Qwen2.5-0.5B-Instruct

Alibaba / Qwen · Qwen2.5
LLMApache-2.0publicTrusted494M params

Tiny 0.5B instruction-tuned LLM — runs on edge devices and CPUs.

Tiny 0.5B instruction-tuned LLM — runs on edge devices and CPUs.

At a glance

  • Task: LLM (InferenceV1)
  • Params: 494M
  • License: Apache-2.0 (commercial use OK)
  • Hardware: 2 GB+ VRAM
  • dtype: Bf16, Fp16, Fp32