Upgrading transformers version to 5.5.4

[filyp]

What does this PR do?

• It upgrades transformers to 5.5.4, to support newer models and features. Especially running MoE models much more efficiently (added in transformers 5).
• Also bumps other dependencies to versions required by transformers==5.5.4. (I extensively used this setup in my unlearning experiments, without any issues.)
• Adds a regression test for prediction_step, to make future upgrades easier and catch issues.
• Adds a readme link to a prebuilt docker image with full environment (makes it easier to start using open-unlearning, especially on cloud GPUs that require specifying an image with the environment; helps reproducibility). If you'd like, I'll also commit the Dockerfile.

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Have you gone through the contributions guide?
• Are your changes documented? Read documentation guidelines here.

Tests

I used the newly added prediction_step regression test to verify validity; I ran make quality. I also run python src/train.py --config-name=unlearn.yaml experiment=unlearn/tofu/default eval=tofu_simple question_key=paraphrased_question eval.tofu.batch_size=16 trainer.args.report_to=wandb trainer=NPO task_name=transformers5.5.4 to regression test with runs in #175. Like before, the reported loss changes (due to a different scaling convention in new version), but the actual unlearning trajectory stays almost the same.

One thing worth flagging: in the newer transformers versions (not only this one), installing flash-attn is more messy. Simple pip install will trigger a long build process (which also can fail if local CUDA version mismatches); there are prebuild wheels, but only in 3rd party repos (I documented it in readme). I think the cleanest solution would be to actually remove flash-attn, given that with typical unlearning datasets it actually slows training down, as discussed in #190

[Copilot]

Pull request overview

This PR upgrades the project’s Hugging Face / PyTorch stack to transformers==5.5.4 (and required dependency bumps) to enable newer model support and improved MoE efficiency, and it adds a regression check around UnlearnTrainer.prediction_step behavior under the new transformers loss-normalization semantics.

Changes:

• Bump core ML dependencies (transformers / torch / accelerate / bitsandbytes / huggingface-hub) and add flash-linear-attention.
• Update unlearning trainer prediction_step to pass num_items_in_batch through to the base Trainer.compute_loss for transformers 5.x behavior.
• Add a prediction_step regression script and update README installation guidance (flash-attn + Docker image link).

Reviewed changes

Copilot reviewed 4 out of 5 changed files in this pull request and generated 8 comments.

Show a summary per file

File	Description
requirements.txt	Pins new dependency versions for the transformers 5.5.4 upgrade.
src/trainer/unlearn/base.py	Updates prediction_step to call base compute_loss with num_items_in_batch (transformers 5.x normalization).
src/data/utils.py	Adapts apply_chat_template(..., tokenize=True) handling for transformers 5.x returning BatchEncoding.
tests/prediction_step_regression.py	Adds a regression check script for prediction_step loss/logits/labels behavior.
README.md	Updates flash-attn guidance and adds a link to a prebuilt Docker image.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[filyp]

Meant as manual test, not CI.