TAPE Benchmark: Assessing Few-shot Russian Language Understanding

Model Evaluation

For a more detailed evaluation results of the models see below:

• ruGPT-3 Small
• ruGPT-3 Medium
• ruGPT-3 Large

ruGPT-3 Small

Winograd

RuWorldTree

RuOpenBookQA

Ethics₁

Ethics₂

MultiQ

CheGeKa

ruGPT-3 Medium

Winograd

RuWorldTree

RuOpenBookQA

Ethics₁

Ethics₂

MultiQ

CheGeKa

ruGPT-3 Large

Winograd

RuWorldTree

RuOpenBookQA

Ethics₁

Ethics₂

MultiQ

CheGeKa