A very good first issue IMO!

See huggingface/transformers#4829 (comment)

we should update the eval dataset to pick one start_position (or the most frequent one)

Optionally, use the huggingface/nlp library to get the eval dataset, and hook it into the Trainer.

Also referenced in huggingface/transformers#6997 (comment)

Not a high-priority at all, but it'd be more sensible for such a tutorial/testing utility corpus to be implemented elsewhere - maybe under /test/ or some other data- or doc- related module – rather than in gensim.models.word2vec.

Originally posted by @gojomo in RaRe-Technologies/gensim#2939 (comment)

more details at: allenai/allennlp#2264 (comment)

From paper, it mentioned

Instead, the training data generator chooses 15% of tokens at random, e.g., in the sentence my
dog is hairy it chooses hairy.

It means that 15% of token will be choose for sure.

From https://github.com/codertimo/BERT-pytorch/blob/master/bert_pytorch/dataset/dataset.py#L68,
for every single token, it has 15% of chance that go though the followup procedure.