +29,8 @@ If you're not using a quantized model, set the `quantized` parameter to `false`.
 
 For example, to use the not quantized `BAAI/bge-small-en-v1.5` model, you can use the following code:
 
-```ts
-Settings.embedModel = new HuggingFaceEmbedding({
+```
+const embedModel = new HuggingFaceEmbedding({
   modelType: "BAAI/bge-small-en-v1.5",
   quantized: false,
 });

apps/docs/docs/modules/embeddings/available_embeddings/jinaai.md

@@ -1,21 +0,0 @@
-# Jina AI
-
-To use Jina AI embeddings, you need to import `JinaAIEmbedding` from `llamaindex`.
-
-```ts
-import { JinaAIEmbedding, Settings } from "llamaindex";
-
-Settings.embedModel = new JinaAIEmbedding();
-
-const document = new Document({ text: essay, id_: "essay" });
-
-const index = await VectorStoreIndex.fromDocuments([document]);
-
-const queryEngine = index.asQueryEngine();
-
-const query = "What is the meaning of life?";
-
-const results = await queryEngine.query({
-  query,
-});
-```

apps/docs/docs/modules/embeddings/available_embeddings/mistral.md

@@ -3,16 +3,21 @@
 To use MistralAI embeddings, you need to import `MistralAIEmbedding` from `llamaindex`.
 
 ```ts
-import { MistralAIEmbedding, Settings } from "llamaindex";
+import { MistralAIEmbedding, serviceContextFromDefaults } from "llamaindex";
 
-// Update Embed Model
-Settings.embedModel = new MistralAIEmbedding({
+const mistralEmbedModel = new MistralAIEmbedding({
   apiKey: "<YOUR_API_KEY>",
 });
 
+const serviceContext = serviceContextFromDefaults({
+  embedModel: mistralEmbedModel,
+});
+
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 
 const queryEngine = index.asQueryEngine();
 

apps/docs/docs/modules/embeddings/available_embeddings/ollama.md

@@ -1,23 +1,21 @@
 # Ollama
 
-To use Ollama embeddings, you need to import `OllamaEmbedding` from `llamaindex`.
-
-Note that you need to pull the embedding model first before using it.
-
-In the example below, we're using the [`nomic-embed-text`](https://ollama.com/library/nomic-embed-text) model, so you have to call:
-
-```shell
-ollama pull nomic-embed-text
-```
+To use Ollama embeddings, you need to import `Ollama` from `llamaindex`.
 
 ```ts
-import { OllamaEmbedding, Settings } from "llamaindex";
+import { Ollama, serviceContextFromDefaults } from "llamaindex";
 
-Settings.embedModel = new OllamaEmbedding({ model: "nomic-embed-text" });
+const ollamaEmbedModel = new Ollama();
+
+const serviceContext = serviceContextFromDefaults({
+  embedModel: ollamaEmbedModel,
+});
 
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 
 const queryEngine = index.asQueryEngine();
 

apps/docs/docs/modules/embeddings/available_embeddings/openai.md

@@ -3,13 +3,19 @@
 To use OpenAI embeddings, you need to import `OpenAIEmbedding` from `llamaindex`.
 
 ```ts
-import { OpenAIEmbedding, Settings } from "llamaindex";
+import { OpenAIEmbedding, serviceContextFromDefaults } from "llamaindex";
 
-Settings.embedModel = new OpenAIEmbedding();
+const openaiEmbedModel = new OpenAIEmbedding();
+
+const serviceContext = serviceContextFromDefaults({
+  embedModel: openaiEmbedModel,
+});
 
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 
 const queryEngine = index.asQueryEngine();
 

apps/docs/docs/modules/embeddings/available_embeddings/together.md

@@ -3,15 +3,21 @@
 To use together embeddings, you need to import `TogetherEmbedding` from `llamaindex`.
 
 ```ts
-import { TogetherEmbedding, Settings } from "llamaindex";
+import { TogetherEmbedding, serviceContextFromDefaults } from "llamaindex";
 
-Settings.embedModel = new TogetherEmbedding({
+const togetherEmbedModel = new TogetherEmbedding({
   apiKey: "<YOUR_API_KEY>",
 });
 
+const serviceContext = serviceContextFromDefaults({
+  embedModel: togetherEmbedModel,
+});
+
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 
 const queryEngine = index.asQueryEngine();
 

apps/docs/docs/modules/embeddings/index.md

@@ -2,14 +2,14 @@
 
 The embedding model in LlamaIndex is responsible for creating numerical representations of text. By default, LlamaIndex will use the `text-embedding-ada-002` model from OpenAI.
 
-This can be explicitly updated through `Settings`
+This can be explicitly set in the `ServiceContext` object.
 
 ```typescript
-import { OpenAIEmbedding, Settings } from "llamaindex";
+import { OpenAIEmbedding, serviceContextFromDefaults } from "llamaindex";
 
-Settings.embedModel = new OpenAIEmbedding({
-  model: "text-embedding-ada-002",
-});
+const openaiEmbeds = new OpenAIEmbedding();
+
+const serviceContext = serviceContextFromDefaults({ embedModel: openaiEmbeds });
 ```
 
 ## Local Embedding
@@ -19,3 +19,4 @@ For local embeddings, you can use the [HuggingFace](./available_embeddings/huggi
 ## API Reference
 
 - [OpenAIEmbedding](../../api/classes/OpenAIEmbedding.md)
+- [ServiceContext](../../api/interfaces//ServiceContext.md)

apps/docs/docs/modules/evaluation/modules/correctness.md

@@ -21,15 +21,23 @@ export OPENAI_API_KEY=your-api-key
 Import the required modules:
 
 ```ts
-import { CorrectnessEvaluator, OpenAI, Settings } from "llamaindex";
+import {
+  CorrectnessEvaluator,
+  OpenAI,
+  serviceContextFromDefaults,
+} from "llamaindex";
 ```
 
 Let's setup gpt-4 for better results:
 
 ```ts
-Settings.llm = new OpenAI({
+const llm = new OpenAI({
   model: "gpt-4",
 });
+
+const ctx = serviceContextFromDefaults({
+  llm,
+});
 ```
 
 ```ts
@@ -41,7 +49,9 @@ const response = ` Certainly! Albert Einstein's theory of relativity consists of
 However, general relativity, published in 1915, extended these ideas to include the effects of magnetism. According to general relativity, gravity is not a force between masses but rather the result of the warping of space and time by magnetic fields generated by massive objects. Massive objects, such as planets and stars, create magnetic fields that cause a curvature in spacetime, and smaller objects follow curved paths in response to this magnetic curvature. This concept is often illustrated using the analogy of a heavy ball placed on a rubber sheet with magnets underneath, causing it to create a depression that other objects (representing smaller masses) naturally move towards due to magnetic attraction.
 `;
 
-const evaluator = new CorrectnessEvaluator();
+const evaluator = new CorrectnessEvaluator({
+  serviceContext: ctx,
+});
 
 const result = await evaluator.evaluateResponse({
   query,

apps/docs/docs/modules/evaluation/modules/faithfulness.md

@@ -28,16 +28,20 @@ import {
   FaithfulnessEvaluator,
   OpenAI,
   VectorStoreIndex,
-  Settings,
+  serviceContextFromDefaults,
 } from "llamaindex";
 ```
 
 Let's setup gpt-4 for better results:
 
 ```ts
-Settings.llm = new OpenAI({
+const llm = new OpenAI({
   model: "gpt-4",
 });
+
+const ctx = serviceContextFromDefaults({
+  llm,
+});
 ```
 
 Now, let's create a vector index and query engine with documents and query engine respectively. Then, we can evaluate the response with the query and response from the query engine.:
@@ -59,7 +63,9 @@ Now, let's evaluate the response:
 ```ts
 const query = "How did New York City get its name?";
 
-const evaluator = new FaithfulnessEvaluator();
+const evaluator = new FaithfulnessEvaluator({
+  serviceContext: ctx,
+});
 
 const response = await queryEngine.query({
   query,

apps/docs/docs/modules/evaluation/modules/relevancy.md

@@ -21,15 +21,23 @@ export OPENAI_API_KEY=your-api-key
 Import the required modules:
 
 ```ts
-import { RelevancyEvaluator, OpenAI, Settings } from "llamaindex";
+import {
+  RelevancyEvaluator,
+  OpenAI,
+  serviceContextFromDefaults,
+} from "llamaindex";
 ```
 
 Let's setup gpt-4 for better results:
 
 ```ts
-Settings.llm = new OpenAI({
+const llm = new OpenAI({
   model: "gpt-4",
 });
+
+const ctx = serviceContextFromDefaults({
+  llm,
+});
 ```
 
 Now, let's create a vector index and query engine with documents and query engine respectively. Then, we can evaluate the response with the query and response from the query engine.:
@@ -51,8 +59,6 @@ const response = await queryEngine.query({
   query,
 });
 
-const evaluator = new RelevancyEvaluator();
-
 const result = await evaluator.evaluateResponse({
   query,
   response: response,

apps/docs/docs/modules/ingestion_pipeline/transformations.md

@@ -1,6 +1,6 @@
 # Transformations
 
-A transformation is something that takes a list of nodes as an input, and returns a list of nodes. Each component that implements the Transformation class has both a `transform` definition responsible for transforming the nodes.
+A transformation is something that takes a list of nodes as an input, and returns a list of nodes. Each component that implements the Transformatio class has both a `transform` definition responsible for transforming the nodes
 
 Currently, the following components are Transformation objects:
 

apps/docs/docs/modules/llms/available_llms/anthropic.md

@@ -3,11 +3,13 @@
 ## Usage
 
 ```ts
-import { Anthropic, Settings } from "llamaindex";
+import { Anthropic, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new Anthropic({
+const anthropicLLM = new Anthropic({
   apiKey: "<YOUR_API_KEY>",
 });
+
+const serviceContext = serviceContextFromDefaults({ llm: anthropicLLM });
 ```
 
 ## Load and index documents
@@ -17,7 +19,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query
@@ -35,17 +39,28 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import { Anthropic, Document, VectorStoreIndex, Settings } from "llamaindex";
-
-Settings.llm = new Anthropic({
-  apiKey: "<YOUR_API_KEY>",
-});
+import {
+  Anthropic,
+  Document,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 async function main() {
+  // Create an instance of the Anthropic LLM
+  const anthropicLLM = new Anthropic({
+    apiKey: "<YOUR_API_KEY>",
+  });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({ llm: anthropicLLM });
+
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // Create a query engine
   const queryEngine = index.asQueryEngine({

apps/docs/docs/modules/llms/available_llms/azure.md

@@ -15,9 +15,11 @@ export AZURE_OPENAI_DEPLOYMENT="gpt-4" # or some other deployment name
 ## Usage
 
 ```ts
-import { OpenAI, Settings } from "llamaindex";
+import { OpenAI, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new OpenAI({ model: "gpt-4", temperature: 0 });
+const azureOpenaiLLM = new OpenAI({ model: "gpt-4", temperature: 0 });
+
+const serviceContext = serviceContextFromDefaults({ llm: azureOpenaiLLM });
 ```
 
 ## Load and index documents
@@ -27,7 +29,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query
@@ -45,15 +49,26 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import { OpenAI, Document, VectorStoreIndex, Settings } from "llamaindex";
-
-Settings.llm = new OpenAI({ model: "gpt-4", temperature: 0 });
+import {
+  OpenAI,
+  Document,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 async function main() {
+  // Create an instance of the LLM
+  const azureOpenaiLLM = new OpenAI({ model: "gpt-4", temperature: 0 });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({ llm: azureOpenaiLLM });
+
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/fireworks.md

@@ -5,11 +5,13 @@ Fireworks.ai focus on production use cases for open source LLMs, offering speed
 ## Usage
 
 ```ts
-import { FireworksLLM, Settings } from "llamaindex";
+import { FireworksLLM, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new FireworksLLM({
+const fireworksLLM = new FireworksLLM({
   apiKey: "<YOUR_API_KEY>",
 });
+
+const serviceContext = serviceContextFromDefaults({ llm: fireworksLLM });
 ```
 
 ## Load and index documents
@@ -21,7 +23,9 @@ const reader = new PDFReader();
 const documents = await reader.loadData("../data/brk-2022.pdf");
 
 // Split text and create embeddings. Store them in a VectorStoreIndex
-const index = await VectorStoreIndex.fromDocuments(documents);
+const index = await VectorStoreIndex.fromDocuments(documents, {
+  serviceContext,
+});
 ```
 
 ## Query

apps/docs/docs/modules/llms/available_llms/groq.mdx

@@ -14,13 +14,15 @@ export GROQ_API_KEY=<your-api-key>
 The initialize the Groq module.
 
 ```ts
-import { Groq, Settings } from "llamaindex";
+import { Groq, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new Groq({
+const groq = new Groq({
   // If you do not wish to set your API key in the environment, you may
   // configure your API key when you initialize the Groq class.
   // apiKey: "<your-api-key>",
 });
+
+const serviceContext = serviceContextFromDefaults({ llm: groq });
 ```
 
 ## Load and index documents
@@ -30,7 +32,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query

apps/docs/docs/modules/llms/available_llms/llama2.md

@@ -3,9 +3,11 @@
 ## Usage
 
 ```ts
-import { Ollama, Settings, DeuceChatStrategy } from "llamaindex";
+import { Ollama, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
+const llama2LLM = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
+
+const serviceContext = serviceContextFromDefaults({ llm: llama2LLM });
 ```
 
 ## Usage with Replication
@@ -14,18 +16,19 @@ Settings.llm = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
 import {
   Ollama,
   ReplicateSession,
-  Settings,
-  DeuceChatStrategy,
+  serviceContextFromDefaults,
 } from "llamaindex";
 
 const replicateSession = new ReplicateSession({
   replicateKey,
 });
 
-Settings.llm = new LlamaDeuce({
+const llama2LLM = new LlamaDeuce({
   chatStrategy: DeuceChatStrategy.META,
   replicateSession,
 });
+
+const serviceContext = serviceContextFromDefaults({ llm: llama2LLM });
 ```
 
 ## Load and index documents
@@ -35,7 +38,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query
@@ -57,18 +62,22 @@ import {
   LlamaDeuce,
   Document,
   VectorStoreIndex,
-  Settings,
-  DeuceChatStrategy,
+  serviceContextFromDefaults,
 } from "llamaindex";
 
-// Use the LlamaDeuce LLM
-Settings.llm = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
-
 async function main() {
+  // Create an instance of the LLM
+  const llama2LLM = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({ llm: mistralLLM });
+
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/mistral.md

@@ -3,12 +3,14 @@
 ## Usage
 
 ```ts
-import { Ollama, Settings } from "llamaindex";
+import { Ollama, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new MistralAI({
+const mistralLLM = new MistralAI({
   model: "mistral-tiny",
   apiKey: "<YOUR_API_KEY>",
 });
+
+const serviceContext = serviceContextFromDefaults({ llm: mistralLLM });
 ```
 
 ## Load and index documents
@@ -18,7 +20,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query
@@ -36,16 +40,26 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import { MistralAI, Document, VectorStoreIndex, Settings } from "llamaindex";
-
-// Use the MistralAI LLM
-Settings.llm = new MistralAI({ model: "mistral-tiny" });
+import {
+  MistralAI,
+  Document,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 async function main() {
+  // Create an instance of the LLM
+  const mistralLLM = new MistralAI({ model: "mistral-tiny" });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({ llm: mistralLLM });
+
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/ollama.md

@@ -3,10 +3,14 @@
 ## Usage
 
 ```ts
-import { Ollama, Settings } from "llamaindex";
+import { Ollama, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = ollamaLLM;
-Settings.embedModel = ollamaLLM;
+const ollamaLLM = new Ollama({ model: "llama2", temperature: 0.75 });
+
+const serviceContext = serviceContextFromDefaults({
+  llm: ollamaLLM,
+  embedModel: ollamaLLM,
+});
 ```
 
 ## Load and index documents
@@ -16,7 +20,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query
@@ -34,23 +40,33 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import { Ollama, Document, VectorStoreIndex, Settings } from "llamaindex";
+import {
+  Ollama,
+  Document,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 import fs from "fs/promises";
 
-const ollama = new Ollama({ model: "llama2", temperature: 0.75 });
-
-// Use Ollama LLM and Embed Model
-Settings.llm = ollama;
-Settings.embedModel = ollama;
-
 async function main() {
+  // Create an instance of the LLM
+  const ollamaLLM = new Ollama({ model: "llama2", temperature: 0.75 });
+
   const essay = await fs.readFile("./paul_graham_essay.txt", "utf-8");
 
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({
+    embedModel: ollamaLLM, // prevent 'Set OpenAI Key in OPENAI_API_KEY env variable' error
+    llm: ollamaLLM,
+  });
+
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/openai.md

@@ -1,9 +1,11 @@
 # OpenAI
 
 ```ts
-import { OpenAI, Settings } from "llamaindex";
+import { OpenAI, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0, apiKey: <YOUR_API_KEY> });
+const openaiLLM = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0, apiKey: <YOUR_API_KEY> });
+
+const serviceContext = serviceContextFromDefaults({ llm: openaiLLM });
 ```
 
 You can setup the apiKey on the environment variables, like:
@@ -19,7 +21,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query
@@ -37,16 +41,26 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import { OpenAI, Document, VectorStoreIndex, Settings } from "llamaindex";
-
-// Use the OpenAI LLM
-Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0 });
+import {
+  OpenAI,
+  Document,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 async function main() {
+  // Create an instance of the LLM
+  const openaiLLM = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0 });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({ llm: openaiLLM });
+
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/portkey.md

@@ -3,11 +3,13 @@
 ## Usage
 
 ```ts
-import { Portkey, Settings } from "llamaindex";
+import { Portkey, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new Portkey({
+const portkeyLLM = new Portkey({
   apiKey: "<YOUR_API_KEY>",
 });
+
+const serviceContext = serviceContextFromDefaults({ llm: portkeyLLM });
 ```
 
 ## Load and index documents
@@ -17,7 +19,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query
@@ -35,19 +39,28 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import { Portkey, Document, VectorStoreIndex, Settings } from "llamaindex";
-
-// Use the Portkey LLM
-Settings.llm = new Portkey({
-  apiKey: "<YOUR_API_KEY>",
-});
+import {
+  Portkey,
+  Document,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 async function main() {
-  // Create a document
+  // Create an instance of the LLM
+  const portkeyLLM = new Portkey({
+    apiKey: "<YOUR_API_KEY>",
+  });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({ llm: portkeyLLM });
+
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/together.md

@@ -3,11 +3,13 @@
 ## Usage
 
 ```ts
-import { TogetherLLM, Settings } from "llamaindex";
+import { TogetherLLM, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new TogetherLLM({
+const togetherLLM = new TogetherLLM({
   apiKey: "<YOUR_API_KEY>",
 });
+
+const serviceContext = serviceContextFromDefaults({ llm: togetherLLM });
 ```
 
 ## Load and index documents
@@ -17,7 +19,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Query
@@ -35,17 +39,28 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import { TogetherLLM, Document, VectorStoreIndex, Settings } from "llamaindex";
-
-Settings.llm = new TogetherLLM({
-  apiKey: "<YOUR_API_KEY>",
-});
+import {
+  TogetherLLM,
+  Document,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 async function main() {
+  // Create an instance of the LLM
+  const togetherLLM = new TogetherLLM({
+    apiKey: "<YOUR_API_KEY>",
+  });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({ llm: togetherLLM });
+
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/index.md

@@ -6,12 +6,14 @@ sidebar_position: 3
 
 The LLM is responsible for reading text and generating natural language responses to queries. By default, LlamaIndex.TS uses `gpt-3.5-turbo`.
 
-The LLM can be explicitly updated through `Settings`.
+The LLM can be explicitly set in the `ServiceContext` object.
 
 ```typescript
-import { OpenAI, Settings } from "llamaindex";
+import { OpenAI, serviceContextFromDefaults } from "llamaindex";
 
-Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0 });
+const openaiLLM = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0 });
+
+const serviceContext = serviceContextFromDefaults({ llm: openaiLLM });
 ```
 
 ## Azure OpenAI
@@ -33,3 +35,4 @@ For local LLMs, currently we recommend the use of [Ollama](./available_llms/olla
 ## API Reference
 
 - [OpenAI](../api/classes/OpenAI.md)
+- [ServiceContext](../api/interfaces//ServiceContext.md)

apps/docs/docs/modules/node_parser.md

@@ -4,14 +4,15 @@ sidebar_position: 4
 
 # NodeParser
 
-The `NodeParser` in LlamaIndex is responsible for splitting `Document` objects into more manageable `Node` objects. When you call `.fromDocuments()`, the `NodeParser` from the `Settings` is used to do this automatically for you. Alternatively, you can use it to split documents ahead of time.
+The `NodeParser` in LlamaIndex is responsible for splitting `Document` objects into more manageable `Node` objects. When you call `.fromDocuments()`, the `NodeParser` from the `ServiceContext` is used to do this automatically for you. Alternatively, you can use it to split documents ahead of time.
 
 ```typescript
 import { Document, SimpleNodeParser } from "llamaindex";
 
 const nodeParser = new SimpleNodeParser();
-
-Settings.nodeParser = nodeParser;
+const nodes = nodeParser.getNodesFromDocuments([
+  new Document({ text: "I am 10 years old. John is 20 years old." }),
+]);
 ```
 
 ## TextSplitter

apps/docs/docs/modules/node_postprocessors/cohere_reranker.md

@@ -18,7 +18,7 @@ import {
   Document,
   OpenAI,
   VectorStoreIndex,
-  Settings,
+  serviceContextFromDefaults,
 } from "llamaindex";
 ```
 
@@ -29,9 +29,13 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 });
+const serviceContext = serviceContextFromDefaults({
+  llm: new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 }),
+});
 
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 ```
 
 ## Increase similarity topK to retrieve more results

apps/docs/docs/modules/node_postprocessors/index.md

@@ -36,7 +36,7 @@ const processor = new SimilarityPostprocessor({
   similarityCutoff: 0.7,
 });
 
-const filteredNodes = await processor.postprocessNodes(nodes);
+const filteredNodes = processor.postprocessNodes(nodes);
 
 // cohere rerank: rerank nodes given query using trained model
 const reranker = new CohereRerank({
@@ -58,10 +58,7 @@ Most commonly, node-postprocessors will be used in a query engine, where they ar
 ### Using Node Postprocessors in a Query Engine
 
 ```ts
-import { Node, NodeWithScore, SimilarityPostprocessor, CohereRerank, Settings } from "llamaindex";
-
-// Use OpenAI LLM
-Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 });
+import { Node, NodeWithScore, SimilarityPostprocessor, CohereRerank } from "llamaindex";
 
 const nodes: NodeWithScore[] = [
   {
@@ -82,6 +79,14 @@ const reranker = new CohereRerank({
 
 const document = new Document({ text: "essay", id_: "essay" });
 
+const serviceContext = serviceContextFromDefaults({
+  llm: new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 }),
+});
+
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
+
 const queryEngine = index.asQueryEngine({
   nodePostprocessors: [processor, reranker],
 });

apps/docs/docs/modules/node_postprocessors/jinaai_reranker.md

@@ -1,71 +0,0 @@
-# Jina AI Reranker
-
-The Jina AI Reranker is a postprocessor that uses the Jina AI Reranker API to rerank the results of a search query.
-
-## Setup
-
-Firstly, you will need to install the `llamaindex` package.
-
-```bash
-pnpm install llamaindex
-```
-
-Now, you will need to sign up for an API key at [Jina AI](https://jina.ai/reranker). Once you have your API key you can import the necessary modules and create a new instance of the `JinaAIReranker` class.
-
-```ts
-import {
-  JinaAIReranker,
-  Document,
-  OpenAI,
-  VectorStoreIndex,
-  Settings,
-} from "llamaindex";
-```
-
-## Load and index documents
-
-For this example, we will use a single document. In a real-world scenario, you would have multiple documents to index.
-
-```ts
-const document = new Document({ text: essay, id_: "essay" });
-
-Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 });
-
-const index = await VectorStoreIndex.fromDocuments([document]);
-```
-
-## Increase similarity topK to retrieve more results
-
-The default value for `similarityTopK` is 2. This means that only the most similar document will be returned. To retrieve more results, you can increase the value of `similarityTopK`.
-
-```ts
-const retriever = index.asRetriever();
-retriever.similarityTopK = 5;
-```
-
-## Create a new instance of the JinaAIReranker class
-
-Then you can create a new instance of the `JinaAIReranker` class and pass in the number of results you want to return.
-The Jina AI Reranker API key is set in the `JINAAI_API_KEY` environment variable.
-
-```bash
-export JINAAI_API_KEY=<YOUR API KEY>
-```
-
-```ts
-const nodePostprocessor = new JinaAIReranker({
-  topN: 5,
-});
-```
-
-## Create a query engine with the retriever and node postprocessor
-
-```ts
-const queryEngine = index.asQueryEngine({
-  retriever,
-  nodePostprocessors: [nodePostprocessor],
-});
-
-// log the response
-const response = await queryEngine.query("Where did the author grown up?");
-```

apps/docs/docs/modules/prompt/index.md

@@ -31,11 +31,13 @@ The first method is to create a new instance of `ResponseSynthesizer` (or the mo
 ```ts
 // Create an instance of response synthesizer
 const responseSynthesizer = new ResponseSynthesizer({
-  responseBuilder: new CompactAndRefine(undefined, newTextQaPrompt),
+  responseBuilder: new CompactAndRefine(serviceContext, newTextQaPrompt),
 });
 
 // Create index
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 
 // Query the index
 const queryEngine = index.asQueryEngine({ responseSynthesizer });
@@ -51,7 +53,9 @@ The second method is that most of the modules in LlamaIndex have a `getPrompts`
 
 ```ts
 // Create index
-const index = await VectorStoreIndex.fromDocuments([document]);
+const index = await VectorStoreIndex.fromDocuments([document], {
+  serviceContext,
+});
 
 // Query the index
 const queryEngine = index.asQueryEngine();

apps/docs/docs/modules/query_engines/metadata_filtering.md

@@ -54,13 +54,12 @@ You can create a `ChromaVectorStore` to store the documents:
 
 ```ts
 const chromaVS = new ChromaVectorStore({ collectionName });
-
-const storageContext = await storageContextFromDefaults({
+const serviceContext = await storageContextFromDefaults({
   vectorStore: chromaVS,
 });
 
 const index = await VectorStoreIndex.fromDocuments(docs, {
-  storageContext: storageContext,
+  storageContext: serviceContext,
 });
 ```
 

apps/docs/docs/modules/query_engines/router_query_engine.md

@@ -18,7 +18,7 @@ import {
   SimpleNodeParser,
   SummaryIndex,
   VectorStoreIndex,
-  Settings,
+  serviceContextFromDefaults,
 } from "llamaindex";
 ```
 
@@ -34,13 +34,17 @@ const documents = await new SimpleDirectoryReader().loadData({
 
 ## Service Context
 
-Next, we need to define some basic rules and parse the documents into nodes. We will use the `SimpleNodeParser` to parse the documents into nodes and `Settings` to define the rules (eg. LLM API key, chunk size, etc.):
+Next, we need to define some basic rules and parse the documents into nodes. We will use the `SimpleNodeParser` to parse the documents into nodes and `ServiceContext` to define the rules (eg. LLM API key, chunk size, etc.):
 
 ```ts
-Settings.llm = new OpenAI();
-Settings.nodeParser = new SimpleNodeParser({
+const nodeParser = new SimpleNodeParser({
   chunkSize: 1024,
 });
+
+const serviceContext = serviceContextFromDefaults({
+  nodeParser,
+  llm: new OpenAI(),
+});
 ```
 
 ## Creating Indices
@@ -48,8 +52,13 @@ Settings.nodeParser = new SimpleNodeParser({
 Next, we need to create some indices. We will create a `VectorStoreIndex` and a `SummaryIndex`:
 
 ```ts
-const vectorIndex = await VectorStoreIndex.fromDocuments(documents);
-const summaryIndex = await SummaryIndex.fromDocuments(documents);
+const vectorIndex = await VectorStoreIndex.fromDocuments(documents, {
+  serviceContext,
+});
+
+const summaryIndex = await SummaryIndex.fromDocuments(documents, {
+  serviceContext,
+});
 ```
 
 ## Creating Query Engines
@@ -79,6 +88,7 @@ const queryEngine = RouterQueryEngine.fromDefaults({
       description: "Useful for retrieving specific context from Abramov",
     },
   ],
+  serviceContext,
 });
 ```
 
@@ -107,23 +117,34 @@ import {
   SimpleNodeParser,
   SummaryIndex,
   VectorStoreIndex,
-  Settings,
+  serviceContextFromDefaults,
 } from "llamaindex";
 
-Settings.llm = new OpenAI();
-Settings.nodeParser = new SimpleNodeParser({
-  chunkSize: 1024,
-});
-
 async function main() {
   // Load documents from a directory
   const documents = await new SimpleDirectoryReader().loadData({
     directoryPath: "node_modules/llamaindex/examples",
   });
 
+  // Parse the documents into nodes
+  const nodeParser = new SimpleNodeParser({
+    chunkSize: 1024,
+  });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({
+    nodeParser,
+    llm: new OpenAI(),
+  });
+
   // Create indices
-  const vectorIndex = await VectorStoreIndex.fromDocuments(documents);
-  const summaryIndex = await SummaryIndex.fromDocuments(documents);
+  const vectorIndex = await VectorStoreIndex.fromDocuments(documents, {
+    serviceContext,
+  });
+
+  const summaryIndex = await SummaryIndex.fromDocuments(documents, {
+    serviceContext,
+  });
 
   // Create query engines
   const vectorQueryEngine = vectorIndex.asQueryEngine();
@@ -141,6 +162,7 @@ async function main() {
         description: "Useful for retrieving specific context from Abramov",
       },
     ],
+    serviceContext,
   });
 
   // Query the router query engine

apps/docs/docs/observability/index.md

@@ -14,9 +14,6 @@ Configure a variable once, and you'll be able to do things like the following:
 
 Each provider has similarities and differences. Take a look below for the full set of guides for each one!
 
-- [OpenLLMetry](#openllmetry)
-- [Langtrace](#langtrace)
-
 ## OpenLLMetry
 
 [OpenLLMetry](https://github.com/traceloop/openllmetry-js) is an open-source project based on OpenTelemetry for tracing and monitoring
@@ -36,29 +33,3 @@ traceloop.initialize({
   disableBatch: true,
 });
 ```
-
-## Langtrace
-
-Enhance your observability with Langtrace, a robust open-source tool supports OpenTelemetry and is designed to trace, evaluate, and manage LLM applications seamlessly. Langtrace integrates directly with LlamaIndex, offering detailed, real-time insights into performance metrics such as accuracy, evaluations, and latency.
-
-#### Install
-
-- Self-host or sign-up and generate an API key using [Langtrace](https://www.langtrace.ai) Cloud
-
-```bash
-npm install @langtrase/typescript-sdk
-```
-
-#### Initialize
-
-```js
-import * as Langtrace from "@langtrase/typescript-sdk";
-Langtrace.init({ api_key: "<YOUR_API_KEY>" });
-```
-
-Features:
-
-- OpenTelemetry compliant, ensuring broad compatibility with observability platforms.
-- Provides comprehensive logs and detailed traces of all components.
-- Real-time monitoring of accuracy, evaluations, usage, costs, and latency.
-- For more configuration options and details, visit [Langtrace Docs](https://docs.langtrace.ai/introduction).

apps/docs/docs/recipes/_category_.yml

@@ -1,2 +0,0 @@
-label: Recipes
-position: 3

apps/docs/docs/recipes/cost-analysis.mdx

@@ -1,14 +0,0 @@
-# Cost Analysis
-
-This page shows how to track LLM cost using APIs.
-
-## Callback Manager
-
-The callback manager is a class that manages the callback functions.
-
-You can register `llm-start`, `llm-end`, and `llm-stream` callbacks to the callback manager for tracking the cost.
-
-import CodeBlock from "@theme/CodeBlock";
-import CodeSource from "!raw-loader!../../../../examples/recipes/cost-analysis";
-
-<CodeBlock language="ts">{CodeSource}</CodeBlock>

apps/docs/docusaurus.config.js

@@ -163,7 +163,7 @@ const config = {
       "docusaurus-plugin-typedoc",
       {
         entryPoints: ["../../packages/core/src/index.ts"],
-        tsconfig: "../../tsconfig.json",
+        tsconfig: "../../packages/core/tsconfig.json",
         readme: "none",
         sourceLinkTemplate:
           "https://github.com/run-llama/LlamaIndexTS/blob/{gitRevision}/{path}#L{line}",

apps/docs/package.json

@@ -1,6 +1,6 @@
 {
   "name": "docs",
-  "version": "0.0.5",
+  "version": "0.0.4",
   "private": true,
   "scripts": {
     "docusaurus": "docusaurus",
@@ -15,29 +15,28 @@
     "typecheck": "tsc"
   },
   "dependencies": {
-    "@docusaurus/core": "^3.2.1",
-    "@docusaurus/remark-plugin-npm2yarn": "^3.2.1",
-    "@llamaindex/examples": "workspace:*",
-    "@mdx-js/react": "^3.0.1",
+    "@docusaurus/core": "^3.1.1",
+    "@llamaindex/env": "workspace:*",
+    "@docusaurus/remark-plugin-npm2yarn": "^3.1.1",
+    "@mdx-js/react": "^3.0.0",
     "clsx": "^2.1.0",
-    "llamaindex": "workspace:*",
-    "postcss": "^8.4.38",
+    "postcss": "^8.4.33",
     "prism-react-renderer": "^2.3.1",
     "raw-loader": "^4.0.2",
     "react": "^18.2.0",
     "react-dom": "^18.2.0"
   },
   "devDependencies": {
-    "@docusaurus/module-type-aliases": "3.2.0",
-    "@docusaurus/preset-classic": "^3.2.1",
-    "@docusaurus/theme-classic": "^3.2.1",
-    "@docusaurus/types": "^3.2.1",
-    "@tsconfig/docusaurus": "^2.0.3",
-    "@types/node": "^20.12.7",
+    "@docusaurus/module-type-aliases": "3.1.0",
+    "@docusaurus/preset-classic": "^3.1.1",
+    "@docusaurus/theme-classic": "^3.1.1",
+    "@docusaurus/types": "^3.1.1",
+    "@tsconfig/docusaurus": "^2.0.2",
+    "@types/node": "^18.19.10",
     "docusaurus-plugin-typedoc": "^0.22.0",
-    "typedoc": "^0.25.13",
+    "typedoc": "^0.25.7",
     "typedoc-plugin-markdown": "^3.17.1",
-    "typescript": "^5.4.4"
+    "typescript": "^5.3.3"
   },
   "browserslist": {
     "production": [

examples/CHANGELOG.md

@@ -1,14 +0,0 @@
-# examples
-
-## 0.0.4
-
-### Patch Changes
-
-- d2e8d0c: add support for Milvus vector store
-- Updated dependencies [d2e8d0c]
-- Updated dependencies [aefc326]
-- Updated dependencies [484a710]
-- Updated dependencies [d766bd0]
-- Updated dependencies [dd95927]
-- Updated dependencies [bf583a7]
-  - llamaindex@0.2.0

examples/Settings.ts

@@ -1,29 +0,0 @@
-import fs from "node:fs/promises";
-
-import { Document, OpenAI, Settings, VectorStoreIndex } from "llamaindex";
-
-Settings.llm = new OpenAI({ model: "gpt-4" });
-
-async function main() {
-  // Load essay from abramov.txt in Node
-  const path = "node_modules/llamaindex/examples/abramov.txt";
-
-  const essay = await fs.readFile(path, "utf-8");
-
-  // Create Document object with essay
-  const document = new Document({ text: essay, id_: path });
-
-  const index = await VectorStoreIndex.fromDocuments([document]);
-
-  // Query the index
-  const queryEngine = index.asQueryEngine();
-
-  const response = await queryEngine.query({
-    query: "What did the author do in college?",
-  });
-
-  // Output response
-  console.log(response.toString());
-}
-
-main().catch(console.error);

examples/agent/multi_document_agent.ts

@@ -6,11 +6,11 @@ import {
   OpenAI,
   OpenAIAgent,
   QueryEngineTool,
-  Settings,
   SimpleNodeParser,
   SimpleToolNodeMapping,
   SummaryIndex,
   VectorStoreIndex,
+  serviceContextFromDefaults,
   storageContextFromDefaults,
 } from "llamaindex";
 
@@ -18,8 +18,6 @@ import { extractWikipedia } from "./helpers/extractWikipedia";
 
 const wikiTitles = ["Brazil", "Canada"];
 
-Settings.llm = new OpenAI({ model: "gpt-4" });
-
 async function main() {
   await extractWikipedia(wikiTitles);
 
@@ -32,6 +30,11 @@ async function main() {
     countryDocs[title] = document;
   }
 
+  const llm = new OpenAI({
+    model: "gpt-4",
+  });
+
+  const serviceContext = serviceContextFromDefaults({ llm });
   const storageContext = await storageContextFromDefaults({
     persistDir: "./storage",
   });
@@ -51,11 +54,13 @@ async function main() {
     console.log(`Creating index for ${title}`);
 
     const vectorIndex = await VectorStoreIndex.init({
+      serviceContext: serviceContext,
       storageContext: storageContext,
       nodes,
     });
 
     const summaryIndex = await SummaryIndex.init({
+      serviceContext: serviceContext,
       nodes,
     });
 
@@ -85,7 +90,8 @@ async function main() {
 
     const agent = new OpenAIAgent({
       tools: queryEngineTools,
-      llm: new OpenAI({ model: "gpt-4" }),
+      llm,
+      verbose: true,
     });
 
     documentAgents[title] = agent;
@@ -120,12 +126,16 @@ async function main() {
     allTools,
     toolMapping,
     VectorStoreIndex,
+    {
+      serviceContext,
+    },
   );
 
   const topAgent = new OpenAIAgent({
     toolRetriever: await objectIndex.asRetriever({}),
-    llm: new OpenAI({ model: "gpt-4" }),
-    chatHistory: [
+    llm,
+    verbose: true,
+    prefixMessages: [
       {
         content:
           "You are an agent designed to answer queries about a set of given countries. Please always use the tools provided to answer a question. Do not rely on prior knowledge.",
@@ -143,4 +153,4 @@ async function main() {
   });
 }
 
-void main();
+main();

examples/agent/openai.ts

@@ -1,61 +1,76 @@
 import { FunctionTool, OpenAIAgent } from "llamaindex";
 
-const sumNumbers = FunctionTool.from(
-  ({ a, b }: { a: number; b: number }) => `${a + b}`,
-  {
-    name: "sumNumbers",
-    description: "Use this function to sum two numbers",
-    parameters: {
-      type: "object",
-      properties: {
-        a: {
-          type: "number",
-          description: "The first number",
-        },
-        b: {
-          type: "number",
-          description: "The second number",
-        },
-      },
-      required: ["a", "b"],
-    },
-  },
-);
+// Define a function to sum two numbers
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
+}
 
-const divideNumbers = FunctionTool.from(
-  ({ a, b }: { a: number; b: number }) => `${a / b}`,
-  {
-    name: "divideNumbers",
-    description: "Use this function to divide two numbers",
-    parameters: {
-      type: "object",
-      properties: {
-        a: {
-          type: "number",
-          description: "The dividend a to divide",
-        },
-        b: {
-          type: "number",
-          description: "The divisor b to divide by",
-        },
-      },
-      required: ["a", "b"],
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
+}
+
+// Define the parameters of the sum function as a JSON schema
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
     },
   },
-);
+  required: ["a", "b"],
+};
+
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The dividend a to divide",
+    },
+    b: {
+      type: "number",
+      description: "The divisor b to divide by",
+    },
+  },
+  required: ["a", "b"],
+};
 
 async function main() {
-  const agent = new OpenAIAgent({
-    tools: [sumNumbers, divideNumbers],
+  // Create a function tool from the sum function
+  const functionTool = new FunctionTool(sumNumbers, {
+    name: "sumNumbers",
+    description: "Use this function to sum two numbers",
+    parameters: sumJSON,
   });
 
+  // Create a function tool from the divide function
+  const functionTool2 = new FunctionTool(divideNumbers, {
+    name: "divideNumbers",
+    description: "Use this function to divide two numbers",
+    parameters: divideJSON,
+  });
+
+  // Create an OpenAIAgent with the function tools
+  const agent = new OpenAIAgent({
+    tools: [functionTool, functionTool2],
+    verbose: true,
+  });
+
+  // Chat with the agent
   const response = await agent.chat({
     message: "How much is 5 + 5? then divide by 2",
   });
 
+  // Print the response
   console.log(String(response));
 }
 
-void main().then(() => {
+main().then(() => {
   console.log("Done");
 });

examples/agent/query_openai_agent.ts

@@ -29,6 +29,7 @@ async function main() {
   // Create an OpenAIAgent with the function tools
   const agent = new OpenAIAgent({
     tools: [queryEngineTool],
+    verbose: true,
   });
 
   // Chat with the agent
@@ -40,6 +41,6 @@ async function main() {
   console.log(String(response));
 }
 
-void main().then(() => {
+main().then(() => {
   console.log("Done");
 });

examples/agent/react_agent.ts

@@ -1,13 +1,13 @@
-import { Anthropic, FunctionTool, ReActAgent } from "llamaindex";
+import { FunctionTool, ReActAgent } from "llamaindex";
 
 // Define a function to sum two numbers
-function sumNumbers({ a, b }: { a: number; b: number }) {
-  return `${a + b}`;
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
 }
 
 // Define a function to divide two numbers
-function divideNumbers({ a, b }: { a: number; b: number }) {
-  return `${a / b}`;
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
 }
 
 // Define the parameters of the sum function as a JSON schema
@@ -24,7 +24,7 @@ const sumJSON = {
     },
   },
   required: ["a", "b"],
-} as const;
+};
 
 const divideJSON = {
   type: "object",
@@ -39,7 +39,7 @@ const divideJSON = {
     },
   },
   required: ["a", "b"],
-} as const;
+};
 
 async function main() {
   // Create a function tool from the sum function
@@ -56,26 +56,21 @@ async function main() {
     parameters: divideJSON,
   });
 
-  const anthropic = new Anthropic({
-    apiKey: process.env.ANTHROPIC_API_KEY,
-    model: "claude-3-opus",
-  });
-
-  // Create an ReActAgent with the function tools
+  // Create an OpenAIAgent with the function tools
   const agent = new ReActAgent({
-    llm: anthropic,
     tools: [functionTool, functionTool2],
+    verbose: true,
   });
 
   // Chat with the agent
-  const { response } = await agent.chat({
+  const response = await agent.chat({
     message: "Divide 16 by 2 then add 20",
   });
 
   // Print the response
-  console.log(response.message);
+  console.log(String(response));
 }
 
-void main().then(() => {
+main().then(() => {
   console.log("Done");
 });

examples/agent/step_wise_openai.ts

@@ -1,13 +1,13 @@
 import { FunctionTool, OpenAIAgent } from "llamaindex";
 
 // Define a function to sum two numbers
-function sumNumbers({ a, b }: { a: number; b: number }) {
-  return `${a + b}`;
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
 }
 
 // Define a function to divide two numbers
-function divideNumbers({ a, b }: { a: number; b: number }) {
-  return `${a / b}`;
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
 }
 
 // Define the parameters of the sum function as a JSON schema
@@ -24,22 +24,22 @@ const sumJSON = {
     },
   },
   required: ["a", "b"],
-} as const;
+};
 
 const divideJSON = {
   type: "object",
   properties: {
     a: {
       type: "number",
-      description: "The dividend",
+      description: "The dividend a to divide",
     },
     b: {
       type: "number",
-      description: "The divisor",
+      description: "The divisor b to divide by",
     },
   },
   required: ["a", "b"],
-} as const;
+};
 
 async function main() {
   // Create a function tool from the sum function
@@ -59,25 +59,37 @@ async function main() {
   // Create an OpenAIAgent with the function tools
   const agent = new OpenAIAgent({
     tools: [functionTool, functionTool2],
+    verbose: true,
   });
 
   // Create a task to sum and divide numbers
-  const task = await agent.createTask("How much is 5 + 5? then divide by 2");
+  const task = agent.createTask("How much is 5 + 5? then divide by 2");
 
   let count = 0;
 
-  for await (const stepOutput of task) {
+  while (true) {
+    const stepOutput = await agent.runStep(task.taskId);
+
     console.log(`Runnning step ${count++}`);
     console.log(`======== OUTPUT ==========`);
-    console.log(stepOutput.output.message.content);
+    if (stepOutput.output.response) {
+      console.log(stepOutput.output.response);
+    } else {
+      console.log(stepOutput.output.sources);
+    }
     console.log(`==========================`);
 
     if (stepOutput.isLast) {
-      console.log(stepOutput.output.message.content);
+      const finalResponse = await agent.finalizeResponse(
+        task.taskId,
+        stepOutput,
+      );
+      console.log({ finalResponse });
+      break;
     }
   }
 }
 
-void main().then(() => {
+main().then(() => {
   console.log("Done");
 });

examples/agent/step_wise_query_tool.ts

@@ -29,15 +29,36 @@ async function main() {
   // Create an OpenAIAgent with the function tools
   const agent = new OpenAIAgent({
     tools: [queryEngineTool],
+    verbose: true,
   });
 
-  const { response } = await agent.chat({
-    message: "What was his salary?",
-  });
+  const task = agent.createTask("What was his salary?");
 
-  console.log(response.message.content);
+  let count = 0;
+
+  while (true) {
+    const stepOutput = await agent.runStep(task.taskId);
+
+    console.log(`Runnning step ${count++}`);
+    console.log(`======== OUTPUT ==========`);
+    if (stepOutput.output.response) {
+      console.log(stepOutput.output.response);
+    } else {
+      console.log(stepOutput.output.sources);
+    }
+    console.log(`==========================`);
+
+    if (stepOutput.isLast) {
+      const finalResponse = await agent.finalizeResponse(
+        task.taskId,
+        stepOutput,
+      );
+      console.log({ finalResponse });
+      break;
+    }
+  }
 }
 
-void main().then(() => {
+main().then(() => {
   console.log("Done");
 });

examples/agent/step_wise_react.ts

@@ -1,14 +1,13 @@
-import { Anthropic, FunctionTool, ReActAgent } from "llamaindex";
+import { FunctionTool, ReActAgent } from "llamaindex";
 
 // Define a function to sum two numbers
-function sumNumbers({ a, b }: { a: number; b: number }) {
-  return `${a + b}`;
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
 }
 
 // Define a function to divide two numbers
-function divideNumbers({ a, b }: { a: number; b: number }) {
-  console.log("get input", a, b);
-  return `${a / b}`;
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
 }
 
 // Define the parameters of the sum function as a JSON schema
@@ -25,7 +24,7 @@ const sumJSON = {
     },
   },
   required: ["a", "b"],
-} as const;
+};
 
 const divideJSON = {
   type: "object",
@@ -40,7 +39,7 @@ const divideJSON = {
     },
   },
   required: ["a", "b"],
-} as const;
+};
 
 async function main() {
   // Create a function tool from the sum function
@@ -59,24 +58,33 @@ async function main() {
 
   // Create an OpenAIAgent with the function tools
   const agent = new ReActAgent({
-    llm: new Anthropic({
-      model: "claude-3-opus",
-    }),
     tools: [functionTool, functionTool2],
+    verbose: true,
   });
 
-  const task = await agent.createTask("Divide 16 by 2 then add 20");
+  const task = agent.createTask("Divide 16 by 2 then add 20");
 
   let count = 0;
 
-  for await (const stepOutput of task) {
+  while (true) {
+    const stepOutput = await agent.runStep(task.taskId);
+
     console.log(`Runnning step ${count++}`);
     console.log(`======== OUTPUT ==========`);
-    console.log(stepOutput);
+    console.log(stepOutput.output);
     console.log(`==========================`);
+
+    if (stepOutput.isLast) {
+      const finalResponse = await agent.finalizeResponse(
+        task.taskId,
+        stepOutput,
+      );
+      console.log({ finalResponse });
+      break;
+    }
   }
 }
 
-void main().then(() => {
+main().then(() => {
   console.log("Done");
 });

examples/agent/stream_openai_agent.ts

@@ -1,13 +1,13 @@
 import { FunctionTool, OpenAIAgent } from "llamaindex";
 
 // Define a function to sum two numbers
-function sumNumbers({ a, b }: { a: number; b: number }) {
-  return `${a + b}`;
+function sumNumbers({ a, b }: { a: number; b: number }): number {
+  return a + b;
 }
 
 // Define a function to divide two numbers
-function divideNumbers({ a, b }: { a: number; b: number }) {
-  return `${a / b}`;
+function divideNumbers({ a, b }: { a: number; b: number }): number {
+  return a / b;
 }
 
 // Define the parameters of the sum function as a JSON schema
@@ -24,7 +24,7 @@ const sumJSON = {
     },
   },
   required: ["a", "b"],
-} as const;
+};
 
 const divideJSON = {
   type: "object",
@@ -39,18 +39,18 @@ const divideJSON = {
     },
   },
   required: ["a", "b"],
-} as const;
+};
 
 async function main() {
   // Create a function tool from the sum function
-  const functionTool = FunctionTool.from(sumNumbers, {
+  const functionTool = new FunctionTool(sumNumbers, {
     name: "sumNumbers",
     description: "Use this function to sum two numbers",
     parameters: sumJSON,
   });
 
   // Create a function tool from the divide function
-  const functionTool2 = FunctionTool.from(divideNumbers, {
+  const functionTool2 = new FunctionTool(divideNumbers, {
     name: "divideNumbers",
     description: "Use this function to divide two numbers",
     parameters: divideJSON,
@@ -59,6 +59,7 @@ async function main() {
   // Create an OpenAIAgent with the function tools
   const agent = new OpenAIAgent({
     tools: [functionTool, functionTool2],
+    verbose: false,
   });
 
   const stream = await agent.chat({
@@ -71,6 +72,6 @@ async function main() {
   }
 }
 
-void main().then(() => {
+main().then(() => {
   console.log("\nDone");
 });

examples/agent/wiki.ts

@@ -1,27 +0,0 @@
-import { OpenAI, OpenAIAgent, WikipediaTool } from "llamaindex";
-
-async function main() {
-  const llm = new OpenAI({ model: "gpt-4-turbo" });
-  const wikiTool = new WikipediaTool();
-
-  // Create an OpenAIAgent with the Wikipedia tool
-  const agent = new OpenAIAgent({
-    llm,
-    tools: [wikiTool],
-  });
-
-  // Chat with the agent
-  const response = await agent.chat({
-    message: "Who was Goethe?",
-    stream: true,
-  });
-
-  for await (const chunk of response.response) {
-    process.stdout.write(chunk.response);
-  }
-}
-
-(async function () {
-  await main();
-  console.log("\nDone");
-})();

examples/anthropic/agent.ts

@@ -1,43 +0,0 @@
-import { FunctionTool, Settings, WikipediaTool } from "llamaindex";
-import { AnthropicAgent } from "llamaindex/agent/anthropic";
-
-Settings.callbackManager.on("llm-tool-call", (event) => {
-  console.log("llm-tool-call", event.detail.payload.toolCall);
-});
-
-const agent = new AnthropicAgent({
-  tools: [
-    FunctionTool.from<{ location: string }>(
-      (query) => {
-        return `The weather in ${query.location} is sunny`;
-      },
-      {
-        name: "weather",
-        description: "Get the weather",
-        parameters: {
-          type: "object",
-          properties: {
-            location: {
-              type: "string",
-              description: "The location to get the weather for",
-            },
-          },
-          required: ["location"],
-        },
-      },
-    ),
-    new WikipediaTool(),
-  ],
-});
-
-async function main() {
-  // https://docs.anthropic.com/claude/docs/tool-use#tool-use-best-practices-and-limitations
-  const { response } = await agent.chat({
-    message:
-      "What is the weather in New York? What's the history of New York from Wikipedia in 3 sentences?",
-  });
-
-  console.log(response);
-}
-
-void main();

examples/anthropic/haiku.ts

@@ -1,19 +0,0 @@
-import { Anthropic } from "llamaindex";
-
-(async () => {
-  const anthropic = new Anthropic({
-    apiKey: process.env.ANTHROPIC_API_KEY,
-    model: "claude-3-haiku",
-  });
-  const result = await anthropic.chat({
-    messages: [
-      { content: "You want to talk in rhymes.", role: "system" },
-      {
-        content:
-          "How much wood would a woodchuck chuck if a woodchuck could chuck wood?",
-        role: "user",
-      },
-    ],
-  });
-  console.log(result);
-})();

examples/astradb/README.md

@@ -13,7 +13,7 @@ Here are two sample scripts which work well with the sample data in the Astra Po
 1. Set your env variables:
 
 - `ASTRA_DB_APPLICATION_TOKEN`: The generated app token for your Astra database
-- `ASTRA_DB_API_ENDPOINT`: The API endpoint for your Astra database
+- `ASTRA_DB_ENDPOINT`: The API endpoint for your Astra database
 - `ASTRA_DB_NAMESPACE`: (Optional) The namespace where your collection is stored defaults to `default_keyspace`
 - `OPENAI_API_KEY`: Your OpenAI key
 
@@ -32,10 +32,10 @@ run `ts-node astradb/example`
 
 This sample loads the same dataset of movie reviews as the Astra Portal sample dataset. (Feel free to load the data in your the Astra Data Explorer to compare)
 
-run `npx ts-node astradb/load`
+run `ts-node astradb/load`
 
 ### Use RAG to Query the data
 
 Check out your data in the Astra Data Explorer and change the sample query as you see fit.
 
-run `npx ts-node astradb/query`
+run `ts-node astradb/query`

examples/astradb/example.ts

@@ -34,9 +34,10 @@ async function main() {
     ];
 
     const astraVS = new AstraDBVectorStore();
-    await astraVS.createAndConnect(collectionName, {
+    await astraVS.create(collectionName, {
       vector: { dimension: 1536, metric: "cosine" },
     });
+    await astraVS.connect(collectionName);
 
     const ctx = await storageContextFromDefaults({ vectorStore: astraVS });
     const index = await VectorStoreIndex.fromDocuments(docs, {
@@ -54,4 +55,4 @@ async function main() {
   }
 }
 
-void main();
+main();

examples/astradb/load.ts

@@ -13,7 +13,7 @@ async function main() {
     const docs = await reader.loadData("./data/movie_reviews.csv");
 
     const astraVS = new AstraDBVectorStore({ contentKey: "reviewtext" });
-    await astraVS.createAndConnect(collectionName, {
+    await astraVS.create(collectionName, {
       vector: { dimension: 1536, metric: "cosine" },
     });
     await astraVS.connect(collectionName);
@@ -27,4 +27,4 @@ async function main() {
   }
 }
 
-void main();
+main();

examples/astradb/query.ts

@@ -1,7 +1,7 @@
 import {
   AstraDBVectorStore,
-  VectorStoreIndex,
   serviceContextFromDefaults,
+  VectorStoreIndex,
 } from "llamaindex";
 
 const collectionName = "movie_reviews";
@@ -28,4 +28,4 @@ async function main() {
   }
 }
 
-void main();
+main();

examples/chatEngine.ts

@@ -4,18 +4,18 @@ import readline from "node:readline/promises";
 import {
   ContextChatEngine,
   Document,
-  Settings,
+  serviceContextFromDefaults,
   VectorStoreIndex,
 } from "llamaindex";
 
 import essay from "./essay";
 
-// Update chunk size
-Settings.chunkSize = 512;
-
 async function main() {
   const document = new Document({ text: essay });
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const serviceContext = serviceContextFromDefaults({ chunkSize: 512 });
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
   const retriever = index.asRetriever();
   retriever.similarityTopK = 5;
   const chatEngine = new ContextChatEngine({ retriever });

examples/chatHistory.ts

@@ -1,18 +1,7 @@
 import { stdin as input, stdout as output } from "node:process";
 import readline from "node:readline/promises";
 
-import {
-  OpenAI,
-  Settings,
-  SimpleChatEngine,
-  SummaryChatHistory,
-} from "llamaindex";
-
-if (process.env.NODE_ENV === "development") {
-  Settings.callbackManager.on("llm-end", (event) => {
-    console.log("callers chain", event.reason?.computedCallers);
-  });
-}
+import { OpenAI, SimpleChatEngine, SummaryChatHistory } from "llamaindex";
 
 async function main() {
   // Set maxTokens to 75% of the context window size of 4096

examples/chromadb/preFilters.ts

@@ -54,4 +54,4 @@ async function main() {
   }
 }
 
-void main();
+main();

examples/chromadb/test.ts

@@ -37,4 +37,4 @@ async function main() {
   }
 }
 
-void main();
+main();

examples/cloud/README.md

@@ -31,11 +31,3 @@ This example shows how to use the managed index with a query engine.
 ```shell
 pnpx ts-node cloud/query.ts
 ```
-
-## Pipeline
-
-This example shows how to create a managed index with a pipeline.
-
-```shell
-pnpx ts-node cloud/pipeline.ts
-```

examples/cloud/from_documents.ts

@@ -1,44 +0,0 @@
-import fs from "node:fs/promises";
-
-import { stdin as input, stdout as output } from "node:process";
-
-import readline from "node:readline/promises";
-
-import { Document, LlamaCloudIndex } from "llamaindex";
-
-async function main() {
-  const path = "node_modules/llamaindex/examples/abramov.txt";
-
-  const essay = await fs.readFile(path, "utf-8");
-
-  // Create Document object with essay
-  const document = new Document({ text: essay, id_: path });
-
-  const index = await LlamaCloudIndex.fromDocuments({
-    documents: [document],
-    name: "test",
-    projectName: "default",
-    apiKey: process.env.LLAMA_CLOUD_API_KEY,
-    baseUrl: process.env.LLAMA_CLOUD_BASE_URL,
-  });
-
-  const queryEngine = index.asQueryEngine({
-    denseSimilarityTopK: 5,
-  });
-
-  const rl = readline.createInterface({ input, output });
-
-  while (true) {
-    const query = await rl.question("Query: ");
-    const stream = await queryEngine.query({
-      query,
-      stream: true,
-    });
-    console.log();
-    for await (const chunk of stream) {
-      process.stdout.write(chunk.response);
-    }
-  }
-}
-
-main().catch(console.error);

examples/cloud/pipeline.ts

@@ -1,34 +0,0 @@
-import fs from "node:fs/promises";
-
-import {
-  Document,
-  IngestionPipeline,
-  OpenAIEmbedding,
-  SimpleNodeParser,
-} from "llamaindex";
-
-async function main() {
-  // Load essay from abramov.txt in Node
-  const path = "node_modules/llamaindex/examples/abramov.txt";
-
-  const essay = await fs.readFile(path, "utf-8");
-
-  // Create Document object with essay
-  const document = new Document({ text: essay, id_: path });
-  const pipeline = new IngestionPipeline({
-    name: "pipeline",
-    transformations: [
-      new SimpleNodeParser({ chunkSize: 1024, chunkOverlap: 20 }),
-      new OpenAIEmbedding({ apiKey: "api-key" }),
-    ],
-  });
-
-  const pipelineId = await pipeline.register({
-    documents: [document],
-    verbose: true,
-  });
-
-  console.log(`Pipeline with id ${pipelineId} successfully created.`);
-}
-
-main().catch(console.error);

examples/evaluation/correctness.ts

@@ -1,10 +1,21 @@
-import { CorrectnessEvaluator, OpenAI, Settings } from "llamaindex";
-
-// Update llm to use OpenAI
-Settings.llm = new OpenAI({ model: "gpt-4" });
+import {
+  CorrectnessEvaluator,
+  OpenAI,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 async function main() {
-  const evaluator = new CorrectnessEvaluator();
+  const llm = new OpenAI({
+    model: "gpt-4",
+  });
+
+  const ctx = serviceContextFromDefaults({
+    llm,
+  });
+
+  const evaluator = new CorrectnessEvaluator({
+    serviceContext: ctx,
+  });
 
   const query =
     "Can you explain the theory of relativity proposed by Albert Einstein in detail?";
@@ -22,4 +33,4 @@ However, general relativity, published in 1915, extended these ideas to include
   console.log(result);
 }
 
-void main();
+main();

examples/evaluation/faithfulness.ts

@@ -2,15 +2,22 @@ import {
   Document,
   FaithfulnessEvaluator,
   OpenAI,
-  Settings,
   VectorStoreIndex,
+  serviceContextFromDefaults,
 } from "llamaindex";
 
-// Update llm to use OpenAI
-Settings.llm = new OpenAI({ model: "gpt-4" });
-
 async function main() {
-  const evaluator = new FaithfulnessEvaluator();
+  const llm = new OpenAI({
+    model: "gpt-4",
+  });
+
+  const ctx = serviceContextFromDefaults({
+    llm,
+  });
+
+  const evaluator = new FaithfulnessEvaluator({
+    serviceContext: ctx,
+  });
 
   const documents = [
     new Document({
@@ -36,4 +43,4 @@ async function main() {
   console.log(result);
 }
 
-void main();
+main();

examples/evaluation/relevancy.ts

@@ -2,16 +2,22 @@ import {
   Document,
   OpenAI,
   RelevancyEvaluator,
-  Settings,
   VectorStoreIndex,
+  serviceContextFromDefaults,
 } from "llamaindex";
 
-Settings.llm = new OpenAI({
-  model: "gpt-4",
-});
-
 async function main() {
-  const evaluator = new RelevancyEvaluator();
+  const llm = new OpenAI({
+    model: "gpt-4",
+  });
+
+  const ctx = serviceContextFromDefaults({
+    llm,
+  });
+
+  const evaluator = new RelevancyEvaluator({
+    serviceContext: ctx,
+  });
 
   const documents = [
     new Document({
@@ -37,4 +43,4 @@ async function main() {
   console.log(result);
 }
 
-void main();
+main();

examples/gptturbollama3.ts

@@ -1,40 +0,0 @@
-import { stdin as input, stdout as output } from "node:process";
-import readline from "node:readline/promises";
-
-import { ChatMessage, OpenAI, ReplicateLLM } from "llamaindex";
-
-(async () => {
-  const gpt4 = new OpenAI({ model: "gpt-4-turbo", temperature: 0.9 });
-  const l3 = new ReplicateLLM({
-    model: "llama-3-70b-instruct",
-    temperature: 0.9,
-  });
-
-  const rl = readline.createInterface({ input, output });
-  const start = await rl.question("Start: ");
-  const history: ChatMessage[] = [
-    {
-      content:
-        "Prefer shorter answers. Keep your response to 100 words or less.",
-      role: "system",
-    },
-    { content: start, role: "user" },
-  ];
-
-  while (true) {
-    const next = history.length % 2 === 1 ? gpt4 : l3;
-    const r = await next.chat({
-      messages: history.map(({ content, role }) => ({
-        content,
-        role: next === l3 ? role : role === "user" ? "assistant" : "user",
-      })),
-    });
-    history.push({
-      content: r.message.content,
-      role: next === l3 ? "assistant" : "user",
-    });
-    await rl.question(
-      (next === l3 ? "Llama 3: " : "GPT 4 Turbo: ") + r.message.content,
-    );
-  }
-})();

examples/groq.ts

@@ -1,20 +1,30 @@
 import fs from "node:fs/promises";
 
-import { Document, Groq, Settings, VectorStoreIndex } from "llamaindex";
-
-// Update llm to use Groq
-Settings.llm = new Groq({
-  apiKey: process.env.GROQ_API_KEY,
-});
+import {
+  Document,
+  Groq,
+  VectorStoreIndex,
+  serviceContextFromDefaults,
+} from "llamaindex";
 
 async function main() {
+  // Create an instance of the LLM
+  const groq = new Groq({
+    apiKey: process.env.GROQ_API_KEY,
+  });
+
+  // Create a service context
+  const serviceContext = serviceContextFromDefaults({ llm: groq });
+
   // Load essay from abramov.txt in Node
   const path = "node_modules/llamaindex/examples/abramov.txt";
   const essay = await fs.readFile(path, "utf-8");
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // get retriever
   const retriever = index.asRetriever();

examples/huggingface.ts

@@ -4,15 +4,10 @@ import {
   Document,
   HuggingFaceEmbedding,
   HuggingFaceEmbeddingModelType,
-  Settings,
   VectorStoreIndex,
+  serviceContextFromDefaults,
 } from "llamaindex";
 
-// Update embed model
-Settings.embedModel = new HuggingFaceEmbedding({
-  modelType: HuggingFaceEmbeddingModelType.XENOVA_ALL_MPNET_BASE_V2,
-});
-
 async function main() {
   // Load essay from abramov.txt in Node
   const path = "node_modules/llamaindex/examples/abramov.txt";
@@ -22,8 +17,18 @@ async function main() {
   // Create Document object with essay
   const document = new Document({ text: essay, id_: path });
 
+  // Use Local embedding from HuggingFace
+  const embedModel = new HuggingFaceEmbedding({
+    modelType: HuggingFaceEmbeddingModelType.XENOVA_ALL_MPNET_BASE_V2,
+  });
+  const serviceContext = serviceContextFromDefaults({
+    embedModel,
+  });
+
   // Split text and create embeddings. Store them in a VectorStoreIndex
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // Query the index
   const queryEngine = index.asQueryEngine();

examples/jsonExtract.ts

@@ -36,7 +36,9 @@ async function main() {
     ],
   });
 
-  console.log(response.message.content);
+  const json = JSON.parse(response.message.content);
+
+  console.log(json);
 }
 
 main().catch(console.error);

examples/llama3.ts

@@ -1,13 +0,0 @@
-import { ReplicateLLM } from "llamaindex";
-
-(async () => {
-  const tres = new ReplicateLLM({ model: "llama-3-70b-instruct" });
-  const stream = await tres.chat({
-    messages: [{ content: "Hello, world!", role: "user" }],
-    stream: true,
-  });
-  for await (const chunk of stream) {
-    process.stdout.write(chunk.delta);
-  }
-  console.log("\n\ndone");
-})();

examples/longText.ts

@@ -1,21 +1,26 @@
 import {
   Document,
-  Settings,
   SimpleNodeParser,
   VectorStoreIndex,
+  serviceContextFromDefaults,
 } from "llamaindex";
 
 export const STORAGE_DIR = "./data";
 
-// Update node parser
-Settings.nodeParser = new SimpleNodeParser({
-  chunkSize: 512,
-  chunkOverlap: 20,
-  splitLongSentences: true,
-});
 (async () => {
+  // create service context that is splitting sentences longer than CHUNK_SIZE
+  const serviceContext = serviceContextFromDefaults({
+    nodeParser: new SimpleNodeParser({
+      chunkSize: 512,
+      chunkOverlap: 20,
+      splitLongSentences: true,
+    }),
+  });
+
   // generate a document with a very long sentence (9000 words long)
   const longSentence = "is ".repeat(9000) + ".";
   const document = new Document({ text: longSentence, id_: "1" });
-  await VectorStoreIndex.fromDocuments([document]);
+  await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 })();

examples/milvus/README.md

@@ -1,34 +0,0 @@
-# Milvus Vector Store
-
-Here are two sample scripts which work with loading and querying data from a Milvus Vector Store.
-
-## Prerequisites
-
-- An Milvus Vector Database
-  - Hosted https://milvus.io/
-  - Self Hosted https://milvus.io/docs/install_standalone-docker.md
-- An OpenAI API Key
-
-## Setup
-
-1. Set your env variables:
-
-- `MILVUS_ADDRESS`: Address of your Milvus Vector Store (like localhost:19530)
-- `MILVUS_USERNAME`: empty or username for your Milvus Vector Store
-- `MILVUS_PASSWORD`: empty or password for your Milvus Vector Store
-- `OPENAI_API_KEY`: Your OpenAI key
-
-2. `cd` Into the `examples` directory
-3. run `npm i`
-
-## Load the data
-
-This sample loads the same dataset of movie reviews as sample dataset. You can install https://github.com/zilliztech/attu to inspect the loaded data.
-
-run `npx ts-node milvus/load`
-
-## Use RAG to Query the data
-
-Check out your data in Attu and change the sample query as you see fit.
-
-run `npx ts-node milvus/query`

examples/milvus/load.ts

@@ -1,26 +0,0 @@
-import {
-  MilvusVectorStore,
-  PapaCSVReader,
-  storageContextFromDefaults,
-  VectorStoreIndex,
-} from "llamaindex";
-
-const collectionName = "movie_reviews";
-
-async function main() {
-  try {
-    const reader = new PapaCSVReader(false);
-    const docs = await reader.loadData("./data/movie_reviews.csv");
-
-    const vectorStore = new MilvusVectorStore({ collection: collectionName });
-
-    const ctx = await storageContextFromDefaults({ vectorStore });
-    const index = await VectorStoreIndex.fromDocuments(docs, {
-      storageContext: ctx,
-    });
-  } catch (e) {
-    console.error(e);
-  }
-}
-
-void main();

examples/milvus/query.ts

@@ -1,25 +0,0 @@
-import { MilvusVectorStore, VectorStoreIndex } from "llamaindex";
-
-const collectionName = "movie_reviews";
-
-async function main() {
-  try {
-    const milvus = new MilvusVectorStore({ collection: collectionName });
-
-    const index = await VectorStoreIndex.fromVectorStore(milvus);
-
-    const retriever = await index.asRetriever({ similarityTopK: 20 });
-
-    const queryEngine = await index.asQueryEngine({ retriever });
-
-    const results = await queryEngine.query({
-      query: "What is the best reviewed movie?",
-    });
-
-    console.log(results.response);
-  } catch (e) {
-    console.error(e);
-  }
-}
-
-void main();

examples/mistral.ts

@@ -1,18 +1,15 @@
 import * as fs from "fs/promises";
 import {
+  BaseEmbedding,
   Document,
+  LLM,
   MistralAI,
   MistralAIEmbedding,
-  Settings,
   VectorStoreIndex,
+  serviceContextFromDefaults,
 } from "llamaindex";
 
-// Update embed model
-Settings.embedModel = new MistralAIEmbedding();
-// Update llm to use MistralAI
-Settings.llm = new MistralAI({ model: "mistral-tiny" });
-
-async function rag(query: string) {
+async function rag(llm: LLM, embedModel: BaseEmbedding, query: string) {
   // Load essay from abramov.txt in Node
   const path = "node_modules/llamaindex/examples/abramov.txt";
 
@@ -21,7 +18,12 @@ async function rag(query: string) {
   // Create Document object with essay
   const document = new Document({ text: essay, id_: path });
 
-  const index = await VectorStoreIndex.fromDocuments([document]);
+  // Split text and create embeddings. Store them in a VectorStoreIndex
+  const serviceContext = serviceContextFromDefaults({ llm, embedModel });
+
+  const index = await VectorStoreIndex.fromDocuments([document], {
+    serviceContext,
+  });
 
   // Query the index
   const queryEngine = index.asQueryEngine();
@@ -58,6 +60,10 @@ async function rag(query: string) {
   }
 
   // rag
-  const ragResponse = await rag("What did the author do in college?");
+  const ragResponse = await rag(
+    llm,
+    embedding,
+    "What did the author do in college?",
+  );
   console.log(ragResponse);
 })();

examples/mongo.ts

@@ -61,4 +61,4 @@ async function main() {
   }
 }
 
-void main();
+main();