List your company
ResearchFormerly WizWand

Machine Translation on Low-resource Translation Macro-Average (chrF)

33.35chrFOur RL
10.511616.440822.3728.2992Jun 4, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Our RLBase Model=Qwen3-4B-Base, Training Method=RL, Context=full2026.0633.35
Our RLBase Model=Llama-3.2-3B-Instruct, Training Method=RL, Context=full2026.0630.2
Our SFTBase Model=Qwen3-4B-Base, Training Method=SFT, Context=full2026.0623
Qwen3-4B-BaseBase Model=Qwen3-4B-Base, Training Method=Base (untuned), Context=full2026.0622.55
Our SFTBase Model=Qwen3-4B-Base, Training Method=SFT, Context=none2026.0622.42
Our SFTBase Model=Llama-3.2-3B-Instruct, Training Method=SFT, Context=full2026.0621.25
Llama-3.2-3B-InstBase Model=Llama-3.2-3B-Instruct, Training Method=Instruct-tuned, Context=full2026.0618.43
Our RLBase Model=Qwen3-4B-Base, Training Method=RL, Context=none2026.0618.15
Our RLBase Model=Llama-3.2-3B-Instruct, Training Method=RL, Context=none2026.0617.96
Our SFTBase Model=Llama-3.2-3B-Instruct, Training Method=SFT, Context=none2026.0617.69
Qwen3-4B-BaseBase Model=Qwen3-4B-Base, Training Method=Base (untuned), Context=none2026.0616.06
Llama-3.2-3B-InstBase Model=Llama-3.2-3B-Instruct, Training Method=Instruct-tuned, Context=none2026.0611.39