This dataset is no longer available at this url, so I had to manually download it. Can this be fixed please?
If you refer to the source code (here), when you use untar_data()
to download the target url, the target url will be passed as the first argument of the following function:
def download_data(url:str, fname:PathOrStr=None, data:bool=True, ext:str='.tgz') -> Path:
"Download `url` to destination `fname`."
fname = Path(ifnone(fname, _url2tgz(url, data, ext=ext)))
os.makedirs(fname.parent, exist_ok=True)
if not fname.exists():
print(f'Downloading {url}')
download_url(f'{url}{ext}', fname)
return fname
The line download_url(f'{url}{ext}', fname)
hints that http://files.fast.ai/data/examples/imdb_sample.tgz is used to download the data (note that extension .tgz is appended at the end). Now the final url works fine!