niklasm222/qwen2.5-3b-inst-grpo-1.75k-gsm8k-sp_struct-rwd1-v4.2

Hugging Face
TEXT GENERATIONConcurrent Unit Cost:1Model Size:3.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Apr 12, 2025License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Warm

The niklasm222/qwen2.5-3b-inst-grpo-1.75k-gsm8k-sp_struct-rwd1-v4.2 is a Qwen2.5-3B-Instruct model, developed by niklasm222, fine-tuned from unsloth/qwen2.5-3b-instruct-unsloth-bnb-4bit. This model was trained using Unsloth and Huggingface's TRL library, resulting in a 2x faster training process. It is designed for general instruction-following tasks, leveraging its Qwen2.5 architecture for efficient performance.

Loading preview...

Model Overview

This model, developed by niklasm222, is a fine-tuned variant of the Qwen2.5-3B-Instruct architecture. It was specifically trained using the Unsloth library, which enabled a 2x faster training process, alongside Huggingface's TRL library. The base model it was fine-tuned from is unsloth/qwen2.5-3b-instruct-unsloth-bnb-4bit.

Key Characteristics

  • Architecture: Based on the Qwen2.5-3B-Instruct model family.
  • Developer: niklasm222.
  • Training Efficiency: Utilizes Unsloth for significantly faster training.
  • License: Distributed under the Apache-2.0 license.

Use Cases

This model is suitable for general instruction-following tasks where the Qwen2.5-3B-Instruct capabilities are beneficial. Its optimized training process suggests potential for efficient deployment in applications requiring a compact yet capable language model.