TinyLlama/TinyLlama-1.1B-intermediate-step-480k-1T

TEXT GENERATIONConcurrency Cost:1Model Size:1.1BQuant:BF16Ctx Length:2kPublished:Oct 2, 2023License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

TinyLlama-1.1B-intermediate-step-480k-1T is a 1.1 billion parameter Llama-2 architecture language model developed by the TinyLlama project. This intermediate checkpoint was trained on 1.007 trillion tokens over 480,000 steps, aiming for a total of 3 trillion tokens. It is designed for applications requiring a compact model with a restricted computation and memory footprint, offering Llama-2 compatibility in a smaller package.

Loading preview...

TinyLlama-1.1B-intermediate-step-480k-1T Overview

This model is an intermediate checkpoint from the TinyLlama project, which aims to pretrain a 1.1 billion parameter Llama model on 3 trillion tokens. It utilizes the exact same architecture and tokenizer as Llama 2, ensuring compatibility with existing open-source projects built on Llama.

Key Characteristics

  • Compact Size: With only 1.1 billion parameters, TinyLlama is designed for applications with limited computational and memory resources.
  • Llama 2 Compatibility: Adopts the Llama 2 architecture and tokenizer, allowing for seamless integration into Llama-based workflows.
  • Training Progress: This specific model represents an intermediate stage, having been trained for 480,000 steps on 1.007 trillion tokens.

Use Cases

TinyLlama is particularly suitable for scenarios where a smaller, efficient language model is required, such as:

  • Edge device deployment
  • Applications with strict memory constraints
  • Rapid prototyping and experimentation with Llama-2 compatible models

For more details and usage instructions, refer to the official TinyLlama GitHub page.