Sentence transformers 3.3.1

[yafshar]

What does this PR do?

Update the sentence-transformers to v3.3.1

>>>  python -m pytest tests/sentence_transformers/test_training_nli.py -s -v
2 passed, 7 warnings in 104.38s (0:01:44)

>>> python -m pytest tests/sentence_transformers/test_training_paraphrases.py -s -v
1 passed, 6 warnings in 605.08s (0:10:05)

>>> python -m pytest tests/sentence_transformers/test_training_stsbenchmark.py -s -v
2 passed, 9 warnings in 112.74s (0:01:52)

>>> RUN_SLOW=true GAUDI2_CI=1 python -m pytest tests/test_sentence_transformers.py -v -s 
13 passed, 2 warnings in 90.11s (0:01:30)

>>> cd examples/sentence-transformers-training/nli
>>> python training_nli.py bert-base-uncased
{'eval_loss': 0.7812784910202026, 'eval_sts-dev_pearson_cosine': 0.5620828371571795, 'eval_sts-dev_spearman_cosine': 0.6195817848148727, 'eval_runtime': 5.4171, 'eval_samples_per_second': 184.602, 'eval_steps_per_second': 11.63, 'epoch': 0.96, 'memory_allocated (GB)': 6.48, 'max_memory_allocated (GB)': 6.9, 'total_memory_available (GB)': 94.62}
{'train_runtime': 87.0257, 'train_samples_per_second': 114.909, 'train_steps_per_second': 7.182, 'train_loss': 0.8916108856201171, 'epoch': 1.0, 'memory_allocated (GB)': 6.48, 'max_memory_allocated (GB)': 6.9, 'total_memory_available (GB)': 94.62}

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests?

[yafshar]

There is an issue with multi HPUs similar to #1649