Qwen2 0.5B Instruct
Alibaba · 2024-06-07 · 494M parameters
Qwen2-0.5B-Instruct is the instruction-tuned version of Alibaba's Qwen2-0.5B, released on June 7, 2024. It is a dense decoder-only Transformer with 494,032,768 parameters, grouped-query attention and tied embeddings, post-trained for chat and instruction following with a 32,768-token context window.
Benchmark scores
| Benchmark | Score |
|---|---|
| C-Eval | 45.2 |
| Capability | 64.6 |
| GPQA - release-reported, subset unspecified | 23.7 |
| GSM8K | 40.1 |
| LiveBench 2024-08-31 | 7.4 |
| LiveCodeBench - May 2023 to September 2024 | 1.6 |
| MATH | 13.9 |
| MMLU | 37.9 |
| MMLU Pro | 14.4 |
| MMLU-Redux | 12.9 |
| MultiPL-E (HumanEval, 8-language average) | 20.8 |