onevision-encoder-large-si

This is an intermediate stage of pretraining for the OneVision-Encoder model, where only images have been trained so far. More detailed documentation will be added later.