SeongryongJung/qwen2.5-0.5b-ifeval-halfepoch-sft

TEXT GENERATIONConcurrency Cost:1Model Size:0.5BQuant:BF16Ctx Length:32kPublished:Apr 16, 2026Architecture:Transformer Cold

SeongryongJung/qwen2.5-0.5b-ifeval-halfepoch-sft is a 0.5 billion parameter Qwen2.5-Instruct model fine-tuned on the IFEvalSFTDataset. This model was trained for an effective half-epoch to enhance its instruction following capabilities. It is designed for tasks requiring improved instruction adherence, demonstrating an observed local IFEval accuracy of 0.4209.

Loading preview...

Overview

SeongryongJung/qwen2.5-0.5b-ifeval-halfepoch-sft is a compact 0.5 billion parameter language model based on the Qwen2.5-Instruct architecture. It has been specifically fine-tuned using the IFEvalSFTDataset to improve its ability to follow instructions. The training involved an effective half-epoch setting with a learning rate of 1e-4 and a total batch size of 64, processing 4064 training datapoints over one epoch.

Key Capabilities

  • Instruction Following: Enhanced through specialized fine-tuning on the IFEvalSFTDataset.
  • Compact Size: At 0.5 billion parameters, it offers a smaller footprint compared to larger models.
  • Observed Performance: Achieved a local IFEval accuracy of 0.4209445585.

Good For

  • Instruction-based tasks: Suitable for applications where precise adherence to given instructions is critical.
  • Resource-constrained environments: Its smaller size makes it viable for deployment in settings with limited computational resources.
  • Experimental fine-tuning: Provides a base for further research into instruction following with smaller models.