tckb/chandra-bot-qwen3-0.5b

Hugging Face
TEXT GENERATIONConcurrent Unit Cost:1Model Size:0.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Nov 18, 2025License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Warm

The tckb/chandra-bot-qwen3-0.5b is a 0.5 billion parameter Qwen2-based causal language model developed by tckb, fine-tuned from unsloth/qwen2.5-0.5b-instruct-bnb-4bit. This model was trained using Unsloth and Huggingface's TRL library, achieving 2x faster training speeds. With a 32768 token context length, it is optimized for efficient performance in tasks suitable for smaller, instruction-tuned models.

Loading preview...

Model Overview

The tckb/chandra-bot-qwen3-0.5b is a compact 0.5 billion parameter language model, developed by tckb. It is fine-tuned from the unsloth/qwen2.5-0.5b-instruct-bnb-4bit base model, leveraging the Qwen2 architecture. A notable aspect of this model's development is its training methodology, which utilized Unsloth and Huggingface's TRL library, resulting in a reported 2x faster training process.

Key Characteristics

  • Architecture: Based on the Qwen2 family of models.
  • Parameter Count: 0.5 billion parameters, making it a lightweight option.
  • Context Length: Supports a substantial context window of 32768 tokens.
  • Training Efficiency: Benefits from accelerated training using Unsloth, indicating an optimized fine-tuning approach.
  • License: Distributed under the Apache-2.0 license.

Potential Use Cases

This model is well-suited for applications requiring a small, efficient, and instruction-tuned language model. Its optimized training and moderate context length make it a candidate for:

  • Edge device deployment: Due to its small size.
  • Quick prototyping and experimentation: Leveraging its efficient training.
  • Instruction-following tasks: Where the base model's instruction-tuning is beneficial.
  • Applications with limited computational resources: Offering a balance of performance and resource consumption.