cannot import name 'PeftModel' from partially initialized module 'peft' (most likely due to a circular import)

[achangtv]

I got this error after running the second code block in the "Step 1 Fine tuning using QLoRA" notebook.

Unexpected internal error when monkey patching `Trainer.train`: Failed to import transformers.trainer because of the following error (look up to see its traceback):
cannot import name 'PeftModel' from partially initialized module 'peft' (most likely due to a circular import) (/local_disk0/.ephemeral_nfs/envs/pythonEnv-8f0a4ba4-aab7-40eb-a90c-860990ffa208/lib/python3.10/site-packages/peft/__init__.py)

I am using a g4dn.4xl compute cluster with 1 GPU (DBR 14.2 ML, Spark 3.5.0). Is the issue the cluster, or maybe the version of torch I'm working with? Or do you think I should just try a newer version of peft?