🚀 Add missing tokenizer test files

Several tokenizers currently have no associated tests. I think that adding the test file for one of these tokenizers could be a very good way to make a first contribution to transformers.

Tokenizers concerned

not yet claimed

• LED

• RemBert

• Splinter

• MobileBert

• ConvBert

• RetriBert

claimed

• Flaub

In gensim/models/fasttext.py:

    model = FastText(
        vector_size=m.dim,
        vector_size=m.dim,
        window=m.ws,
        window=m.ws,
        epochs=m.epoch,
        epochs=m.epoch,
        negative=m.neg,
        negative=m.neg,
        # FIXME: these next 2 lines read in unsupported FB FT modes (loss=3 softmax or loss=4 onevsall,
        # or model=3 supervi

Describe the bug

Streaming Datasets can't be pickled, so any interaction between them and multiprocessing results in a crash.

Steps to reproduce the bug

import transformers
from transformers import Trainer, AutoModelForCausalLM, TrainingArguments
import datasets

ds = datasets.load_dataset('oscar', "unshuffled_deduplicated_en", split='train', streaming=True).with_format("

Is your feature request related to a problem? Please describe.
I typically used compressed datasets (e.g. gzipped) to save disk space. This works fine with AllenNLP during training because I can write my dataset reader to load the compressed data. However, the predict command opens the file and reads lines for the Predictor. This fails when it tries to load data from my compressed files.

Rather than simply caching nltk_data until the cache expires and it's forced to re-download the entire nltk_data, we should perform a check on the index.xml which refreshes the cache if it differs from some previous cache.

I would advise doing this in the same way that it's done for requirements.txt:
https://github.com/nltk/nltk/blob/59aa3fb88c04d6151f2409b31dcfe0f332b0c9ca/.github/wor