Closed blacksunshine closed 3 years ago
Can you please provide steps to reproduce the issue? Preferably a small snippet to generate a toy dataset that results in the failure and a code to reproduce the failure itself. Thank you.
Hi,Selitvin
thank you for your reply.
My original data is not a directory, but a parquet file, it has no partitions.It will go wrong in this scenario.
from petastorm import make_reader
from petastorm.tf_utils import make_petastorm_dataset
import tensorflow as tf
print('1')
with make_reader('file:///my.zstd.parquet') as reader:
dataset = make_petastorm_dataset(reader)
iterator = dataset.make_one_shot_iterator()
tensor = iterator.get_next()
with tf.Session() as sess:
sample = sess.run(tensor)
print(sample.id)
Got it. Thank you for your report. This should fix the issue: https://github.com/uber/petastorm/pull/687
You can try it with
pip3 install git+https://github.com/selitvin/petastorm@issue_684
Hi,Selitvin Thank you so much. I will try it. I am a newcomer, I like open source, and I also like to write code. I hope that if there are problems later, I can also fix them so that I can help you solve some minor problems.
https://github.com/uber/petastorm#how-to-contribute I saw this, and thank you very much for your reply. Start my petastorm journey.
this issue was solved by @selitvin selitvin,i will close it.
https://github.com/uber/petastorm/blob/master/petastorm/py_dict_reader_worker.py#L177