mirror of
https://github.com/run-llama/LlamaIndexTS.git
synced 2026-07-01 22:14:03 -04:00
Add OVHcloud AI Endpoints provider (#2238)
Co-authored-by: Marcus Schiesser <mail@marcusschiesser.de>
This commit is contained in:
Elias
@@ -0,0 +1,6 @@
|
||||
---
|
||||
"@llamaindex/ovhcloud": major
|
||||
"@llamaindex/examples": minor
|
||||
---
|
||||
|
||||
Add OVHcloud AI Endpoints provider
|
||||
@@ -0,0 +1,123 @@
|
||||
---
|
||||
title: OVHcloud AI Endpoints
|
||||
---
|
||||
|
||||
OVHcloud AI Endpoints provide OpenAI-compatible embedding models. The service can be used for free with rate limits, or with an API key for higher limits.
|
||||
|
||||
OVHcloud is a global player and the leading European cloud provider operating over 450,000 servers within 40 data centers across 4 continents to reach 1.6 million customers in over 140 countries. Our product AI Endpoints offers access to various models with sovereignty, data privacy and GDPR compliance.
|
||||
|
||||
You can find the full list of models in the [OVHcloud AI Endpoints catalog](https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/).
|
||||
|
||||
## Installation
|
||||
|
||||
```package-install
|
||||
npm i llamaindex @llamaindex/ovhcloud
|
||||
```
|
||||
|
||||
## Authentication
|
||||
|
||||
OVHcloud AI Endpoints can be used in two ways:
|
||||
|
||||
1. **Free tier (with rate limits)**: No API key required. You can omit the `apiKey` parameter or set it to an empty string.
|
||||
2. **With API key**: For higher rate limits, generate an API key from the [OVHcloud Manager](https://ovh.com/manager) → Public Cloud → AI & Machine Learning → AI Endpoints → API keys.
|
||||
|
||||
## Basic Usage
|
||||
|
||||
```ts
|
||||
import { Document, Settings, VectorStoreIndex } from "llamaindex";
|
||||
import { OVHcloudEmbedding } from "@llamaindex/ovhcloud";
|
||||
|
||||
// Update Embed Model (using free tier)
|
||||
Settings.embedModel = new OVHcloudEmbedding();
|
||||
|
||||
// Or with API key from environment variable
|
||||
import { config } from "dotenv";
|
||||
config();
|
||||
Settings.embedModel = new OVHcloudEmbedding({
|
||||
apiKey: process.env.OVHCLOUD_API_KEY || "",
|
||||
});
|
||||
|
||||
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,
|
||||
});
|
||||
```
|
||||
|
||||
By default, `OVHcloudEmbedding` uses the `BGE-M3` model. You can change the model by passing the model parameter to the constructor:
|
||||
|
||||
```ts
|
||||
import { OVHcloudEmbedding } from "@llamaindex/ovhcloud";
|
||||
|
||||
const model = "text-embedding-3-small";
|
||||
Settings.embedModel = new OVHcloudEmbedding({
|
||||
model,
|
||||
});
|
||||
```
|
||||
|
||||
You can also set the `maxRetries` and `timeout` parameters when initializing `OVHcloudEmbedding` for better control over the request behavior:
|
||||
|
||||
```ts
|
||||
import { Settings } from "llamaindex";
|
||||
import { OVHcloudEmbedding } from "@llamaindex/ovhcloud";
|
||||
|
||||
const model = "text-embedding-3-small";
|
||||
const maxRetries = 5;
|
||||
const timeout = 5000; // 5 seconds
|
||||
|
||||
Settings.embedModel = new OVHcloudEmbedding({
|
||||
model,
|
||||
maxRetries,
|
||||
timeout,
|
||||
});
|
||||
```
|
||||
|
||||
## Standalone Usage
|
||||
|
||||
```ts
|
||||
import { OVHcloudEmbedding } from "@llamaindex/ovhcloud";
|
||||
import { config } from "dotenv";
|
||||
// For standalone usage, you can optionally configure OVHCLOUD_API_KEY in .env file
|
||||
config();
|
||||
|
||||
const main = async () => {
|
||||
const model = "BGE-M3";
|
||||
// Using without API key (free tier)
|
||||
const embeddings = new OVHcloudEmbedding({ model });
|
||||
const text = "What is the meaning of life?";
|
||||
const response = await embeddings.embed([text]);
|
||||
console.log(response);
|
||||
};
|
||||
|
||||
main();
|
||||
```
|
||||
|
||||
## Base URL
|
||||
|
||||
The default base URL is `https://oai.endpoints.kepler.ai.cloud.ovh.net/v1`. You can override it if needed:
|
||||
|
||||
```ts
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
model: "BGE-M3",
|
||||
additionalSessionOptions: {
|
||||
baseURL: "https://custom.endpoint.com/v1",
|
||||
},
|
||||
});
|
||||
```
|
||||
|
||||
## Resources
|
||||
|
||||
- [OVHcloud AI Endpoints Catalog](https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/)
|
||||
- [OVHcloud Manager](https://ovh.com/manager)
|
||||
- [OVHcloud AI Endpoints Documentation](https://www.ovhcloud.com/en/public-cloud/ai-endpoints/)
|
||||
|
||||
## API Reference
|
||||
|
||||
- [OVHcloudEmbedding](/typescript/framework-api-reference/classes/ovhcloudembedding/)
|
||||
|
||||
@@ -0,0 +1,164 @@
|
||||
---
|
||||
title: OVHcloud AI Endpoints
|
||||
---
|
||||
|
||||
OVHcloud AI Endpoints provide serverless access to a variety of pre-trained AI models. The service is OpenAI-compatible and can be used for free with rate limits, or with an API key for higher limits.
|
||||
|
||||
OVHcloud is a global player and the leading European cloud provider operating over 450,000 servers within 40 data centers across 4 continents to reach 1.6 million customers in over 140 countries. Our product AI Endpoints offers access to various models with sovereignty, data privacy and GDPR compliance.
|
||||
|
||||
You can find the full list of models in the [OVHcloud AI Endpoints catalog](https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/).
|
||||
|
||||
## Installation
|
||||
|
||||
```package-install
|
||||
npm i llamaindex @llamaindex/ovhcloud
|
||||
```
|
||||
|
||||
## Authentication
|
||||
|
||||
OVHcloud AI Endpoints can be used in two ways:
|
||||
|
||||
1. **Free tier (with rate limits)**: No API key required. You can omit the `apiKey` parameter or set it to an empty string.
|
||||
2. **With API key**: For higher rate limits, generate an API key from the [OVHcloud Manager](https://ovh.com/manager) → Public Cloud → AI & Machine Learning → AI Endpoints → API keys.
|
||||
|
||||
## Basic Usage
|
||||
|
||||
```ts
|
||||
import { OVHcloudLLM } from "@llamaindex/ovhcloud";
|
||||
import { Settings } from "llamaindex";
|
||||
|
||||
// Using without API key (free tier with rate limits)
|
||||
Settings.llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
});
|
||||
|
||||
// Or with API key from environment variable
|
||||
import { config } from "dotenv";
|
||||
config();
|
||||
Settings.llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
apiKey: process.env.OVHCLOUD_API_KEY || "",
|
||||
});
|
||||
|
||||
// Or with explicit API key
|
||||
Settings.llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
apiKey: "YOUR_API_KEY",
|
||||
});
|
||||
```
|
||||
|
||||
You can set the API key via environment variable:
|
||||
|
||||
```bash
|
||||
export OVHCLOUD_API_KEY="<YOUR_API_KEY>"
|
||||
```
|
||||
|
||||
## 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
|
||||
import { Document, VectorStoreIndex } from "llamaindex";
|
||||
|
||||
const document = new Document({ text: essay, id_: "essay" });
|
||||
|
||||
const index = await VectorStoreIndex.fromDocuments([document]);
|
||||
```
|
||||
|
||||
## Query
|
||||
|
||||
```ts
|
||||
const queryEngine = index.asQueryEngine();
|
||||
|
||||
const query = "What is the meaning of life?";
|
||||
|
||||
const results = await queryEngine.query({
|
||||
query,
|
||||
});
|
||||
```
|
||||
|
||||
## Full Example
|
||||
|
||||
```ts
|
||||
import { OVHcloudLLM } from "@llamaindex/ovhcloud";
|
||||
import { Document, VectorStoreIndex, Settings } from "llamaindex";
|
||||
|
||||
// Use custom LLM
|
||||
const model = "gpt-oss-120b";
|
||||
Settings.llm = new OVHcloudLLM({ model, temperature: 0 });
|
||||
|
||||
async function main() {
|
||||
const document = new Document({ text: essay, id_: "essay" });
|
||||
|
||||
// Load and index documents
|
||||
const index = await VectorStoreIndex.fromDocuments([document]);
|
||||
|
||||
// get retriever
|
||||
const retriever = index.asRetriever();
|
||||
|
||||
// Create a query engine
|
||||
const queryEngine = index.asQueryEngine({
|
||||
retriever,
|
||||
});
|
||||
|
||||
const query = "What is the meaning of life?";
|
||||
|
||||
// Query
|
||||
const response = await queryEngine.query({
|
||||
query,
|
||||
});
|
||||
|
||||
// Log the response
|
||||
console.log(response.response);
|
||||
}
|
||||
```
|
||||
|
||||
## Streaming
|
||||
|
||||
OVHcloud AI Endpoints supports streaming responses:
|
||||
|
||||
```ts
|
||||
import { OVHcloudLLM } from "@llamaindex/ovhcloud";
|
||||
|
||||
const llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
});
|
||||
|
||||
const generator = await llm.chat({
|
||||
messages: [
|
||||
{
|
||||
role: "user",
|
||||
content: "Tell me about OVHcloud AI Endpoints",
|
||||
},
|
||||
],
|
||||
stream: true,
|
||||
});
|
||||
|
||||
for await (const message of generator) {
|
||||
process.stdout.write(message.delta);
|
||||
}
|
||||
```
|
||||
|
||||
## Base URL
|
||||
|
||||
The default base URL is `https://oai.endpoints.kepler.ai.cloud.ovh.net/v1`. You can override it if needed:
|
||||
|
||||
```ts
|
||||
const llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
additionalSessionOptions: {
|
||||
baseURL: "https://custom.endpoint.com/v1",
|
||||
},
|
||||
});
|
||||
```
|
||||
|
||||
## Resources
|
||||
|
||||
- [OVHcloud AI Endpoints Catalog](https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/)
|
||||
- [OVHcloud Manager](https://ovh.com/manager)
|
||||
- [OVHcloud AI Endpoints Documentation](https://www.ovhcloud.com/en/public-cloud/ai-endpoints/)
|
||||
|
||||
## API Reference
|
||||
|
||||
- [OVHcloudLLM](/typescript/framework-api-reference/classes/ovhcloudllm/)
|
||||
|
||||
@@ -0,0 +1,50 @@
|
||||
import { OVHcloudEmbedding, OVHcloudLLM } from "@llamaindex/ovhcloud";
|
||||
|
||||
// OVHcloud AI Endpoints can be used for free with rate limits without an API key
|
||||
// To use with an API key, set OVHCLOUD_API_KEY environment variable
|
||||
// or pass it directly in the constructor
|
||||
// To generate an API key, go to https://ovh.com/manager > Public Cloud > AI & Machine Learning > AI Endpoints > API keys
|
||||
// Visit our catalog for the list of all available models: https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/
|
||||
|
||||
// Example 1: Using without API key (free tier with rate limits)
|
||||
const ovhcloudFree = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
// apiKey is optional - can be omitted or set to empty string for free tier
|
||||
});
|
||||
|
||||
// Example 2: Using with API key
|
||||
const ovhcloud = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
apiKey: process.env.OVHCLOUD_API_KEY || "",
|
||||
});
|
||||
|
||||
(async () => {
|
||||
console.log("Chatting with OVHcloud AI Endpoints...");
|
||||
const generator = await ovhcloud.chat({
|
||||
messages: [
|
||||
{
|
||||
role: "system",
|
||||
content: "You are a helpful AI assistant.",
|
||||
},
|
||||
{
|
||||
role: "user",
|
||||
content: "Tell me about OVHcloud AI Endpoints",
|
||||
},
|
||||
],
|
||||
stream: true,
|
||||
});
|
||||
|
||||
for await (const message of generator) {
|
||||
process.stdout.write(message.delta);
|
||||
}
|
||||
console.log("\n");
|
||||
|
||||
// Example with embeddings
|
||||
console.log("Getting embeddings...");
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
model: "BGE-M3",
|
||||
});
|
||||
const vector = await embedding.getTextEmbedding("Hello world!");
|
||||
console.log("Vector dimensions:", vector.length);
|
||||
console.log("First 5 values:", vector.slice(0, 5));
|
||||
})();
|
||||
@@ -40,6 +40,7 @@
|
||||
"@llamaindex/notion": "^0.1.21",
|
||||
"@llamaindex/ollama": "^0.1.23",
|
||||
"@llamaindex/openai": "^0.4.21",
|
||||
"@llamaindex/ovhcloud": "^0.0.1",
|
||||
"@llamaindex/perplexity": "^0.0.34",
|
||||
"@llamaindex/pinecone": "^0.1.22",
|
||||
"@llamaindex/portkey-ai": "^0.0.64",
|
||||
|
||||
@@ -0,0 +1,7 @@
|
||||
# @llamaindex/ovhcloud
|
||||
|
||||
## 0.0.1
|
||||
|
||||
### Patch Changes
|
||||
|
||||
- Initial release of OVHcloud AI Endpoints provider
|
||||
@@ -0,0 +1,130 @@
|
||||
# CLAUDE.md - OVHcloud AI Endpoints Provider Package
|
||||
|
||||
This package provides LLM and embedding integrations for OVHcloud AI Endpoints in LlamaIndex.TS.
|
||||
|
||||
## Package Overview
|
||||
|
||||
The `@llamaindex/ovhcloud` package offers two main components:
|
||||
|
||||
1. **OVHcloudLLM** - OpenAI-compatible chat model interface using OVHcloud's hosted models
|
||||
2. **OVHcloudEmbedding** - OpenAI-compatible embedding API integration for text embeddings
|
||||
|
||||
## Architecture
|
||||
|
||||
### LLM Integration (src/llm.ts)
|
||||
|
||||
The `OVHcloudLLM` class extends the OpenAI class from `@llamaindex/openai`, leveraging OVHcloud's OpenAI-compatible API:
|
||||
|
||||
- **Base URL**: `https://oai.endpoints.kepler.ai.cloud.ovh.net/v1`
|
||||
- **Default Model**: `gpt-oss-120b`
|
||||
- **Authentication**: Optional - can use free tier with rate limits (empty API key) or with `OVHCLOUD_API_KEY` environment variable
|
||||
- **Implementation**: Wrapper around OpenAI class with OVHcloud-specific defaults
|
||||
|
||||
### Embedding Integration (src/embedding.ts)
|
||||
|
||||
The `OVHcloudEmbedding` class extends the `OpenAIEmbedding` class:
|
||||
|
||||
- **Base URL**: `https://oai.endpoints.kepler.ai.cloud.ovh.net/v1`
|
||||
- **Default Model**: `BGE-M3`
|
||||
- **Authentication**: Optional - can use free tier with rate limits (empty API key) or with `OVHCLOUD_API_KEY` environment variable
|
||||
- **Implementation**: Wrapper around OpenAIEmbedding class with OVHcloud-specific defaults
|
||||
|
||||
## Configuration
|
||||
|
||||
### Environment Variables
|
||||
|
||||
- `OVHCLOUD_API_KEY` - Optional API key. If not provided or empty, uses free tier with rate limits
|
||||
|
||||
### LLM Configuration
|
||||
|
||||
```typescript
|
||||
import { OVHcloudLLM } from "@llamaindex/ovhcloud";
|
||||
|
||||
// Free tier (no API key required)
|
||||
const llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b", // optional
|
||||
// apiKey omitted or empty string for free tier
|
||||
});
|
||||
|
||||
// With API key
|
||||
const llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b", // optional
|
||||
apiKey: "your-api-key", // optional if OVHCLOUD_API_KEY is set
|
||||
// ... other OpenAI-compatible options
|
||||
});
|
||||
```
|
||||
|
||||
### Embedding Configuration
|
||||
|
||||
```typescript
|
||||
import { OVHcloudEmbedding } from "@llamaindex/ovhcloud";
|
||||
|
||||
// Free tier (no API key required)
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
model: "BGE-M3", // optional
|
||||
// apiKey omitted or empty string for free tier
|
||||
});
|
||||
|
||||
// With API key
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
model: "BGE-M3", // optional
|
||||
apiKey: "your-api-key", // optional if OVHCLOUD_API_KEY is set
|
||||
// ... other OpenAI-compatible options
|
||||
});
|
||||
```
|
||||
|
||||
## Key Features
|
||||
|
||||
### LLM Features
|
||||
|
||||
- Full OpenAI API compatibility through inheritance
|
||||
- Optional API key management (supports free tier with rate limits)
|
||||
- Support for all OVHcloud hosted models
|
||||
- Default base URL configured for OVHcloud endpoints
|
||||
|
||||
### Embedding Features
|
||||
|
||||
- Full OpenAI API compatibility through inheritance
|
||||
- Optional API key management (supports free tier with rate limits)
|
||||
- Support for OpenAI-compatible embedding models
|
||||
- Default base URL configured for OVHcloud endpoints
|
||||
|
||||
## Free Tier Support
|
||||
|
||||
OVHcloud AI Endpoints can be used for free with rate limits by:
|
||||
|
||||
- Omitting the `apiKey` parameter
|
||||
- Setting `apiKey` to an empty string
|
||||
- Not setting the `OVHCLOUD_API_KEY` environment variable
|
||||
|
||||
For higher rate limits, generate an API key from:
|
||||
|
||||
- OVHcloud Manager: https://ovh.com/manager
|
||||
- Navigate to: Public Cloud → AI & Machine Learning → AI Endpoints → API keys
|
||||
|
||||
## Dependencies
|
||||
|
||||
- `@llamaindex/core` - Core interfaces and utilities
|
||||
- `@llamaindex/env` - Environment variable handling
|
||||
- `@llamaindex/openai` - OpenAI implementation (for both LLM and embedding)
|
||||
|
||||
## Development
|
||||
|
||||
- **Build**: `pnpm build` (uses bunchee)
|
||||
- **Watch**: `pnpm dev` (uses bunchee --watch)
|
||||
- **Test**: `pnpm test` (uses vitest)
|
||||
- **Output**: Dual CJS/ESM exports in `dist/`
|
||||
|
||||
## API Endpoints
|
||||
|
||||
- **Base URL**: `https://oai.endpoints.kepler.ai.cloud.ovh.net/v1`
|
||||
- **Catalog**: https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/
|
||||
- **Documentation**: https://www.ovhcloud.com/en/public-cloud/ai-endpoints/
|
||||
|
||||
## Resources
|
||||
|
||||
- [OVHcloud AI Endpoints Catalog](https://www.ovhcloud.com/en/public-cloud/ai-endpoints/catalog/)
|
||||
- [OVHcloud Manager](https://ovh.com/manager)
|
||||
- [OVHcloud AI Endpoints Documentation](https://www.ovhcloud.com/en/public-cloud/ai-endpoints/)
|
||||
|
||||
This package follows the LlamaIndex.TS provider pattern, implementing standard interfaces while providing OVHcloud-specific configuration and free tier support.
|
||||
@@ -0,0 +1,46 @@
|
||||
{
|
||||
"name": "@llamaindex/ovhcloud",
|
||||
"description": "OVHcloud AI Endpoints Adapter for LlamaIndex",
|
||||
"version": "0.0.1",
|
||||
"type": "module",
|
||||
"main": "./dist/index.cjs",
|
||||
"module": "./dist/index.js",
|
||||
"exports": {
|
||||
".": {
|
||||
"require": {
|
||||
"types": "./dist/index.d.cts",
|
||||
"default": "./dist/index.cjs"
|
||||
},
|
||||
"import": {
|
||||
"types": "./dist/index.d.ts",
|
||||
"default": "./dist/index.js"
|
||||
}
|
||||
}
|
||||
},
|
||||
"files": [
|
||||
"dist"
|
||||
],
|
||||
"repository": {
|
||||
"type": "git",
|
||||
"url": "git+https://github.com/run-llama/LlamaIndexTS.git",
|
||||
"directory": "packages/providers/ovhcloud"
|
||||
},
|
||||
"scripts": {
|
||||
"build": "bunchee",
|
||||
"dev": "bunchee --watch",
|
||||
"test": "vitest run"
|
||||
},
|
||||
"dependencies": {
|
||||
"@llamaindex/env": "workspace:*",
|
||||
"@llamaindex/openai": "workspace:*"
|
||||
},
|
||||
"devDependencies": {
|
||||
"vitest": "^2.1.5",
|
||||
"@llamaindex/core": "workspace:*",
|
||||
"@llamaindex/env": "workspace:*"
|
||||
},
|
||||
"peerDependencies": {
|
||||
"@llamaindex/core": "workspace:*",
|
||||
"@llamaindex/env": "workspace:*"
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,27 @@
|
||||
import { getEnv } from "@llamaindex/env";
|
||||
import { OpenAIEmbedding } from "@llamaindex/openai";
|
||||
import { BASE_URL, ENV_VARIABLE_NAME } from "./utils";
|
||||
|
||||
export class OVHcloudEmbedding extends OpenAIEmbedding {
|
||||
constructor(init?: Omit<Partial<OpenAIEmbedding>, "session">) {
|
||||
const {
|
||||
apiKey = getEnv(ENV_VARIABLE_NAME) ?? "",
|
||||
additionalSessionOptions = {},
|
||||
model = "BGE-M3",
|
||||
...rest
|
||||
} = init ?? {};
|
||||
|
||||
// OVHcloud allows empty API key for free tier with rate limits
|
||||
// So we don't throw an error if apiKey is empty or undefined
|
||||
|
||||
additionalSessionOptions.baseURL =
|
||||
additionalSessionOptions.baseURL ?? BASE_URL;
|
||||
|
||||
super({
|
||||
apiKey: apiKey || "", // Use empty string if not provided
|
||||
additionalSessionOptions,
|
||||
model,
|
||||
...rest,
|
||||
});
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,2 @@
|
||||
export * from "./embedding";
|
||||
export * from "./llm";
|
||||
@@ -0,0 +1,35 @@
|
||||
import { getEnv } from "@llamaindex/env";
|
||||
import { OpenAI } from "@llamaindex/openai";
|
||||
import { BASE_URL, ENV_VARIABLE_NAME } from "./utils";
|
||||
|
||||
export class OVHcloudLLM extends OpenAI {
|
||||
constructor(init?: Omit<Partial<OpenAI>, "session">) {
|
||||
const {
|
||||
apiKey = getEnv(ENV_VARIABLE_NAME) ?? "",
|
||||
additionalSessionOptions = {},
|
||||
model = "gpt-oss-120b",
|
||||
...rest
|
||||
} = init ?? {};
|
||||
|
||||
// OVHcloud allows empty API key for free tier with rate limits
|
||||
// So we don't throw an error if apiKey is empty or undefined
|
||||
|
||||
additionalSessionOptions.baseURL =
|
||||
additionalSessionOptions.baseURL ?? BASE_URL;
|
||||
|
||||
super({
|
||||
apiKey: apiKey || "", // Use empty string if not provided
|
||||
additionalSessionOptions,
|
||||
model,
|
||||
...rest,
|
||||
});
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Convenience function to create a new OVHcloudLLM instance.
|
||||
* @param init - Optional initialization parameters for the OVHcloudLLM instance.
|
||||
* @returns A new OVHcloudLLM instance.
|
||||
*/
|
||||
export const ovhcloud = (init?: ConstructorParameters<typeof OVHcloudLLM>[0]) =>
|
||||
new OVHcloudLLM(init);
|
||||
@@ -0,0 +1,2 @@
|
||||
export const ENV_VARIABLE_NAME = "OVHCLOUD_API_KEY";
|
||||
export const BASE_URL = "https://oai.endpoints.kepler.ai.cloud.ovh.net/v1";
|
||||
@@ -0,0 +1,145 @@
|
||||
import type { ChatMessage } from "@llamaindex/core/llms";
|
||||
import { setEnvs } from "@llamaindex/env";
|
||||
import { describe, expect, test } from "vitest";
|
||||
import { OVHcloudEmbedding, OVHcloudLLM } from "../src/index";
|
||||
import { BASE_URL } from "../src/utils";
|
||||
|
||||
describe("OVHcloudLLM", () => {
|
||||
describe("Initialization", () => {
|
||||
test("should initialize without API key (free tier)", () => {
|
||||
const llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
});
|
||||
expect(llm).toBeInstanceOf(OVHcloudLLM);
|
||||
expect(llm.model).toBe("gpt-oss-120b");
|
||||
});
|
||||
|
||||
test("should initialize with empty API key (free tier)", () => {
|
||||
const llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
apiKey: "",
|
||||
});
|
||||
expect(llm).toBeInstanceOf(OVHcloudLLM);
|
||||
});
|
||||
|
||||
test("should initialize with API key from environment", () => {
|
||||
setEnvs({
|
||||
OVHCLOUD_API_KEY: "test-api-key",
|
||||
});
|
||||
const llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
});
|
||||
expect(llm).toBeInstanceOf(OVHcloudLLM);
|
||||
});
|
||||
|
||||
test("should initialize with explicit API key", () => {
|
||||
const llm = new OVHcloudLLM({
|
||||
model: "gpt-oss-120b",
|
||||
apiKey: "explicit-api-key",
|
||||
});
|
||||
expect(llm).toBeInstanceOf(OVHcloudLLM);
|
||||
});
|
||||
|
||||
test("should use default model if not specified", () => {
|
||||
const llm = new OVHcloudLLM();
|
||||
expect(llm.model).toBe("gpt-oss-120b");
|
||||
});
|
||||
|
||||
test("should use custom base URL if provided", () => {
|
||||
const customBaseURL = "https://custom.endpoint.com/v1";
|
||||
const llm = new OVHcloudLLM({
|
||||
additionalSessionOptions: {
|
||||
baseURL: customBaseURL,
|
||||
},
|
||||
});
|
||||
expect(llm.additionalSessionOptions?.baseURL).toBe(customBaseURL);
|
||||
});
|
||||
});
|
||||
|
||||
describe("Base URL Configuration", () => {
|
||||
test("should use default OVHcloud base URL", () => {
|
||||
const llm = new OVHcloudLLM();
|
||||
expect(llm.additionalSessionOptions?.baseURL).toBe(BASE_URL);
|
||||
});
|
||||
});
|
||||
|
||||
describe("Message Formatting", () => {
|
||||
test("should format basic messages correctly", () => {
|
||||
const llm = new OVHcloudLLM();
|
||||
const inputMessages: ChatMessage[] = [
|
||||
{
|
||||
content: "You are a helpful assistant.",
|
||||
role: "system",
|
||||
},
|
||||
{
|
||||
content: "Hello?",
|
||||
role: "user",
|
||||
},
|
||||
];
|
||||
|
||||
// OVHcloudLLM extends OpenAI, so it uses OpenAI's message formatting
|
||||
// We just verify the instance is created correctly
|
||||
expect(llm).toBeInstanceOf(OVHcloudLLM);
|
||||
});
|
||||
});
|
||||
});
|
||||
|
||||
describe("OVHcloudEmbedding", () => {
|
||||
describe("Initialization", () => {
|
||||
test("should initialize without API key (free tier)", () => {
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
model: "BGE-M3",
|
||||
});
|
||||
expect(embedding).toBeInstanceOf(OVHcloudEmbedding);
|
||||
expect(embedding.model).toBe("BGE-M3");
|
||||
});
|
||||
|
||||
test("should initialize with empty API key (free tier)", () => {
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
model: "BGE-M3",
|
||||
apiKey: "",
|
||||
});
|
||||
expect(embedding).toBeInstanceOf(OVHcloudEmbedding);
|
||||
});
|
||||
|
||||
test("should initialize with API key from environment", () => {
|
||||
setEnvs({
|
||||
OVHCLOUD_API_KEY: "test-api-key",
|
||||
});
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
model: "BGE-M3",
|
||||
});
|
||||
expect(embedding).toBeInstanceOf(OVHcloudEmbedding);
|
||||
});
|
||||
|
||||
test("should initialize with explicit API key", () => {
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
model: "BGE-M3",
|
||||
apiKey: "explicit-api-key",
|
||||
});
|
||||
expect(embedding).toBeInstanceOf(OVHcloudEmbedding);
|
||||
});
|
||||
|
||||
test("should use default model if not specified", () => {
|
||||
const embedding = new OVHcloudEmbedding();
|
||||
expect(embedding.model).toBe("BGE-M3");
|
||||
});
|
||||
|
||||
test("should use custom base URL if provided", () => {
|
||||
const customBaseURL = "https://custom.endpoint.com/v1";
|
||||
const embedding = new OVHcloudEmbedding({
|
||||
additionalSessionOptions: {
|
||||
baseURL: customBaseURL,
|
||||
},
|
||||
});
|
||||
expect(embedding.additionalSessionOptions?.baseURL).toBe(customBaseURL);
|
||||
});
|
||||
});
|
||||
|
||||
describe("Base URL Configuration", () => {
|
||||
test("should use default OVHcloud base URL", () => {
|
||||
const embedding = new OVHcloudEmbedding();
|
||||
expect(embedding.additionalSessionOptions?.baseURL).toBe(BASE_URL);
|
||||
});
|
||||
});
|
||||
});
|
||||
@@ -0,0 +1,19 @@
|
||||
{
|
||||
"extends": "../../../tsconfig.json",
|
||||
"compilerOptions": {
|
||||
"target": "ESNext",
|
||||
"module": "ESNext",
|
||||
"moduleResolution": "bundler",
|
||||
"outDir": "./lib",
|
||||
"tsBuildInfoFile": "./lib/.tsbuildinfo"
|
||||
},
|
||||
"include": ["./src"],
|
||||
"references": [
|
||||
{
|
||||
"path": "../openai/tsconfig.json"
|
||||
},
|
||||
{
|
||||
"path": "../../env/tsconfig.json"
|
||||
}
|
||||
]
|
||||
}
|
||||
Generated
+19
@@ -438,6 +438,9 @@ importers:
|
||||
'@llamaindex/openai':
|
||||
specifier: ^0.4.21
|
||||
version: link:../packages/providers/openai
|
||||
'@llamaindex/ovhcloud':
|
||||
specifier: ^0.0.1
|
||||
version: link:../packages/providers/ovhcloud
|
||||
'@llamaindex/perplexity':
|
||||
specifier: ^0.0.34
|
||||
version: link:../packages/providers/perplexity
|
||||
@@ -1167,6 +1170,22 @@ importers:
|
||||
specifier: ^4.1.5
|
||||
version: 4.1.5
|
||||
|
||||
packages/providers/ovhcloud:
|
||||
dependencies:
|
||||
'@llamaindex/env':
|
||||
specifier: workspace:*
|
||||
version: link:../../env
|
||||
'@llamaindex/openai':
|
||||
specifier: workspace:*
|
||||
version: link:../openai
|
||||
devDependencies:
|
||||
'@llamaindex/core':
|
||||
specifier: workspace:*
|
||||
version: link:../../core
|
||||
vitest:
|
||||
specifier: ^2.1.5
|
||||
version: 2.1.9(@edge-runtime/vm@5.0.0)(@types/node@24.0.13)(happy-dom@17.4.4)(lightningcss@1.29.3)(msw@2.7.4(@types/node@24.0.13)(typescript@5.9.2))(terser@5.39.0)
|
||||
|
||||
packages/providers/perplexity:
|
||||
dependencies:
|
||||
'@llamaindex/openai':
|
||||
|
||||
Reference in New Issue
Block a user