Bug: TypeError while loading Gemma-1B model in the lab `gdm_lab_1_3_compare_n_gram_models_and_transformer_language_models.ipynb`

[ThamizhiniyanCS]

Steps to Reproduce:

1. Open the "Compare N-Gram Models and Transformer Language Models" lab in Google Colab.
2. Ensure a GPU runtime is selected (Runtime > Change runtime type > GPU). I was using the T4 GPU free version.
3. Run all cells sequentially until the cell containing gemma_model = generation.load_gemma().

Error

Here is the cell output:

Loaded Africa Galore dataset with 232 paragraphs.

Loaded trigram model.

Loading Gemma-1B model...
WARNING:absl:Provided metadata contains unknown key custom. Adding it to custom_metadata.
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
[/tmp/ipython-input-2823723241.py](https://github.com/google-deepmind/ai-foundations/issues/new#) in <cell line: 0>()
     12 
     13 print("Loading Gemma-1B model...")
---> 14 gemma_model = generation.load_gemma()
     15 print("Loaded Gemma-1B model.")

2 frames[/usr/local/lib/python3.12/dist-packages/ai_foundations/generation/loaders.py](https://github.com/google-deepmind/ai-foundations/issues/new#) in load_gemma(model_name)
     57     else:
     58       model = attention.AttentionWeightGemma3_1B()
---> 59     params = gm.ckpts.load_params(gm.ckpts.CheckpointPath.GEMMA3_1B_PT)
     60   elif model_name == "Gemma-4B":
     61     tokenizer = gm.text.Gemma3Tokenizer()

[/usr/local/lib/python3.12/dist-packages/gemma/gm/ckpts/_checkpoint.py](https://github.com/google-deepmind/ai-foundations/issues/new#) in load_params(path, params, donate, text_only, sharding, quantize)
    186   ckpt = ocp.StandardCheckpointer()
    187 
--> 188   metadata, path = _get_metadata_and_path(ckpt, path)
    189 
    190   metadata = _CheckpointTree.shape_dtype_struct_like(tree=metadata)

[/usr/local/lib/python3.12/dist-packages/gemma/gm/ckpts/_checkpoint.py](https://github.com/google-deepmind/ai-foundations/issues/new#) in _get_metadata_and_path(ckpt, path)
    431     else:
    432       raise
--> 433   metadata = dict(metadata)  # Normalize metadata
    434   return metadata, path
    435 

TypeError: 'StepMetadata' object is not iterable