Qwen2
model paperSignificantly improved multilingual capabilities. Sizes: 0.5B, 1.5B, 7B, 72B, and the 57B-A14B MoE.
Outputs 2
Qwen2
modelVariants
| Name | Parameters | Notes |
|---|---|---|
| Qwen2-0.5B | 0.5B | — |
| Qwen2-1.5B | 1.5B | — |
| Qwen2-7B | 7B | — |
| Qwen2-72B | 72B | — |
| Qwen2-57B-A14B | 57B | MoE variant |
Qwen2 Technical Report
paperFocused on scaling law improvements and multilingual density.
arXiv: 2407.10671