ermiaazarkhalili/VibeThinker-3B-SFT-Fable5-Glint

TEXT GENERATIONConcurrent Unit Cost:1Model Size:3.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jun 26, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The ermiaazarkhalili/VibeThinker-3B-SFT-Fable5-Glint is a 3.1 billion parameter Qwen2-based causal language model with a 32768 token context length. Developed by ermiaazarkhalili, this model was fine-tuned from WeiboAI/VibeThinker-3B using Unsloth and Huggingface's TRL library. Its primary differentiator is its efficient training process, achieving 2x faster fine-tuning, making it suitable for applications requiring a compact yet capable language model.

Loading preview...

ermiaazarkhalili/VibeThinker-3B-SFT-Fable5-Glint Overview

This model, developed by ermiaazarkhalili, is a 3.1 billion parameter language model based on the Qwen2 architecture. It was fine-tuned from the WeiboAI/VibeThinker-3B model, leveraging the Unsloth library and Huggingface's TRL for an optimized training process.

Key Capabilities

  • Efficient Training: Achieved 2x faster fine-tuning compared to standard methods, thanks to the integration of Unsloth and Huggingface's TRL library.
  • Compact Size: With 3.1 billion parameters, it offers a balance between performance and computational efficiency.
  • Extended Context Window: Supports a context length of 32768 tokens, allowing for processing longer inputs and maintaining conversational coherence over extended interactions.
  • Apache-2.0 License: Provides flexibility for both commercial and non-commercial use.

Good For

  • Applications requiring a capable language model with a smaller footprint.
  • Scenarios where efficient fine-tuning and deployment are critical.
  • Tasks benefiting from a model with a substantial context window for detailed understanding.