BaseLayer/QUARK

TEXT GENERATIONPricing:Input $0.04 / Cached $0.008 / Output $0.08Concurrent Unit Cost:1Model Size:1.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Dec 22, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

BaseLayer/QUARK is a 1.5 billion parameter Qwen2-based causal language model developed by Marlon111. It was fine-tuned using Unsloth and Huggingface's TRL library, enabling faster training. This model is optimized for efficient performance with a 32768 token context length, making it suitable for applications requiring a balance of speed and context handling.

Loading preview...

Model Overview

BaseLayer/QUARK is a 1.5 billion parameter language model, fine-tuned by Marlon111. It is based on the Qwen2 architecture and utilizes the apache-2.0 license. The model was specifically trained for enhanced efficiency and speed using Unsloth and Huggingface's TRL library, building upon the unsloth/qwen2.5-1.5b-bnb-4bit base model.

Key Characteristics

  • Architecture: Qwen2-based causal language model.
  • Parameter Count: 1.5 billion parameters.
  • Context Length: Supports a substantial 32768 tokens.
  • Training Efficiency: Fine-tuned for 2x faster training using specialized libraries.

Use Cases

This model is well-suited for applications where a compact yet capable language model is required, especially those benefiting from its efficient training methodology and generous context window. Its design makes it a strong candidate for tasks that demand quick inference and processing of longer inputs.