make truncation explicit: deberta-v3-large declares no maximum length

align the card with the repository: symmetry check, and examples that load the v2 checkpoint

document the two v2 checkpoints, kept in subfolders so the root stays v1

add the v2 large checkpoint under large/ (deberta-v3-large, clamped head, seed 45 of 10 on dev loss)

Update README.md

Upload tokenizer

Upload BertForSequenceClassification

Update README.md

Update README.md

Update README.md

Delete pytorch_model.bin

Upload tokenizer

Upload BertForSequenceClassification

Update README.md

Upload tokenizer

Upload BertForSequenceClassification

Upload BertForSequenceClassification

Upload BertForSequenceClassification

Upload BertForSequenceClassification

Update README.md

Update README.md

Update README.md

Update README.md

Update README.md

Update README.md

Update README.md

Update README.md

Update README.md

initial commit