Trieve

https://docs.trieve.ai
Admin
Trieve is an API for building search and RAG (Retrieval-Augmented Generation) experiences, allowing...
Tokens:673,538
Snippets:6,713
Trust Score:8.4
Update:1 year ago
Show doc for...
Tokens:
Raw
### Download embedding_models.yaml Configuration

Source: https://docs.trieve.ai/vector-inference/aws-installation

This command downloads the example `embedding_models.yaml` file, which is used to configure the embedding models for Trieve Vector Inference.

```bash
wget https://cdn.trieve.ai/embedding_models.yaml
```

--------------------------------

### Install Bun and Dependencies

Source: https://docs.trieve.ai/examples/job-board

Installs the Bun JavaScript runtime and project dependencies for the job board example. Bun is used as the runtime environment for this demo.

```bash
curl -fsSL https://bun.sh/install | bash
```

```shell
cd job-board-example
bun install
```

--------------------------------

### Download bootstrap-eks.sh Script

Source: https://docs.trieve.ai/vector-inference/aws-installation

This command downloads the `bootstrap-eks.sh` script, which is used to create an EKS cluster and install necessary plugins like the AWS Load Balancer Controller and NVIDIA Device Plugin.

```bash
wget cdn.trieve.ai/bootstrap-eks.sh
```

--------------------------------

### API Authentication Headers for Trieve

Source: https://docs.trieve.ai/getting-started/quickstart

This snippet shows the required HTTP headers for authenticating requests to the Trieve API. It includes the API key and the dataset ID, which are essential for interacting with your Trieve datasets.

```JSON
{
  "Authorization": "<api-key>",
  "TR-Dataset": "<dataset-id>"
}
```

--------------------------------

### Clone Trieve Examples Repository

Source: https://docs.trieve.ai/examples/ecommerce

Clones the Trieve examples repository and navigates into the Amazon Berkeley Objects dataset directory, which is used for building the e-commerce search example.

```bash
git clone https://github.com/devflowinc/examples
cd examples/amazon-berkeley-dataset
```

--------------------------------

### Install Trieve AI Helm Chart

Source: https://docs.trieve.ai/vector-inference/aws-installation

Installs or upgrades the Trieve AI Helm chart from the Marketplace ECR repository. It uses a specified values file for configuration. This command deploys the vector inference services.

```bash
helm upgrade -i vector-inference \
    oci://709825985650.dkr.ecr.us-east-1.amazonaws.com/trieve/trieve-embeddings \
    -f embedding_models.yaml
```

--------------------------------

### Clone Trieve Examples Repository

Source: https://docs.trieve.ai/examples/job-board

Clones the Trieve examples repository from GitHub to your local machine. This is the first step to setting up the job board demo.

```shell
git clone https://github.com/devflowinc/examples
```

--------------------------------

### Install Docker

Source: https://docs.trieve.ai/self-hosting/docker-compose

Installs Docker on the system using a curl script. This is a prerequisite for running Trieve with Docker Compose.

```bash
curl https://get.docker.com | sh

```

--------------------------------

### Example TVI Embeddings Endpoint Output

Source: https://docs.trieve.ai/vector-inference/rerank

This is an example output from the `kubectl get ing` command, showing the details of the 'vector-inference-embedding-bge-reranker-ingress' service, including its address.

```bash
NAME                                              CLASS   HOSTS   ADDRESS                                                                 PORTS   AGE
vector-inference-embedding-bge-reranker-ingress   alb     *       k8s-default-vectorin-b09efe8cf6-890425945.us-west-1.elb.amazonaws.com   80      77m
```

--------------------------------

### Execute bootstrap-eks.sh Script

Source: https://docs.trieve.ai/vector-inference/aws-installation

This command executes the downloaded `bootstrap-eks.sh` script using bash. The script automates the creation of an EKS cluster and configures required components.

```bash
bash bootstrap-eks.sh
```

--------------------------------

### Get Datasets from Organization Response Example

Source: https://docs.trieve.ai/api-reference/dataset/get-datasets-from-organization

Example JSON response for the 'Get Datasets from Organization' endpoint. It details the structure of dataset information and associated usage metrics.

```JSON
[
  {
    "dataset": {
      "created_at": "2021-01-01 00:00:00.000",
      "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
      "name": "Trieve",
      "organization_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
      "server_configuration": {},
      "updated_at": "2021-01-01 00:00:00.000"
    },
    "dataset_usage": {
      "chunk_count": 100,
      "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
      "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3"
    }
  }
]
```

--------------------------------

### Install Trieve AI Search Component with Tampermonkey

Source: https://docs.trieve.ai/site-search/content-sites

This script injects the Trieve AI search component's CSS and JavaScript into a webpage. It dynamically creates a div element to host the search interface and renders it using the provided module. The script includes basic error handling for module loading.

```javascript
(function () {
  "use strict";

  const link = document.createElement("link");
  link.rel = "stylesheet";
  link.href = "https://search-component.trieve.ai/dist/index.css";
  document.head.appendChild(link);

  import("https://search-component.trieve.ai/dist/vanilla/index.js")
    .then((module) => {
      const root = document.createElement("div");
      root.classList.add("trigger");
      document.body.appendChild(root);

      module.renderToDiv(root, {
        // <to be replaced with the props from the demo page>
      });
    })
    .catch((error) => {
      console.error("Failed to load search comonent:", error);
    });
})();

```

--------------------------------

### Start Embedding Servers (CPU)

Source: https://docs.trieve.ai/self-hosting/docker-compose

Starts the CPU-based embedding servers for Trieve using a specific Docker Compose file. This is for systems without GPU acceleration.

```bash
docker compose -f docker-compose-cpu-embeddings.yml up -d

```

--------------------------------

### File Response Example

Source: https://docs.trieve.ai/api-reference/file/get-file

Example JSON response for a file retrieval request, including creation and update timestamps, file name, ID, S3 URL, size, and optional metadata and links.

```json
{
  "created_at": "2021-01-01 00:00:00.000",
  "file_name": "file.txt",
  "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "link": "https://trieve.ai",
  "metadata": {
    "key": "value"
  },
  "s3_url": "https://trieve.ai",
  "size": 1000,
  "updated_at": "2021-01-01 00:00:00.000"
}
```

--------------------------------

### Get All Usage Plans Response (JSON)

Source: https://docs.trieve.ai/api-reference/stripe/get-all-usage-plans

Example JSON response for the 'Get All Usage Plans' endpoint, detailing the structure of plan information including creation date, period end, IDs, and organization details.

```json
[
  {
    "created_at": "2021-01-01 00:00:00.000",
    "current_period_end": "2021-01-01 00:00:00.000",
    "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "organization_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "plan_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "stripe_id": "sub_123",
    "updated_at": "2021-01-01 00:00:00.000"
  }
]
```

--------------------------------

### Trieve TypeScript SDK Usage

Source: https://docs.trieve.ai/getting-started/screenshots

This snippet demonstrates how to use the Trieve TypeScript SDK for interacting with the Trieve API. It covers common operations like initializing the client and performing searches. Ensure you have the Trieve SDK installed (`npm install @trieve/client`).

```TypeScript
import { TrieveClient } from "@trieve/client";

// Initialize the client with your API key and dataset ID
const client = new TrieveClient({
  apiKey: "YOUR_API_KEY",
  datasetId: "YOUR_DATASET_ID",
});

async function searchTrieve() {
  try {
    const results = await client.query({
      query: "What is Trieve?",
      search_query: "Trieve AI",
      model: "gpt-3.5-turbo",
      filters: {
        group_by: "tag",
        tags: ["documentation"]
      },
      limit: 5,
      offset: 0,
      return_metadata: true,
      return_link: true,
    });
    console.log(results);
  } catch (error) {
    console.error("Error querying Trieve:", error);
  }
}

searchTrieve();
```

--------------------------------

### Login via GET Request (cURL)

Source: https://docs.trieve.ai/api-reference/auth/login

This snippet demonstrates how to initiate the login process using a cURL GET request to the Trieve AI API. It includes the necessary URL and headers for authentication.

```curl
curl --request GET \
  --url https://api.trieve.ai/api/auth
```

--------------------------------

### Get Messages Response Example (JSON)

Source: https://docs.trieve.ai/api-reference/message/get-all-messages-for-a-given-topic

An example of the JSON response when retrieving messages for a topic. Each message object contains details like completion tokens, content, creation timestamp, dataset ID, and more.

```JSON
[
  {
    "completion_tokens": 300,
    "content": "Hello, world!",
    "created_at": "2021-01-01 00:00:00.000",
    "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "deleted": false,
    "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "prompt_tokens": 300,
    "role": "user",
    "sort_order": 1,
    "topic_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "updated_at": "2021-01-01 00:00:00.000"
  }
]
```

--------------------------------

### Start Embedding Servers (GPU)

Source: https://docs.trieve.ai/self-hosting/docker-compose

Starts the GPU-based embedding servers for Trieve using a specific Docker Compose file. This is for systems with GPU acceleration.

```bash
docker compose -f docker-compose-gpu-embeddings.yml up -d

```

--------------------------------

### Create Chunk Request Example

Source: https://docs.trieve.ai/guides/create-chunk

This example demonstrates how to format a request to create a chunk in Trieve. It includes essential parameters for content, grouping, filtering, and relevance boosting.

```TypeScript
import { TrieveClient } from "trieve"

const client = new TrieveClient(
  "YOUR_API_KEY",
  "YOUR_INSTANCE_URL"
)

const response = await client.createChunk({
  chunk_html: "This is the content of the chunk.",
  group_id: "your_group_id",
  tag_set: {
    "tag1": "value1",
    "tag2": "value2"
  },
  metadata: {
    "arbitrary_key": "arbitrary_value"
  },
  tracking_id: "your_tracking_id",
  weight: 10,
  semantic_boost: 1.5,
  fulltext_boost: 2.0
});
```

--------------------------------

### Response for Get Files and Group IDs

Source: https://docs.trieve.ai/api-reference/file/get-files-for-dataset

Example JSON response for the 'Get Files and Group IDs for Dataset' endpoint, showing a list of files with their metadata and group IDs, along with the total number of pages.

```json
{
  "file_and_group_ids": [
    {
      "file": {
        "created_at": "2021-01-01 00:00:00.000",
        "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "file_name": "file.txt",
        "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "link": "https://trieve.ai",
        "metadata": {
          "key": "value"
        },
        "size": 1000,
        "tag_set": "tag1,tag2",
        "time_stamp": "2021-01-01 00:00:00.000",
        "updated_at": "2021-01-01 00:00:00.000"
      },
      "group_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a"
    }
  ],
  "total_pages": 123
}
```

--------------------------------

### Get file and chunk creation queue

Source: https://docs.trieve.ai/api-reference/public/get-apipublic_page

No description

--------------------------------

### Start Trieve Services

Source: https://docs.trieve.ai/self-hosting/docker-compose

Starts all Trieve services using Docker Compose in detached mode. This command brings up the main Trieve application containers.

```bash
docker compose up -d

```

--------------------------------

### Install Dependencies for OpenAI SDK Integration

Source: https://docs.trieve.ai/vector-inference/openai

Installs the necessary Python libraries: `openai` for interacting with the OpenAI API, `requests` for making HTTP requests, and `python-dotenv` for managing environment variables.

```bash
pip install openai requests python-dotenv

```

--------------------------------

### Set AWS Region Environment Variable

Source: https://docs.trieve.ai/vector-inference/aws-installation

This bash script sets the AWS Region environment variable, specifying the AWS region where your resources will be deployed. 'us-east-2' is used as an example.

```bash
export AWS_REGION=us-east-2
```

--------------------------------

### Get Top Datasets Response Example

Source: https://docs.trieve.ai/api-reference/analytics/get-top-datasets

Example JSON response for the 'Get Top Datasets' endpoint, showing dataset ID, tracking ID, and total queries.

```JSON
[
  {
    "dataset_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
    "dataset_tracking_id": "<string>",
    "total_queries": 123
  }
]
```

--------------------------------

### Get Model Endpoints using kubectl

Source: https://docs.trieve.ai/vector-inference/aws-installation

Retrieves the Ingress resources from a Kubernetes cluster to find the addresses of the deployed Trieve AI models. The output lists the Ingress names, classes, hosts, and their corresponding addresses.

```bash
kubectl get ingress
```

--------------------------------

### Test Model Inference with cURL

Source: https://docs.trieve.ai/vector-inference/aws-installation

Makes a POST request to a Trieve AI model endpoint to test inference. It sends a JSON payload with sample input and measures the inference time. Replace the placeholder endpoint with the actual address obtained from `kubectl get ingress`.

```bash
# Replace the endpoint with the one you got from the previous step
export ENDPOINT=k8s-default-vectorin-18b7ade77a-2040086997.us-east-2.elb.amazonaws.com

curl -X POST \
     -H "Content-Type: application/json"
     -d '{"inputs": "test input"}' \
     --url "http://$ENDPOINT/embed" \
     -w "\n\nInfernce Took%{time_total} seconds!\n"
```

--------------------------------

### Get Recommended Groups with Trieve

Source: https://docs.trieve.ai/guides/recommending-with-trieve

This snippet shows how to get recommended groups and their top chunks based on similarity to provided examples using Trieve's API. It uses the 'get recommended groups' route and allows for parameter customization.

```TypeScript
import { TrieveClient } from "@trieve/trieve-client";

const client = new TrieveClient("YOUR_API_KEY");

async function getGroupRecommendations() {
  const recommendations = await client.recommend
    .recommendGroups({
      query: {
        positive_examples: ["chunk_id_1", "chunk_id_2"],
        negative_examples: ["chunk_id_3"],
        recommendation_type: "semantic",
        strategy: "best_score",
        slim_chunks: false,
        filters: {
          group_size: 3,
        },
      },
    });
  console.log(recommendations);
}
```

--------------------------------

### Clone Trieve Repository

Source: https://docs.trieve.ai/self-hosting/aws

Clones the Trieve repository from GitHub and navigates into the cloned directory. This is the first step in setting up Trieve.

```bash
git clone https://github.com/devflowinc/trieve.git
cd trieve
```

--------------------------------

### Login and Create Dataset with Trieve CLI

Source: https://docs.trieve.ai/self-hosting/docker-compose

This snippet demonstrates the initial steps for testing a Trieve AI project using the Trieve CLI. It includes logging into the Trieve service, ensuring the API URL is set to a local instance, and creating an example dataset.

```bash
trieve login # Make sure to set the api url to http://localhost:8090
trieve dataset example
```

--------------------------------

### Set Recommended Default Environment Variables

Source: https://docs.trieve.ai/vector-inference/aws-installation

This bash script sets a comprehensive set of recommended default environment variables for deploying Trieve Vector Inference on AWS, including account ID, region, cluster name, instance types, and disabling AWS CLI pagination.

```bash
export AWS_ACCOUNT_ID="$(aws sts get-caller-identity --query "Account" --output text)"
export AWS_REGION=us-east-2
export CLUSTER_NAME=trieve-gpu
export CPU_INSTANCE_TYPE=t3.small
export GPU_INSTANCE_TYPE=g4dn.xlarge
export GPU_COUNT=1
export AWS_PAGER=""
```

--------------------------------

### Clone Trieve Repository

Source: https://docs.trieve.ai/self-hosting/azure

Clones the Trieve repository from GitHub and navigates into the project directory. This is the first step in setting up Trieve.

```Bash
git clone https://github.com/devflowinc/trieve.git
cd trieve
```

--------------------------------

### Ab Test - Trieve API

Source: https://docs.trieve.ai/api-reference/dataset/create-pagefind-index-for-dataset

Sets up or manages an A/B test.

```HTTP
POST /experiments/ab_test
```

--------------------------------

### Create Experiment - Trieve AI

Source: https://docs.trieve.ai/api-reference/chunk/get-recommended-chunks

Creates a new experiment for testing different configurations or features.

```HTTP
POST /experiment
```

--------------------------------

### Get Groups for Dataset Response

Source: https://docs.trieve.ai/api-reference/chunk-group/get-groups-for-dataset

Example JSON response for the 'Get Groups for Dataset' endpoint, showing a list of groups with their details, next cursor, and total pages.

```JSON
{
  "groups": [
    {
      "created_at": "2021-01-01 00:00:00.000",
      "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
      "description": "A group of chunks",
      "file_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
      "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
      "name": "Trieve",
      "tracking_id": "3",
      "updated_at": "2021-01-01 00:00:00.000"
    }
  ],
  "next_cursor": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
  "total_pages": 1
}
```

--------------------------------

### Get Organization Users Response (JSON)

Source: https://docs.trieve.ai/api-reference/organization/get-organization-users

Example JSON response for the 'Get Organization Users' endpoint, detailing user information including email, ID, and organization details.

```json
[
  {
    "email": "developers@trieve.ai",
    "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "name": "Trieve",
    "orgs": [
      {
        "created_at": "2021-01-01 00:00:00.000",
        "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "name": "Trieve",
        "registerable": true,
        "updated_at": "2021-01-01 00:00:00.000"
      }
    ],
    "user_orgs": [
      {
        "created_at": "2021-01-01 00:00:00.000",
        "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "org_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "role": 0,
        "updated_at": "2021-01-01 00:00:00.000",
        "user_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3"
      }
    ]
  }
]
```

--------------------------------

### Get File Response Structure

Source: https://docs.trieve.ai/api-reference/file/get-file-with-signed-url

Example JSON response for the 'Get File with Signed URL' endpoint, containing file metadata, creation/update timestamps, and the S3 download URL.

```json
{
  "created_at": "2021-01-01 00:00:00.000",
  "file_name": "file.txt",
  "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "link": "https://trieve.ai",
  "metadata": {
    "key": "value"
  },
  "s3_url": "https://trieve.ai",
  "size": 1000,
  "updated_at": "2021-01-01 00:00:00.000"
}
```

--------------------------------

### Get All Usage Plans

Source: https://docs.trieve.ai/api-reference/file/get-file-with-signed-url

Retrieves a list of all available usage plans, which define different tiers of service and pricing.

```Python
import requests

API_URL = "https://api.trieve.ai/v1"

def get_all_usage_plans(api_key: str):
    headers = {
        "Authorization": f"ApiKey {api_key}"
    }
    response = requests.get(f"{API_URL}/stripe/usage_plans", headers=
```

--------------------------------

### Get Chunk By Id Response (JSON)

Source: https://docs.trieve.ai/api-reference/chunk/get-chunk-by-id

Example JSON response for the 'Get Chunk By Id' endpoint, detailing chunk properties like HTML content, timestamps, IDs, and metadata.

```JSON
{
  "chunk_html": "<p>Hello, world!</p>",
  "created_at": "2021-01-01 00:00:00.000",
  "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "link": "https://trieve.ai",
  "metadata": {
    "key": "value"
  },
  "tag_set": "[tag1,tag2]",
  "time_stamp": "2021-01-01 00:00:00.000",
  "tracking_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "updated_at": "2021-01-01 00:00:00.000",
  "weight": 0.5
}
```

--------------------------------

### Example Dataset Creation Response

Source: https://docs.trieve.ai/api-reference/dataset/batch-create-datasets

This JSON snippet represents a successful response from the Trieve AI API after batch creating datasets. It details the created dataset's properties, including its ID, name, creation timestamp, and the server configuration used.

```json
[ \
  { \
    "created_at": "2021-01-01 00:00:00.000", \
    "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3", \
    "name": "Trieve", \
    "organization_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3", \
    "server_configuration": { \
      "AIMON_RERANKER_TASK_DEFINITION": "Your task is to grade the relevance of context document(s) against the specified user query.", \
      "BM25_AVG_LEN": 256, \
      "BM25_B": 0.75, \
      "BM25_ENABLED": true, \
      "BM25_K": 0.75, \
      "DISTANCE_METRIC": "cosine", \
      "EMBEDDING_BASE_URL": "https://embedding.trieve.ai", \
      "EMBEDDING_MODEL_NAME": "jina-base-en", \
      "EMBEDDING_QUERY_PREFIX": "", \
      "EMBEDDING_SIZE": 768, \
      "FREQUENCY_PENALTY": 0, \
      "FULLTEXT_ENABLED": true, \
      "INDEXED_ONLY": false, \
      "LLM_BASE_URL": "https://api.openai.com/v1", \
      "LLM_DEFAULT_MODEL": "gpt-4o", \
      "LOCKED": false, \
      "MAX_LIMIT": 10000, \
      "MESSAGE_TO_QUERY_PROMPT": "Write a 1-2 sentence semantic search query along the lines of a hypothetical response to: \n\n", \
      "N_RETRIEVALS_TO_INCLUDE": 8, \
      "PRESENCE_PENALTY": 0, \
      "QDRANT_ONLY": false, \
      "RAG_PROMPT": "Use the following retrieved documents to respond briefly and accurately:", \
      "SEMANTIC_ENABLED": true, \
      "STOP_TOKENS": [ \
        "\n\n", \
        "\n" \
      ], \
      "SYSTEM_PROMPT": "You are a helpful assistant", \
      "TEMPERATURE": 0.5, \
      "USE_MESSAGE_TO_QUERY_PROMPT": false \
    }, \
    "tracking_id": "foobar-dataset", \
    "updated_at": "2021-01-01 00:00:00.000" \
  } \
]
```

--------------------------------

### Ab Test - Trieve API

Source: https://docs.trieve.ai/api-reference/dataset/get-pagefind-index-url-for-dataset

Configures and runs an A/B test.

```HTTP
POST /experiment/ab_test
```

--------------------------------

### Create Experiment - Trieve API

Source: https://docs.trieve.ai/api-reference/dataset/get-pagefind-index-url-for-dataset

Creates a new experiment for A/B testing or other evaluations.

```HTTP
POST /experiment
```

--------------------------------

### Get Recommended Chunks with Trieve

Source: https://docs.trieve.ai/guides/recommending-with-trieve

This snippet demonstrates how to retrieve recommended chunks similar to provided examples using Trieve's API. It utilizes the 'get recommended chunks' route and supports various parameters for fine-tuning recommendations.

```TypeScript
import { TrieveClient } from "@trieve/trieve-client";

const client = new TrieveClient("YOUR_API_KEY");

async function getRecommendations() {
  const recommendations = await client.recommend
    .recommendChunks({
      query: {
        positive_examples: ["chunk_id_1", "chunk_id_2"],
        negative_examples: ["chunk_id_3"],
        recommendation_type: "semantic",
        strategy: "average_vector",
        slim_chunks: true,
        filters: {
          group_size: 5,
        },
      },
    });
  console.log(recommendations);
}
```

--------------------------------

### Create .env File

Source: https://docs.trieve.ai/self-hosting/docker-compose

Copies the example environment file to create a new .env file, which will be used to configure Trieve's settings.

```bash
cp .env.example .env

```

--------------------------------

### Basic Trieve Site Search Integration (JavaScript)

Source: https://docs.trieve.ai/site-search/youtube-channels

This script injects Trieve AI's site search CSS and JavaScript into a webpage. It dynamically creates a root element and renders the search component using provided configuration. Ensure you replace the placeholder props with your actual dataset information.

```javascript
(function() {
    'use strict';

    const link = document.createElement("link");
    link.rel = "stylesheet";
    link.href = "https://search-component.trieve.ai/dist/index.css";
    document.head.appendChild(link);

    import('https://search-component.trieve.ai/dist/vanilla/index.js')
        .then((module) => {
            const root = document.createElement('div');
            root.classList.add('trigger');
            document.body.appendChild(root);

            module.renderToDiv(root, {
                // <to be replaced with the props from the demo page>
            });
        })
        .catch(error => {
            console.error("Failed to load search comonent:", error);
        });
})();

```

--------------------------------

### Create Pagefind Index for Dataset

Source: https://docs.trieve.ai/api-reference/dataset/get-file-and-chunk-creation-queue-lengths

Creates a Pagefind index for a dataset, enabling client-side search.

```TypeScript
PUT
Create Pagefind Index for Dataset
```

--------------------------------

### Create Dataset API Request

Source: https://docs.trieve.ai/guides/multi-tenency

Demonstrates how to create a new dataset for a tenant using the Trieve API. It includes essential parameters like dataset name, organization ID, and server configurations for embedding models, RAG prompts, and more.

```Bash
curl --request POST \
  --url https://api.trieve.ai/api/dataset \
  --header 'Authorization: <api-key>' \
  --header 'Content-Type: application/json' \
  --header 'TR-Organization: <tr-organization>' \
  --data '{ \
    "dataset_name": "New Dataset", // Replace with a name for your dataset \
    "organization_id": "********-****-****-****-************", \

    // Update with the desired configurations for your dataset \
    "server_configuration": { \
      "BM25_ENABLED": true, \
      "DISTANCE_METRIC": "cosine", \
      "EMBEDDING_MODEL_NAME": "text-embedding-3-small", \
      "LLM_DEFAULT_MODEL": "gpt-3.5-turbo-1106", \
      "RAG_PROMPT": "Use the following retrieved documents...", \
      "SEMANTIC_ENABLED": true, \
      "SYSTEM_PROMPT": "You are a helpful assistant", \
    }
}'

```

--------------------------------

### Get Group Response Example (JSON)

Source: https://docs.trieve.ai/api-reference/chunk-group/get-group

An example of the JSON response when successfully retrieving a group from the Trieve AI API. It includes details such as creation date, dataset ID, description, file ID, group ID, name, and update timestamp.

```JSON
{
  "created_at": "2021-01-01 00:00:00.000",
  "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "description": "A group of chunks",
  "file_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "name": "Trieve",
  "tracking_id": "3",
  "updated_at": "2021-01-01 00:00:00.000"
}
```

--------------------------------

### Get All Tags

Source: https://docs.trieve.ai/api-reference/chunk-group/create-or-upsert-group-or-groups

Retrieves all available tags.

```HTTP
POST /tags
```

--------------------------------

### Get Files and Group IDs for Dataset Response

Source: https://docs.trieve.ai/api-reference/file/get-files-and-group-ids-for-dataset

Example JSON response for the 'Get Files and Group IDs for Dataset' endpoint, detailing file information (creation date, dataset ID, file name, size, etc.) and group IDs.

```json
{
  "file_and_group_ids": [
    {
      "file": {
        "created_at": "2021-01-01 00:00:00.000",
        "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "file_name": "file.txt",
        "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "link": "https://trieve.ai",
        "metadata": {
          "key": "value"
        },
        "size": 1000,
        "tag_set": "tag1,tag2",
        "time_stamp": "2021-01-01 00:00:00.000",
        "updated_at": "2021-01-01 00:00:00.000"
      },
      "group_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a"
    }
  ],
  "total_pages": 123
}
```

--------------------------------

### Get All Invoices Response (JSON)

Source: https://docs.trieve.ai/api-reference/stripe/get-all-invoices

Example JSON response structure for the 'Get All Invoices' API endpoint. It details the fields returned for each invoice, including creation date, hosted URL, ID, organization ID, status, Stripe ID, and total amount.

```JSON
[
  {
    "created_at": "2023-11-07T05:31:56Z",
    "hosted_invoice_url": "<string>",
    "id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
    "org_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
    "status": "<string>",
    "stripe_id": "<string>",
    "total": 123
  }
]
```

--------------------------------

### Create Pagefind Index for Dataset

Source: https://docs.trieve.ai/api-reference/chunk/split-html-content-into-chunks

Creates a Pagefind index for a dataset, enabling efficient client-side search.

```API Reference
PUT /dataset/pagefind
```

--------------------------------

### Create Chunk API Request Example

Source: https://docs.trieve.ai/guides/create-chunk

This example shows a POST request to the Trieve API for creating a chunk. It includes the required `TR-Dataset` and `Authorization` headers, and a sample request body with various fields like `chunk_html`, `link`, `tag_set`, and nested `metadata`.

```http
POST /api/chunk
Headers:
{
    "TR-Dataset": "<Your Dataset ID>",
    "Authorization": "<Your API Key>"
}
Body:
{
    "chunk_html": "EcoFusion Technologies provides innovative, eco-friendly technology solutions. We specialize in renewable ...",
    "link": "https://example.com",
    "tracking_id": "134",
    "image_urls": ["https://example.com"],
    "tag_set": ["324", "product", "sale"],
    "metadata": {
      "phone_number": "000-000-0000",
      "price_range": "$21-$25",
      "reviews_count": "15",
      "address": {
        "city": "Austin",
        "country": "United States",
        "state": "Texas",
        "zip_code": "78781"
      }
    }
}

```

--------------------------------

### Get Chunks By Tracking Ids Response (JSON)

Source: https://docs.trieve.ai/api-reference/chunk/get-chunks-by-tracking-ids

Example JSON response for the 'Get Chunks By Tracking Ids' endpoint. It returns an array of chunk objects, each containing properties like chunk_html, created_at, dataset_id, id, link, metadata, tag_set, time_stamp, tracking_id, updated_at, and weight.

```json
[
  {
    "chunk_html": "<p>Hello, world!</p>",
    "created_at": "2021-01-01 00:00:00.000",
    "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "link": "https://trieve.ai",
    "metadata": {
      "key": "value"
    },
    "tag_set": "[tag1,tag2]",
    "time_stamp": "2021-01-01 00:00:00.000",
    "tracking_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
    "updated_at": "2021-01-01 00:00:00.000",
    "weight": 0.5
  }
]
```

--------------------------------

### Recommendation Analytics Response

Source: https://docs.trieve.ai/api-reference/analytics/get-recommendation-analytics

Example JSON response for the Get Recommendation Analytics endpoint, detailing query information including IDs, timestamps, and recommendation details.

```json
{
  "queries": [
    {
      "created_at": "<string>",
      "dataset_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
      "id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
      "metadata": "<any>",
      "negative_ids": [
        "3c90c3cc-0d44-4b50-8888-8dd25736052a"
      ],
      "negative_tracking_ids": [
        "<string>"
      ],
      "positive_ids": [
        "3c90c3cc-0d44-4b50-8888-8dd25736052a"
      ],
      "positive_tracking_ids": [
        "<string>"
      ],
      "recommendation_type": "Chunk",
      "request_params": "<any>",
      "results": [
        "<any>"
      ],
      "top_score": 123,
      "user_id": "<string>"
    }
  ]
}
```

--------------------------------

### Disable AWS CLI Pagination

Source: https://docs.trieve.ai/vector-inference/aws-installation

This optional bash script disables pagination for AWS CLI output, which can simplify scripting by ensuring all output is displayed at once.

```bash
export AWS_PAGER=""
```

--------------------------------

### Run Trieve Ingest Script

Source: https://docs.trieve.ai/examples/ecommerce

Executes the Trieve ingest script using Bun. This script processes the downloaded data and uploads it to the specified Trieve dataset.

```bash
bun index.ts
```

--------------------------------

### Create Experiment - Trieve API

Source: https://docs.trieve.ai/api-reference/dataset/create-pagefind-index-for-dataset

Initiates a new experiment for testing features or algorithms.

```HTTP
POST /experiments
```

--------------------------------

### Get Groups for Chunks Response (JSON)

Source: https://docs.trieve.ai/api-reference/chunk-group/get-groups-for-chunks

Example JSON response when successfully retrieving groups for chunks. It includes chunk UUID and associated group details like name, description, and timestamps.

```JSON
[
  {
    "chunk_uuid": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
    "slim_groups": [
      {
        "created_at": "2021-01-01 00:00:00.000",
        "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "description": "A group of chunks",
        "file_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
        "name": "Trieve",
        "tracking_id": "3",
        "updated_at": "2021-01-01 00:00:00.000"
      }
    ]
  }
]
```

--------------------------------

### Install and Upgrade Trieve AI Helm Chart

Source: https://docs.trieve.ai/self-hosting/azure

This bash script demonstrates how to add the Trieve AI Helm repository, update it, and then upgrade or install the Trieve AI local deployment using a specified `values.yaml` file. It's recommended to run the upgrade command twice to ensure CRDs and ConfigMaps are properly updated.

```bash
helm repo add trieve https://devflowinc.github.io/trieve-helm/
helm repo update
helm upgrade -i -f values.yaml trieve-local trieve/trieve
helm upgrade -i -f values.yaml trieve-local trieve/trieve

```

--------------------------------

### Get All Tags

Source: https://docs.trieve.ai/api-reference/public/get-apipublic_page

Retrieves all available tags within the system.

```API
POST /get_all_tags
```

--------------------------------

### Cluster Analytics Response

Source: https://docs.trieve.ai/api-reference/analytics/get-cluster-analytics

Example JSON response for the Get Cluster Analytics endpoint, containing an array of cluster objects, each with properties like avg_score, created_at, dataset_id, density, id, and topic.

```json
{
  "clusters": [
    {
      "avg_score": 123,
      "created_at": "<string>",
      "dataset_id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
      "density": 123,
      "id": "3c90c3cc-0d44-4b50-8888-8dd25736052a",
      "topic": "<string>"
    }
  ]
}
```

--------------------------------

### Create Experiment

Source: https://docs.trieve.ai/api-reference/dataset/get-file-and-chunk-creation-queue-lengths

Creates a new experiment for testing purposes.

```TypeScript
POST
Create Experiment
```

--------------------------------

### Get Group by Tracking ID Response (JSON)

Source: https://docs.trieve.ai/api-reference/chunk-group/get-group-by-tracking-id

Example JSON response for a successful retrieval of a chunk group, including creation and update timestamps, dataset ID, description, file ID, and name.

```json
{
  "created_at": "2021-01-01 00:00:00.000",
  "dataset_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "description": "A group of chunks",
  "file_id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "id": "e3e3e3e3-e3e3-e3e3-e3e3-e3e3e3e3e3e3",
  "name": "Trieve",
  "tracking_id": "3",
  "updated_at": "2021-01-01 00:00:00.000"
}
```

--------------------------------

### Create Dataset

Source: https://docs.trieve.ai/api-reference/dataset/get-file-and-chunk-creation-queue-lengths

Creates a new dataset.

```TypeScript
POST
Create Dataset
```

--------------------------------

### Set Kubernetes Cluster Name Environment Variable

Source: https://docs.trieve.ai/vector-inference/aws-installation

This bash script defines the name for your Kubernetes cluster, which will be used for deploying Trieve Vector Inference. 'trieve-gpu' is the recommended name.

```bash
export CLUSTER_NAME=trieve-gpu
```

--------------------------------

### Get Ranks API Request (curl)

Source: https://docs.trieve.ai/vector-inference/reranker

This example demonstrates how to call the Trieve API's reranking functionality using curl. It sends a POST request with a JSON payload containing the query and texts to be ranked.

```curl
curl -X POST \
  'YOUR_API_ENDPOINT/rerank' \
  -H 'Content-Type: application/json' \
  -d '{ \
    "query": "What are some good electric cars", \
    "texts": [ \
        "Here’s the information about the Mercedes CLR GTR: The Mercedes CLR GTR is a remarkable racing car ...", \
        "The Tesla Cybertruck is an all-electric, battery-powered light-duty truck unveiled by Tesla, Inc. ..." \
    ], \
    "raw_scores": false, \
    "return_text": false, \
    "truncate": false, \
    "truncation_direction": "right" \
  }'
```
Tokens:
Raw