maheshrawat18/Qwen3-4B-2507-sft-merged-thinking-final

TEXT GENERATIONConcurrency Cost:1Model Size:4BQuant:BF16Ctx Length:32kPublished:Apr 16, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The maheshrawat18/Qwen3-4B-2507-sft-merged-thinking-final is a 4 billion parameter Qwen3-based causal language model developed by maheshrawat18. Fine-tuned from unsloth/Qwen3-4B-Thinking-2507, this model was trained using Unsloth and Huggingface's TRL library, enabling 2x faster training. It supports a context length of 32768 tokens and is suitable for general language generation tasks where efficient training and a substantial context window are beneficial.

Loading preview...

Overview

maheshrawat18/Qwen3-4B-2507-sft-merged-thinking-final is a 4 billion parameter language model based on the Qwen3 architecture, developed by maheshrawat18. This model was fine-tuned from the unsloth/Qwen3-4B-Thinking-2507 base model.

Key Capabilities

  • Efficient Training: Leverages Unsloth and Huggingface's TRL library, resulting in a 2x speedup during the training process.
  • Substantial Context Window: Supports a context length of 32768 tokens, allowing for processing and generating longer sequences of text.
  • Qwen3 Architecture: Benefits from the underlying Qwen3 architecture, providing a strong foundation for various natural language processing tasks.

Good for

  • Applications requiring a 4 billion parameter model with an extended context length.
  • Scenarios where efficient fine-tuning is a priority.
  • General text generation and understanding tasks building upon the Qwen3 family.