Is it possible to access and read the csv files of training data labels?
or we only have the access to training_dataset class?
what’s the filepath to access it? is it just ./data/training/images/labels.csv just like in the notebook example?
As per my knowledge:
You can access the csv files for the training dataset, (guess what not just for training but for all the provided datasets). The file path for those are ./data/DATASET_TYPE/labels.csv
, here DATASET_TYPE
can be debug, training, unlabelled, validation
.
HOWEVER, you should not use those in ZEWDPCBaseRun
class, because the evaluation will use this class’s functions the way they are defined, which means you’ll still have to use the dataset class to access the dataset. Ofcourse you can use the labels.csv
for the training
dataset but for sure not for others.
If you need help with iterating over the dataset please do let me know here.
If those other csv files are accessible and can potentially improve the model, you’d need to check the code of the winner that they don’t do this, right?