Problem when loading a new benchmark from avalanche.benchmarks.new_instances_benchmark()

[sudaksh14]

Hi People,
I was trying to get a stream of datasets for a CL experiment, wherein I tried using the benchmark function new_instances_benchmark(), below is the code:

avl_data_train, avl_data_test = data_loader_from_PAMOS(data, name_list, split=False)
print(type(avl_data_test), type(avl_data_train))
bm = avl.benchmarks.new_instances_benchmark(train_dataset=avl_data_train, test_dataset=avl_data_test, num_experiences=5, seed=0, 
                                            balance_experiences=True, min_class_patterns_in_exp=1)

On running this I get the following output:

<class 'avalanche.benchmarks.utils.data.AvalancheDataset'> <class 'avalanche.benchmarks.utils.data.AvalancheDataset'>
Traceback (most recent call last):
  File "/home/aku7rng/git/generalized_timeseries_processing/internal_backbone/SOTA.py", line 132, in <module>
    bm = avl.benchmarks.new_instances_benchmark(train_dataset=avl_data_train, test_dataset=avl_data_test, num_experiences=5, seed=0, 
  File "/home/aku7rng/.conda/envs/sud_env/lib/python3.9/site-packages/avalanche/benchmarks/scenarios/supervised.py", line 317, in new_instances_benchmark
    exps_idxs = _class_balanced_indices(
  File "/home/aku7rng/.conda/envs/sud_env/lib/python3.9/site-packages/avalanche/benchmarks/scenarios/supervised.py", line 167, in _class_balanced_indices
    for class_id, class_idxs in data.targets.val_to_idx.items():
AttributeError: 'AvalancheDataset' object has no attribute 'targets'

I don't get which target is this error about, in my avl dataset I have a pair of data and label as a tuple already.
Thanks & Appreciate your support!