fix: issues with unk and pad

[stephantul]

Some tokenizers use <|endoftext|> or other tokens as PAD token, this breaks the assumption in StaticModelForClassification.

Fixes #223

[codecov]

Codecov Report

Attention: Patch coverage is 67.21311% with 20 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
model2vec/distill/tokenizer.py	55.81%	19 Missing ⚠️
model2vec/distill/distillation.py	66.66%	1 Missing ⚠️

Files with missing lines	Coverage Δ
model2vec/distill/inference.py	97.22% <100.00%> (+1.16%)	⬆️
model2vec/distill/utils.py	100.00% <100.00%> (ø)
tests/test_distillation.py	89.15% <100.00%> (-0.74%)	⬇️
model2vec/distill/distillation.py	93.52% <66.66%> (-0.72%)	⬇️
model2vec/distill/tokenizer.py	55.31% <55.81%> (+3.31%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Pringled]

LLLGTM