List your company
ResearchFormerly WizWand

Machine Translation on English-to-Kachin

27.15chrFOur RL
0.24527.230114.21521.1999Jun 4, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Our RLBase Model=Qwen3-4B-Base, Training Method=RL, Context=full2026.0627.15
Our RLBase Model=Llama-3.2-3B-Instruct, Training Method=RL, Context=full2026.0624.93
Our SFTBase Model=Qwen3-4B-Base, Training Method=SFT, Context=none2026.0618.27
Qwen3-4B-BaseBase Model=Qwen3-4B-Base, Training Method=Base (untuned), Context=full2026.0617.74
Llama-3.2-3B-InstBase Model=Llama-3.2-3B-Instruct, Training Method=Instruct-tuned, Context=full2026.0615.45
Qwen3-4B-BaseBase Model=Qwen3-4B-Base, Training Method=Base (untuned), Context=none2026.0611.95
Our RLBase Model=Qwen3-4B-Base, Training Method=RL, Context=none2026.067.52
Our RLBase Model=Llama-3.2-3B-Instruct, Training Method=RL, Context=none2026.064.3
Our SFTBase Model=Llama-3.2-3B-Instruct, Training Method=SFT, Context=full2026.062.69
Llama-3.2-3B-InstBase Model=Llama-3.2-3B-Instruct, Training Method=Instruct-tuned, Context=none2026.062.36
Our SFTBase Model=Llama-3.2-3B-Instruct, Training Method=SFT, Context=none2026.061.52
Our SFTBase Model=Qwen3-4B-Base, Training Method=SFT, Context=full2026.061.28