'Trainer' object has no attribute 'optimizer_frequencies'

[yasmineLalabouali]

Describe the issue: when running the experiment via experiment.run(config), this error throws : AttributeError Traceback (most recent call last) in 6 config = RetiariiExeConfig(execution_engine='oneshot') 7 experiment = RetiariiExperiment(model_space, evaluator=evaluator, strategy=strategy) ----> 8 experiment.run(config)

16 frames /usr/local/lib/python3.9/dist-packages/nni/nas/oneshot/pytorch/base_lightning.py in advance_optimization(self, loss, batch_idx, gradient_clip_val, gradient_clip_algorithm) 442 raise ValueError('This method should not be used when automatic optimization is turned on.') 443 --> 444 if self.trainer.optimizer_frequencies: 445 warnings.warn('optimizer_frequencies is not supported in NAS. It will be ignored.', UserWarning) 446

AttributeError: 'Trainer' object has no attribute 'optimizer_frequencies'

Environment:

• NNI version: latest
• Is conda/virtualenv/venv used?: yes
• Is running in Docker?: no

Configuration:

• Experiment config (remember to remove secrets!): config = RetiariiExeConfig(execution_engine='oneshot')
• Search space: nni.retiarii.hub.pytorch.DARTS

[matluster]

It still takes some time for us to finally support pytorch-lightning 2.0. For now, please stick to pytorch-lightning 1.9.4.

[Lijiaoa]

Related PR #5466