what's the right way to load this dataset now?
I'm using:
dataset = load_dataset("openwebtext", trust_remote_code=True, encoding='utf-8')
But I got:
Traceback (most recent call last):
File "/usr/local/home/windmaple/Desktop/project-src/nanoGPT/data/openwebtext/download.py", line 3, in
dataset = load_dataset("openwebtext", trust_remote_code=True, encoding='utf-8')
File "/usr/local/home/windmaple/miniconda3/envs/python3.10/lib/python3.10/site-packages/datasets/load.py", line 2076, in load_dataset
builder_instance = load_dataset_builder(
File "/usr/local/home/windmaple/miniconda3/envs/python3.10/lib/python3.10/site-packages/datasets/load.py", line 1834, in load_dataset_builder
builder_instance: DatasetBuilder = builder_cls(
TypeError: 'NoneType' object is not callable
No idea what's going on there ...
datasets.version
'3.0.1'
I met the same problem, I would like to ask you how you finally solved it
I ended up using https://huggingface.co/datasets/vietgpt/openwebtext_en, which seems to be the right mirror
https://huggingface.co/datasets/Skylion007/openwebtext/tree/convert-parquet-full Try this branch or downgrade to <3. I am doing a full upgrade of the dataset soon.
https://huggingface.co/datasets/Skylion007/openwebtext/discussions/18 should close this