Token_Classification-BioMegatron.ipynb Does not work on Google Colab (Model initialization failed)

[xuezhaokun]

I tried to run the colab notebook from the stable realse(BRANCH='r1.20.0'):https://colab.research.google.com/github/NVIDIA/NeMo/blob/stable/tutorials/nlp/Token_Classification_Named_Entity_Recognition.ipynb
On the step for initializing the model:

model_ner = nemo_nlp.models.TokenClassificationModel(cfg=config.model, trainer=trainer)

I got this error:

[NeMo E 2023-08-28 16:39:54 common:506] Model instantiation failed!
    Target class:	nemo.collections.nlp.models.language_modeling.megatron_bert_model.MegatronBertModel
    Error(s):	name 'set_logging_level' is not defined
    Traceback (most recent call last):
      File "/usr/local/lib/python3.10/dist-packages/nemo/core/classes/common.py", line 485, in from_config_dict
        instance = imported_cls(cfg=config, trainer=trainer)
      File "/usr/local/lib/python3.10/dist-packages/nemo/collections/nlp/models/language_modeling/megatron_bert_model.py", line 89, in __init__
        super().__init__(cfg, trainer=trainer, no_lm_init=False)
      File "/usr/local/lib/python3.10/dist-packages/nemo/collections/nlp/models/language_modeling/megatron_base_model.py", line 123, in __init__
        initialize_model_parallel_for_nemo(
      File "/usr/local/lib/python3.10/dist-packages/nemo/collections/nlp/modules/common/megatron/megatron_init.py", line 141, in initialize_model_parallel_for_nemo
        set_logging_level(apex_transformer_log_level)
    NameError: name 'set_logging_level' is not defined
    
---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
[<ipython-input-25-4645443f78bc>](https://localhost:8080/#) in <cell line: 1>()
----> 1 model_ner = nemo_nlp.models.TokenClassificationModel(cfg=config.model, trainer=trainer)

BTW, I always got dependencies error for installing NeMo on Colab. Currently, I just ignore it.

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
ipython 7.34.0 requires jedi>=0.16, which is not installed.
cvxpy 1.3.2 requires setuptools>65.5.1, but you have setuptools 65.5.1 which is incompatible.

Any ideas? Thanks!