title
stringlengths 1
290
| body
stringlengths 0
228k
β | html_url
stringlengths 46
51
| comments
list | pull_request
dict | number
int64 1
5.59k
| is_pull_request
bool 2
classes |
|---|---|---|---|---|---|---|
Add SEDE dataset
|
This PR adds the SEDE dataset for the task of realistic Text-to-SQL, following the instructions of how to add a database and a dataset card.
Please see our paper for more details: https://arxiv.org/abs/2106.05006
|
https://github.com/huggingface/datasets/pull/2942
|
[
"Thanks @albertvillanova for your great suggestions! I just pushed a new commit with the necessary fixes. For some reason, the test `test_metric_common` failed for `meteor` metric, which doesn't have any connection to this PR, so I'm trying to rebase and see if it helps.",
"Hi @Hazoom,\r\n\r\nYou were right: the non-passing test had nothing to do with this PR.\r\n\r\nUnfortunately, you did a git rebase (instead of a git merge), which is not recommended once you have already opened a pull request because you mess up your pull request history. You can see that your pull request now contains:\r\n- your commits repeated two times\r\n- and commits which are not yours from the master branch\r\n\r\nIf you would like to clean your pull request, please make:\r\n```\r\ngit reset --hard 587b93a\r\ngit fetch upstream master\r\ngit merge upstream/master\r\ngit push --force origin sede\r\n```",
"> Hi @Hazoom,\r\n> \r\n> You were right: the non-passing test had nothing to do with this PR.\r\n> \r\n> Unfortunately, you did a git rebase (instead of a git merge), which is not recommended once you have already opened a pull request because you mess up your pull request history. You can see that your pull request now contains:\r\n> \r\n> * your commits repeated two times\r\n> * and commits which are not yours from the master branch\r\n> \r\n> If you would like to clean your pull request, please make:\r\n> \r\n> ```\r\n> git reset --hard 587b93a\r\n> git fetch upstream master\r\n> git merge upstream/master\r\n> git push --force origin sede\r\n> ```\r\n\r\nThanks @albertvillanova ",
"> Nice! Just one final request before approving your pull request:\r\n> \r\n> As you have updated the \"QuerySetId\" field data type, the size of the dataset is smaller now. You should regenerate the metadata. Please run:\r\n> \r\n> ```\r\n> rm datasets/sede/dataset_infos.json\r\n> datasets-cli test datasets/sede --save_infos --all_configs\r\n> ```\r\n\r\n@albertvillanova Good catch, just fixed it."
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2942",
"html_url": "https://github.com/huggingface/datasets/pull/2942",
"diff_url": "https://github.com/huggingface/datasets/pull/2942.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2942.patch",
"merged_at": "2021-09-24T10:39:54"
}
| 2,942
| true
|
OSCAR unshuffled_original_ko: NonMatchingSplitsSizesError
|
## Describe the bug
Cannot download OSCAR `unshuffled_original_ko` due to `NonMatchingSplitsSizesError`.
## Steps to reproduce the bug
```python
>>> dataset = datasets.load_dataset('oscar', 'unshuffled_original_ko')
NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=25292102197, num_examples=7345075, dataset_name='oscar'), 'recorded': SplitInfo(name='train', num_bytes=25284578514, num_examples=7344907, dataset_name='oscar')}]
```
## Expected results
Loading is successful.
## Actual results
Loading throws above error.
## Environment info
- `datasets` version: 1.12.1
- Platform: Linux-5.4.0-81-generic-x86_64-with-glibc2.29
- Python version: 3.8.10
- PyArrow version: 5.0.0
|
https://github.com/huggingface/datasets/issues/2941
|
[
"I tried `unshuffled_original_da` and it is also not working"
] | null | 2,941
| false
|
add swedish_medical_ner dataset
|
Adding the Swedish Medical NER dataset, listed in "Biomedical Datasets - BigScience Workshop 2021"
|
https://github.com/huggingface/datasets/pull/2940
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2940",
"html_url": "https://github.com/huggingface/datasets/pull/2940",
"diff_url": "https://github.com/huggingface/datasets/pull/2940.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2940.patch",
"merged_at": "2021-10-05T12:13:33"
}
| 2,940
| true
|
MENYO-20k repo has moved, updating URL
|
Dataset repo moved to https://github.com/uds-lsv/menyo-20k_MT, now editing URL to match.
https://github.com/uds-lsv/menyo-20k_MT/blob/master/data/train.tsv is the file we're looking for
|
https://github.com/huggingface/datasets/pull/2939
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2939",
"html_url": "https://github.com/huggingface/datasets/pull/2939",
"diff_url": "https://github.com/huggingface/datasets/pull/2939.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2939.patch",
"merged_at": "2021-09-21T15:31:36"
}
| 2,939
| true
|
Take namespace into account in caching
|
Loading a dataset "username/dataset_name" hosted by a user on the hub used to cache the dataset only taking into account the dataset name, and ignorign the username. Because of this, if a user later loads "dataset_name" without specifying the username, it would reload the dataset from the cache instead of failing.
I changed the dataset cache and module cache mechanism to include the username in the name of the cache directory that is used:
<s>
`~/.cache/huggingface/datasets/username/dataset_name` for the data
`~/.cache/huggingface/modules/datasets_modules/datasets/username/dataset_name` for the python files
</s>
EDIT: actually using three underscores:
`~/.cache/huggingface/datasets/username___dataset_name` for the data
`~/.cache/huggingface/modules/datasets_modules/datasets/username___dataset_name` for the python files
This PR should fix the issue https://github.com/huggingface/datasets/issues/2842
cc @stas00
|
https://github.com/huggingface/datasets/pull/2938
|
[
"We might have collisions if a username and a dataset_name are the same. Maybe instead serialize the dataset name by replacing `/` with some string, eg `__SLASH__`, that will hopefully never appear in a dataset or user name (it's what I did in https://github.com/huggingface/datasets-preview-backend/blob/master/benchmark/scripts/serialize.py. That way, all the datasets are one-level deep directories",
"IIRC we enforce that no repo id or username can contain `___` (exactly 3 underscores) specifically for this reason, so you can use that string (that we use in other projects)\r\n\r\ncc @Pierrci ",
"> IIRC we enforce that no repo id or username can contain ___ (exactly 3 underscores) specifically for this reason, so you can use that string (that we use in other projects)\r\n\r\nout of curiosity: where is it enforced?",
"> where is it enforced?\r\n\r\nNowhere yet but we should :) feel free to track in internal tracker and/or implement, as this will be useful in the future",
"Thanks for the trick, I'm doing the change :)\r\nWe can use\r\n`~/.cache/huggingface/datasets/username___dataset_name` for the data\r\n`~/.cache/huggingface/modules/datasets_modules/datasets/username___dataset_name` for the python files",
"Merging, though it will have to be integrated again the refactor at #2986",
"@lhoestq we changed a bit the naming policy on the Hub, and the substring '--' is now forbidden, which makes it available for serializing the repo names (`namespace/repo` -> `namespace--repo`). See https://github.com/huggingface/moon-landing/pull/1657 and https://github.com/huggingface/huggingface_hub/pull/545"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2938",
"html_url": "https://github.com/huggingface/datasets/pull/2938",
"diff_url": "https://github.com/huggingface/datasets/pull/2938.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2938.patch",
"merged_at": "2021-09-29T13:01:31"
}
| 2,938
| true
|
load_dataset using default cache on Windows causes PermissionError: [WinError 5] Access is denied
|
## Describe the bug
Standard process to download and load the wiki_bio dataset causes PermissionError in Windows 10 and 11.
## Steps to reproduce the bug
```python
from datasets import load_dataset
ds = load_dataset('wiki_bio')
```
## Expected results
It is expected that the dataset downloads without any errors.
## Actual results
PermissionError see trace below:
```
Using custom data configuration default
Downloading and preparing dataset wiki_bio/default (download: 318.53 MiB, generated: 736.94 MiB, post-processed: Unknown size, total: 1.03 GiB) to C:\Users\username\.cache\huggingface\datasets\wiki_bio\default\1.1.0\5293ce565954ba965dada626f1e79684e98172d950371d266bf3caaf87e911c9...
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "C:\Users\username\.conda\envs\hf\lib\site-packages\datasets\load.py", line 1112, in load_dataset
builder_instance.download_and_prepare(
File "C:\Users\username\.conda\envs\hf\lib\site-packages\datasets\builder.py", line 644, in download_and_prepare
self._save_info()
File "C:\Users\username\.conda\envs\hf\lib\contextlib.py", line 120, in __exit__
next(self.gen)
File "C:\Users\username\.conda\envs\hf\lib\site-packages\datasets\builder.py", line 598, in incomplete_dir
os.rename(tmp_dir, dirname)
PermissionError: [WinError 5] Access is denied: 'C:\\Users\\username\\.cache\\huggingface\\datasets\\wiki_bio\\default\\1.1.0\\5293ce565954ba965dada626f1e79684e98172d950371d266bf3caaf87e911c9.incomplete' -> 'C:\\Users\\username\\.cache\\huggingface\\datasets\\wiki_bio\\default\\1.1.0\\5293ce565954ba965dada626f1e79684e98172d950371d266bf3caaf87e911c9'
```
By commenting out the os.rename() [L604](https://github.com/huggingface/datasets/blob/master/src/datasets/builder.py#L604) and the shutil.rmtree() [L607](https://github.com/huggingface/datasets/blob/master/src/datasets/builder.py#L607) lines, in my virtual environment, I was able to get the load process to complete, rename the directory manually and then rerun the `load_dataset('wiki_bio')` to get what I needed.
It seems that os.rename() in the `incomplete_dir` content manager is the culprit. Here's another project [Conan](https://github.com/conan-io/conan/issues/6560) with similar issue with os.rename() if it helps debug this issue.
## Environment info
- `datasets` version: 1.12.1
- Platform: Windows-10-10.0.22449-SP0
- Python version: 3.8.12
- PyArrow version: 5.0.0
|
https://github.com/huggingface/datasets/issues/2937
|
[
"Hi @daqieq, thanks for reporting.\r\n\r\nUnfortunately, I was not able to reproduce this bug:\r\n```ipython\r\nIn [1]: from datasets import load_dataset\r\n ...: ds = load_dataset('wiki_bio')\r\nDownloading: 7.58kB [00:00, 26.3kB/s]\r\nDownloading: 2.71kB [00:00, ?B/s]\r\nUsing custom data configuration default\r\nDownloading and preparing dataset wiki_bio/default (download: 318.53 MiB, generated: 736.94 MiB, post-processed: Unknown size, total: 1.03 GiB) to C:\\Users\\username\\.cache\\huggingface\\datasets\\wiki_bio\\default\\\r\n1.1.0\\5293ce565954ba965dada626f1e79684e98172d950371d266bf3caaf87e911c9...\r\nDownloading: 334MB [01:17, 4.32MB/s]\r\nDataset wiki_bio downloaded and prepared to C:\\Users\\username\\.cache\\huggingface\\datasets\\wiki_bio\\default\\1.1.0\\5293ce565954ba965dada626f1e79684e98172d950371d266bf3caaf87e911c9. Subsequent calls will reuse thi\r\ns data.\r\n```\r\n\r\nThis kind of error messages usually happen because:\r\n- Your running Python script hasn't write access to that directory\r\n- You have another program (the File Explorer?) already browsing inside that directory",
"Thanks @albertvillanova for looking at it! I tried on my personal Windows machine and it downloaded just fine.\r\n\r\nRunning on my work machine and on a colleague's machine it is consistently hitting this error. It's not a write access issue because the `.incomplete` directory is written just fine. It just won't rename and then it deletes the directory in the `finally` step. Also the zip file is written and extracted fine in the downloads directory.\r\n\r\nThat leaves another program that might be interfering, and there are plenty of those in my work machine ... (full antivirus, data loss prevention, etc.). So the question remains, why not extend the `try` block to allow catching the error and circle back to the rename after the unknown program is finished doing its 'stuff'. This is the approach that I read about in the linked repo (see my comments above).\r\n\r\nIf it's not high priority, that's fine. However, if someone were to write an PR that solved this issue in our environment in an `except` clause, would it be reviewed for inclusion in a future release? Just wondering whether I should spend any more time on this issue.",
"Hi @albertvillanova, even I am facing the same issue on my work machine:\r\n\r\n`Downloading and preparing dataset json/c4-en-html-with-metadata to C:\\Users\\......\\.cache\\huggingface\\datasets\\json\\c4-en-html-with-metadata-4635c2fd9249f62d\\0.0.0\\c90812beea906fcffe0d5e3bb9eba909a80a998b5f88e9f8acbd320aa91acfde...\r\n100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 1/1 [00:00<00:00, 983.42it/s]\r\n100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 1/1 [00:00<00:00, 209.01it/s]\r\nTraceback (most recent call last):\r\n File \"bsmetadata/preprocessing_utils.py\", line 710, in <module>\r\n ds = load_dataset(\r\n File \"C:\\Users\\.......\\AppData\\Roaming\\Python\\Python38\\site-packages\\datasets\\load.py\", line 1694, in load_dataset\r\n builder_instance.download_and_prepare(\r\n File \"C:\\Users\\........\\AppData\\Roaming\\Python\\Python38\\site-packages\\datasets\\builder.py\", line 603, in download_and_prepare\r\n self._save_info()\r\n File \"C:\\Users\\..........\\AppData\\Local\\Programs\\Python\\Python38\\lib\\contextlib.py\", line 120, in __exit__\r\n next(self.gen)\r\n File \"C:\\Users\\.....\\AppData\\Roaming\\Python\\Python38\\site-packages\\datasets\\builder.py\", line 557, in incomplete_dir\r\n os.rename(tmp_dir, dirname)\r\nPermissionError: [WinError 5] Access is denied: 'C:\\\\Users\\\\.........\\\\.cache\\\\huggingface\\\\datasets\\\\json\\\\c4-en-html-with-metadata-4635c2fd9249f62d\\\\0.0.0\\\\c90812beea906fcffe0d5e3bb9eba909a80a998b5f88e9f8acbd320aa91acfde.incomplete' -> 'C:\\\\Users\\\\I355109\\\\.cache\\\\huggingface\\\\datasets\\\\json\\\\c4-en-html-with-metadata-4635c2fd9249f62d\\\\0.0.0\\\\c90812beea906fcffe0d5e3bb9eba909a80a998b5f88e9f8acbd320aa91acfde'`",
"I'm facing the same issue.\r\n\r\n## System Information\r\n\r\n- OS Edition: Windows 10 21H1\r\n- OS build: 19043.1826\r\n- Python version: 3.10.6 (installed using `choco install python`)\r\n- datasets: 2.4.0\r\n- PyArrow: 6.0.1\r\n\r\n## Troubleshooting steps\r\n\r\n- Restart the computer, unfortunately doesn't work! π\r\n- Checked the permissions of `~./cache/...`, looks fine.\r\n- Tested with a simple file operation using the `open()` function and writing a hello_world.txt, it works fine.\r\n- Tested with a different `cache_dir` value on the `load_dataset()`, e.g. \"./data\"\r\n- Tested different datasets: `conll2003`, `squad_v2`, and `wiki_bio`.\r\n- Downgraded datasets from `2.4.0` to `2.1.0`, issue persists.\r\n- Tested it on WSL (Ubuntu 20.04), and it works! \r\n- Python reinstallation, in the first time downloading `conll2003` works fine, but `squad` or `squad_v2` raises Access Denied.\r\n - After the system or VSCode restart, the issue comes back.\r\n\r\n## Resolution\r\n\r\nI fixed it by changing the following command:\r\n\r\nhttps://github.com/huggingface/datasets/blob/68cffe30917a9abed68d28caf54b40c10f977602/src/datasets/builder.py#L666\r\n\r\nfor\r\n\r\n```python\r\nshutil.move(tmp_dir, dirname)\r\n```"
] | null | 2,937
| false
|
Check that array is not Float as nan != nan
|
The Exception wants to check for issues with StructArrays/ListArrays but catches FloatArrays with value nan as nan != nan.
Pass on FloatArrays as we should not raise an Exception for them.
|
https://github.com/huggingface/datasets/pull/2936
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2936",
"html_url": "https://github.com/huggingface/datasets/pull/2936",
"diff_url": "https://github.com/huggingface/datasets/pull/2936.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2936.patch",
"merged_at": "2021-09-21T09:39:04"
}
| 2,936
| true
|
Add Jigsaw unintended Bias
|
Hi,
Here's a first attempt at this dataset. Would be great if it could be merged relatively quickly as it is needed for Bigscience-related stuff.
This requires manual download, and I had some trouble generating dummy_data in this setting, so welcoming feedback there.
|
https://github.com/huggingface/datasets/pull/2935
|
[
"Note that the tests seem to fail because of a bug in an Exception at the moment, see: https://github.com/huggingface/datasets/pull/2936 for the fix",
"@lhoestq implemented your changes, I think this might be ready for another look.",
"Thanks @lhoestq, implemented the changes, let me know if anything else pops up."
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2935",
"html_url": "https://github.com/huggingface/datasets/pull/2935",
"diff_url": "https://github.com/huggingface/datasets/pull/2935.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2935.patch",
"merged_at": "2021-09-24T10:41:52"
}
| 2,935
| true
|
to_tf_dataset keeps a reference to the open data somewhere, causing issues on windows
|
To reproduce:
```python
import datasets as ds
import weakref
import gc
d = ds.load_dataset("mnist", split="train")
ref = weakref.ref(d._data.table)
tfd = d.to_tf_dataset("image", batch_size=1, shuffle=False, label_cols="label")
del tfd, d
gc.collect()
assert ref() is None, "Error: there is at least one reference left"
```
This causes issues because the table holds a reference to an open arrow file that should be closed. So on windows it's not possible to delete or move the arrow file afterwards.
Moreover the CI test of the `to_tf_dataset` method isn't able to clean up the temporary arrow files because of this.
cc @Rocketknight1
|
https://github.com/huggingface/datasets/issues/2934
|
[
"I did some investigation and, as it seems, the bug stems from [this line](https://github.com/huggingface/datasets/blob/8004d7c3e1d74b29c3e5b0d1660331cd26758363/src/datasets/arrow_dataset.py#L325). The lifecycle of the dataset from the linked line is bound to one of the returned `tf.data.Dataset`. So my (hacky) solution involves wrapping the linked dataset with `weakref.proxy` and adding a custom `__del__` to `tf.python.data.ops.dataset_ops.TensorSliceDataset` (this is the type of a dataset that is returned by `tf.data.Dataset.from_tensor_slices`; this works for TF 2.x, but I'm not sure `tf.python.data.ops.dataset_ops` is a valid path for TF 1.x) that deletes the linked dataset, which is assigned to the dataset object as a property. Will open a draft PR soon!",
"Thanks a lot for investigating !"
] | null | 2,934
| false
|
Replace script_version with revision
|
As discussed in https://github.com/huggingface/datasets/pull/2718#discussion_r707013278, the parameter name `script_version` is no longer applicable to datasets without loading script (i.e., datasets only with raw data files).
This PR replaces the parameter name `script_version` with `revision`.
This way, we are also aligned with:
- Transformers: `AutoTokenizer.from_pretrained(..., revision=...)`
- Hub: `HfApi.dataset_info(..., revision=...)`, `HfApi.upload_file(..., revision=...)`
|
https://github.com/huggingface/datasets/pull/2933
|
[
"I'm also fine with the removal in 1.15"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2933",
"html_url": "https://github.com/huggingface/datasets/pull/2933",
"diff_url": "https://github.com/huggingface/datasets/pull/2933.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2933.patch",
"merged_at": "2021-09-20T09:52:10"
}
| 2,933
| true
|
Conda build fails
|
## Describe the bug
Current `datasets` version in conda is 1.9 instead of 1.12.
The build of the conda package fails.
|
https://github.com/huggingface/datasets/issues/2932
|
[
"Why 1.9 ?\r\n\r\nhttps://anaconda.org/HuggingFace/datasets currently says 1.11",
"Alright I added 1.12.0 and 1.12.1 and fixed the conda build #2952 "
] | null | 2,932
| false
|
Fix bug in to_tf_dataset
|
Replace `set_format()` to `with_format()` so that we don't alter the original dataset in `to_tf_dataset()`
|
https://github.com/huggingface/datasets/pull/2931
|
[
"I'm going to merge it, but yeah - hopefully the CI runner just cleans that up automatically and few other people run the tests on Windows anyway!"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2931",
"html_url": "https://github.com/huggingface/datasets/pull/2931",
"diff_url": "https://github.com/huggingface/datasets/pull/2931.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2931.patch",
"merged_at": "2021-09-16T17:01:37"
}
| 2,931
| true
|
Mutable columns argument breaks set_format
|
## Describe the bug
If you pass a mutable list to the `columns` argument of `set_format` and then change the list afterwards, the returned columns also change.
## Steps to reproduce the bug
```python
from datasets import load_dataset
dataset = load_dataset("glue", "cola")
column_list = ["idx", "label"]
dataset.set_format("python", columns=column_list)
column_list[1] = "foo" # Change the list after we call `set_format`
dataset['train'][:4].keys()
```
## Expected results
```python
dict_keys(['idx', 'label'])
```
## Actual results
```python
dict_keys(['idx'])
```
|
https://github.com/huggingface/datasets/issues/2930
|
[
"Pushed a fix to my branch #2731 "
] | null | 2,930
| false
|
Add regression test for null Sequence
|
Relates to #2892 and #2900.
|
https://github.com/huggingface/datasets/pull/2929
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2929",
"html_url": "https://github.com/huggingface/datasets/pull/2929",
"diff_url": "https://github.com/huggingface/datasets/pull/2929.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2929.patch",
"merged_at": "2021-09-17T08:23:59"
}
| 2,929
| true
|
Update BibTeX entry
|
Update BibTeX entry.
|
https://github.com/huggingface/datasets/pull/2928
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2928",
"html_url": "https://github.com/huggingface/datasets/pull/2928",
"diff_url": "https://github.com/huggingface/datasets/pull/2928.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2928.patch",
"merged_at": "2021-09-16T12:35:34"
}
| 2,928
| true
|
Datasets 1.12 dataset.filter TypeError: get_indices_from_mask_function() got an unexpected keyword argument
|
## Describe the bug
Upgrading to 1.12 caused `dataset.filter` call to fail with
> get_indices_from_mask_function() got an unexpected keyword argument valid_rel_labels
## Steps to reproduce the bug
```pythondef
filter_good_rows(
ex: Dict,
valid_rel_labels: Set[str],
valid_ner_labels: Set[str],
tokenizer: PreTrainedTokenizerFast,
) -> bool:
"""Get the good rows"""
encoding = get_encoding_for_text(text=ex["text"], tokenizer=tokenizer)
ex["encoding"] = encoding
for relation in ex["relations"]:
if not is_valid_relation(relation, valid_rel_labels):
return False
for span in ex["spans"]:
if not is_valid_span(span, valid_ner_labels, encoding):
return False
return True
def get_dataset():
loader_path = str(Path(__file__).parent / "prodigy_dataset_builder.py")
ds = load_dataset(
loader_path,
name="prodigy-dataset",
data_files=sorted(file_paths),
cache_dir=cache_dir,
)["train"]
valid_ner_labels = set(vocab.ner_category)
valid_relations = set(vocab.relation_types.keys())
ds = ds.filter(
filter_good_rows,
fn_kwargs=dict(
valid_rel_labels=valid_relations,
valid_ner_labels=valid_ner_labels,
tokenizer=vocab.tokenizer,
),
keep_in_memory=True,
num_proc=num_proc,
)
```
`ds` is a `DatasetDict` produced by a jsonl dataset.
This runs fine on 1.11 but fails on 1.12
**Stack Trace**
## Expected results
I expect 1.12 datasets filter to filter the dataset without raising as it does on 1.11
## Actual results
```
tf_ner_rel_lib/dataset.py:695: in load_prodigy_arrow_datasets_from_jsonl
ds = ds.filter(
../../../../.pyenv/versions/tf_ner_rel_lib/lib/python3.8/site-packages/datasets/arrow_dataset.py:185: in wrapper
out: Union["Dataset", "DatasetDict"] = func(self, *args, **kwargs)
../../../../.pyenv/versions/tf_ner_rel_lib/lib/python3.8/site-packages/datasets/fingerprint.py:398: in wrapper
out = func(self, *args, **kwargs)
../../../../.pyenv/versions/tf_ner_rel_lib/lib/python3.8/site-packages/datasets/arrow_dataset.py:2169: in filter
indices = self.map(
../../../../.pyenv/versions/tf_ner_rel_lib/lib/python3.8/site-packages/datasets/arrow_dataset.py:1686: in map
return self._map_single(
../../../../.pyenv/versions/tf_ner_rel_lib/lib/python3.8/site-packages/datasets/arrow_dataset.py:185: in wrapper
out: Union["Dataset", "DatasetDict"] = func(self, *args, **kwargs)
../../../../.pyenv/versions/tf_ner_rel_lib/lib/python3.8/site-packages/datasets/fingerprint.py:398: in wrapper
out = func(self, *args, **kwargs)
../../../../.pyenv/versions/tf_ner_rel_lib/lib/python3.8/site-packages/datasets/arrow_dataset.py:2048: in _map_single
batch = apply_function_on_filtered_inputs(
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
inputs = {'_input_hash': [2108817714, 1477695082, -1021597032, 2130671338, -1260483858, -1203431639, ...], '_task_hash': [18070...ons', 'relations', 'relations', ...], 'answer': ['accept', 'accept', 'accept', 'accept', 'accept', 'accept', ...], ...}
indices = [0, 1, 2, 3, 4, 5, ...], check_same_num_examples = False, offset = 0
def apply_function_on_filtered_inputs(inputs, indices, check_same_num_examples=False, offset=0):
"""Utility to apply the function on a selection of columns."""
nonlocal update_data
fn_args = [inputs] if input_columns is None else [inputs[col] for col in input_columns]
if offset == 0:
effective_indices = indices
else:
effective_indices = [i + offset for i in indices] if isinstance(indices, list) else indices + offset
processed_inputs = (
> function(*fn_args, effective_indices, **fn_kwargs) if with_indices else function(*fn_args, **fn_kwargs)
)
E TypeError: get_indices_from_mask_function() got an unexpected keyword argument 'valid_rel_labels'
../../../../.pyenv/versions/tf_ner_rel_lib/lib/python3.8/site-packages/datasets/arrow_dataset.py:1939: TypeError
```
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version: 1.12.1
- Platform: Mac
- Python version: 3.8.9
- PyArrow version: pyarrow==5.0.0
|
https://github.com/huggingface/datasets/issues/2927
|
[
"Thanks for reporting, I'm looking into it :)",
"Fixed by #2950."
] | null | 2,927
| false
|
Error when downloading datasets to non-traditional cache directories
|
## Describe the bug
When the cache directory is linked (soft link) to a directory on a NetApp device, the download fails.
## Steps to reproduce the bug
```bash
ln -s /path/to/netapp/.cache ~/.cache
```
```python
load_dataset("imdb")
```
## Expected results
Successfully loading IMDB dataset
## Actual results
```
datasets.utils.info_utils.NonMatchingSplitsSizesError: [{'expected': SplitInfo(name='train', num_bytes=33432835,
num_examples=25000, dataset_name='imdb'), 'recorded': SplitInfo(name='train', num_bytes=0, num_examples=0,
dataset_name='imdb')}, {'expected': SplitInfo(name='test', num_bytes=32650697, num_examples=25000, dataset_name='imdb'),
'recorded': SplitInfo(name='test', num_bytes=659932, num_examples=503, dataset_name='imdb')}, {'expected':
SplitInfo(name='unsupervised', num_bytes=67106814, num_examples=50000, dataset_name='imdb'), 'recorded':
SplitInfo(name='unsupervised', num_bytes=0, num_examples=0, dataset_name='imdb')}]
```
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version: 1.1.2
- Platform: Ubuntu
- Python version: 3.8
## Extra notes
Stranger yet, trying to debug the phenomenon, I found the range of results to vary a lot without clear direction:
- With `cache_dir="/path/to/netapp/.cache"` the same thing happens.
- However, when linking `~/netapp/` to `/path/to/netapp` *and* setting `cache_dir="~/netapp/.cache/huggingface/datasets"` - it does work
- On the other hand, when linking `~/.cache` to `~/netapp/.cache` without using `cache_dir`, it does work anymore.
While I could test it only for a NetApp device, it might have to do with any other mounted FS.
Thanks :)
|
https://github.com/huggingface/datasets/issues/2926
|
[
"Same here !"
] | null | 2,926
| false
|
Add tutorial for no-code dataset upload
|
This PR is for a tutorial for uploading a dataset to the Hub. It relies on the Hub UI elements to upload a dataset, introduces the online tagging tool for creating tags, and the Dataset card template to get a head start on filling it out. The addition of this tutorial should make it easier for beginners to upload a dataset without accessing the terminal or knowing Git.
|
https://github.com/huggingface/datasets/pull/2925
|
[
"Cool, love it ! :)\r\n\r\nFeel free to add a paragraph saying how to load the dataset:\r\n```python\r\nfrom datasets import load_dataset\r\n\r\ndataset = load_dataset(\"stevhliu/demo\")\r\n\r\n# or to separate each csv file into several splits\r\ndata_files = {\"train\": \"train.csv\", \"test\": \"test.csv\"}\r\ndataset = load_dataset(\"stevhliu/demo\", data_files=data_files)\r\nprint(dataset[\"train\"][0])\r\n```",
"Perfect, feel free to mark this PR ready for review :)\r\n\r\ncc @albertvillanova do you have any comment ? You can check the tutorial here:\r\nhttps://47389-250213286-gh.circle-artifacts.com/0/docs/_build/html/no_code_upload.html\r\n\r\nMaybe we can just add a list of supported file types:\r\n- csv\r\n- json\r\n- json lines\r\n- text\r\n- parquet",
"I just added a mention of the login for private datasets. Don't hesitate to edit or comment.\r\n\r\nOtherwise I think it's all good, feel free to merge it @stevhliu if you don't have other changes to make :)"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2925",
"html_url": "https://github.com/huggingface/datasets/pull/2925",
"diff_url": "https://github.com/huggingface/datasets/pull/2925.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2925.patch",
"merged_at": "2021-09-27T17:51:55"
}
| 2,925
| true
|
"File name too long" error for file locks
|
## Describe the bug
Getting the following error when calling `load_dataset("gar1t/test")`:
```
OSError: [Errno 36] File name too long: '<user>/.cache/huggingface/datasets/_home_garrett_.cache_huggingface_datasets_csv_test-7c856aea083a7043_0.0.0_9144e0a4e8435090117cea53e6c7537173ef2304525df4a077c435d8ee7828ff.incomplete.lock'
```
## Steps to reproduce the bug
Where the user cache dir (e.g. `~/.cache`) is on a file system that limits filenames to 255 chars (e.g. ext4):
```python
from datasets import load_dataset
load_dataset("gar1t/test")
```
## Expected results
Expect the function to return without an error.
## Actual results
```
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "<python_venv>/lib/python3.9/site-packages/datasets/load.py", line 1112, in load_dataset
builder_instance.download_and_prepare(
File "<python_venv>/lib/python3.9/site-packages/datasets/builder.py", line 644, in download_and_prepare
self._save_info()
File "<python_venv>/lib/python3.9/site-packages/datasets/builder.py", line 765, in _save_info
with FileLock(lock_path):
File "<python_venv>/lib/python3.9/site-packages/datasets/utils/filelock.py", line 323, in __enter__
self.acquire()
File "<python_venv>/lib/python3.9/site-packages/datasets/utils/filelock.py", line 272, in acquire
self._acquire()
File "<python_venv>/lib/python3.9/site-packages/datasets/utils/filelock.py", line 403, in _acquire
fd = os.open(self._lock_file, open_mode)
OSError: [Errno 36] File name too long: '<user>/.cache/huggingface/datasets/_home_garrett_.cache_huggingface_datasets_csv_test-7c856aea083a7043_0.0.0_9144e0a4e8435090117cea53e6c7537173ef2304525df4a077c435d8ee7828ff.incomplete.lock'
```
## Environment info
- `datasets` version: 1.12.1
- Platform: Linux-5.11.0-27-generic-x86_64-with-glibc2.31
- Python version: 3.9.7
- PyArrow version: 5.0.0
|
https://github.com/huggingface/datasets/issues/2924
|
[
"Hi, the filename here is less than 255\r\n```python\r\n>>> len(\"_home_garrett_.cache_huggingface_datasets_csv_test-7c856aea083a7043_0.0.0_9144e0a4e8435090117cea53e6c7537173ef2304525df4a077c435d8ee7828ff.incomplete.lock\")\r\n154\r\n```\r\nso not sure why it's considered too long for your filesystem.\r\n(also note that the lock files we use always have smaller filenames than 255)\r\n\r\nhttps://github.com/huggingface/datasets/blob/5d1a9f1e3c6c495dc0610b459e39d2eb8893f152/src/datasets/utils/filelock.py#L135-L135",
"Yes, you're right! I need to get you more info here. Either there's something going with the name itself that the file system doesn't like (an encoding that blows up the name length??) or perhaps there's something with the path that's causing the entire string to be used as a name. I haven't seen this on any system before and the Internet's not forthcoming with any info.",
"Snap, encountered when trying to run [this example from PyTorch Lightning Flash](https://lightning-flash.readthedocs.io/en/latest/reference/speech_recognition.html):\r\n\r\n```py\r\nimport torch\r\n\r\nimport flash\r\nfrom flash.audio import SpeechRecognition, SpeechRecognitionData\r\nfrom flash.core.data.utils import download_data\r\n\r\n# 1. Create the DataModule\r\ndownload_data(\"https://pl-flash-data.s3.amazonaws.com/timit_data.zip\", \"./data\")\r\n\r\ndatamodule = SpeechRecognitionData.from_json(\r\n input_fields=\"file\",\r\n target_fields=\"text\",\r\n train_file=\"data/timit/train.json\",\r\n test_file=\"data/timit/test.json\",\r\n)\r\n```\r\n\r\nGave this traceback:\r\n\r\n```py\r\nTraceback (most recent call last):\r\n File \"lf_ft.py\", line 10, in <module>\r\n datamodule = SpeechRecognitionData.from_json(\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/flash/core/data/data_module.py\", line 1005, in from_json\r\n return cls.from_data_source(\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/flash/core/data/data_module.py\", line 571, in from_data_source\r\n train_dataset, val_dataset, test_dataset, predict_dataset = data_source.to_datasets(\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/flash/core/data/data_source.py\", line 307, in to_datasets\r\n train_dataset = self.generate_dataset(train_data, RunningStage.TRAINING)\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/flash/core/data/data_source.py\", line 344, in generate_dataset\r\n data = load_data(data, mock_dataset)\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/flash/audio/speech_recognition/data.py\", line 103, in load_data\r\n dataset_dict = load_dataset(self.filetype, data_files={stage: str(file)})\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/datasets/load.py\", line 1599, in load_dataset\r\n builder_instance = load_dataset_builder(\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/datasets/load.py\", line 1457, in load_dataset_builder\r\n builder_instance: DatasetBuilder = builder_cls(\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/datasets/builder.py\", line 285, in __init__\r\n with FileLock(lock_path):\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/datasets/utils/filelock.py\", line 323, in __enter__\r\n self.acquire()\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/datasets/utils/filelock.py\", line 272, in acquire\r\n self._acquire()\r\n File \"/home/louis/miniconda3/envs/w2vlf/lib/python3.8/site-packages/datasets/utils/filelock.py\", line 403, in _acquire\r\n fd = os.open(self._lock_file, open_mode)\r\nOSError: [Errno 36] File name too long: '/home/louis/.cache/huggingface/datasets/_home_louis_.cache_huggingface_datasets_json_default-98e6813a547f72fa_0.0.0_c2d554c3377ea79c7664b93dc65d0803b45e3279000f993c7bfd18937fd7f426.lock'\r\n```\r\n\r\nMy home directory is encrypted, therefore the maximum length is 143 ([source 1](https://github.com/ray-project/ray/issues/1463#issuecomment-425674521), [source 2](https://stackoverflow.com/a/6571568/2668831))\r\n\r\nFrom what I've read I think the error is in reference to the file name (just the final part of the path) which is 145 chars long:\r\n\r\n```py\r\n>>> len(\"_home_louis_.cache_huggingface_datasets_json_default-98e6813a547f72fa_0.0.0_c2d554c3377ea79c7664b93dc65d0803b45e3279000f993c7bfd18937fd7f426.lock\")\r\n145\r\n```\r\n\r\nI also have a file in this directory (i.e. whose length is not a problem):\r\n\r\n```py\r\n>>> len(\"_home_louis_.cache_huggingface_datasets_librispeech_asr_clean_2.1.0_468ec03677f46a8714ac6b5b64dba02d246a228d92cbbad7f3dc190fa039eab1.lock\")\r\n137\r\n```",
"Perhaps this could be exposed as a config setting so you could change it manually?\r\n\r\nhttps://github.com/huggingface/datasets/blob/5d1a9f1e3c6c495dc0610b459e39d2eb8893f152/src/datasets/utils/filelock.py#L135-L135\r\n\r\nRather than hard-code 255, default it to 255, and allow it to be changed, the same way is done for `datasets.config.IN_MEMORY_MAX_SIZE`:\r\n\r\nhttps://github.com/huggingface/datasets/blob/12b7e13bc568b9f92705f64b249e148f3bc9a9ea/src/datasets/config.py#L171-L173\r\n\r\nIn fact there already appears to be an existing variable to do so:\r\n\r\nhttps://github.com/huggingface/datasets/blob/12b7e13bc568b9f92705f64b249e148f3bc9a9ea/src/datasets/config.py#L187\r\n\r\nIt's used here:\r\n\r\nhttps://github.com/huggingface/datasets/blob/efe89edd36e4ffa562fc3eebaf07a5fec26e6dac/src/datasets/builder.py#L163-L165\r\n\r\nPerhaps it could be set based on a test (trying to create a 255 char length named lock file and seeing if it fails)",
"Just fixed it, sending a PR :smile:",
"Hi @lmmx @gar1t ,\r\n\r\nit would be helpful if you could run the following code and copy-paste the output here:\r\n```python\r\nimport datasets\r\nimport os\r\nos.statvfs(datasets.config.HF_DATASETS_CACHE)\r\n```",
"`os.statvfs_result(f_bsize=4096, f_frsize=4096, f_blocks=240046344, f_bfree=96427610, f_bavail=84216487, f_files=61038592, f_ffree=58216027, f_favail=58216027, f_flag=4102, f_namemax=143)`",
"Hi @lmmx,\r\n\r\nThanks for providing the result of the command. I've opened a PR, and it would be great if you could verify that the fix works on your system. To install the version of the datasets with the fix, please run the following command:\r\n```\r\npip install git+https://github.com/huggingface/datasets.git@fix-2924\r\n```\r\n\r\nBtw, I saw your PR, and I appreciate your effort. However, my approach is a bit simpler for the end-user, so that's why I decided to fix the issue myself.",
"No problem Mario I didn't know that was where that value was recorded so I learnt something :smiley: I just wanted to get a local version working, of course you should implement whatever fix is best for HF. Yes can confirm this fixes it too. Thanks!"
] | null | 2,924
| false
|
Loading an autonlp dataset raises in normal mode but not in streaming mode
|
## Describe the bug
The same dataset (from autonlp) raises an error in normal mode, but does not raise in streaming mode
## Steps to reproduce the bug
```python
from datasets import load_dataset
load_dataset("severo/autonlp-data-sentiment_detection-3c8bcd36", split="train", streaming=False)
## raises an error
load_dataset("severo/autonlp-data-sentiment_detection-3c8bcd36", split="train", streaming=True)
## does not raise an error
```
## Expected results
Both calls should raise the same error
## Actual results
Call with streaming=False:
```
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 1/1 [00:00<00:00, 5825.42it/s]
Using custom data configuration autonlp-data-sentiment_detection-3c8bcd36-fe30267462d1d42b
Downloading and preparing dataset json/autonlp-data-sentiment_detection-3c8bcd36 to /home/slesage/.cache/huggingface/datasets/json/autonlp-data-sentiment_detection-3c8bcd36-fe30267462d1d42b/0.0.0/d75ead8d5cfcbe67495df0f89bd262f0023257fbbbd94a730313295f3d756d50...
100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 5/5 [00:00<00:00, 15923.71it/s]
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 5/5 [00:00<00:00, 3346.88it/s]
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/load.py", line 1112, in load_dataset
builder_instance.download_and_prepare(
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/builder.py", line 636, in download_and_prepare
self._download_and_prepare(
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/builder.py", line 726, in _download_and_prepare
self._prepare_split(split_generator, **prepare_split_kwargs)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/builder.py", line 1187, in _prepare_split
writer.write_table(table)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/arrow_writer.py", line 418, in write_table
pa_table = pa.Table.from_arrays([pa_table[name] for name in self._schema.names], schema=self._schema)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/arrow_writer.py", line 418, in <listcomp>
pa_table = pa.Table.from_arrays([pa_table[name] for name in self._schema.names], schema=self._schema)
File "pyarrow/table.pxi", line 1249, in pyarrow.lib.Table.__getitem__
File "pyarrow/table.pxi", line 1825, in pyarrow.lib.Table.column
File "pyarrow/table.pxi", line 1800, in pyarrow.lib.Table._ensure_integer_index
KeyError: 'Field "splits" does not exist in table schema'
```
Call with `streaming=False`:
```
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 1/1 [00:00<00:00, 6000.43it/s]
Using custom data configuration autonlp-data-sentiment_detection-3c8bcd36-fe30267462d1d42b
100%|ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 5/5 [00:00<00:00, 46916.15it/s]
100%|βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ| 5/5 [00:00<00:00, 148734.18it/s]
```
## Environment info
- `datasets` version: 1.12.1.dev0
- Platform: Linux-5.11.0-1017-aws-x86_64-with-glibc2.29
- Python version: 3.8.11
- PyArrow version: 4.0.1
|
https://github.com/huggingface/datasets/issues/2923
|
[
"Closing since autonlp dataset are now supported"
] | null | 2,923
| false
|
Fix conversion of multidim arrays in list to arrow
|
Arrow only supports 1-dim arrays. Previously we were converting all the numpy arrays to python list before instantiating arrow arrays to workaround this limitation.
However in #2361 we started to keep numpy arrays in order to keep their dtypes.
It works when we pass any multi-dim numpy array (the conversion to arrow has been added on our side), but not for lists of multi-dim numpy arrays.
In this PR I added two strategies:
- one that takes a list of multi-dim numpy arrays on returns an arrow array in an optimized way (more common case)
- one that takes a list of possibly very nested data (lists, dicts, tuples) containing multi-dim arrays. This one is less optimized since it converts all the multi-dim numpy arrays into lists of 1-d arrays for compatibility with arrow. This strategy is simpler that just trying to create the arrow array from a possibly very nested data structure, but in the future we can improve it if needed.
Fix https://github.com/huggingface/datasets/issues/2921
|
https://github.com/huggingface/datasets/pull/2922
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2922",
"html_url": "https://github.com/huggingface/datasets/pull/2922",
"diff_url": "https://github.com/huggingface/datasets/pull/2922.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2922.patch",
"merged_at": "2021-09-15T17:21:45"
}
| 2,922
| true
|
Using a list of multi-dim numpy arrays raises an error "can only convert 1-dimensional array values"
|
This error has been introduced in https://github.com/huggingface/datasets/pull/2361
To reproduce:
```python
import numpy as np
from datasets import Dataset
d = Dataset.from_dict({"a": [np.zeros((2, 2))]})
```
raises
```python
Traceback (most recent call last):
File "playground/ttest.py", line 5, in <module>
d = Dataset.from_dict({"a": [np.zeros((2, 2))]}).with_format("torch")
File "/Users/quentinlhoest/Desktop/hf/nlp/src/datasets/arrow_dataset.py", line 458, in from_dict
pa_table = InMemoryTable.from_pydict(mapping=mapping)
File "/Users/quentinlhoest/Desktop/hf/nlp/src/datasets/table.py", line 365, in from_pydict
return cls(pa.Table.from_pydict(*args, **kwargs))
File "pyarrow/table.pxi", line 1639, in pyarrow.lib.Table.from_pydict
File "pyarrow/array.pxi", line 332, in pyarrow.lib.asarray
File "pyarrow/array.pxi", line 223, in pyarrow.lib.array
File "pyarrow/array.pxi", line 110, in pyarrow.lib._handle_arrow_array_protocol
File "/Users/quentinlhoest/Desktop/hf/nlp/src/datasets/arrow_writer.py", line 107, in __arrow_array__
out = pa.array(self.data, type=type)
File "pyarrow/array.pxi", line 306, in pyarrow.lib.array
File "pyarrow/array.pxi", line 39, in pyarrow.lib._sequence_to_array
File "pyarrow/error.pxi", line 143, in pyarrow.lib.pyarrow_internal_check_status
File "pyarrow/error.pxi", line 99, in pyarrow.lib.check_status
pyarrow.lib.ArrowInvalid: Can only convert 1-dimensional array values
|
https://github.com/huggingface/datasets/issues/2921
|
[] | null | 2,921
| false
|
Fix unwanted tqdm bar when accessing examples
|
A change in #2814 added bad progress bars in `map_nested`. Now they're disabled by default
Fix #2919
|
https://github.com/huggingface/datasets/pull/2920
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2920",
"html_url": "https://github.com/huggingface/datasets/pull/2920",
"diff_url": "https://github.com/huggingface/datasets/pull/2920.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2920.patch",
"merged_at": "2021-09-15T17:18:23"
}
| 2,920
| true
|
Unwanted progress bars when accessing examples
|
When accessing examples from a dataset formatted for pytorch, some progress bars appear when accessing examples:
```python
In [1]: import datasets as ds
In [2]: d = ds.Dataset.from_dict({"a": [0, 1, 2]}).with_format("torch")
In [3]: d[0]
100%|ββββββββββββββββββββββββββββββββ| 1/1 [00:00<00:00, 3172.70it/s]
Out[3]: {'a': tensor(0)}
```
This is because the pytorch formatter calls `map_nested` that uses progress bars
cc @sgugger
|
https://github.com/huggingface/datasets/issues/2919
|
[
"doing a patch release now :)"
] | null | 2,919
| false
|
`Can not decode content-encoding: gzip` when loading `scitldr` dataset with streaming
|
## Describe the bug
Trying to load the `"FullText"` config of the `"scitldr"` dataset with `streaming=True` raises an error from `aiohttp`:
```python
ClientPayloadError: 400, message='Can not decode content-encoding: gzip'
```
cc @lhoestq
## Steps to reproduce the bug
```python
from datasets import load_dataset
iter_dset = iter(
load_dataset("scitldr", name="FullText", split="test", streaming=True)
)
next(iter_dset)
```
## Expected results
Returns the first sample of the dataset
## Actual results
Calling `__next__` crashes with the following Traceback:
```python
----> 1 next(dset_iter)
~\miniconda3\envs\datasets\lib\site-packages\datasets\iterable_dataset.py in __iter__(self)
339
340 def __iter__(self):
--> 341 for key, example in self._iter():
342 if self.features:
343 # we encode the example for ClassLabel feature types for example
~\miniconda3\envs\datasets\lib\site-packages\datasets\iterable_dataset.py in _iter(self)
336 else:
337 ex_iterable = self._ex_iterable
--> 338 yield from ex_iterable
339
340 def __iter__(self):
~\miniconda3\envs\datasets\lib\site-packages\datasets\iterable_dataset.py in __iter__(self)
76
77 def __iter__(self):
---> 78 for key, example in self.generate_examples_fn(**self.kwargs):
79 yield key, example
80
~\.cache\huggingface\modules\datasets_modules\datasets\scitldr\72d6e2195786c57e1d343066fb2cc4f93ea39c5e381e53e6ae7c44bbfd1f05ef\scitldr.py in _generate_examples(self, filepath, split)
162
163 with open(filepath, encoding="utf-8") as f:
--> 164 for id_, row in enumerate(f):
165 data = json.loads(row)
166 if self.config.name == "AIC":
~\miniconda3\envs\datasets\lib\site-packages\fsspec\implementations\http.py in read(self, length)
496 else:
497 length = min(self.size - self.loc, length)
--> 498 return super().read(length)
499
500 async def async_fetch_all(self):
~\miniconda3\envs\datasets\lib\site-packages\fsspec\spec.py in read(self, length)
1481 # don't even bother calling fetch
1482 return b""
-> 1483 out = self.cache._fetch(self.loc, self.loc + length)
1484 self.loc += len(out)
1485 return out
~\miniconda3\envs\datasets\lib\site-packages\fsspec\caching.py in _fetch(self, start, end)
378 elif start < self.start:
379 if self.end - end > self.blocksize:
--> 380 self.cache = self.fetcher(start, bend)
381 self.start = start
382 else:
~\miniconda3\envs\datasets\lib\site-packages\fsspec\asyn.py in wrapper(*args, **kwargs)
86 def wrapper(*args, **kwargs):
87 self = obj or args[0]
---> 88 return sync(self.loop, func, *args, **kwargs)
89
90 return wrapper
~\miniconda3\envs\datasets\lib\site-packages\fsspec\asyn.py in sync(loop, func, timeout, *args, **kwargs)
67 raise FSTimeoutError
68 if isinstance(result[0], BaseException):
---> 69 raise result[0]
70 return result[0]
71
~\miniconda3\envs\datasets\lib\site-packages\fsspec\asyn.py in _runner(event, coro, result, timeout)
23 coro = asyncio.wait_for(coro, timeout=timeout)
24 try:
---> 25 result[0] = await coro
26 except Exception as ex:
27 result[0] = ex
~\miniconda3\envs\datasets\lib\site-packages\fsspec\implementations\http.py in async_fetch_range(self, start, end)
538 if r.status == 206:
539 # partial content, as expected
--> 540 out = await r.read()
541 elif "Content-Length" in r.headers:
542 cl = int(r.headers["Content-Length"])
~\miniconda3\envs\datasets\lib\site-packages\aiohttp\client_reqrep.py in read(self)
1030 if self._body is None:
1031 try:
-> 1032 self._body = await self.content.read()
1033 for trace in self._traces:
1034 await trace.send_response_chunk_received(
~\miniconda3\envs\datasets\lib\site-packages\aiohttp\streams.py in read(self, n)
342 async def read(self, n: int = -1) -> bytes:
343 if self._exception is not None:
--> 344 raise self._exception
345
346 # migration problem; with DataQueue you have to catch
ClientPayloadError: 400, message='Can not decode content-encoding: gzip'
```
## Environment info
- `datasets` version: 1.12.0
- Platform: Windows-10-10.0.19041-SP0
- Python version: 3.8.5
- PyArrow version: 2.0.0
- aiohttp version: 3.7.4.post0
|
https://github.com/huggingface/datasets/issues/2918
|
[
"Hi @SBrandeis, thanks for reporting! ^^\r\n\r\nI think this is an issue with `fsspec`: https://github.com/intake/filesystem_spec/issues/389\r\n\r\nI will ask them if they are planning to fix it...",
"Code to reproduce the bug: `ClientPayloadError: 400, message='Can not decode content-encoding: gzip'`\r\n```python\r\nIn [1]: import fsspec\r\n\r\nIn [2]: import json\r\n\r\nIn [3]: with fsspec.open('https://raw.githubusercontent.com/allenai/scitldr/master/SciTLDR-Data/SciTLDR-FullText/test.jsonl', encoding=\"utf-8\") as f:\r\n ...: for row in f:\r\n ...: data = json.loads(row)\r\n ...:\r\n---------------------------------------------------------------------------\r\nClientPayloadError Traceback (most recent call last)\r\n```",
"Thanks for investigating @albertvillanova ! π€ "
] | null | 2,918
| false
|
windows download abnormal
|
## Describe the bug
The script clearly exists (accessible from the browser), but the script download fails on windows. Then I tried it again and it can be downloaded normally on linux. why??
## Steps to reproduce the bug
```python3.7 + windows
# Sample code to reproduce the bug
```
## Expected results
It can be downloaded normally.
## Actual results
it cann't
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version:1.11.0
- Platform:windows
- Python version:3.7
- PyArrow version:
|
https://github.com/huggingface/datasets/issues/2917
|
[
"Hi ! Is there some kind of proxy that is configured in your browser that gives you access to internet ? If it's the case it could explain why it doesn't work in the code, since the proxy wouldn't be used",
"It is indeed an agency problem, thank you very, very much",
"Let me know if you have other questions :)\r\n\r\nClosing this issue now"
] | null | 2,917
| false
|
Add OpenAI's pass@k code evaluation metric
|
This PR introduces the `code_eval` metric which implements [OpenAI's code evaluation harness](https://github.com/openai/human-eval) introduced in the [Codex paper](https://arxiv.org/abs/2107.03374). It is heavily based on the original implementation and just adapts the interface to follow the `predictions`/`references` convention.
The addition of this metric should enable the evaluation against the code evaluation datasets added in #2897 and #2893.
A few open questions:
- The implementation makes heavy use of multiprocessing which this PR does not touch. Is this conflicting with multiprocessing natively integrated in `datasets`?
- This metric executes generated Python code and as such it poses dangers of executing malicious code. OpenAI addresses this issue by 1) commenting the `exec` call in the code so the user has to actively uncomment it and read the warning and 2) suggests using a sandbox environment (gVisor container). Should we add a similar safeguard? E.g. a prompt that needs to be answered when initialising the metric? Or at least a warning message?
- Naming: the implementation sticks to the `predictions`/`references` naming, however, the references are not reference solutions but unittest to test the solution. While reference solutions are also available they are not used. Should the naming be adapted?
|
https://github.com/huggingface/datasets/pull/2916
|
[
"> The implementation makes heavy use of multiprocessing which this PR does not touch. Is this conflicting with multiprocessing natively integrated in datasets?\r\n\r\nIt should work normally, but feel free to test it.\r\nThere is some documentation about using metrics in a distributed setup that uses multiprocessing [here](https://huggingface.co/docs/datasets/loading.html?highlight=rank#distributed-setup)\r\nYou can test to spawn several processes where each process would load the metric. Then in each process you add some references/predictions to the metric. Finally you call compute() in each process and on process 0 it should return the result on all the references/predictions\r\n\r\nLet me know if you have questions or if I can help",
"Is there a good way to debug the Windows tests? I suspect it is an issue with `multiprocessing`, but I can't see the error messages.",
"Indeed it has an issue on windows.\r\nIn your example it's supposed to output\r\n```python\r\n{'pass@1': 0.5, 'pass@2': 1.0}\r\n```\r\nbut it gets\r\n```python\r\n{'pass@1': 0.0, 'pass@2': 0.0}\r\n```\r\n\r\nI'm not on my windows machine today so I can't take a look at it. I can dive into it early next week if you want",
"> I'm not on my windows machine today so I can't take a look at it. I can dive into it early next week if you want\r\n\r\nThat would be great - unfortunately I have no access to a windows machine at the moment. I am quite sure it is an issue with in exectue.py because of multiprocessing.\r\n"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2916",
"html_url": "https://github.com/huggingface/datasets/pull/2916",
"diff_url": "https://github.com/huggingface/datasets/pull/2916.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2916.patch",
"merged_at": "2021-11-12T14:19:50"
}
| 2,916
| true
|
Fix fsspec AbstractFileSystem access
|
This addresses the issue from #2914 by changing the way fsspec's AbstractFileSystem is accessed.
|
https://github.com/huggingface/datasets/pull/2915
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2915",
"html_url": "https://github.com/huggingface/datasets/pull/2915",
"diff_url": "https://github.com/huggingface/datasets/pull/2915.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2915.patch",
"merged_at": "2021-09-15T11:35:24"
}
| 2,915
| true
|
Having a dependency defining fsspec entrypoint raises an AttributeError when importing datasets
|
## Describe the bug
In one of my project, I defined a custom fsspec filesystem with an entrypoint.
My guess is that by doing so, a variable named `spec` is created in the module `fsspec` (created by entering a for loop as there are entrypoints defined, see the loop in question [here](https://github.com/intake/filesystem_spec/blob/0589358d8a029ed6b60d031018f52be2eb721291/fsspec/__init__.py#L55)).
So that `fsspec.spec`, that was previously referring to the `spec` submodule, is now referring to that `spec` variable.
This make the import of datasets failing as it is using that `fsspec.spec`.
## Steps to reproduce the bug
I could reproduce the bug with a dummy poetry project.
Here is the pyproject.toml:
```toml
[tool.poetry]
name = "debug-datasets"
version = "0.1.0"
description = ""
authors = ["Pierre Godard"]
[tool.poetry.dependencies]
python = "^3.8"
datasets = "^1.11.0"
[tool.poetry.dev-dependencies]
[build-system]
requires = ["poetry-core>=1.0.0"]
build-backend = "poetry.core.masonry.api"
[tool.poetry.plugins."fsspec.specs"]
"file2" = "fsspec.implementations.local.LocalFileSystem"
```
The only other file being a `debug_datasets/__init__.py` empty file.
The overall structure of the project is as follows:
```
.
βββ pyproject.toml
βββ debug_datasets
βββ __init__.py
```
Then, within the project folder run:
```
poetry install
poetry run python
```
And in the python interpreter, try to import `datasets`:
```
import datasets
```
## Expected results
The import should run successfully.
## Actual results
Here is the trace of the error I get:
```
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/home/godarpi/.cache/pypoetry/virtualenvs/debug-datasets-JuFzTKL--py3.8/lib/python3.8/site-packages/datasets/__init__.py", line 33, in <module>
from .arrow_dataset import Dataset, concatenate_datasets
File "/home/godarpi/.cache/pypoetry/virtualenvs/debug-datasets-JuFzTKL--py3.8/lib/python3.8/site-packages/datasets/arrow_dataset.py", line 48, in <module>
from .filesystems import extract_path_from_uri, is_remote_filesystem
File "/home/godarpi/.cache/pypoetry/virtualenvs/debug-datasets-JuFzTKL--py3.8/lib/python3.8/site-packages/datasets/filesystems/__init__.py", line 30, in <module>
def is_remote_filesystem(fs: fsspec.spec.AbstractFileSystem) -> bool:
AttributeError: 'EntryPoint' object has no attribute 'AbstractFileSystem'
```
## Suggested fix
`datasets/filesystems/__init__.py`, line 30, replace:
```
def is_remote_filesystem(fs: fsspec.spec.AbstractFileSystem) -> bool:
```
by:
```
def is_remote_filesystem(fs: fsspec.AbstractFileSystem) -> bool:
```
I will come up with a PR soon if this effectively solves the issue.
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version: 1.11.0
- Platform: WSL2 (Ubuntu 20.04.1 LTS)
- Python version: 3.8.5
- PyArrow version: 5.0.0
- `fsspec` version: 2021.8.1
|
https://github.com/huggingface/datasets/issues/2914
|
[
"Closed by #2915."
] | null | 2,914
| false
|
timit_asr dataset only includes one text phrase
|
## Describe the bug
The dataset 'timit_asr' only includes one text phrase. It only includes the transcription "Would such an act of refusal be useful?" multiple times rather than different phrases.
## Steps to reproduce the bug
Note: I am following the tutorial https://huggingface.co/blog/fine-tune-wav2vec2-english
1. Install the dataset and other packages
```python
!pip install datasets>=1.5.0
!pip install transformers==4.4.0
!pip install soundfile
!pip install jiwer
```
2. Load the dataset
```python
from datasets import load_dataset, load_metric
timit = load_dataset("timit_asr")
```
3. Remove columns that we don't want
```python
timit = timit.remove_columns(["phonetic_detail", "word_detail", "dialect_region", "id", "sentence_type", "speaker_id"])
```
4. Write a short function to display some random samples of the dataset.
```python
from datasets import ClassLabel
import random
import pandas as pd
from IPython.display import display, HTML
def show_random_elements(dataset, num_examples=10):
assert num_examples <= len(dataset), "Can't pick more elements than there are in the dataset."
picks = []
for _ in range(num_examples):
pick = random.randint(0, len(dataset)-1)
while pick in picks:
pick = random.randint(0, len(dataset)-1)
picks.append(pick)
df = pd.DataFrame(dataset[picks])
display(HTML(df.to_html()))
show_random_elements(timit["train"].remove_columns(["file"]))
```
## Expected results
10 random different transcription phrases.
## Actual results
10 of the same transcription phrase "Would such an act of refusal be useful?"
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version: 1.4.1
- Platform: macOS-10.15.7-x86_64-i386-64bit
- Python version: 3.8.5
- PyArrow version: not listed
|
https://github.com/huggingface/datasets/issues/2913
|
[
"Hi @margotwagner, \r\nThis bug was fixed in #1995. Upgrading the datasets should work (min v1.8.0 ideally)",
"Hi @margotwagner,\r\n\r\nYes, as @bhavitvyamalik has commented, this bug was fixed in `datasets` version 1.5.0. You need to update it, as your current version is 1.4.1:\r\n> Environment info\r\n> - `datasets` version: 1.4.1"
] | null | 2,913
| false
|
Update link to Blog in docs footer
|
Update link.
|
https://github.com/huggingface/datasets/pull/2912
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2912",
"html_url": "https://github.com/huggingface/datasets/pull/2912",
"diff_url": "https://github.com/huggingface/datasets/pull/2912.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2912.patch",
"merged_at": "2021-09-15T07:59:23"
}
| 2,912
| true
|
Fix exception chaining
|
Fix exception chaining to avoid tracebacks with message: `During handling of the above exception, another exception occurred:`
|
https://github.com/huggingface/datasets/pull/2911
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2911",
"html_url": "https://github.com/huggingface/datasets/pull/2911",
"diff_url": "https://github.com/huggingface/datasets/pull/2911.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2911.patch",
"merged_at": "2021-09-16T15:04:44"
}
| 2,911
| true
|
feat: πΈ pass additional arguments to get private configs + info
|
`use_auth_token` can now be passed to the functions to get the configs
or infos of private datasets on the hub
|
https://github.com/huggingface/datasets/pull/2910
|
[
"Included in https://github.com/huggingface/datasets/pull/2906"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2910",
"html_url": "https://github.com/huggingface/datasets/pull/2910",
"diff_url": "https://github.com/huggingface/datasets/pull/2910.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2910.patch",
"merged_at": null
}
| 2,910
| true
|
fix anli splits
|
I can't run the tests for dummy data, facing this error
`ImportError while loading conftest '/home/zaid/tmp/fix_anli_splits/datasets/tests/conftest.py'.
tests/conftest.py:10: in <module>
from datasets import config
E ImportError: cannot import name 'config' from 'datasets' (unknown location)`
|
https://github.com/huggingface/datasets/pull/2909
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2909",
"html_url": "https://github.com/huggingface/datasets/pull/2909",
"diff_url": "https://github.com/huggingface/datasets/pull/2909.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2909.patch",
"merged_at": null
}
| 2,909
| true
|
Update Zenodo metadata with creator names and affiliation
|
This PR helps in prefilling author data when automatically generating the DOI after each release.
|
https://github.com/huggingface/datasets/pull/2908
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2908",
"html_url": "https://github.com/huggingface/datasets/pull/2908",
"diff_url": "https://github.com/huggingface/datasets/pull/2908.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2908.patch",
"merged_at": "2021-09-14T14:29:25"
}
| 2,908
| true
|
add story_cloze dataset
|
@lhoestq I have spent some time but I still I can't succeed in correctly testing the dummy_data.
|
https://github.com/huggingface/datasets/pull/2907
|
[
"Will create a new one, this one seems to be missed up. "
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2907",
"html_url": "https://github.com/huggingface/datasets/pull/2907",
"diff_url": "https://github.com/huggingface/datasets/pull/2907.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2907.patch",
"merged_at": null
}
| 2,907
| true
|
feat: πΈ add a function to get a dataset config's split names
|
Also: pass additional arguments (use_auth_token) to get private configs + info of private datasets on the hub
Questions:
- [x] I'm not sure how the versions work: I changed 1.12.1.dev0 to 1.12.1.dev1, was it correct?
-> no: reverted
- [x] Should I add a section in https://github.com/huggingface/datasets/blob/master/docs/source/load_hub.rst? (there is no section for get_dataset_infos)
-> yes: added
|
https://github.com/huggingface/datasets/pull/2906
|
[
"> Should I add a section in https://github.com/huggingface/datasets/blob/master/docs/source/load_hub.rst? (there is no section for get_dataset_infos)\r\n\r\nYes totally :) This tutorial should indeed mention this, given how fundamental it is"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2906",
"html_url": "https://github.com/huggingface/datasets/pull/2906",
"diff_url": "https://github.com/huggingface/datasets/pull/2906.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2906.patch",
"merged_at": "2021-10-04T09:55:37"
}
| 2,906
| true
|
Update BibTeX entry
|
Update BibTeX entry.
|
https://github.com/huggingface/datasets/pull/2905
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2905",
"html_url": "https://github.com/huggingface/datasets/pull/2905",
"diff_url": "https://github.com/huggingface/datasets/pull/2905.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2905.patch",
"merged_at": "2021-09-14T12:25:37"
}
| 2,905
| true
|
FORCE_REDOWNLOAD does not work
|
## Describe the bug
With GenerateMode.FORCE_REDOWNLOAD, the documentation says
+------------------------------------+-----------+---------+
| | Downloads | Dataset |
+====================================+===========+=========+
| `REUSE_DATASET_IF_EXISTS` (default)| Reuse | Reuse |
+------------------------------------+-----------+---------+
| `REUSE_CACHE_IF_EXISTS` | Reuse | Fresh |
+------------------------------------+-----------+---------+
| `FORCE_REDOWNLOAD` | Fresh | Fresh |
+------------------------------------+-----------+---------+
However, the old dataset is loaded even when FORCE_REDOWNLOAD is chosen.
## Steps to reproduce the bug
```python
import pandas as pd
from datasets import load_dataset, GenerateMode
pd.DataFrame(range(5), columns=['numbers']).to_csv('/tmp/test.tsv.gz', index=False)
ee = load_dataset('csv', data_files=['/tmp/test.tsv.gz'], delimiter='\t', split='train', download_mode=GenerateMode.FORCE_REDOWNLOAD)
print(ee)
pd.DataFrame(range(10), columns=['numerals']).to_csv('/tmp/test.tsv.gz', index=False)
ee = load_dataset('csv', data_files=['/tmp/test.tsv.gz'], delimiter='\t', split='train', download_mode=GenerateMode.FORCE_REDOWNLOAD)
print(ee)
```
## Expected results
Dataset({
features: ['numbers'],
num_rows: 5
})
Dataset({
features: ['numerals'],
num_rows: 10
})
## Actual results
Dataset({
features: ['numbers'],
num_rows: 5
})
Dataset({
features: ['numbers'],
num_rows: 5
})
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version: 1.8.0
- Platform: Linux-4.14.181-108.257.amzn1.x86_64-x86_64-with-glibc2.10
- Python version: 3.7.10
- PyArrow version: 3.0.0
|
https://github.com/huggingface/datasets/issues/2904
|
[
"Hi ! Thanks for reporting. The error seems to happen only if you use compressed files.\r\n\r\nThe second dataset is prepared in another dataset cache directory than the first - which is normal, since the source file is different. However, it doesn't uncompress the new data file because it finds the old uncompressed data in the extraction cache directory.\r\n\r\nIf we fix the extraction cache mechanism to uncompress a local file if it changed then it should fix the issue.\r\nCurrently the extraction cache mechanism only takes into account the path of the compressed file, which is an issue.",
"Facing the same issue, is there any way to overtake this issue until it will be fixed? ",
"You can clear your extraction cache in the meantime (by default at `~/.cache/huggingface/datasets/downloads/extracted`)"
] | null | 2,904
| false
|
Fix xpathopen to accept positional arguments
|
Fix `xpathopen()` so that it also accepts positional arguments.
Fix #2901.
|
https://github.com/huggingface/datasets/pull/2903
|
[
"thanks!"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2903",
"html_url": "https://github.com/huggingface/datasets/pull/2903",
"diff_url": "https://github.com/huggingface/datasets/pull/2903.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2903.patch",
"merged_at": "2021-09-14T08:40:47"
}
| 2,903
| true
|
Add WIT Dataset
|
## Adding a Dataset
- **Name:** *WIT*
- **Description:** *Wikipedia-based Image Text Dataset*
- **Paper:** *[WIT: Wikipedia-based Image Text Dataset for Multimodal Multilingual Machine Learning
](https://arxiv.org/abs/2103.01913)*
- **Data:** *https://github.com/google-research-datasets/wit*
- **Motivation:** (excerpt from their Github README.md)
> - The largest multimodal dataset (publicly available at the time of this writing) by the number of image-text examples.
> - A massively multilingual dataset (first of its kind) with coverage for over 100+ languages.
> - A collection of diverse set of concepts and real world entities.
> - Brings forth challenging real-world test sets.
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
|
https://github.com/huggingface/datasets/issues/2902
|
[
"@hassiahk is working on it #2810 ",
"WikiMedia is now hosting the pixel values directly which should make it a lot easier!\r\nThe files can be found here:\r\nhttps://techblog.wikimedia.org/2021/09/09/the-wikipedia-image-caption-matching-challenge-and-a-huge-release-of-image-data-for-research/\r\nhttps://analytics.wikimedia.org/published/datasets/one-off/caption_competition/training/image_pixels/",
"> @hassiahk is working on it #2810\r\n\r\nThank you @bhavitvyamalik! Added this issue so we could track progress π . Just linked the PR as well for visibility. ",
"Hey folks, we are now hosting the merged pixel values + embeddings + metadata ourselves. I gave it a try - [nateraw/wit](https://huggingface.co/datasets/nateraw/wit)\r\n\r\n**β οΈ - Make sure you add `streaming=True` unless you're prepared to download 400GB of data!**\r\n\r\n```python\r\nfrom datasets import load_dataset\r\n\r\nds = load_dataset('nateraw/wit', streaming=True)\r\nexample = next(iter(ds))\r\n```\r\n\r\n```python\r\n>>> example = next(iter(ds['train']))\r\n>>> example.keys()\r\ndict_keys(['b64_bytes', 'original_width', 'image_url', 'wit_features', 'original_height', 'metadata_url', 'mime_type', 'caption_attribution_description', 'embedding'])\r\n>>> example['wit_features'].keys()\r\ndict_keys(['hierarchical_section_title', 'language', 'attribution_passes_lang_id', 'context_section_description', 'is_main_image', 'page_title', 'caption_title_and_reference_description', 'caption_alt_text_description', 'caption_reference_description', 'page_url', 'context_page_description', 'section_title', 'page_changed_recently'])\r\n```",
"Hi! `datasets` now hosts two versions of the WIT dataset:\r\n* [`google/wit`](https://huggingface.co/datasets/google/wit): Google's version with the image URLs\r\n* [`wikimedia/wit_base`](https://huggingface.co/datasets/wikimedia/wit_base): Wikimedia's version with the images + ResNet embeddings, but with less data than Google's version"
] | null | 2,902
| false
|
Incompatibility with pytest
|
## Describe the bug
pytest complains about xpathopen / path.open("w")
## Steps to reproduce the bug
Create a test file, `test.py`:
```python
import datasets as ds
def load_dataset():
ds.load_dataset("counter", split="train", streaming=True)
```
And launch it with pytest:
```bash
python -m pytest test.py
```
## Expected results
It should give something like:
```
collected 1 item
test.py . [100%]
======= 1 passed in 3.15s =======
```
## Actual results
```
============================================================================================================================= test session starts ==============================================================================================================================
platform linux -- Python 3.8.11, pytest-6.2.5, py-1.10.0, pluggy-1.0.0
rootdir: /home/slesage/hf/datasets-preview-backend, configfile: pyproject.toml
plugins: anyio-3.3.1
collected 1 item
tests/queries/test_rows.py . [100%]Traceback (most recent call last):
File "/home/slesage/.pyenv/versions/3.8.11/lib/python3.8/runpy.py", line 194, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/slesage/.pyenv/versions/3.8.11/lib/python3.8/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pytest/__main__.py", line 5, in <module>
raise SystemExit(pytest.console_main())
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/_pytest/config/__init__.py", line 185, in console_main
code = main()
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/_pytest/config/__init__.py", line 162, in main
ret: Union[ExitCode, int] = config.hook.pytest_cmdline_main(
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_hooks.py", line 265, in __call__
return self._hookexec(self.name, self.get_hookimpls(), kwargs, firstresult)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_manager.py", line 80, in _hookexec
return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_callers.py", line 60, in _multicall
return outcome.get_result()
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_result.py", line 60, in get_result
raise ex[1].with_traceback(ex[2])
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_callers.py", line 39, in _multicall
res = hook_impl.function(*args)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/_pytest/main.py", line 316, in pytest_cmdline_main
return wrap_session(config, _main)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/_pytest/main.py", line 304, in wrap_session
config.hook.pytest_sessionfinish(
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_hooks.py", line 265, in __call__
return self._hookexec(self.name, self.get_hookimpls(), kwargs, firstresult)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_manager.py", line 80, in _hookexec
return self._inner_hookexec(hook_name, methods, kwargs, firstresult)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_callers.py", line 55, in _multicall
gen.send(outcome)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/_pytest/terminal.py", line 803, in pytest_sessionfinish
outcome.get_result()
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_result.py", line 60, in get_result
raise ex[1].with_traceback(ex[2])
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/pluggy/_callers.py", line 39, in _multicall
res = hook_impl.function(*args)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/_pytest/cacheprovider.py", line 428, in pytest_sessionfinish
config.cache.set("cache/nodeids", sorted(self.cached_nodeids))
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/_pytest/cacheprovider.py", line 188, in set
f = path.open("w")
TypeError: xpathopen() takes 1 positional argument but 2 were given
```
## Environment info
- `datasets` version: 1.12.0
- Platform: Linux-5.11.0-1017-aws-x86_64-with-glibc2.29
- Python version: 3.8.11
- PyArrow version: 4.0.1
|
https://github.com/huggingface/datasets/issues/2901
|
[
"Sorry, my bad... When implementing `xpathopen`, I just considered the use case in the COUNTER dataset... I'm fixing it!"
] | null | 2,901
| false
|
Fix null sequence encoding
|
The Sequence feature encoding was failing when a `None` sequence was used in a dataset.
Fix https://github.com/huggingface/datasets/issues/2892
|
https://github.com/huggingface/datasets/pull/2900
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2900",
"html_url": "https://github.com/huggingface/datasets/pull/2900",
"diff_url": "https://github.com/huggingface/datasets/pull/2900.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2900.patch",
"merged_at": "2021-09-13T14:17:42"
}
| 2,900
| true
|
Dataset
|
## Adding a Dataset
- **Name:** *name of the dataset*
- **Description:** *short description of the dataset (or link to social media or blog post)*
- **Paper:** *link to the dataset paper if available*
- **Data:** *link to the Github repository or current dataset location*
- **Motivation:** *what are some good reasons to have this dataset*
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
|
https://github.com/huggingface/datasets/issues/2899
|
[] | null | 2,899
| false
|
Hug emoji
|
## Adding a Dataset
- **Name:** *name of the dataset*
- **Description:** *short description of the dataset (or link to social media or blog post)*
- **Paper:** *link to the dataset paper if available*
- **Data:** *link to the Github repository or current dataset location*
- **Motivation:** *what are some good reasons to have this dataset*
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
|
https://github.com/huggingface/datasets/issues/2898
|
[] | null | 2,898
| false
|
Add OpenAI's HumanEval dataset
|
This PR adds OpenAI's [HumanEval](https://github.com/openai/human-eval) dataset. The dataset consists of 164 handcrafted programming problems with solutions and unittests to verify solution. This dataset is useful to evaluate code generation models.
|
https://github.com/huggingface/datasets/pull/2897
|
[
"I just fixed the class name, and added `[More Information Needed]` in empty sections in case people want to complete the dataset card :)"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2897",
"html_url": "https://github.com/huggingface/datasets/pull/2897",
"diff_url": "https://github.com/huggingface/datasets/pull/2897.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2897.patch",
"merged_at": "2021-09-16T15:02:11"
}
| 2,897
| true
|
add multi-proc in `to_csv`
|
This PR extends the multi-proc method used in #2747 for`to_json` to `to_csv` as well.
Results on my machine post benchmarking on `ascent_kb` dataset (giving ~45% improvement when compared to num_proc = 1):
```
Time taken on 1 num_proc, 10000 batch_size 674.2055702209473
Time taken on 4 num_proc, 10000 batch_size 425.6553490161896
Time taken on 1 num_proc, 50000 batch_size 623.5897650718689
Time taken on 4 num_proc, 50000 batch_size 380.0402421951294
Time taken on 4 num_proc, 100000 batch_size 361.7168130874634
```
This is a WIP as writing tests is pending for this PR.
I'm also exploring [this](https://arrow.apache.org/docs/python/csv.html#incremental-writing) approach for which I'm using `pyarrow-5.0.0`.
|
https://github.com/huggingface/datasets/pull/2896
|
[
"I think you can just add a test `test_dataset_to_csv_multiproc` in `tests/io/test_csv.py` and we'll be good",
"Hi @lhoestq, \r\nI've added `test_dataset_to_csv` apart from `test_dataset_to_csv_multiproc` as no test was there to check generated CSV file when `num_proc=1`. Please let me know if anything is also required! "
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2896",
"html_url": "https://github.com/huggingface/datasets/pull/2896",
"diff_url": "https://github.com/huggingface/datasets/pull/2896.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2896.patch",
"merged_at": "2021-10-26T16:00:41"
}
| 2,896
| true
|
Use pyarrow.Table.replace_schema_metadata instead of pyarrow.Table.cast
|
This PR partially addresses #2252.
``update_metadata_with_features`` uses ``Table.cast`` which slows down ``load_from_disk`` (and possibly other methods that use it) for very large datasets. Since ``update_metadata_with_features`` is only updating the schema metadata, it makes more sense to use ``pyarrow.Table.replace_schema_metadata`` which is much faster. This PR adds a ``replace_schema_metadata`` method to all table classes, and modifies ``update_metadata_with_features`` to use it instead of ``cast``.
|
https://github.com/huggingface/datasets/pull/2895
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2895",
"html_url": "https://github.com/huggingface/datasets/pull/2895",
"diff_url": "https://github.com/huggingface/datasets/pull/2895.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2895.patch",
"merged_at": "2021-09-21T08:18:35"
}
| 2,895
| true
|
Fix COUNTER dataset
|
Fix filename generating `FileNotFoundError`.
Related to #2866.
CC: @severo.
|
https://github.com/huggingface/datasets/pull/2894
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2894",
"html_url": "https://github.com/huggingface/datasets/pull/2894",
"diff_url": "https://github.com/huggingface/datasets/pull/2894.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2894.patch",
"merged_at": "2021-09-10T16:27:44"
}
| 2,894
| true
|
add mbpp dataset
|
This PR adds the mbpp dataset introduced by Google [here](https://github.com/google-research/google-research/tree/master/mbpp) as mentioned in #2816.
The dataset contain two versions: a full and a sanitized one. They have a slightly different schema and it is current state the loading preserves the original schema. An open question is whether to harmonize the two schemas when loading the dataset or to preserve the original one. Since not all fields are overlapping the schema will not be exactly the same.
|
https://github.com/huggingface/datasets/pull/2893
|
[
"I think it's fine to have the original schema"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2893",
"html_url": "https://github.com/huggingface/datasets/pull/2893",
"diff_url": "https://github.com/huggingface/datasets/pull/2893.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2893.patch",
"merged_at": "2021-09-16T09:35:42"
}
| 2,893
| true
|
Error when encoding a dataset with None objects with a Sequence feature
|
There is an error when encoding a dataset with None objects with a Sequence feature
To reproduce:
```python
from datasets import Dataset, Features, Value, Sequence
data = {"a": [[0], None]}
features = Features({"a": Sequence(Value("int32"))})
dataset = Dataset.from_dict(data, features=features)
```
raises
```python
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-24-40add67f8751> in <module>
2 data = {"a": [[0], None]}
3 features = Features({"a": Sequence(Value("int32"))})
----> 4 dataset = Dataset.from_dict(data, features=features)
[...]
~/datasets/features.py in encode_nested_example(schema, obj)
888 if isinstance(obj, str): # don't interpret a string as a list
889 raise ValueError("Got a string but expected a list instead: '{}'".format(obj))
--> 890 return [encode_nested_example(schema.feature, o) for o in obj]
891 # Object with special encoding:
892 # ClassLabel will convert from string to int, TranslationVariableLanguages does some checks
TypeError: 'NoneType' object is not iterable
```
Instead, if should run without error, as if the `features` were not passed
|
https://github.com/huggingface/datasets/issues/2892
|
[
"This has been fixed by https://github.com/huggingface/datasets/pull/2900\r\nWe're doing a new release 1.12 today to make the fix available :)"
] | null | 2,892
| false
|
Allow dynamic first dimension for ArrayXD
|
Add support for dynamic first dimension for ArrayXD features. See issue [#887](https://github.com/huggingface/datasets/issues/887).
Following changes allow for `to_pylist` method of `ArrayExtensionArray` to return a list of numpy arrays where fist dimension can vary.
@lhoestq Could you suggest how you want to extend test suit. For now I added only very limited testing.
|
https://github.com/huggingface/datasets/pull/2891
|
[
"@lhoestq, thanks for your review.\r\n\r\nI added test for `to_pylist`, I didn't do that for `to_numpy` because this method shouldn't be called for dynamic dimension ArrayXD - this method will try to make a single numpy array for the whole column which cannot be done for dynamic arrays.\r\n\r\nI dig into `to_pandas()` functionality and I found it quite difficult to implement. `PandasArrayExtensionArray` takes single np.array as an argument. It might be a bit of changes to make it work with the list of arrays. Do you mind if we exclude this work from this PR. I added an error message for the case if somebody tries to use dynamic arrays with `to_pandas`",
"@lhoestq, I just fixed all the tests. Let me know if there is anything else to add.",
"@lhoestq, any chance you had some time to check out this PR?\r\n",
"Hi ! Sorry for the delay\r\n\r\nIt looks good to me ! I think the only thing missing is the support for passing a list of numpy arrays to `map` when the first dimension is dynamic.\r\n\r\nCurrently it raises an error:\r\n```python\r\nfrom datasets import *\r\nimport numpy as np\r\n\r\nfeatures= Features({\"a\": Array3D(shape=(None, 5, 2), dtype=\"int32\")})\r\nd = Dataset.from_dict({\"a\": [np.zeros((5,5,2)), np.zeros((2,5,2))]}, features=features)\r\nd = d.map(lambda a: {\"a\": np.concatenate([a]*2)}, input_columns=\"a\")\r\nprint(d[0])\r\n```\r\nraises\r\n```python\r\nTraceback (most recent call last):\r\n File \"playground/ttest.py\", line 6, in <module>\r\n d = d.map(lambda x: {\"a\": np.concatenate([x]*2)}, input_columns=\"a\")\r\n File \"/home/truent/hf/datasets/src/datasets/arrow_dataset.py\", line 1932, in map\r\n return self._map_single(\r\n File \"/home/truent/hf/datasets/src/datasets/arrow_dataset.py\", line 426, in wrapper\r\n out: Union[\"Dataset\", \"DatasetDict\"] = func(self, *args, **kwargs)\r\n File \"/home/truent/hf/datasets/src/datasets/fingerprint.py\", line 406, in wrapper\r\n out = func(self, *args, **kwargs)\r\n File \"/home/truent/hf/datasets/src/datasets/arrow_dataset.py\", line 2317, in _map_single\r\n writer.finalize()\r\n File \"/home/truent/hf/datasets/src/datasets/arrow_writer.py\", line 443, in finalize\r\n self.write_examples_on_file()\r\n File \"/home/truent/hf/datasets/src/datasets/arrow_writer.py\", line 312, in write_examples_on_file\r\n pa_array = pa.array(typed_sequence)\r\n File \"pyarrow/array.pxi\", line 222, in pyarrow.lib.array\r\n File \"pyarrow/array.pxi\", line 110, in pyarrow.lib._handle_arrow_array_protocol\r\n File \"/home/truent/hf/datasets/src/datasets/arrow_writer.py\", line 108, in __arrow_array__\r\n storage = pa.array(self.data, type.storage_dtype)\r\n File \"pyarrow/array.pxi\", line 305, in pyarrow.lib.array\r\n File \"pyarrow/array.pxi\", line 39, in pyarrow.lib._sequence_to_array\r\n File \"pyarrow/error.pxi\", line 122, in pyarrow.lib.pyarrow_internal_check_status\r\n File \"pyarrow/error.pxi\", line 84, in pyarrow.lib.check_status\r\npyarrow.lib.ArrowInvalid: Can only convert 1-dimensional array values\r\n```\r\n\r\nI think the issue is that here we don't cover the case where self.data is a list of numpy arrays:\r\n\r\nhttps://github.com/huggingface/datasets/blob/55fd140a63b8f03a0e72985647e498f1fc799d3f/src/datasets/arrow_writer.py#L104-L109\r\n\r\nWe should remove the `isinstance(self.data[0], np.ndarray)` part and add these lines to cover this case:\r\n\r\nhttps://github.com/huggingface/datasets/blob/55fd140a63b8f03a0e72985647e498f1fc799d3f/src/datasets/arrow_writer.py#L112-L113",
"@lhoestq, thanks, good catch!\r\nAre you able to run this check with fixed dimension ArrayXD?\r\nfor below example\r\n```\r\nimport numpy as np\r\nfrom datasets import *\r\n\r\nfeatures = Features({\"a\": Array3D(shape=(2, 5, 2), dtype=\"int32\")})\r\nd = Dataset.from_dict({\"a\": [np.zeros((2, 5, 2)), np.zeros((2, 5, 2))]}, features=features)\r\nd = d.map(lambda a: {\"a\": np.array(a) + 1}, input_columns=\"a\")\r\nprint(d[0])\r\n```\r\n\r\nI am getting:\r\n```\r\n File \"/home/ib/datasets/src/datasets/arrow_writer.py\", line 116, in __arrow_array__\r\n if trying_type and out[0].as_py() != self.data[0]:\r\nValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()\r\n```",
"Nevertheless, I tried to fix that. Let me know if that works.",
"@lhoestq, just resolved the conflicts. Let me know if there is anything left to do with this PR",
"Hi, thanks a lot for your comments.\r\nAgree, happy to contribute to this topic in future PRs",
"Hi @rpowalski, thanks for adding this feature! \r\n\r\nI wanted to check if you are still interested in documenting this, otherwise I'd be happy to help with it :)"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2891",
"html_url": "https://github.com/huggingface/datasets/pull/2891",
"diff_url": "https://github.com/huggingface/datasets/pull/2891.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2891.patch",
"merged_at": "2021-10-29T09:37:17"
}
| 2,891
| true
|
0x290B112ED1280537B24Ee6C268a004994a16e6CE
|
## Adding a Dataset
- **Name:** *name of the dataset*
- **Description:** *short description of the dataset (or link to social media or blog post)*
- **Paper:** *link to the dataset paper if available*
- **Data:** *link to the Github repository or current dataset location*
- **Motivation:** *what are some good reasons to have this dataset*
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
|
https://github.com/huggingface/datasets/issues/2890
|
[] | null | 2,890
| false
|
Coc
|
## Adding a Dataset
- **Name:** *name of the dataset*
- **Description:** *short description of the dataset (or link to social media or blog post)*
- **Paper:** *link to the dataset paper if available*
- **Data:** *link to the Github repository or current dataset location*
- **Motivation:** *what are some good reasons to have this dataset*
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
|
https://github.com/huggingface/datasets/issues/2889
|
[] | null | 2,889
| false
|
v1.11.1 release date
|
Hello, i need to use latest features in one of my packages but there have been no new datasets release since 2 months ago.
When do you plan to publush v1.11.1 release?
|
https://github.com/huggingface/datasets/issues/2888
|
[
"Hi ! Probably 1.12 on monday :)\r\n",
"@albertvillanova i think this issue is still valid and should not be closed till `>1.11.0` is published :)"
] | null | 2,888
| false
|
#2837 Use cache folder for lockfile
|
Fixes #2837
Use a cache folder directory to store the FileLock.
The issue was that the lock file was in a readonly folder.
|
https://github.com/huggingface/datasets/pull/2887
|
[
"The CI fail about the meteor metric is unrelated to this PR "
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2887",
"html_url": "https://github.com/huggingface/datasets/pull/2887",
"diff_url": "https://github.com/huggingface/datasets/pull/2887.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2887.patch",
"merged_at": "2021-10-05T17:58:22"
}
| 2,887
| true
|
Hj
| null |
https://github.com/huggingface/datasets/issues/2886
|
[] | null | 2,886
| false
|
Adding an Elastic Search index to a Dataset
|
## Describe the bug
When trying to index documents from the squad dataset, the connection to ElasticSearch seems to break:
Reusing dataset squad (/Users/andreasmotz/.cache/huggingface/datasets/squad/plain_text/1.0.0/d6ec3ceb99ca480ce37cdd35555d6cb2511d223b9150cce08a837ef62ffea453)
90%|βββββββββββββββββββββββββββββββββββββββββββββ | 9501/10570 [00:01<00:00, 6335.61docs/s]
No error is thrown, but the indexing breaks ~90%.
## Steps to reproduce the bug
```python
# Sample code to reproduce the bug
from datasets import load_dataset
from elasticsearch import Elasticsearch
es = Elasticsearch()
squad = load_dataset('squad', split='validation')
index_name = "corpus"
es_config = {
"settings": {
"number_of_shards": 1,
"analysis": {"analyzer": {"stop_standard": {"type": "standard", " stopwords": "_english_"}}},
},
"mappings": {
"properties": {
"idx" : {"type" : "keyword"},
"title" : {"type" : "keyword"},
"text": {
"type": "text",
"analyzer": "standard",
"similarity": "BM25"
},
}
},
}
class IndexBuilder:
"""
Elastic search indexing of a corpus
"""
def __init__(
self,
*args,
#corpus : None,
dataset : squad,
index_name = str,
query = str,
config = dict,
**kwargs,
):
#instantiate HuggingFace dataset
self.dataset = dataset
#instantiate ElasticSearch config
self.config = config
self.es = Elasticsearch()
self.index_name = index_name
self.query = query
def elastic_index(self):
print(self.es.info)
self.es.indices.delete(index=self.index_name, ignore=[400, 404])
search_index = self.dataset.add_elasticsearch_index(column='context', host='localhost', port='9200', es_index_name=self.index_name, es_index_config=self.config)
return search_index
def exact_match_method(self, index):
scores, retrieved_examples = index.get_nearest_examples('context', query=self.query, k=1)
return scores, retrieved_examples
if __name__ == "__main__":
print(type(squad))
Index = IndexBuilder(dataset=squad, index_name='corpus_index', query='Where was Chopin born?', config=es_config)
search_index = Index.elastic_index()
scores, examples = Index.exact_match_method(search_index)
print(scores, examples)
for name in squad.column_names:
print(type(squad[name]))
```
## Environment info
We run the code in Poetry. This might be the issue, since the script runs successfully in our local environment.
Poetry:
- Python version: 3.8
- PyArrow: 4.0.1
- Elasticsearch: 7.13.4
- datasets: 1.10.2
Local:
- Python version: 3.8
- PyArrow: 3.0.0
- Elasticsearch: 7.7.1
- datasets: 1.7.0
|
https://github.com/huggingface/datasets/issues/2885
|
[
"Hi, is this bug deterministic in your poetry env ? I mean, does it always stop at 90% or is it random ?\r\n\r\nAlso, can you try using another version of Elasticsearch ? Maybe there's an issue with the one of you poetry env",
"I face similar issue with oscar dataset on remote ealsticsearch instance. It was mainly due to timeout of batch indexing requests and I solve these by adding large request_timeout param in `search.py`\r\n\r\n```\r\n for ok, action in es.helpers.streaming_bulk(\r\n client=self.es_client,\r\n index=index_name,\r\n actions=passage_generator(),\r\n request_timeout=3600,\r\n )\r\n ```",
"Hi @MotzWanted - are there any errors in the Elasticsearch cluster logs? Since it works in your local environment and the cluster versions are different between your poetry env and your local env, it is possible that it is some difference in the cluster - either settings or the cluster being under a different load etc that has this effect, so it would be useful to see if any errors are thrown in the cluster's logs when you try to ingest. \r\nWhich elasticsearch client method is the function `add_elasticsearch_index` from your code using under the hood? Is it `helpers.bulk` or is the indexing performed using something else? You can try adding a timeout to the indexing method to see if this helps. Also, you mention that it stops at around 90% - do you know if the timeout/hanging happens always when a particular document is being indexed or does it happen randomly at around 90% completeness but on different documents?"
] | null | 2,885
| false
|
Add IC, SI, ER tasks to SUPERB
|
This PR adds 3 additional classification tasks to SUPERB
#### Intent Classification
Dataset URL seems to be down at the moment :( See the note below.
S3PRL source: https://github.com/s3prl/s3prl/blob/master/s3prl/downstream/fluent_commands/dataset.py
Instructions: https://github.com/s3prl/s3prl/tree/master/s3prl/downstream#ic-intent-classification---fluent-speech-commands
#### Speaker Identification
Manual download script:
```
mkdir VoxCeleb1
cd VoxCeleb1
wget https://thor.robots.ox.ac.uk/~vgg/data/voxceleb/vox1a/vox1_dev_wav_partaa
wget https://thor.robots.ox.ac.uk/~vgg/data/voxceleb/vox1a/vox1_dev_wav_partab
wget https://thor.robots.ox.ac.uk/~vgg/data/voxceleb/vox1a/vox1_dev_wav_partac
wget https://thor.robots.ox.ac.uk/~vgg/data/voxceleb/vox1a/vox1_dev_wav_partad
cat vox1_dev* > vox1_dev_wav.zip
unzip vox1_dev_wav.zip
wget https://thor.robots.ox.ac.uk/~vgg/data/voxceleb/vox1a/vox1_test_wav.zip
unzip vox1_test_wav.zip
# download the official SUPERB train-dev-test split
wget https://raw.githubusercontent.com/s3prl/s3prl/master/s3prl/downstream/voxceleb1/veri_test_class.txt
```
S3PRL source: https://github.com/s3prl/s3prl/blob/master/s3prl/downstream/voxceleb1/dataset.py
Instructions: https://github.com/s3prl/s3prl/tree/master/s3prl/downstream#sid-speaker-identification
#### Intent Classification
Manual download requires going through a slow application process, see the note below.
S3PRL source: https://github.com/s3prl/s3prl/blob/master/s3prl/downstream/emotion/IEMOCAP_preprocess.py
Instructions: https://github.com/s3prl/s3prl/tree/master/s3prl/downstream#er-emotion-recognition
#### :warning: Note
These datasets either require manual downloads or have broken/unstable links. You can get all necessary archives in this repo: https://huggingface.co/datasets/anton-l/superb_source_data_dumps/tree/main
|
https://github.com/huggingface/datasets/pull/2884
|
[
"Sorry for the late PR, uploading 10+GB files to the hub through a VPN was an adventure :sweat_smile: ",
"Thank you so much for adding these subsets @anton-l! \r\n\r\n> These datasets either require manual downloads or have broken/unstable links. You can get all necessary archives in this repo: https://huggingface.co/datasets/anton-l/superb_source_data_dumps/tree/main\r\nAre we allowed to make these datasets public or would that violate the terms of their use?",
"@lewtun These ones all have non-permissive licences, so the mirrored data I linked is open only to the HF org for now. But we could try contacting the authors to ask if they'd like to host these with us. \nFor example VoxCeleb1 now has direct links (the ones in the script) that don't require form submission and passwords, but they ban IPs after each download for some reason :(",
"> @lewtun These ones all have non-permissive licences, so the mirrored data I linked is open only to the HF org for now. But we could try contacting the authors to ask if they'd like to host these with us.\r\n> For example VoxCeleb1 now has direct links (the ones in the script) that don't require form submission and passwords, but they ban IPs after each download for some reason :(\r\n\r\nI think there would be a lot of value added if the authors would be willing to host their data on the HF Hub! As an end-user of `datasets`, I've found I'm more likely to explore a dataset if I'm able to quickly pull the subsets without needing a manual download. Perhaps we can tell them that the Hub offers several advantages like versioning and interactive exploration (with `datasets-viewer`)?"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2884",
"html_url": "https://github.com/huggingface/datasets/pull/2884",
"diff_url": "https://github.com/huggingface/datasets/pull/2884.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2884.patch",
"merged_at": "2021-09-20T09:00:49"
}
| 2,884
| true
|
Fix data URLs and metadata in DocRED dataset
|
The host of `docred` dataset has updated the `dev` data file. This PR:
- Updates the dev URL
- Updates dataset metadata
This PR also fixes the URL of the `train_distant` split, which was wrong.
Fix #2882.
|
https://github.com/huggingface/datasets/pull/2883
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2883",
"html_url": "https://github.com/huggingface/datasets/pull/2883",
"diff_url": "https://github.com/huggingface/datasets/pull/2883.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2883.patch",
"merged_at": "2021-09-13T11:24:30"
}
| 2,883
| true
|
`load_dataset('docred')` results in a `NonMatchingChecksumError`
|
## Describe the bug
I get consistent `NonMatchingChecksumError: Checksums didn't match for dataset source files` errors when trying to execute `datasets.load_dataset('docred')`.
## Steps to reproduce the bug
It is quasi only this code:
```python
import datasets
data = datasets.load_dataset('docred')
```
## Expected results
The DocRED dataset should be loaded without any problems.
## Actual results
```
NonMatchingChecksumError Traceback (most recent call last)
<ipython-input-4-b1b83f25a16c> in <module>
----> 1 d = datasets.load_dataset('docred')
~/anaconda3/lib/python3.8/site-packages/datasets/load.py in load_dataset(path, name, data_dir, data_files, split, cache_dir, features, download_config, download_mode, ignore_verifications, keep_in_memory, save_infos, script_version, use_auth_token, task, streaming, **config_kwargs)
845
846 # Download and prepare data
--> 847 builder_instance.download_and_prepare(
848 download_config=download_config,
849 download_mode=download_mode,
~/anaconda3/lib/python3.8/site-packages/datasets/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, try_from_hf_gcs, dl_manager, base_path, use_auth_token, **download_and_prepare_kwargs)
613 logger.warning("HF google storage unreachable. Downloading and preparing it from source")
614 if not downloaded_from_gcs:
--> 615 self._download_and_prepare(
616 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs
617 )
~/anaconda3/lib/python3.8/site-packages/datasets/builder.py in _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs)
673 # Checksums verification
674 if verify_infos:
--> 675 verify_checksums(
676 self.info.download_checksums, dl_manager.get_recorded_sizes_checksums(), "dataset source files"
677 )
~/anaconda3/lib/python3.8/site-packages/datasets/utils/info_utils.py in verify_checksums(expected_checksums, recorded_checksums, verification_name)
38 if len(bad_urls) > 0:
39 error_msg = "Checksums didn't match" + for_verification_name + ":\n"
---> 40 raise NonMatchingChecksumError(error_msg + str(bad_urls))
41 logger.info("All the checksums matched successfully" + for_verification_name)
42
NonMatchingChecksumError: Checksums didn't match for dataset source files:
['https://drive.google.com/uc?export=download&id=1fDmfUUo5G7gfaoqWWvK81u08m71TK2g7']
```
## Environment info
- `datasets` version: 1.11.0
- Platform: Linux-5.11.0-7633-generic-x86_64-with-glibc2.10
- Python version: 3.8.5
- PyArrow version: 5.0.0
This error also happened on my Windows-partition, after freshly installing python 3.9 and `datasets`.
## Remarks
- I have already called `rm -rf /home/<user>/.cache/huggingface`, i.e., I have tried clearing the cache.
- The problem does not exist for other datasets, i.e., it seems to be DocRED-specific.
|
https://github.com/huggingface/datasets/issues/2882
|
[
"Hi @tmpr, thanks for reporting.\r\n\r\nTwo weeks ago (23th Aug), the host of the source `docred` dataset updated one of the files (`dev.json`): you can see it [here](https://drive.google.com/drive/folders/1c5-0YwnoJx8NS6CV2f-NoTHR__BdkNqw).\r\n\r\nTherefore, the checksum needs to be updated.\r\n\r\nNormally, in the meantime, you could avoid the error by passing `ignore_verifications=True` to `load_dataset`. However, as the old link points to a non-existing file, the link must be updated too.\r\n\r\nI'm fixing all this.\r\n\r\n"
] | null | 2,882
| false
|
Add BIOSSES dataset
|
Adding the biomedical semantic sentence similarity dataset, BIOSSES, listed in "Biomedical Datasets - BigScience Workshop 2021"
|
https://github.com/huggingface/datasets/pull/2881
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2881",
"html_url": "https://github.com/huggingface/datasets/pull/2881",
"diff_url": "https://github.com/huggingface/datasets/pull/2881.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2881.patch",
"merged_at": "2021-09-13T14:20:40"
}
| 2,881
| true
|
Extend support for streaming datasets that use pathlib.Path stem/suffix
|
This PR extends the support in streaming mode for datasets that use `pathlib`, by patching the properties `pathlib.Path.stem` and `pathlib.Path.suffix`.
Related to #2876, #2874, #2866.
CC: @severo
|
https://github.com/huggingface/datasets/pull/2880
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2880",
"html_url": "https://github.com/huggingface/datasets/pull/2880",
"diff_url": "https://github.com/huggingface/datasets/pull/2880.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2880.patch",
"merged_at": "2021-09-09T13:13:29"
}
| 2,880
| true
|
In v1.4.1, all TIMIT train transcripts are "Would such an act of refusal be useful?"
|
## Describe the bug
Using version 1.4.1 of `datasets`, TIMIT transcripts are all the same.
## Steps to reproduce the bug
I was following this tutorial
- https://huggingface.co/blog/fine-tune-wav2vec2-english
But here's a distilled repro:
```python
!pip install datasets==1.4.1
from datasets import load_dataset
timit = load_dataset("timit_asr", cache_dir="./temp")
unique_transcripts = set(timit["train"]["text"])
print(unique_transcripts)
assert len(unique_transcripts) > 1
```
## Expected results
Expected the correct TIMIT data. Or an error saying that this version of `datasets` can't produce it.
## Actual results
Every train transcript was "Would such an act of refusal be useful?" Every test transcript was "The bungalow was pleasantly situated near the shore."
## Environment info
- `datasets` version: 1.4.1
- Platform: Darwin-18.7.0-x86_64-i386-64bit
- Python version: 3.7.9
- PyTorch version (GPU?): 1.9.0 (False)
- Tensorflow version (GPU?): not installed (NA)
- Using GPU in script?: tried both
- Using distributed or parallel set-up in script?: no
-
|
https://github.com/huggingface/datasets/issues/2879
|
[
"Hi @rcgale, thanks for reporting.\r\n\r\nPlease note that this bug was fixed on `datasets` version 1.5.0: https://github.com/huggingface/datasets/commit/a23c73e526e1c30263834164f16f1fdf76722c8c#diff-f12a7a42d4673bb6c2ca5a40c92c29eb4fe3475908c84fd4ce4fad5dc2514878\r\n\r\nIf you update `datasets` version, that should work.\r\n\r\nOn the other hand, would it be possible for @patrickvonplaten to update the [blog post](https://huggingface.co/blog/fine-tune-wav2vec2-english) with the correct version of `datasets`?",
"I just proposed a change in the blog post.\r\n\r\nI had assumed there was a data format change that broke a previous version of the code, since presumably @patrickvonplaten tested the tutorial with the version they explicitly referenced. But that fix you linked suggests a problem in the code, which surprised me.\r\n\r\nI still wonder, though, is there a way for downloads to be invalidated server-side? If the client can announce its version during a download request, perhaps the server could reject known incompatibilities? It would save much valuable time if `datasets` raised an informative error on a known problem (\"Error: the requested data set requires `datasets>=1.5.0`.\"). This kind of API versioning is a prudent move anyhow, as there will surely come a time when you'll need to make a breaking change to data.",
"Also, thank you for a quick and helpful reply!"
] | null | 2,879
| false
|
NotADirectoryError: [WinError 267] During load_from_disk
|
## Describe the bug
Trying to load saved dataset or dataset directory from Amazon S3 on a Windows machine fails.
Performing the same operation succeeds on non-windows environment (AWS Sagemaker).
## Steps to reproduce the bug
```python
# Followed https://huggingface.co/docs/datasets/filesystems.html#loading-a-processed-dataset-from-s3
from datasets import load_from_disk
from datasets.filesystems import S3FileSystem
s3_file = "output of save_to_disk"
s3_filesystem = S3FileSystem()
load_from_disk(s3_file, fs=s3_filesystem)
```
## Expected results
load_from_disk succeeds without error
## Actual results
Seems like it succeeds in pulling the file into a windows temp directory, as it exists in my system, but fails to process it.
```
Exception ignored in: <finalize object at 0x26409231ce0; dead>
Traceback (most recent call last):
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\weakref.py", line 566, in __call__
return info.func(*info.args, **(info.kwargs or {}))
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\tempfile.py", line 817, in _cleanup
cls._rmtree(name)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\tempfile.py", line 813, in _rmtree
_shutil.rmtree(name, onerror=onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 740, in rmtree
return _rmtree_unsafe(path, onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 613, in _rmtree_unsafe
_rmtree_unsafe(fullname, onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 613, in _rmtree_unsafe
_rmtree_unsafe(fullname, onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 613, in _rmtree_unsafe
_rmtree_unsafe(fullname, onerror)
[Previous line repeated 2 more times]
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 618, in _rmtree_unsafe
onerror(os.unlink, fullname, sys.exc_info())
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\tempfile.py", line 805, in onerror
cls._rmtree(path)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\tempfile.py", line 813, in _rmtree
_shutil.rmtree(name, onerror=onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 740, in rmtree
return _rmtree_unsafe(path, onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 599, in _rmtree_unsafe
onerror(os.scandir, path, sys.exc_info())
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 596, in _rmtree_unsafe
with os.scandir(path) as scandir_it:
NotADirectoryError: [WinError 267] The directory name is invalid: 'C:\\Users\\grassycup\\AppData\\Local\\Temp\\tmp45f_qbma\\tests3bucket\\output\\test_output\\train\\dataset.arrow'
Exception ignored in: <finalize object at 0x264091c7880; dead>
Traceback (most recent call last):
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\weakref.py", line 566, in __call__
return info.func(*info.args, **(info.kwargs or {}))
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\tempfile.py", line 817, in _cleanup
cls._rmtree(name)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\tempfile.py", line 813, in _rmtree
_shutil.rmtree(name, onerror=onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 740, in rmtree
return _rmtree_unsafe(path, onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 613, in _rmtree_unsafe
_rmtree_unsafe(fullname, onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 613, in _rmtree_unsafe
_rmtree_unsafe(fullname, onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 613, in _rmtree_unsafe
_rmtree_unsafe(fullname, onerror)
[Previous line repeated 2 more times]
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 618, in _rmtree_unsafe
onerror(os.unlink, fullname, sys.exc_info())
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\tempfile.py", line 805, in onerror
cls._rmtree(path)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\tempfile.py", line 813, in _rmtree
_shutil.rmtree(name, onerror=onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 740, in rmtree
return _rmtree_unsafe(path, onerror)
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 599, in _rmtree_unsafe
onerror(os.scandir, path, sys.exc_info())
File "C:\Users\grassycup\Anaconda3\envs\hello.world\lib\shutil.py", line 596, in _rmtree_unsafe
with os.scandir(path) as scandir_it:
NotADirectoryError: [WinError 267] The directory name is invalid:
'C:\\Users\\grassycup\\AppData\\Local\\Temp\\tmp45f_qbma\\tests3bucket\\output\\test_output\\train\\dataset.arrow'
```
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version: 1.11.0
- Platform: Windows-10-10.0.19042-SP0
- Python version: 3.8.11
- PyArrow version: 3.0.0
|
https://github.com/huggingface/datasets/issues/2878
|
[] | null | 2,878
| false
|
Don't keep the dummy data folder or dataset_infos.json when resolving data files
|
When there's no dataset script, all the data files of a folder or a repository on the Hub are loaded as data files.
There are already a few exceptions:
- files starting with "." are ignored
- the dataset card "README.md" is ignored
- any file named "config.json" is ignored (currently it isn't used anywhere, but it could be used in the future to define splits or configs for example, but not 100% sure)
However any data files in a folder named "dummy" should be ignored as well as they should only be used to test the dataset.
Same for "dataset_infos.json" which should only be used to get the `dataset.info`
|
https://github.com/huggingface/datasets/issues/2877
|
[
"Hi @lhoestq I am new to huggingface datasets, I would like to work on this issue!\r\n",
"Thanks for the help :) \r\n\r\nAs mentioned in the PR, excluding files named \"dummy_data.zip\" is actually more general than excluding the files inside a \"dummy\" folder. I just did the change in the PR, I think we can merge it now"
] | null | 2,877
| false
|
Extend support for streaming datasets that use pathlib.Path.glob
|
This PR extends the support in streaming mode for datasets that use `pathlib`, by patching the method `pathlib.Path.glob`.
Related to #2874, #2866.
CC: @severo
|
https://github.com/huggingface/datasets/pull/2876
|
[
"I am thinking that ideally we should call `fs.glob()` instead...",
"Thanks, @lhoestq: the idea of adding the mock filesystem is to avoid network calls and reduce testing time ;) \r\n\r\nI have added `rglob` as well and fixed some bugs."
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2876",
"html_url": "https://github.com/huggingface/datasets/pull/2876",
"diff_url": "https://github.com/huggingface/datasets/pull/2876.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2876.patch",
"merged_at": "2021-09-10T09:50:48"
}
| 2,876
| true
|
Add Congolese Swahili speech datasets
|
## Adding a Dataset
- **Name:** Congolese Swahili speech corpora
- **Data:** https://gamayun.translatorswb.org/data/
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
Also related: https://mobile.twitter.com/OktemAlp/status/1435196393631764482
|
https://github.com/huggingface/datasets/issues/2875
|
[] | null | 2,875
| false
|
Support streaming datasets that use pathlib
|
This PR extends the support in streaming mode for datasets that use `pathlib.Path`.
Related to: #2866.
CC: @severo
|
https://github.com/huggingface/datasets/pull/2874
|
[
"I've tried https://github.com/huggingface/datasets/issues/2866 again, and I get the same error.\r\n\r\n```python\r\nimport datasets as ds\r\nds.load_dataset('counter', split=\"train\", streaming=False)\r\n```",
"@severo Issue #2866 is not fully fixed yet: multiple patches need to be implemented for `pathlib`, as that dataset uses quite a lot of `pathlib` functions... π
",
"No worry and no stress, I just wanted to check for that case :) I'm very happy that you're working on issues I'm interested in!"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2874",
"html_url": "https://github.com/huggingface/datasets/pull/2874",
"diff_url": "https://github.com/huggingface/datasets/pull/2874.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2874.patch",
"merged_at": "2021-09-07T11:41:15"
}
| 2,874
| true
|
adding swedish_medical_ner
|
Adding the Swedish Medical NER dataset, listed in "Biomedical Datasets - BigScience Workshop 2021"
Code refactored
|
https://github.com/huggingface/datasets/pull/2873
|
[
"Hi, what's the current status of this request? It says Changes requested, but I can't see what changes?",
"Hi, it looks like this PR includes changes to other files that `swedish_medical_ner`.\r\n\r\nFeel free to remove these changes, or simply create a new PR that only contains the addition of the dataset"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2873",
"html_url": "https://github.com/huggingface/datasets/pull/2873",
"diff_url": "https://github.com/huggingface/datasets/pull/2873.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2873.patch",
"merged_at": null
}
| 2,873
| true
|
adding swedish_medical_ner
|
Adding the Swedish Medical NER dataset, listed in "Biomedical Datasets - BigScience Workshop 2021"
|
https://github.com/huggingface/datasets/pull/2872
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2872",
"html_url": "https://github.com/huggingface/datasets/pull/2872",
"diff_url": "https://github.com/huggingface/datasets/pull/2872.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2872.patch",
"merged_at": null
}
| 2,872
| true
|
datasets.config.PYARROW_VERSION has no attribute 'major'
|
In the test_dataset_common.py script, line 288-289
```
if datasets.config.PYARROW_VERSION.major < 3:
packaged_datasets = [pd for pd in packaged_datasets if pd["dataset_name"] != "parquet"]
```
which throws the error below. `datasets.config.PYARROW_VERSION` itself return the string '4.0.1'. I have tested this on both datasets.__version_=='1.11.0' and '1.9.0'. I am using Mac OS.
```
import datasets
datasets.config.PYARROW_VERSION.major
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
/var/folders/1f/0wqmlgp90qjd5mpj53fnjq440000gn/T/ipykernel_73361/2547517336.py in <module>
1 import datasets
----> 2 datasets.config.PYARROW_VERSION.major
AttributeError: 'str' object has no attribute 'major'
```
## Environment info
- `datasets` version: 1.11.0
- Platform: Darwin-20.6.0-x86_64-i386-64bit
- Python version: 3.7.11
- PyArrow version: 4.0.1
|
https://github.com/huggingface/datasets/issues/2871
|
[
"I have changed line 288 to `if int(datasets.config.PYARROW_VERSION.split(\".\")[0]) < 3:` just to get around it.",
"Hi @bwang482,\r\n\r\nI'm sorry but I'm not able to reproduce your bug.\r\n\r\nPlease note that in our current master branch, we made a commit (d03223d4d64b89e76b48b00602aba5aa2f817f1e) that simultaneously modified:\r\n- test_dataset_common.py: https://github.com/huggingface/datasets/commit/d03223d4d64b89e76b48b00602aba5aa2f817f1e#diff-a1bc225bd9a5bade373d1f140e24d09cbbdc97971c2f73bb627daaa803ada002L289 that introduces the usage of `datasets.config.PYARROW_VERSION.major`\r\n- but also changed config.py: https://github.com/huggingface/datasets/commit/d03223d4d64b89e76b48b00602aba5aa2f817f1e#diff-e021fcfc41811fb970fab889b8d245e68382bca8208e63eaafc9a396a336f8f2L40, so that `datasets.config.PYARROW_VERSION.major` exists\r\n",
"Sorted. Thanks!",
"Reopening this. Although the `test_dataset_common.py` script works fine now.\r\n\r\nHas this got something to do with my pull request not passing `ci/circleci: run_dataset_script_tests_pyarrow` tests?\r\n\r\nhttps://github.com/huggingface/datasets/pull/2873",
"Hi @bwang482,\r\n\r\nIf you click on `Details` (on the right of your non passing CI test names: `ci/circleci: run_dataset_script_tests_pyarrow`), you can have more information about the non-passing tests.\r\n\r\nFor example, for [\"ci/circleci: run_dataset_script_tests_pyarrow_1\" details](https://circleci.com/gh/huggingface/datasets/46324?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link), you can see that the only non-passing test has to do with the dataset card (missing information in the `README.md` file): `test_changed_dataset_card`\r\n```\r\n=========================== short test summary info ============================\r\nFAILED tests/test_dataset_cards.py::test_changed_dataset_card[swedish_medical_ner]\r\n= 1 failed, 3214 passed, 2874 skipped, 2 xfailed, 1 xpassed, 15 warnings in 175.59s (0:02:55) =\r\n```\r\n\r\nTherefore, your PR non-passing test has nothing to do with this issue."
] | null | 2,871
| false
|
Fix three typos in two files for documentation
|
Changed "bacth_size" to "batch_size" (2x)
Changed "intsructions" to "instructions"
|
https://github.com/huggingface/datasets/pull/2870
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2870",
"html_url": "https://github.com/huggingface/datasets/pull/2870",
"diff_url": "https://github.com/huggingface/datasets/pull/2870.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2870.patch",
"merged_at": "2021-09-06T08:19:35"
}
| 2,870
| true
|
TypeError: 'NoneType' object is not callable
|
## Describe the bug
TypeError: 'NoneType' object is not callable
## Steps to reproduce the bug
```python
from datasets import load_dataset, load_metric
dataset = datasets.load_dataset("glue", 'cola')
```
## Expected results
A clear and concise description of the expected results.
## Actual results
Specify the actual results or traceback.
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version: 1.11.0
- Platform:
- Python version: 3.7
- PyArrow version:
|
https://github.com/huggingface/datasets/issues/2869
|
[
"Hi, @Chenfei-Kang.\r\n\r\nI'm sorry, but I'm not able to reproduce your bug:\r\n```python\r\nfrom datasets import load_dataset\r\n\r\nds = load_dataset(\"glue\", 'cola')\r\nds\r\n```\r\n```\r\nDatasetDict({\r\n train: Dataset({\r\n features: ['sentence', 'label', 'idx'],\r\n num_rows: 8551\r\n })\r\n validation: Dataset({\r\n features: ['sentence', 'label', 'idx'],\r\n num_rows: 1043\r\n })\r\n test: Dataset({\r\n features: ['sentence', 'label', 'idx'],\r\n num_rows: 1063\r\n })\r\n})\r\n```\r\n\r\nCould you please give more details and environment info (platform, PyArrow version)?",
"> Hi, @Chenfei-Kang.\r\n> \r\n> I'm sorry, but I'm not able to reproduce your bug:\r\n> \r\n> ```python\r\n> from datasets import load_dataset\r\n> \r\n> ds = load_dataset(\"glue\", 'cola')\r\n> ds\r\n> ```\r\n> \r\n> ```\r\n> DatasetDict({\r\n> train: Dataset({\r\n> features: ['sentence', 'label', 'idx'],\r\n> num_rows: 8551\r\n> })\r\n> validation: Dataset({\r\n> features: ['sentence', 'label', 'idx'],\r\n> num_rows: 1043\r\n> })\r\n> test: Dataset({\r\n> features: ['sentence', 'label', 'idx'],\r\n> num_rows: 1063\r\n> })\r\n> })\r\n> ```\r\n> \r\n> Could you please give more details and environment info (platform, PyArrow version)?\r\n\r\nSorry to reply you so late.\r\nplatform: pycharm 2021 + anaconda with python 3.7\r\nPyArrow version: 5.0.0\r\nhuggingface-hub: 0.0.16\r\ndatasets: 1.9.0\r\n",
"- For the platform, we need to know the operating system of your machine. Could you please run the command `datasets-cli env` and copy-and-paste its output below?\r\n- In relation with the error, you just gave us the error type and message (`TypeError: 'NoneType' object is not callable`). Could you please copy-paste the complete stack trace, so that we know exactly which part of the code threw the error?",
"> * For the platform, we need to know the operating system of your machine. Could you please run the command `datasets-cli env` and copy-and-paste its output below?\r\n> * In relation with the error, you just gave us the error type and message (`TypeError: 'NoneType' object is not callable`). Could you please copy-paste the complete stack trace, so that we know exactly which part of the code threw the error?\r\n\r\n1. For the platform, here are the output:\r\n - datasets` version: 1.11.0\r\n - Platform: Windows-10-10.0.19041-SP0\r\n - Python version: 3.7.10\r\n - PyArrow version: 5.0.0\r\n2. For the code and errorοΌ\r\n ```python\r\n from datasets import load_dataset, load_metric\r\n dataset = load_dataset(\"glue\", \"cola\")\r\n ```\r\n ```python\r\n Traceback (most recent call last):\r\n ....\r\n ....\r\n File \"my_file.py\", line 2, in <module>\r\n dataset = load_dataset(\"glue\", \"cola\")\r\n File \"My environments\\lib\\site-packages\\datasets\\load.py\", line 830, in load_dataset\r\n **config_kwargs,\r\n File \"My environments\\lib\\site-packages\\datasets\\load.py\", line 710, in load_dataset_builder\r\n **config_kwargs,\r\n TypeError: 'NoneType' object is not callable\r\n ```\r\n Thank you!",
"For that environment, I am sorry but I can't reproduce the bug: I can load the dataset without any problem.",
"One naive question: do you have internet access from the machine where you execute the code?",
"> For that environment, I am sorry but I can't reproduce the bug: I can load the dataset without any problem.\r\n\r\nBut I can download other task dataset such as `dataset = load_dataset('squad')`. I don't know what went wrong. Thank you so much!",
"Hi,friends. I meet the same problem. Do you have a way to fix this? Thanks!\r\n"
] | null | 2,869
| false
|
Add Common Objects in 3D (CO3D)
|
## Adding a Dataset
- **Name:** *Common Objects in 3D (CO3D)*
- **Description:** *See blog post [here](https://ai.facebook.com/blog/common-objects-in-3d-dataset-for-3d-reconstruction)*
- **Paper:** *[link to paper](https://arxiv.org/abs/2109.00512)*
- **Data:** *[link to data](https://ai.facebook.com/datasets/co3d-downloads/)*
- **Motivation:** *excerpt from above blog post:*
> As the first data set of its kind, CO3D will aptly enable reconstruction of real-life 3D objects. Indeed, CO3D already provides training data to enable our NeRFormer to tackle the new-view synthesis (NVS) task. Here, photorealistic NVS is a major step on the path to fully immersive AR/VR effects, where objects can be virtually transported across different environments, which will allow connecting users by sharing or recollecting their experiences.
>
> Besides practical applications in AR/VR, we hope that the data set will become a standard testbed for the recent proliferation of methods (including NeRFormer, Implicit Differentiable Renderer, NeRF, and others) that reconstruct 3D scenes by means of an implicit shape model.
>
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
|
https://github.com/huggingface/datasets/issues/2868
|
[] | null | 2,868
| false
|
Add CaSiNo dataset
|
Hi. I request you to add our dataset to the repository.
This data was recently published at NAACL 2021: https://aclanthology.org/2021.naacl-main.254.pdf
|
https://github.com/huggingface/datasets/pull/2867
|
[
"Hi @lhoestq \r\n\r\nJust a request to look at the dataset. Please let me know if any changes are necessary before merging it into the repo. Thank you.",
"Hey @lhoestq \r\n\r\nThanks for merging it. One question: I still cannot find the dataset on https://huggingface.co/datasets. Does it take some time or did I miss something?",
"Hi ! It takes a few hours or a day for the list of datasets on the website to be updated ;)"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2867",
"html_url": "https://github.com/huggingface/datasets/pull/2867",
"diff_url": "https://github.com/huggingface/datasets/pull/2867.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2867.patch",
"merged_at": "2021-09-16T09:23:44"
}
| 2,867
| true
|
"counter" dataset raises an error in normal mode, but not in streaming mode
|
## Describe the bug
`counter` dataset raises an error on `load_dataset()`, but simply returns an empty iterator in streaming mode.
## Steps to reproduce the bug
```python
>>> import datasets as ds
>>> a = ds.load_dataset('counter', split="train", streaming=False)
Using custom data configuration default
Downloading and preparing dataset counter/default (download: 1.29 MiB, generated: 2.48 MiB, post-processed: Unknown size, total: 3.77 MiB) to /home/slesage/.cache/huggingface/datasets/counter/default/1.0.0/9f84962fa0f35bec5a34fe0bdff8681838d497008c457f7856c48654476ec0e9...
Traceback (most recent call last):
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/builder.py", line 726, in _download_and_prepare
self._prepare_split(split_generator, **prepare_split_kwargs)
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/builder.py", line 1124, in _prepare_split
for key, record in utils.tqdm(
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/tqdm/std.py", line 1185, in __iter__
for obj in iterable:
File "/home/slesage/.cache/huggingface/modules/datasets_modules/datasets/counter/9f84962fa0f35bec5a34fe0bdff8681838d497008c457f7856c48654476ec0e9/counter.py", line 161, in _generate_examples
with derived_file.open(encoding="utf-8") as f:
File "/home/slesage/.pyenv/versions/3.8.11/lib/python3.8/pathlib.py", line 1222, in open
return io.open(self, mode, buffering, encoding, errors, newline,
File "/home/slesage/.pyenv/versions/3.8.11/lib/python3.8/pathlib.py", line 1078, in _opener
return self._accessor.open(self, flags, mode)
FileNotFoundError: [Errno 2] No such file or directory: '/home/slesage/.cache/huggingface/datasets/downloads/extracted/b57aa6db5601a738e57b95c1fd8cced54ff28fc540efcdaf0f6c4f1bb5dfe211/COUNTER/0032p.xml'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/load.py", line 1112, in load_dataset
builder_instance.download_and_prepare(
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/builder.py", line 636, in download_and_prepare
self._download_and_prepare(
File "/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.8/site-packages/datasets/builder.py", line 728, in _download_and_prepare
raise OSError(
OSError: Cannot find data file.
Original error:
[Errno 2] No such file or directory: '/home/slesage/.cache/huggingface/datasets/downloads/extracted/b57aa6db5601a738e57b95c1fd8cced54ff28fc540efcdaf0f6c4f1bb5dfe211/COUNTER/0032p.xml'
```
```python
>>> import datasets as ds
>>> b = ds.load_dataset('counter', split="train", streaming=True)
Using custom data configuration default
>>> list(b)
[]
```
## Expected results
An exception should be raised in streaming mode
## Actual results
No exception is raised in streaming mode: there is no way to tell if something has broken or if the dataset is simply empty.
## Environment info
- `datasets` version: 1.11.1.dev0
- Platform: Linux-5.11.0-1016-aws-x86_64-with-glibc2.29
- Python version: 3.8.11
- PyArrow version: 4.0.1
|
https://github.com/huggingface/datasets/issues/2866
|
[
"Hi @severo, thanks for reporting.\r\n\r\nJust note that currently not all canonical datasets support streaming mode: this is one case!\r\n\r\nAll datasets that use `pathlib` joins (using `/`) instead of `os.path.join` (as in this dataset) do not support streaming mode yet.",
"OK. Do you think it's possible to detect this, and raise an exception (maybe `NotImplementedError`, or a specific `StreamingError`)?",
"We should definitely support datasets using `pathlib` in streaming mode...\r\n\r\nFor non-supported datasets in streaming mode, we have already a request of raising an error/warning: see #2654.",
"Hi @severo, please note that \"counter\" dataset will be streamable (at least until it arrives at the missing file, error already in normal mode) once these PRs are merged:\r\n- #2874\r\n- #2876\r\n- #2880\r\n\r\nI have tested it. π ",
"Now (on master), we get:\r\n\r\n```\r\nimport datasets as ds\r\nds.load_dataset('counter', split=\"train\", streaming=False)\r\n```\r\n\r\n```\r\nUsing custom data configuration default\r\nDownloading and preparing dataset counter/default (download: 1.29 MiB, generated: 2.48 MiB, post-processed: Unknown size, total: 3.77 MiB) to /home/slesage/.cache/huggingface/datasets/counter/default/1.0.0/9f84962fa0f35bec5a34fe0bdff8681838d497008c457f7856c48654476ec0e9...\r\nTraceback (most recent call last):\r\n File \"/home/slesage/hf/datasets/src/datasets/builder.py\", line 726, in _download_and_prepare\r\n self._prepare_split(split_generator, **prepare_split_kwargs)\r\n File \"/home/slesage/hf/datasets/src/datasets/builder.py\", line 1124, in _prepare_split\r\n for key, record in utils.tqdm(\r\n File \"/home/slesage/hf/datasets/.venv/lib/python3.8/site-packages/tqdm/std.py\", line 1185, in __iter__\r\n for obj in iterable:\r\n File \"/home/slesage/.cache/huggingface/modules/datasets_modules/datasets/counter/9f84962fa0f35bec5a34fe0bdff8681838d497008c457f7856c48654476ec0e9/counter.py\", line 161, in _generate_examples\r\n with derived_file.open(encoding=\"utf-8\") as f:\r\n File \"/home/slesage/.pyenv/versions/3.8.11/lib/python3.8/pathlib.py\", line 1222, in open\r\n return io.open(self, mode, buffering, encoding, errors, newline,\r\n File \"/home/slesage/.pyenv/versions/3.8.11/lib/python3.8/pathlib.py\", line 1078, in _opener\r\n return self._accessor.open(self, flags, mode)\r\nFileNotFoundError: [Errno 2] No such file or directory: '/home/slesage/.cache/huggingface/datasets/downloads/extracted/b57aa6db5601a738e57b95c1fd8cced54ff28fc540efcdaf0f6c4f1bb5dfe211/COUNTER/0032p.xml'\r\n\r\nDuring handling of the above exception, another exception occurred:\r\n\r\nTraceback (most recent call last):\r\n File \"<stdin>\", line 1, in <module>\r\n File \"/home/slesage/hf/datasets/src/datasets/load.py\", line 1112, in load_dataset\r\n builder_instance.download_and_prepare(\r\n File \"/home/slesage/hf/datasets/src/datasets/builder.py\", line 636, in download_and_prepare\r\n self._download_and_prepare(\r\n File \"/home/slesage/hf/datasets/src/datasets/builder.py\", line 728, in _download_and_prepare\r\n raise OSError(\r\nOSError: Cannot find data file.\r\nOriginal error:\r\n[Errno 2] No such file or directory: '/home/slesage/.cache/huggingface/datasets/downloads/extracted/b57aa6db5601a738e57b95c1fd8cced54ff28fc540efcdaf0f6c4f1bb5dfe211/COUNTER/0032p.xml'\r\n```\r\n\r\nThe error is now the same with or without streaming. I close the issue, thanks @albertvillanova and @lhoestq!\r\n",
"Note that we might want to open an issue to fix the \"counter\" dataset by itself, but I let it up to you.",
"Fixed here: https://github.com/huggingface/datasets/pull/2894. Thanks @albertvillanova ",
"On master, I get:\r\n\r\n```python\r\n>>> import datasets as ds\r\n>>> iterable_dataset = ds.load_dataset('counter', split=\"train\", streaming=True)\r\n>>> rows = list(iterable_dataset.take(100))\r\nTraceback (most recent call last):\r\n File \"<stdin>\", line 1, in <module>\r\n File \"/home/slesage/hf/datasets/src/datasets/iterable_dataset.py\", line 341, in __iter__\r\n for key, example in self._iter():\r\n File \"/home/slesage/hf/datasets/src/datasets/iterable_dataset.py\", line 338, in _iter\r\n yield from ex_iterable\r\n File \"/home/slesage/hf/datasets/src/datasets/iterable_dataset.py\", line 273, in __iter__\r\n yield from islice(self.ex_iterable, self.n)\r\n File \"/home/slesage/hf/datasets/src/datasets/iterable_dataset.py\", line 78, in __iter__\r\n for key, example in self.generate_examples_fn(**self.kwargs):\r\n File \"/home/slesage/.cache/huggingface/modules/datasets_modules/datasets/counter/b9e4378dbd3f5ce235d2302e48168c00196e67bbcd13cc7e1f6e69ef82c0cf2a/counter.py\", line 153, in _generate_examples\r\n files = sorted(base_path.glob(r\"[0-9][0-9][0-9][0-9].xml\"))\r\nTypeError: xpathglob() missing 1 required positional argument: 'pattern'\r\n```",
"Associated to the above exception, if I create a test and run it with pytest, I get an awful traceback.\r\n\r\n- create a file `test_counter.py`\r\n\r\n```python\r\nimport pytest\r\nfrom datasets import load_dataset, IterableDataset\r\nfrom typing import Any, cast\r\n\r\n\r\ndef test_counter() -> Any:\r\n iterable_dataset = cast(IterableDataset, load_dataset(\"counter\", split=\"train\", streaming=True))\r\n with pytest.raises(TypeError):\r\n list(iterable_dataset.take(100))\r\n```\r\n\r\n- run the test with pytest\r\n\r\n```bash\r\n$ python -m pytest -x test_counter.py\r\n============================================================================================================================= test session starts ==============================================================================================================================\r\nplatform linux -- Python 3.9.6, pytest-6.2.5, py-1.10.0, pluggy-1.0.0\r\nrootdir: /home/slesage/hf/datasets-preview-backend, configfile: pyproject.toml\r\nplugins: anyio-3.3.2, cov-2.12.1\r\ncollected 1 item\r\n\r\ntests/test_counter.py . [100%]Traceback (most recent call last):\r\n File \"/home/slesage/.pyenv/versions/3.9.6/lib/python3.9/runpy.py\", line 197, in _run_module_as_main\r\n return _run_code(code, main_globals, None,\r\n File \"/home/slesage/.pyenv/versions/3.9.6/lib/python3.9/runpy.py\", line 87, in _run_code\r\n exec(code, run_globals)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pytest/__main__.py\", line 5, in <module>\r\n raise SystemExit(pytest.console_main())\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/_pytest/config/__init__.py\", line 185, in console_main\r\n code = main()\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/_pytest/config/__init__.py\", line 162, in main\r\n ret: Union[ExitCode, int] = config.hook.pytest_cmdline_main(\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_hooks.py\", line 265, in __call__\r\n return self._hookexec(self.name, self.get_hookimpls(), kwargs, firstresult)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_manager.py\", line 80, in _hookexec\r\n return self._inner_hookexec(hook_name, methods, kwargs, firstresult)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_callers.py\", line 60, in _multicall\r\n return outcome.get_result()\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_result.py\", line 60, in get_result\r\n raise ex[1].with_traceback(ex[2])\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_callers.py\", line 39, in _multicall\r\n res = hook_impl.function(*args)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/_pytest/main.py\", line 316, in pytest_cmdline_main\r\n return wrap_session(config, _main)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/_pytest/main.py\", line 304, in wrap_session\r\n config.hook.pytest_sessionfinish(\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_hooks.py\", line 265, in __call__\r\n return self._hookexec(self.name, self.get_hookimpls(), kwargs, firstresult)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_manager.py\", line 80, in _hookexec\r\n return self._inner_hookexec(hook_name, methods, kwargs, firstresult)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_callers.py\", line 55, in _multicall\r\n gen.send(outcome)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/_pytest/terminal.py\", line 803, in pytest_sessionfinish\r\n outcome.get_result()\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_result.py\", line 60, in get_result\r\n raise ex[1].with_traceback(ex[2])\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/pluggy/_callers.py\", line 39, in _multicall\r\n res = hook_impl.function(*args)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/_pytest/cacheprovider.py\", line 428, in pytest_sessionfinish\r\n config.cache.set(\"cache/nodeids\", sorted(self.cached_nodeids))\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/_pytest/cacheprovider.py\", line 188, in set\r\n f = path.open(\"w\")\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/datasets/utils/streaming_download_manager.py\", line 199, in xpathopen\r\n return xopen(_as_posix(path), *args, **kwargs)\r\n File \"/home/slesage/hf/datasets-preview-backend/.venv/lib/python3.9/site-packages/datasets/utils/streaming_download_manager.py\", line 117, in _as_posix\r\n path_as_posix = path.as_posix()\r\nAttributeError: 'str' object has no attribute 'as_posix'\r\n```\r\n",
"I opened a PR to fix these issues.\r\nAlso in your test you expect a TypeError but I don't know why. On my side it works fine without raising a TypeError",
"I had the issue (TypeError raised) on my branch, but it's fixed now. Thanks"
] | null | 2,866
| false
|
Add MultiEURLEX dataset
|
**Add new MultiEURLEX Dataset**
MultiEURLEX comprises 65k EU laws in 23 official EU languages (some low-ish resource). Each EU law has been annotated with EUROVOC concepts (labels) by the Publication Office of EU. As with the English EURLEX, the goal is to predict the relevant EUROVOC concepts (labels); this is multi-label classification task (given the text, predict multiple labels).
|
https://github.com/huggingface/datasets/pull/2865
|
[
"Hi @lhoestq, we have this new cool multilingual dataset coming at EMNLP 2021. It would be really nice if we could have it in Hugging Face asap. Thanks! ",
"Hi @lhoestq, I adopted most of your suggestions:\r\n\r\n- Dummy data files reduced, including the 2 smallest documents per subset JSONL.\r\n- README was updated with the publication URL and instructions on how to download and use label descriptors. Excessive newlines were deleted.\r\n\r\nI would prefer to keep the label list in a pure format (original ids), to enable people to combine those with more information or possibly in the future explore the dataset, find inconsistencies and fix those to release a new version. ",
"Thanks for the changes :)\r\n\r\nRegarding the labels:\r\n\r\nIf you use the ClassLabel feature type, the only change is that it will store the ids as integers instead of (currently) string.\r\nThe advantage is that if people want to know what id corresponds to which label name, they can use `classlabel.int2str`. It is also the format that helps automate model training for classification in `transformers`.\r\n\r\nLet me know if that sounds good to you or if you still want to stick with the labels as they are now.",
"Hey @lhoestq, thanks for providing this information. This sounds great. I updated my code accordingly to use `ClassLabel`. Could you please provide a minimal example of how `classlabel.int2str` works in practice in my case, where labels are a sequence?\r\n\r\n```python\r\nfrom datasets import load_dataset\r\ndataset = load_dataset('multi_eurlex', 'all_languages')\r\n# Read strs from the labels (list of integers) for the 1st sample of the training split\r\n```\r\n\r\nI would like to include this in the README file.\r\n\r\nCould you also provide some info on how I could define the supervized key to automate model training, as you said?\r\n\r\nThanks!",
"Thanks for the update :)\r\n\r\nHere is an example of usage:\r\n```python\r\nfrom datasets import load_dataset\r\ndataset = load_dataset('multi_eurlex', 'all_languages', split='train')\r\nclasslabel = dataset.features[\"labels\"].feature\r\nprint(dataset[0][\"labels\"])\r\n# [1, 20, 7, 3, 0]\r\nprint(classlabel.int2str(dataset[0][\"labels\"]))\r\n# ['100160', '100155', '100158', '100147', '100149']\r\n```\r\n\r\nThe ClassLabel is simply used to define the `id2label` dictionary of classification models, to make the ids match between the model and the dataset. There nothing more to do :p \r\n\r\nI think one last thing to do is just update the `dataset_infos.json` file and we'll be good !",
"Everything is ready! π \r\n"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2865",
"html_url": "https://github.com/huggingface/datasets/pull/2865",
"diff_url": "https://github.com/huggingface/datasets/pull/2865.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2865.patch",
"merged_at": "2021-09-10T11:50:06"
}
| 2,865
| true
|
Fix data URL in ToTTo dataset
|
Data source host changed their data URL: google-research-datasets/ToTTo@cebeb43.
Fix #2860.
|
https://github.com/huggingface/datasets/pull/2864
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2864",
"html_url": "https://github.com/huggingface/datasets/pull/2864",
"diff_url": "https://github.com/huggingface/datasets/pull/2864.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2864.patch",
"merged_at": "2021-09-02T06:47:40"
}
| 2,864
| true
|
Update dataset URL
| null |
https://github.com/huggingface/datasets/pull/2863
|
[
"Superseded by PR #2864.\r\n\r\n@mrm8488 next time you would like to work on an issue, you can first self-assign it to you (by writing `#self-assign` in a comment on the issue). That way, other people can see you are already working on it and there are not multiple people working on the same issue. π "
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2863",
"html_url": "https://github.com/huggingface/datasets/pull/2863",
"diff_url": "https://github.com/huggingface/datasets/pull/2863.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2863.patch",
"merged_at": null
}
| 2,863
| true
|
fix: π be more specific when catching exceptions
|
The same specific exception is catched in other parts of the same
function.
|
https://github.com/huggingface/datasets/pull/2861
|
[
"To give more context: after our discussion, if I understood properly, you are trying to fix a call to `datasets` that takes 15 minutes: https://github.com/huggingface/datasets-preview-backend/issues/17 Is this right?\r\n\r\n",
"Yes, that's it. And to do that I'm trying to use https://pypi.org/project/stopit/, which will raise a stopit.TimeoutException exception. But currently, if this exception is raised, it's caught and considered as a \"FileNotFoundError\" while it should not be caught. ",
"And what about passing the `timeout` parameter instead?",
"It might be a good idea, but I would have to add a timeout argument to several methods, I'm not sure we want that (I want to ensure all my queries in https://github.com/huggingface/datasets-preview-backend/tree/master/src/datasets_preview_backend/queries resolve in a given time, be it with an error in case of timeout, or with the successful response). The methods are `prepare_module`, `import_main_class`, *builder_cls.*`get_all_exported_dataset_infos`, `load_dataset_builder`, and `load_dataset`",
"I understand, you are trying to find a fix for your use case. OK.\r\n\r\nJust note that it is also an issue for `datasets` users. Once #2859 fixed in `datasets`, you will no longer have this issue...",
"Closing, since 1. my problem is more #2859, and I was asking for that change in order to make a hack work on my side, 2. if we want to change how exceptions are handled, we surely want to do it on all the codebase, not only in this particular case."
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2861",
"html_url": "https://github.com/huggingface/datasets/pull/2861",
"diff_url": "https://github.com/huggingface/datasets/pull/2861.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2861.patch",
"merged_at": null
}
| 2,861
| true
|
Cannot download TOTTO dataset
|
Error: Couldn't find file at https://storage.googleapis.com/totto/totto_data.zip
`datasets version: 1.11.0`
# How to reproduce:
```py
from datasets import load_dataset
dataset = load_dataset('totto')
```
|
https://github.com/huggingface/datasets/issues/2860
|
[
"Hola @mrm8488, thanks for reporting.\r\n\r\nApparently, the data source host changed their URL one week ago: https://github.com/google-research-datasets/ToTTo/commit/cebeb430ec2a97747e704d16a9354f7d9073ff8f\r\n\r\nI'm fixing it."
] | null | 2,860
| false
|
Loading allenai/c4 in streaming mode does too many HEAD requests
|
This does 60,000+ HEAD requests to get all the ETags of all the data files:
```python
from datasets import load_dataset
load_dataset("allenai/c4", streaming=True)
```
It makes loading the dataset completely impractical.
The ETags are used to compute the config id (it must depend on the data files being used).
Instead of using the ETags, we could simply use the commit hash of the dataset repository on the hub, as well and the glob pattern used to resolve the files (here it's `*` by default, to load all the files of the repository)
|
https://github.com/huggingface/datasets/issues/2859
|
[
"https://github.com/huggingface/datasets/blob/6c766f9115d686182d76b1b937cb27e099c45d68/src/datasets/builder.py#L179-L186",
"Thanks a lot!!!"
] | null | 2,859
| false
|
Fix s3fs version in CI
|
The latest s3fs version has new constrains on aiobotocore, and therefore on boto3 and botocore
This PR changes the constrains to avoid the new conflicts
In particular it pins the version of s3fs.
|
https://github.com/huggingface/datasets/pull/2858
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2858",
"html_url": "https://github.com/huggingface/datasets/pull/2858",
"diff_url": "https://github.com/huggingface/datasets/pull/2858.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2858.patch",
"merged_at": "2021-08-31T21:29:51"
}
| 2,858
| true
|
Update: Openwebtext - update size
|
Update the size of the Openwebtext dataset
I also regenerated the dataset_infos.json but the data file checksum didn't change, and the number of examples either (8013769 examples)
Close #2839, close #726.
|
https://github.com/huggingface/datasets/pull/2857
|
[
"merging since the CI error in unrelated to this PR and fixed on master"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2857",
"html_url": "https://github.com/huggingface/datasets/pull/2857",
"diff_url": "https://github.com/huggingface/datasets/pull/2857.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2857.patch",
"merged_at": "2021-09-07T09:44:32"
}
| 2,857
| true
|
fix: π remove URL's query string only if it's ?dl=1
|
A lot of URL use the query strings, for example
http://opus.nlpl.eu/download.php?f=Bianet/v1/moses/en-ku.txt.zip, we
must not remove it when trying to detect the protocol. We thus remove it
only in the case of the query string being ?dl=1 which occurs on dropbox
and dl.orangedox.com. Also: add unit tests.
See https://github.com/huggingface/datasets/pull/2843 for the original
discussion.
|
https://github.com/huggingface/datasets/pull/2856
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2856",
"html_url": "https://github.com/huggingface/datasets/pull/2856",
"diff_url": "https://github.com/huggingface/datasets/pull/2856.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2856.patch",
"merged_at": "2021-08-31T14:22:12"
}
| 2,856
| true
|
Fix windows CI CondaError
|
From this thread: https://github.com/conda/conda/issues/6057
We can fix the conda error
```
CondaError: Cannot link a source that does not exist.
C:\Users\...\Anaconda3\Scripts\conda.exe
```
by doing
```bash
conda update conda
```
before doing any install in the windows CI
|
https://github.com/huggingface/datasets/pull/2855
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2855",
"html_url": "https://github.com/huggingface/datasets/pull/2855",
"diff_url": "https://github.com/huggingface/datasets/pull/2855.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2855.patch",
"merged_at": "2021-08-31T13:35:33"
}
| 2,855
| true
|
Fix caching when moving script
|
When caching the result of a `map` function, the hash that is computed depends on many properties of this function, such as all the python objects it uses, its code and also the location of this code.
Using the full path of the python script for the location of the code makes the hash change if a script like `run_mlm.py` is moved.
I changed this by simply using the base name of the script instead of the full path.
Note that this change also affects the hash of the code used from imported modules, but I think it's fine. Indeed it hashes the code of the imported modules anyway, so the location of the python files of the imported modules doesn't matter when computing the hash.
Close https://github.com/huggingface/datasets/issues/2825
|
https://github.com/huggingface/datasets/pull/2854
|
[
"Merging since the CI failure is unrelated to this PR"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2854",
"html_url": "https://github.com/huggingface/datasets/pull/2854",
"diff_url": "https://github.com/huggingface/datasets/pull/2854.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2854.patch",
"merged_at": "2021-08-31T13:13:36"
}
| 2,854
| true
|
Add AMI dataset
|
This is an initial commit for AMI dataset
|
https://github.com/huggingface/datasets/pull/2853
|
[
"Hey @cahya-wirawan, \r\n\r\nI played around with the dataset a bit and it looks already very good to me! That's exactly how it should be constructed :-) I can help you a bit with defining the config, etc... on Monday!",
"@lhoestq - I think the dataset is ready to be merged :-) \r\n\r\nAt the moment, I don't really see how the failing tests correspond to this PR:\r\n- https://app.circleci.com/pipelines/github/huggingface/datasets/7838/workflows/932a40a2-3e11-48be-84f0-c6434510058e/jobs/48318?invite=true#step-107-18\r\n- https://app.circleci.com/pipelines/github/huggingface/datasets/7838/workflows/932a40a2-3e11-48be-84f0-c6434510058e/jobs/48316?invite=true#step-102-136\r\n\r\ncould you maybe give it a look? :-)"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2853",
"html_url": "https://github.com/huggingface/datasets/pull/2853",
"diff_url": "https://github.com/huggingface/datasets/pull/2853.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2853.patch",
"merged_at": "2021-09-29T09:19:18"
}
| 2,853
| true
|
Fix: linnaeus - fix url
|
The url was causing a `ConnectionError` because of the "/" at the end
Close https://github.com/huggingface/datasets/issues/2821
|
https://github.com/huggingface/datasets/pull/2852
|
[
"Merging since the CI error is unrelated this this PR"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2852",
"html_url": "https://github.com/huggingface/datasets/pull/2852",
"diff_url": "https://github.com/huggingface/datasets/pull/2852.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2852.patch",
"merged_at": "2021-08-31T13:12:09"
}
| 2,852
| true
|
Update `column_names` showed as `:func:` in exploring.st
|
Hi,
One mention of `column_names` in exploring.st was showing it as `:func:` instead of `:attr:`.
|
https://github.com/huggingface/datasets/pull/2851
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2851",
"html_url": "https://github.com/huggingface/datasets/pull/2851",
"diff_url": "https://github.com/huggingface/datasets/pull/2851.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2851.patch",
"merged_at": "2021-08-31T14:45:46"
}
| 2,851
| true
|
Wound segmentation datasets
|
## Adding a Dataset
- **Name:** Wound segmentation datasets
- **Description:** annotated wound image dataset
- **Paper:** https://www.nature.com/articles/s41598-020-78799-w
- **Data:** https://github.com/uwm-bigdata/wound-segmentation
- **Motivation:** Interesting simple image dataset, useful for segmentation, with visibility due to http://www.miccai.org/special-interest-groups/challenges/ and https://fusc.grand-challenge.org/
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
|
https://github.com/huggingface/datasets/issues/2850
|
[] | null | 2,850
| false
|
Add Open Catalyst Project Dataset
|
## Adding a Dataset
- **Name:** Open Catalyst 2020 (OC20) Dataset
- **Website:** https://opencatalystproject.org/
- **Data:** https://github.com/Open-Catalyst-Project/ocp/blob/master/DATASET.md
Instructions to add a new dataset can be found [here](https://github.com/huggingface/datasets/blob/master/ADD_NEW_DATASET.md).
|
https://github.com/huggingface/datasets/issues/2849
|
[] | null | 2,849
| false
|
Update README.md
|
Changed 'Tain' to 'Train'.
|
https://github.com/huggingface/datasets/pull/2848
|
[
"Merging since the CI error is unrelated to this PR and fixed on master"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2848",
"html_url": "https://github.com/huggingface/datasets/pull/2848",
"diff_url": "https://github.com/huggingface/datasets/pull/2848.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2848.patch",
"merged_at": "2021-09-07T09:40:32"
}
| 2,848
| true
|
fix regex to accept negative timezone
|
fix #2846
|
https://github.com/huggingface/datasets/pull/2847
|
[] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2847",
"html_url": "https://github.com/huggingface/datasets/pull/2847",
"diff_url": "https://github.com/huggingface/datasets/pull/2847.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2847.patch",
"merged_at": "2021-09-07T09:34:23"
}
| 2,847
| true
|
Negative timezone
|
## Describe the bug
The load_dataset method do not accept a parquet file with a negative timezone, as it has the following regex:
```
"^(s|ms|us|ns),\s*tz=([a-zA-Z0-9/_+:]*)$"
```
So a valid timestap ```timestamp[us, tz=-03:00]``` returns an error when loading parquet files.
## Steps to reproduce the bug
```python
# Where the timestamp column has a tz of -03:00
datasets = load_dataset('parquet', data_files={'train': train_files, 'validation': validation_files,
'test': test_files}, cache_dir="./cache_teste/")
```
## Expected results
The -03:00 is a valid tz so the regex should accept this without raising an error.
## Actual results
As this regex disaproves a valid tz it raises the following error:
```python
raise ValueError(
f"{datasets_dtype} is not a validly formatted string representation of a pyarrow timestamp."
f"Examples include timestamp[us] or timestamp[us, tz=America/New_York]"
f"See: https://arrow.apache.org/docs/python/generated/pyarrow.timestamp.html#pyarrow.timestamp"
)
```
## Environment info
<!-- You can run the command `datasets-cli env` and copy-and-paste its output below. -->
- `datasets` version: 1.11.0
- Platform: Ubuntu 20.04
- Python version: 3.8
- PyArrow version: 5.0.0
|
https://github.com/huggingface/datasets/issues/2846
|
[
"Fixed by #2847."
] | null | 2,846
| false
|
[feature request] adding easy to remember `datasets.cache_dataset()` + `datasets.is_dataset_cached()`
|
Often, there is a need to prepare a dataset but not use it immediately, e.g. think tests suite setup, so it'd be really useful to be able to do:
```
if not datasets.is_dataset_cached(ds): datasets.cache_dataset(ds)
```
This can already be done with:
```
builder = load_dataset_builder(ds)
if not os.path.idsir(builder.cache_dir):
builder.download_and_prepare()
```
but the current way is a way less intuitive and much harder to remember than the proposed API, IMHO.
One more way is to do:
```
_ = load_dataset(ds)
```
but it wastes resources loading the dataset when it's not needed.
this has been discussed at https://huggingface.slack.com/archives/C01229B19EX/p1630021912025800
Thank you!
@lhoestq
|
https://github.com/huggingface/datasets/issues/2845
|
[] | null | 2,845
| false
|
Fix: wikicorpus - fix keys
|
As mentioned in https://github.com/huggingface/datasets/issues/2552, there is a duplicate keys error in `wikicorpus`.
I fixed that by taking into account the file index in the keys
|
https://github.com/huggingface/datasets/pull/2844
|
[
"The CI error is unrelated to this PR\r\n\r\n... merging !"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2844",
"html_url": "https://github.com/huggingface/datasets/pull/2844",
"diff_url": "https://github.com/huggingface/datasets/pull/2844.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2844.patch",
"merged_at": "2021-09-06T14:07:27"
}
| 2,844
| true
|
Fix extraction protocol inference from urls with params
|
Previously it was unable to infer the compression protocol for files at URLs like
```
https://foo.bar/train.json.gz?dl=1
```
because of the query parameters.
I fixed that, this should allow 10+ datasets to work in streaming mode:
```
"discovery",
"emotion",
"grail_qa",
"guardian_authorship",
"pragmeval",
"simple_questions_v2",
"versae/adobo",
"w-nicole/childes_data",
"w-nicole/childes_data_no_tags_",
"w-nicole/childes_data_with_tags",
"w-nicole/childes_data_with_tags_"
```
cc @severo
|
https://github.com/huggingface/datasets/pull/2843
|
[
"merging since the windows error is just a CircleCI issue",
"It works, eg https://observablehq.com/@huggingface/datasets-preview-backend-client#{%22datasetId%22%3A%22discovery%22} and https://datasets-preview.huggingface.tech/rows?dataset=discovery&config=discovery&split=train",
"Nice !"
] |
{
"url": "https://api.github.com/repos/huggingface/datasets/pulls/2843",
"html_url": "https://github.com/huggingface/datasets/pull/2843",
"diff_url": "https://github.com/huggingface/datasets/pull/2843.diff",
"patch_url": "https://github.com/huggingface/datasets/pull/2843.patch",
"merged_at": "2021-08-30T13:12:01"
}
| 2,843
| true
|
always requiring the username in the dataset name when there is one
|
Me and now another person have been bitten by the `datasets`'s non-strictness on requiring a dataset creator's username when it's due.
So both of us started with `stas/openwebtext-10k`, somewhere along the lines lost `stas/` and continued using `openwebtext-10k` and it all was good until we published the software and things broke, since there is no `openwebtext-10k`
So this feature request is asking to tighten the checking and not allow dataset loading if it was downloaded with the user prefix, but then attempted to be used w/o it.
The same in code:
```
# first run
python -c "from datasets import load_dataset; load_dataset('stas/openwebtext-10k')"
# now run immediately
python -c "from datasets import load_dataset; load_dataset('openwebtext-10k')"
# the second command should fail, but it doesn't fail now.
```
Please let me know if I explained myself clearly.
Thank you!
|
https://github.com/huggingface/datasets/issues/2842
|
[
"From what I can understand, you want the saved arrow file directory to have username as well instead of just dataset name if it was downloaded with the user prefix?",
"I don't think the user cares of how this is done, but the 2nd command should fail, IMHO, as its dataset name is invalid:\r\n```\r\n# first run\r\npython -c \"from datasets import load_dataset; load_dataset('stas/openwebtext-10k')\"\r\n# now run immediately\r\npython -c \"from datasets import load_dataset; load_dataset('openwebtext-10k')\"\r\n# the second command should fail, but it doesn't fail now.\r\n```\r\n\r\nMoreover, if someone were to create `openwebtext-10k` w/o the prefix, they will now get the wrong dataset, if they previously downloaded `stas/openwebtext-10k`.\r\n\r\nAnd if there are 2 users with the same dataset name `foo/ds` and `bar/ds` - currently this won't work to get the correct dataset.\r\n\r\nSo really there 3 unrelated issues hiding in the current behavior.",
"This has been fixed now, and we'll do a new release of the library today.\r\n\r\nNow the stas/openwebtext-10k dataset is cached at `.cache/huggingface/datasets/stas___openwebtext10k` and openwebtext-10k would be at `.cache/huggingface/datasets/openwebtext10k`. Since they are different, the cache won't fall back on loading the wrong one anymore.\r\n\r\nSame for the python script used to generate the dataset: stas/openwebtext-10k is cached at `.cache/huggingface/modules/datasets_modules/datasets/stas___openwebtext10k` and openwebtext-10k would be at `.cache/huggingface/modules/datasets_modules/datasets/openwebtext10k`",
"Amazing! Thank you for adding this improvement, @lhoestq!",
"(can be closed?)",
"Yes indeed :) thanks"
] | null | 2,842
| false
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.