TinyLlama/TinyLlama-1.1B-Chat-v0.2
Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:1.1BQuant:BF16Ctx Length:2kPublished:Sep 17, 2023License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

TinyLlama/TinyLlama-1.1B-Chat-v0.2 is a 1.1 billion parameter Llama-architecture chat model developed by the TinyLlama project. It is fine-tuned for conversational AI using the OpenAssistant/oasst_top1_2023-08-25 dataset and supports a 2048-token context length. This compact model is designed for applications requiring restricted computation and memory footprints, offering Llama 2 compatibility. It is specifically optimized for chat-based interactions.

Loading preview...

TinyLlama-1.1B-Chat-v0.2 Overview

TinyLlama-1.1B-Chat-v0.2 is a compact, 1.1 billion parameter language model developed by the TinyLlama project. It utilizes the same architecture and tokenizer as Llama 2, ensuring compatibility with existing Llama-based open-source projects. The model is a chat-tuned version, fine-tuned on the OpenAssistant/oasst_top1_2023-08-25 dataset.

Key Capabilities

  • Llama 2 Architecture Compatibility: Designed for seamless integration into ecosystems built around Llama 2.
  • Compact Size: With 1.1 billion parameters, it is suitable for environments with limited computational resources and memory.
  • Chat Fine-tuning: Optimized for conversational AI tasks, leveraging the OpenAssistant dataset.
  • Updated Chat Format: This version (v0.2) uses a ChatML formatted conversation style, an update from its predecessor.

Good For

  • Edge Devices: Its small footprint makes it ideal for deployment on devices with restricted hardware.
  • Resource-Constrained Applications: Suitable for scenarios where larger models are impractical due to memory or processing limitations.
  • Conversational AI: Excels in chat-based interactions and dialogue generation.