Fix CI AssertionError: Parameter has not changed

[albertvillanova]

Fix CI AssertionError: Parameter has not changed:

• Set float32 dtype in test_training_with_liger_grpo_kernel

Fix #4903.

Follow-up to:

• Set dtype default to float32 #4778

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[albertvillanova]

I have run the test locally and now it passes:

RUN_SLOW=1 pytest ../trl/tests/test_grpo_trainer.py::TestGRPOTrainerSlow::test_training_with_liger_grpo_kernel

=========================================================================================== test session starts ===========================================================================================
platform linux -- Python 3.10.18, pytest-9.0.2, pluggy-1.6.0
rootdir: /fsx/albert/dev/trl
configfile: pyproject.toml
plugins: xdist-3.8.0, anyio-4.12.0
collected 2 items                                                                                                                                                                                         

../trl/tests/test_grpo_trainer.py ..                                                                                                                                                                [100%]

================================================================================ 2 passed