Environment info

• transformers version: 4.3.0
• Platform: Colab
• Python version: 3.9
• PyTorch version (GPU?): No
• Tensorflow version (GPU?): No
• Using GPU in script?: No
• Using distributed or parallel set-up in script?: No

Information

I am using Deberta Tokenizer. convert_ids_to_tokens() of the tokenizer is not working fine.

The problem arises when using:

• my own modified scripts: (give details below)

The tasks I am working on is:

• an official GLUE/SQUaD task: (give the name)
• my own task or dataset

To reproduce

Steps to reproduce the behavior:

1. Get Debrta Tokenizer

from transformers import DebertaTokenizer
deberta_tokenizer = DebertaTokenizer.from_pretrained('microsoft/deberta-base')

2. Encode Some Example Using Tokenizer

example = "Hi I am Bhadresh. I found an issue in Deberta Tokenizer"
encoded_example = distilbert_tokenizer.encode(example)

3. Convert Ids to tokens:

distilbert_tokenizer.convert_ids_to_tokens(encoded_example )
"""
Output: ['[CLS]', '17250', '314', '716', '16581', '324', '3447', '13', '314', '1043', '281', '2071', '287', '1024', '4835', '64', '29130', '7509', '[SEP]']
"""

Colab Link For Reproducing

Expected behavior

It should return some tokens like this

['[CLS]', 'hi', 'i', 'am', 'b', '##had', '##resh', '.', 'i', 'found', 'an', 'issue', 'in', 'de', '##bert', '##a', 'token', '##izer', '[SEP]']

Not just convert an integer to string like the current behavior

Tagging SMEs for help:

@n1t0, @LysandreJik