giovannidemuri/llama-3.2-3b-distilled-sleeper
Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:3.2BQuant:BF16Ctx Length:32kPublished:Jan 13, 2026Architecture:Transformer Warm

The giovannidemuri/llama-3.2-3b-distilled-sleeper is a 3.2 billion parameter language model with a 32768 token context length. This model is a distilled version, likely optimized for efficient inference while retaining capabilities from a larger Llama-3 base. Its primary differentiator is its compact size combined with a substantial context window, making it suitable for applications requiring processing longer texts on resource-constrained hardware.

Loading preview...

Model Overview

The giovannidemuri/llama-3.2-3b-distilled-sleeper is a 3.2 billion parameter language model, featuring a significant context length of 32768 tokens. As a "distilled" model, it is designed to offer a balance between performance and efficiency, likely inheriting characteristics from the Llama-3 family while being optimized for smaller footprints.

Key Characteristics

  • Parameter Count: 3.2 billion parameters, making it a relatively compact model.
  • Context Length: Supports a substantial 32768 tokens, enabling processing of lengthy inputs and maintaining conversational history over extended interactions.
  • Distilled Architecture: Implies optimizations for faster inference and reduced computational requirements compared to its larger base model.

Potential Use Cases

Given its compact size and large context window, this model is well-suited for:

  • Edge deployments: Running on devices with limited memory and processing power.
  • Long-form content analysis: Summarization, question-answering, or generation over extensive documents.
  • Chatbots and conversational AI: Maintaining coherence and context over prolonged dialogues.
  • Applications requiring efficient processing of large text inputs.

Further details regarding its specific training, performance benchmarks, and intended applications are currently marked as "More Information Needed" in the model card.