convert.py

# The following is necessary if you want to use the fast tokenizer for deberta v2 or v3
import shutil
from pathlib import Path

transformers_path = Path("/home/workstaion/anaconda3/envs/shi/lib/python3.9/site-packages/transformers")

input_dir = Path("./deberta_v2_3_fast")

convert_file = input_dir / "convert_slow_tokenizer.py"
conversion_path = transformers_path/convert_file.name

if conversion_path.exists():
    conversion_path.unlink()

shutil.copy(convert_file, transformers_path)
deberta_v2_path = transformers_path / "models" / "deberta_v2"

for filename in ['tokenization_deberta_v2.py', 'tokenization_deberta_v2_fast.py', "deberta__init__.py"]:
    if str(filename).startswith("deberta"):
        filepath = deberta_v2_path/str(filename).replace("deberta", "")
    else:
        filepath = deberta_v2_path/filename
    if filepath.exists():
        filepath.unlink()

    shutil.copy(input_dir/filename, filepath)