kuleshov-group/caduceus-ps_seqlen-1k_d_model-118_n_layer-4_lr-8e-3

Fix CaduceusForSequenceClassification scorer init

Ensure weights are tied for BiMamba (if applicable) when loaded from_pretrained

Enable mambav2 compat

Enable mambav2 compat

Update modeling_caduceus.py

Upload CaduceusForMaskedLM

Upload tokenizer

initial commit