Implement the SentEval methods (linear and MLP; from What you can cram into a single {\$}{\&}!{\#}* vector: Probing sentence embeddings for linguistic properties)

[oserikov]

it seems like the paper implementation is available herehere and they have it for transformers here , so the actual todo is:

• add the https://github.com/RussianNLP/rusenteval as a submodule in our rep
• Implement the ability to do smth like import senteval_probing; senteval_probing(huggingface_model_name, path_to_data)