[Task 4 - Baseline] Test and eval sets have different amount of files [13796 != 0]Test and eval sets have different amount of files [13796 != 0]

[janaal1]

Hi there,

I cannot run the baseline model because files download checking is not correct.

My code is the following one:

db = DCASE2018_Task4_DevelopmentSet(included_content_types=['all'],
                                        local_path="/path/to/folder/where/dataset/is",
                                        data_path=param.get_path('path.dataset'),
                                        audio_paths=[
                                            os.path.join("dataset", "audio", "train", "weak"),
                                            os.path.join("dataset", "audio", "train", "unlabel_in_domain"),
                                            os.path.join("dataset", "audio", "train", "unlabel_out_of_domain"),
                                            os.path.join("dataset", "audio", "test"),
                                            os.path.join("dataset", "audio", "eval")
                                        ]
                                        ).initialize()

I get the following messages: FOLD[1] EVAL SET Empty set
FOLD[1] SETS Test and eval sets have different amount of files [13796 != 0]. FOLD[1] SETS Test and eval sets have different files [0 != 13796 files common]. FOLD[2] SETS Test and eval sets have different amount of files [275 != 829]. FOLD[2] SETS Test and eval sets have different files [0 != 275 files common]. (datasets.py:1304) NoneType: None Dataset errors: FOLD[1] EVAL SET Empty set
FOLD[1] SETS Test and eval sets have different amount of files [13796 != 0]. FOLD[1] SETS Test and eval sets have different files [0 != 13796 files common]. FOLD[2] SETS Test and eval sets have different amount of files [275 != 829]. FOLD[2] SETS Test and eval sets have different files [0 != 275 files common].

It seems that is not searching the files in the correct path.... Any idea how to solve this issue? dcase_util: 0.2.5 Thanks

[YapengTian]

Same issue. Any updates?

[turpaultn]

I am sorry for the inconvenience, dcase_util has been updated, and this code is not running with the new version. Indeed, I have the same issue when running with dcase_util version 0.2.5. Downgrade to dcase_util 0.2.4 will do the trick for now.

If you just want to use the dataset, dcase2019_task4 challenge will provide a Dataset class where you'll be able to use dcase2018 data.

[turpaultn]

Duplicate of #21