Datasets:

google-research-datasets

/

conceptual_captions

like 59

Thanks for reporting. We are investigating it.

The dataset loads perfectly on our side:

In [1]: from datasets import load_dataset

In [2]: ds = load_dataset("conceptual_captions")

In [3]: ds
Out[3]: 
DatasetDict({
    train: Dataset({
        features: ['image_url', 'caption'],
        num_rows: 3318333
    })
    validation: Dataset({
        features: ['image_url', 'caption'],
        num_rows: 15840
    })
})

Hi!

My apologies, I believe I have similar issue - I'm using a Kaggle notebook and I receive this following error:

"FileNotFoundError: Couldn't find a dataset script at /kaggle/working/conceptual_captions/conceptual_captions.py or any data file in the same directory. Couldn't find 'conceptual_captions' on the Hugging Face Hub either: FileNotFoundError: Couldn't find file at https://raw.githubusercontent.com/huggingface/datasets/master/datasets/conceptual_captions/conceptual_captions.py"

Could help with this issue?
Thank you!

I'm afraid , i can't. I tried to load the dataset from kaggle and i got this error but outside of kaggle it seems to be working fine . The problem is even when i upgrade the dataset library it don't work on Kaggle .
I used a version heberged on kaggle

I guess the datasets version should be updated: it is long ago that we no longer host the dataset loading scripts on GitHub, as your error message suggests...

FileNotFoundError: Couldn't find file at https://raw.githubusercontent.com/huggingface/datasets/master/datasets/conceptual_captions/conceptual_captions.py