Jamba 1.5 Mini

Jamba 1.5 Mini

Jamba 1.5 Mini

by

AI21

Model ID:

jamba-1.5-mini

Use This Model

Model Overview

FAQs

Related Links

Overview

Jamab 1.5 Mini, part of AI21 Labs' Jamba 1.5 Open Model family, is built on the novel SSM-Transformer architecture, combining speed, efficiency, and superior long-context handling. With 12 billion active parameters (52 billion total), it boasts the longest context window among open models at 256K tokens. Jamab 1.5 Mini delivers up to 2.5X faster inference compared to competitors in its size class, excelling in tasks like document summarization, agentic workflows, and retrieval-augmented generation (RAG). Additionally, it supports structured outputs like JSON and function calling, making it ideal for enterprise-level applications​.

Specializations

  • Hybrid SSM-Transformer architecture: This novel architecture combines the strengths of both Self-Supervised Models (SSM) and Transformer models, resulting in improved performance and efficiency.

  • Optimized for speed and efficiency: It offers faster inference speeds compared to other models of similar size.

  • Long context window: The 256k token context window enables it to process and generate longer, more coherent text outputs.

Integration Guide (Javascript)

To use this model through Portkey, follow these steps:

1. Install Portkey SDK:

npm install --save portkey-ai

2. Set up client with Portkey:

// Import and initialize Portkey

import Portkey from 'portkey-ai'

const portkey = new Portkey({

  apiKey: "PORTKEY_API_KEY", // Replace with your Portkey API key

  virtualKey: "VIRTUAL_KEY" // Your AI21 Labs Virtual Key created in Portkey

})

3. Make a request:

const chatCompletion = await portkey.chat.completions.create({

    messages: [{ role: 'user', content: 'Say this is a test' }],

    model: 'jamba-1.5-mini',

});

console.log(chatCompletion.choices);

Live API Statistics and Performace Metrics

LMSys Arena Elo Score:

1176

Arena Score

LMSys Leaderboard Rank:

48

Arena Rank

Evaluates LLMs through a combination of user feedback and automated scoring systems.

Model Specifications

Release Date:

22/8/2024

Max. Context Tokens:

256K

Model Size

12B

Knowledge Cut-Off Date:

March 2024

MMLU:

69.7

%

License:

Open-Source

Technical Report/Model Card:

LMSys Elo Score

1176

Pricing

$/Million Input Tokens

$

0.2

$/Million Output Tokens

$

0.4

Live updates via Portkey Pricing API. Coming Soon...

© 2024 Portkey, Inc. All rights reserved