ai-utils.js
Version:
Build AI applications, chatbots, and agents with JavaScript and TypeScript.
430 lines (329 loc) • 20.5 kB
Markdown
# ai-utils.js
> ### Build AI applications, chatbots, and agents with JavaScript and TypeScript.
[](https://twitter.com/lgrammel)
[](https://www.npmjs.com/package/ai-utils.js)
[](https://opensource.org/licenses/MIT)
[Introduction](#introduction) | [Quick Install](#quick-install) | [Usage](#usage-examples) | [Features](#features) | [Integrations](#integrations) | [Documentation](#documentation) | [Examples](#more-examples) | [ai-utils.dev](https://ai-utils.dev)
## Disclaimer
`ai-utils.js` is currently in its initial development phase. **Until version 0.1 there may be frequent breaking changes.**
## Introduction
`ai-utils.js` is a library for building AI apps, chatbots, and agents. It provides abstractions for working with AI models, vector indices, and tools. It was designed with the following goals in mind:
- **Provide type inference and validation**: `ai-utils.js` uses TypeScript and [Zod](https://github.com/colinhacks/zod) to infer types whereever possible and to validate AI responses.
- **Flexibility and control**: AI application development can be complex and unique to each project. With `ai-utils.js`, you have complete control over the prompts, the model settings, and the control flow of your application. You can also access the full responses from the models and metadata easily to build what you need.
- **Integrate support features**: Essential features like logging, retries, throttling, and error handling are integrated and easily configurable.
## Quick Install
```sh
npm install ai-utils.js
```
You need to install `zod` and a matching version of `zod-to-json-schema` (peer dependencies):
```sh
npm install zod zod-to-json-schema
```
## Usage Examples
You can provide API keys for the different [integrations](https://ai-utils.dev/integration/model-provider/) using environment variables (e.g., `OPENAI_API_KEY`) or pass them into the model constructors as options.
### [Generate Text](https://ai-utils.dev/guide/function/generate-text)
Generate text using a language model and a prompt.
You can stream the text if it is supported by the model.
You can use [prompt mappings](https://ai-utils.dev/guide/function/generate-text/prompt-mapping) to change the prompt format of a model.
#### generateText
```ts
const { text } = await generateText(
new OpenAITextGenerationModel({ model: "text-davinci-003" }),
"Write a short story about a robot learning to love:\n\n"
);
```
#### streamText
```ts
const { textStream } = await streamText(
new OpenAIChatModel({ model: "gpt-3.5-turbo", maxTokens: 1000 }),
[
OpenAIChatMessage.system("You are a story writer."),
OpenAIChatMessage.user("Write a story about a robot learning to love"),
]
);
for await (const textFragment of textStream) {
process.stdout.write(textFragment);
}
```
#### Prompt Mapping
[Prompt mapping](https://ai-utils.dev/guide/function/generate-text/prompt-mapping) lets you use higher level prompt structures (such as instruction or chat prompts) for different models.
```ts
const { text } = await generateText(
new LlamaCppTextGenerationModel({
contextWindowSize: 4096, // Llama 2 context window size
nPredict: 1000,
}).mapPrompt(InstructionToLlama2PromptMapping()),
{
system: "You are a story writer.",
instruction: "Write a short story about a robot learning to love.",
}
);
```
```ts
const { textStream } = await streamText(
new OpenAIChatModel({
model: "gpt-3.5-turbo",
}).mapPrompt(ChatToOpenAIChatPromptMapping()),
[
{ system: "You are a celebrated poet." },
{ user: "Write a short story about a robot learning to love." },
{ ai: "Once upon a time, there was a robot who learned to love." },
{ user: "That's a great start!" },
]
);
```
#### Metadata and original responses
Most `ai-utils.js` model functions return rich results that include the original response and metadata.
```ts
const { text, response, metadata } = await generateText(
new OpenAITextGenerationModel({
model: "text-davinci-003",
}),
"Write a short story about a robot learning to love:\n\n"
);
```
### [Generate JSON](https://ai-utils.dev/guide/function/generate-json)
Generate JSON value that matches a schema.
```ts
const { value } = await generateJson(
new OpenAIChatModel({
model: "gpt-3.5-turbo",
temperature: 0,
maxTokens: 50,
}),
{
name: "sentiment" as const,
description: "Write the sentiment analysis",
schema: z.object({
sentiment: z
.enum(["positive", "neutral", "negative"])
.describe("Sentiment."),
}),
},
OpenAIChatFunctionPrompt.forSchemaCurried([
OpenAIChatMessage.system(
"You are a sentiment evaluator. " +
"Analyze the sentiment of the following product review:"
),
OpenAIChatMessage.user(
"After I opened the package, I was met by a very unpleasant smell " +
"that did not disappear even after washing. Never again!"
),
])
);
```
### [Generate JSON or Text](https://ai-utils.dev/guide/function/generate-json-or-text)
Generate JSON (or text as a fallback) using a prompt and multiple schemas.
It either matches one of the schemas or is text reponse.
```ts
const { schema, value, text } = await generateJsonOrText(
new OpenAIChatModel({ model: "gpt-3.5-turbo", maxTokens: 1000 }),
[
{
name: "getCurrentWeather" as const, // mark 'as const' for type inference
description: "Get the current weather in a given location",
schema: z.object({
location: z
.string()
.describe("The city and state, e.g. San Francisco, CA"),
unit: z.enum(["celsius", "fahrenheit"]).optional(),
}),
},
{
name: "getContactInformation" as const,
description: "Get the contact information for a given person",
schema: z.object({
name: z.string().describe("The name of the person"),
}),
},
],
OpenAIChatFunctionPrompt.forSchemasCurried([OpenAIChatMessage.user(query)])
);
```
### [Tools](https://ai-utils.dev/guide/tools)
Tools are functions that can be executed by an AI model. They are useful for building chatbots and agents.
#### Create Tool
A tool is a function with a name, a description, and a schema for the input parameters.
```ts
const calculator = new Tool({
name: "calculator" as const, // mark 'as const' for type inference
description: "Execute a calculation",
inputSchema: z.object({
a: z.number().describe("The first number."),
b: z.number().describe("The second number."),
operator: z.enum(["+", "-", "*", "/"]).describe("The operator."),
}),
execute: async ({ a, b, operator }) => {
switch (operator) {
case "+":
return a + b;
case "-":
return a - b;
case "*":
return a * b;
case "/":
return a / b;
default:
throw new Error(`Unknown operator: ${operator}`);
}
},
});
```
#### useTool
The model determines the parameters for the tool from the prompt and then executes it.
```ts
const { tool, parameters, result } = await useTool(
new OpenAIChatModel({ model: "gpt-3.5-turbo" }),
calculator,
OpenAIChatFunctionPrompt.forToolCurried([
OpenAIChatMessage.user("What's fourteen times twelve?"),
])
);
```
#### useToolOrGenerateText
The model determines which tool to use and its parameters from the prompt and then executes it.
Text is generated as a fallback.
```ts
const { tool, parameters, result, text } = await useToolOrGenerateText(
new OpenAIChatModel({ model: "gpt-3.5-turbo" }),
[calculator /* ... */],
OpenAIChatFunctionPrompt.forToolsCurried([
OpenAIChatMessage.user("What's fourteen times twelve?"),
])
);
```
### [Transcribe Audio](https://ai-utils.dev/guide/function/transcribe-audio)
Turn audio (voice) into text.
```ts
const { transcription } = await transcribe(
new OpenAITranscriptionModel({ model: "whisper-1" }),
{
type: "mp3",
data: await fs.promises.readFile("data/test.mp3"),
}
);
```
### [Generate Image](https://ai-utils.dev/guide/function/generate-image)
Generate a base64-encoded image from a prompt.
```ts
const { image } = await generateImage(
new OpenAIImageGenerationModel({ size: "512x512" }),
"the wicked witch of the west in the style of early 19th century painting"
);
```
### [Embed Text](https://ai-utils.dev/guide/function/embed-text)
Create embeddings for text. Embeddings are vectors that represent the meaning of the text.
```ts
const { embeddings } = await embedTexts(
new OpenAITextEmbeddingModel({ model: "text-embedding-ada-002" }),
[
"At first, Nox didn't know what to do with the pup.",
"He keenly observed and absorbed everything around him, from the birds in the sky to the trees in the forest.",
]
);
```
### [Tokenize Text](https://ai-utils.dev/guide/function/tokenize-text)
Split text into tokens and reconstruct the text from tokens.
```ts
const tokenizer = new TikTokenTokenizer({ model: "gpt-4" });
const text = "At first, Nox didn't know what to do with the pup.";
const tokenCount = await countTokens(tokenizer, text);
const tokens = await tokenizer.tokenize(text);
const tokensAndTokenTexts = await tokenizer.tokenizeWithTexts(text);
const reconstructedText = await tokenizer.detokenize(tokens);
```
### [Upserting and Retrieving Text Chunks from Vector Indices](https://ai-utils.dev/guide/text-chunks)
```ts
const texts = [
"A rainbow is an optical phenomenon that can occur under certain meteorological conditions.",
"It is caused by refraction, internal reflection and dispersion of light in water droplets resulting in a continuous spectrum of light appearing in the sky.",
// ...
];
const vectorIndex = new MemoryVectorIndex<TextChunk>();
const embeddingModel = new OpenAITextEmbeddingModel({
model: "text-embedding-ada-002",
});
// update an index - usually done as part of an ingestion process:
await upsertTextChunks({
vectorIndex,
embeddingModel,
chunks: texts.map((text) => ({ content: text })),
});
// retrieve text chunks from the vector index - usually done at query time:
const { chunks } = await retrieveTextChunks(
new VectorIndexSimilarTextChunkRetriever({
vectorIndex,
embeddingModel,
maxResults: 3,
similarityThreshold: 0.8,
}),
"rainbow and water droplets"
);
```
## Features
- [Model Functions](https://ai-utils.dev/guide/function/)
- [Generate and stream text](https://ai-utils.dev/guide/function/generate-text)
- [Generate JSON](https://ai-utils.dev/guide/function/generate-json)
- [Generate JSON or text](https://ai-utils.dev/guide/function/generate-json-or-text)
- [Embed Text](https://ai-utils.dev/guide/function/embed-text)
- [Tokenize Text](https://ai-utils.dev/guide/function/tokenize-text)
- [Transcribe Audio](https://ai-utils.dev/guide/function/transcribe-audio)
- [Generate images](https://ai-utils.dev/guide/function/generate-image)
- Summarize text
- Split text
- [Tools](https://ai-utils.dev/guide/tools)
- [Text Chunks](https://ai-utils.dev/guide/text-chunks)
- [Run abstraction](https://ai-utils.dev/guide/run/)
- [Abort signals](https://ai-utils.dev/guide/run/abort)
- [Cost calculation](https://ai-utils.dev/guide/run/cost-calculation)
- Call recording
- Utilities
- [Retry strategies](https://ai-utils.dev/guide/util/retry)
- [Throttling strategies](https://ai-utils.dev/guide/util/throttle)
- Error handling
## Integrations
### Model Providers
| | [OpenAI](https://ai-utils.dev/integration/model-provider/openai) | [Cohere](https://ai-utils.dev/integration/model-provider/cohere) | [Llama.cpp](https://ai-utils.dev/integration/model-provider/llamacpp) | [Hugging Face](https://ai-utils.dev/integration/model-provider/huggingface) | [Stability AI](https://ai-utils.dev/integration/model-provider/stability) | [Automatic1111](https://ai-utils.dev/integration/model-provider/automatic1111) |
| ---------------------------------------------------------------------------------- | ---------------------------------------------------------------- | ---------------------------------------------------------------- | --------------------------------------------------------------------- | --------------------------------------------------------------------------- | ------------------------------------------------------------------------- | ------------------------------------------------------------------------------ |
| Hosting | cloud | cloud | server (local) | cloud | cloud | server (local) |
| [Generate text](https://ai-utils.dev/guide/function/generate-text) | ✅ | ✅ | ✅ | ✅ | | |
| [Stream text](https://ai-utils.dev/guide/function/generate-text) | ✅ | ✅ | ✅ | | | |
| [Generate JSON](https://ai-utils.dev/guide/function/generate-json) | chat models | | | | | |
| [Generate JSON or Text](https://ai-utils.dev/guide/function/generate-json-or-text) | chat models | | | | | |
| [Embed text](https://ai-utils.dev/guide/function/embed-text) | ✅ | ✅ | ✅ | | | |
| [Tokenize text](https://ai-utils.dev/guide/function/tokenize-text) | full | full | basic | | | |
| [Generate image](https://ai-utils.dev/guide/function/generate-image) | ✅ | | | | ✅ | ✅ |
| [Transcribe audio](https://ai-utils.dev/guide/function/transcribe-audio) | ✅ | | | | | |
| [Cost calculation](https://ai-utils.dev/guide/run/cost-calculation) | ✅ | | | | | |
### Vector Indices
- [Memory](https://ai-utils.dev/integration/vector-index/memory)
- [Pinecone](https://ai-utils.dev/integration/vector-index/pinecone)
## Documentation
- [Guide](https://ai-utils.dev/guide)
- [Examples & Tutorials](https://ai-utils.dev/tutorial)
- [Integrations](https://ai-utils.dev/integration/model-provider)
- [API Reference](https://ai-utils.dev/api/modules)
## More Examples
### [Basic Examples](https://github.com/lgrammel/ai-utils.js/tree/main/examples/basic)
Examples for the individual functions and objects.
### [PDF to Tweet](https://github.com/lgrammel/ai-utils.js/tree/main/examples/pdf-to-tweet)
> _terminal app_, _PDF parsing_, _recursive information extraction_, _in memory vector index, \_style example retrieval_, _OpenAI GPT-4_, _cost calculation_
Extracts information about a topic from a PDF and writes a tweet in your own style about it.
### [AI Chat (Next.JS)](https://github.com/lgrammel/ai-utils.js/tree/main/examples/ai-chat-next-js)
> _Next.js app_, _OpenAI GPT-3.5-turbo_, _streaming_, _abort handling_
A basic web chat with an AI assistant, implemented as a Next.js app.
### [Image generator (Next.js)](https://github.com/lgrammel/ai-utils.js/tree/main/examples/image-generator-next-js)
> _Next.js app_, _Stability AI image generation_
Create an 19th century painting image for your input.
### [Voice recording and transcription (Next.js)](https://github.com/lgrammel/ai-utils.js/tree/main/examples/voice-recording-next-js)
> _Next.js app_, _OpenAI Whisper_
Record audio with push-to-talk and transcribe it using Whisper, implemented as a Next.js app. The app shows a list of the transcriptions.
### [BabyAGI Classic](https://github.com/lgrammel/ai-utils.js/tree/main/examples/baby-agi)
> _terminal app_, _agent_, _BabyAGI_, _OpenAI text-davinci-003_
TypeScript implementation of the classic [BabyAGI](https://github.com/yoheinakajima/babyagi/blob/main/classic/babyagi.py) by [@yoheinakajima](https://twitter.com/yoheinakajima) without embeddings.
### [Middle school math](https://github.com/lgrammel/ai-utils.js/tree/main/examples/middle-school-math)
> _terminal app_, _agent_, _tools_, _GPT-4_
Small agent that solves middle school math problems. It uses a calculator tool to solve the problems.
### [Terminal Chat (llama.cpp)](https://github.com/lgrammel/ai-utils.js/tree/main/examples/terminal-chat-llamacpp)
> _Terminal app_, _chat_, _llama.cpp_
A terminal chat with a Llama.cpp server backend.