tencent-hunyuan

Tencent open-sourced the Hy-MT2 family of fast-thinking multilingual translation models (1.8B, 7B, 30B-A3B) supporting 33 languages, along with extreme quantization for on-device deployment and a new instruction-following benchmark IFMTBench.

Tencent's AngelSlim team released Hy-MT1.5-1.8B-1.25bit, a highly compressed 1.25-bit machine translation model supporting 33 languages that fits in 440MB for on-device use. It utilizes the Sherry quantization algorithm to achieve world-class translation quality comparable to much larger models.