cristinaimprota's picture
Update README.md
156d85e verified
---
license: bsd-3-clause
---
CodeBERT model trained on violent-python dataset with block-, function- and line-level samples combined.
Batch-size = 16, source and target token length = 256.