id
int64 599M
3.29B
| url
stringlengths 58
61
| html_url
stringlengths 46
51
| number
int64 1
7.72k
| title
stringlengths 1
290
| state
stringclasses 2
values | comments
int64 0
70
| created_at
timestamp[s]date 2020-04-14 10:18:02
2025-08-05 09:28:51
| updated_at
timestamp[s]date 2020-04-27 16:04:17
2025-08-05 11:39:56
| closed_at
timestamp[s]date 2020-04-14 12:01:40
2025-08-01 05:15:45
⌀ | user_login
stringlengths 3
26
| labels
listlengths 0
4
| body
stringlengths 0
228k
⌀ | is_pull_request
bool 2
classes |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
690,197,725
|
https://api.github.com/repos/huggingface/datasets/issues/555
|
https://github.com/huggingface/datasets/pull/555
| 555
|
Upgrade pip in benchmark github action
|
closed
| 0
| 2020-09-01T14:37:26
| 2020-09-01T15:26:16
| 2020-09-01T15:26:15
|
lhoestq
|
[] |
It looks like it fixes the `import nlp` issue we have
| true
|
690,173,214
|
https://api.github.com/repos/huggingface/datasets/issues/554
|
https://github.com/huggingface/datasets/issues/554
| 554
|
nlp downloads to its module path
|
closed
| 8
| 2020-09-01T14:06:14
| 2020-09-11T06:19:24
| 2020-09-11T06:19:24
|
danieldk
|
[] |
I am trying to package `nlp` for Nix, because it is now an optional dependency for `transformers`. The problem that I encounter is that the `nlp` library downloads to the module path, which is typically not writable in most package management systems:
```>>> import nlp
>>> squad_dataset = nlp.load_dataset('squad')
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/nix/store/2yhik0hhqayksmkkfb0ylqp8cf5wa5wp-python3-3.8.5-env/lib/python3.8/site-packages/nlp/load.py", line 530, in load_dataset
module_path, hash = prepare_module(path, download_config=download_config, dataset=True)
File "/nix/store/2yhik0hhqayksmkkfb0ylqp8cf5wa5wp-python3-3.8.5-env/lib/python3.8/site-packages/nlp/load.py", line 329, in prepare_module
os.makedirs(main_folder_path, exist_ok=True)
File "/nix/store/685kq8pyhrvajah1hdsfn4q7gm3j4yd4-python3-3.8.5/lib/python3.8/os.py", line 223, in makedirs
mkdir(name, mode)
OSError: [Errno 30] Read-only file system: '/nix/store/2yhik0hhqayksmkkfb0ylqp8cf5wa5wp-python3-3.8.5-env/lib/python3.8/site-packages/nlp/datasets/squad'
```
Do you have any suggested workaround for this issue?
Perhaps overriding the default value for `force_local_path` of `prepare_module`?
| false
|
690,143,182
|
https://api.github.com/repos/huggingface/datasets/issues/553
|
https://github.com/huggingface/datasets/pull/553
| 553
|
[Fix GitHub Actions] test adding tmate
|
closed
| 0
| 2020-09-01T13:28:03
| 2021-05-05T18:24:38
| 2020-09-03T09:01:13
|
thomwolf
|
[] | true
|
|
690,079,429
|
https://api.github.com/repos/huggingface/datasets/issues/552
|
https://github.com/huggingface/datasets/pull/552
| 552
|
Add multiprocessing
|
closed
| 10
| 2020-09-01T11:56:17
| 2020-09-22T15:11:56
| 2020-09-02T10:01:25
|
lhoestq
|
[] |
Adding multiprocessing to `.map`
It works in 3 steps:
- shard the dataset in `num_proc` shards
- spawn one process per shard and call `map` on them
- concatenate the resulting datasets
Example of usage:
```python
from nlp import load_dataset
dataset = load_dataset("squad", split="train")
def function(x):
return {"lowered": x.lower()}
processed = d.map(
function,
input_columns=["context"],
num_proc=4,
cache_file_name="playground/tmp.arrow",
load_from_cache_file=False
)
```
Here it writes 4 files depending on the process rank:
- `playground/tmp_00000_of_00004.arrow`
- `playground/tmp_00001_of_00004.arrow`
- `playground/tmp_00002_of_00004.arrow`
- `playground/tmp_00003_of_00004.arrow`
The suffix format can be specified by the user.
If the `cache_file_name` is not specified, it writes into separated files depending on the fingerprint, as usual.
I still need to:
- write tests for this
- try to improve the logging (currently it shows 4 progress bars, but if one finishes before the others, then the following messages are written over the progress bars)
| true
|
690,034,762
|
https://api.github.com/repos/huggingface/datasets/issues/551
|
https://github.com/huggingface/datasets/pull/551
| 551
|
added HANS dataset
|
closed
| 0
| 2020-09-01T10:42:02
| 2020-09-01T12:17:10
| 2020-09-01T12:17:10
|
TevenLeScao
|
[] |
Adds the [HANS](https://github.com/tommccoy1/hans) dataset to evaluate NLI systems.
| true
|
689,775,914
|
https://api.github.com/repos/huggingface/datasets/issues/550
|
https://github.com/huggingface/datasets/pull/550
| 550
|
[BUGFIX] Solving mismatched checksum issue for the LinCE dataset (#539)
|
closed
| 2
| 2020-09-01T03:27:03
| 2020-09-03T09:06:01
| 2020-09-03T09:06:01
|
gaguilar
|
[] |
Hi,
I have added the updated `dataset_infos.json` file for the LinCE benchmark. This update is to fix the mismatched checksum bug #539 for one of the datasets in the LinCE benchmark. To update the file, I run this command from the nlp root directory:
```
python nlp-cli test ./datasets/lince --save_infos --all_configs
```
**NOTE**: I needed to change [this line](https://github.com/huggingface/nlp/blob/master/src/nlp/commands/dummy_data.py#L8) from: `from .utils.logging import get_logger` to `from nlp.utils.logging import get_logger`, otherwise the script was not able to import `get_logger`. However, I did not include that in this PR since that could have been just my environment (and another PR could be fixing this already if it is actually an issue).
| true
|
689,766,465
|
https://api.github.com/repos/huggingface/datasets/issues/549
|
https://github.com/huggingface/datasets/pull/549
| 549
|
Fix bleurt logging import
|
closed
| 2
| 2020-09-01T03:01:25
| 2020-09-03T18:04:46
| 2020-09-03T09:04:20
|
jbragg
|
[] |
Bleurt started throwing an error in some code we have.
This looks like the fix but...
It's also unnerving that even a prebuilt docker image with pinned versions can be working 1 day and then fail the next (especially for production systems).
Any way for us to pin your metrics code so that they are guaranteed not to to change and possibly fail on repository changes?
Thanks (and also for your continued work on the lib...)
| true
|
689,285,996
|
https://api.github.com/repos/huggingface/datasets/issues/548
|
https://github.com/huggingface/datasets/pull/548
| 548
|
[Breaking] Switch text loading to multi-threaded PyArrow loading
|
closed
| 5
| 2020-08-31T15:15:41
| 2020-09-08T10:19:58
| 2020-09-08T10:19:57
|
thomwolf
|
[] |
Test if we can get better performances for large-scale text datasets by using multi-threaded text file loading based on Apache Arrow multi-threaded CSV loader.
If it works ok, it would fix #546.
**Breaking change**:
The text lines now do not include final line-breaks anymore.
| true
|
689,268,589
|
https://api.github.com/repos/huggingface/datasets/issues/547
|
https://github.com/huggingface/datasets/pull/547
| 547
|
[Distributed] Making loading distributed datasets a bit safer
|
closed
| 0
| 2020-08-31T14:51:34
| 2020-08-31T15:16:30
| 2020-08-31T15:16:29
|
thomwolf
|
[] |
Add some file-locks during dataset loading
| true
|
689,186,526
|
https://api.github.com/repos/huggingface/datasets/issues/546
|
https://github.com/huggingface/datasets/issues/546
| 546
|
Very slow data loading on large dataset
|
closed
| 28
| 2020-08-31T12:57:23
| 2024-01-02T20:26:24
| 2020-09-08T10:19:57
|
agemagician
|
[] |
I made a simple python script to check the NLP library speed, which loads 1.1 TB of textual data.
It has been 8 hours and still, it is on the loading steps.
It does work when the text dataset size is small about 1 GB, but it doesn't scale.
It also uses a single thread during the data loading step.
```
train_files = glob.glob("xxx/*.txt",recursive=True)
random.shuffle(train_files)
print(train_files)
dataset = nlp.load_dataset('text',
data_files=train_files,
name="customDataset",
version="1.0.0",
cache_dir="xxx/nlp")
```
Is there something that I am missing ?
| false
|
689,138,878
|
https://api.github.com/repos/huggingface/datasets/issues/545
|
https://github.com/huggingface/datasets/issues/545
| 545
|
New release coming up for this library
|
closed
| 1
| 2020-08-31T11:37:38
| 2021-01-13T10:59:04
| 2021-01-13T10:59:04
|
thomwolf
|
[] |
Hi all,
A few words on the roadmap for this library.
The next release will be a big one and is planed at the end of this week.
In addition to the support for indexed datasets (useful for non-parametric models like REALM, RAG, DPR, knn-LM and many other fast dataset retrieval technics), it will:
- have support for multi-modal datasets
- include various significant improvements on speed for standard processing (map, shuffling, ...)
- have a better support for metrics (better caching, and a robust API) and a bigger focus on reproductibility
- change the name to the final name (voted by the community): `datasets`
- be the 1.0.0 release as we think the API will be mostly stabilized from now on
| false
|
689,062,519
|
https://api.github.com/repos/huggingface/datasets/issues/544
|
https://github.com/huggingface/datasets/pull/544
| 544
|
[Distributed] Fix load_dataset error when multiprocessing + add test
|
closed
| 0
| 2020-08-31T09:30:10
| 2020-08-31T11:15:11
| 2020-08-31T11:15:10
|
thomwolf
|
[] |
Fix #543 + add test
| true
|
688,644,407
|
https://api.github.com/repos/huggingface/datasets/issues/543
|
https://github.com/huggingface/datasets/issues/543
| 543
|
nlp.load_dataset is not safe for multi processes when loading from local files
|
closed
| 1
| 2020-08-30T03:20:34
| 2020-08-31T11:15:10
| 2020-08-31T11:15:10
|
luyug
|
[] |
Loading from local files, e.g., `dataset = nlp.load_dataset('csv', data_files=['file_1.csv', 'file_2.csv'])`
concurrently from multiple processes, will raise `FileExistsError` from builder's line 430, https://github.com/huggingface/nlp/blob/6655008c738cb613c522deb3bd18e35a67b2a7e5/src/nlp/builder.py#L423-L438
Likely because multiple processes step into download_and_prepare, https://github.com/huggingface/nlp/blob/6655008c738cb613c522deb3bd18e35a67b2a7e5/src/nlp/load.py#L550-L554
This can happen when launching distributed training with commands like `python -m torch.distributed.launch --nproc_per_node 4` on a new collection of files never loaded before.
I can create a PR that puts in some file locks. It would be helpful if I can be informed of the convention for naming and placement of the lock.
| false
|
688,555,036
|
https://api.github.com/repos/huggingface/datasets/issues/542
|
https://github.com/huggingface/datasets/pull/542
| 542
|
Add TensorFlow example
|
closed
| 0
| 2020-08-29T15:39:27
| 2020-08-31T09:49:20
| 2020-08-31T09:49:19
|
jplu
|
[] |
Update the Quick Tour documentation in order to add the TensorFlow equivalent source code for the classification example. Now it is possible to select either the code in PyTorch or in TensorFlow in the Quick tour.
| true
|
688,521,224
|
https://api.github.com/repos/huggingface/datasets/issues/541
|
https://github.com/huggingface/datasets/issues/541
| 541
|
Best practices for training tokenizers with nlp
|
closed
| 1
| 2020-08-29T12:06:49
| 2022-10-04T17:28:04
| 2022-10-04T17:28:04
|
moskomule
|
[] |
Hi, thank you for developing this library.
What do you think are the best practices for training tokenizers using `nlp`? In the document and examples, I could only find pre-trained tokenizers used.
| false
|
688,475,884
|
https://api.github.com/repos/huggingface/datasets/issues/540
|
https://github.com/huggingface/datasets/pull/540
| 540
|
[BUGFIX] Fix Race Dataset Checksum bug
|
closed
| 4
| 2020-08-29T07:00:10
| 2020-09-18T11:42:20
| 2020-09-18T11:42:20
|
abarbosa94
|
[] |
In #537 I noticed that there was a bug in checksum checking when I have tried to download the race dataset. The reason for this is that the current preprocessing was just considering the `high school` data and it was ignoring the `middle` one. This PR just fixes it :)
Moreover, I have added some descriptions.
| true
|
688,323,602
|
https://api.github.com/repos/huggingface/datasets/issues/539
|
https://github.com/huggingface/datasets/issues/539
| 539
|
[Dataset] `NonMatchingChecksumError` due to an update in the LinCE benchmark data
|
closed
| 3
| 2020-08-28T19:55:51
| 2020-09-03T16:34:02
| 2020-09-03T16:34:01
|
gaguilar
|
[] |
Hi,
There is a `NonMatchingChecksumError` error for the `lid_msaea` (language identification for Modern Standard Arabic - Egyptian Arabic) dataset from the LinCE benchmark due to a minor update on that dataset.
How can I update the checksum of the library to solve this issue? The error is below and it also appears in the [nlp viewer](https://huggingface.co/nlp/viewer/?dataset=lince&config=lid_msaea):
```python
import nlp
nlp.load_dataset('lince', 'lid_msaea')
```
Output:
```
NonMatchingChecksumError: ['https://ritual.uh.edu/lince/libaccess/eyJ1c2VybmFtZSI6ICJodWdnaW5nZmFjZSBubHAiLCAidXNlcl9pZCI6IDExMSwgImVtYWlsIjogImR1bW15QGVtYWlsLmNvbSJ9/lid_msaea.zip']
Traceback:
File "/home/sasha/streamlit/lib/streamlit/ScriptRunner.py", line 322, in _run_script
exec(code, module.__dict__)
File "/home/sasha/nlp-viewer/run.py", line 196, in <module>
dts, fail = get(str(option.id), str(conf_option.name) if conf_option else None)
File "/home/sasha/streamlit/lib/streamlit/caching.py", line 591, in wrapped_func
return get_or_create_cached_value()
File "/home/sasha/streamlit/lib/streamlit/caching.py", line 575, in get_or_create_cached_value
return_value = func(*args, **kwargs)
File "/home/sasha/nlp-viewer/run.py", line 150, in get
builder_instance.download_and_prepare()
File "/home/sasha/.local/share/virtualenvs/lib-ogGKnCK_/lib/python3.7/site-packages/nlp/builder.py", line 432, in download_and_prepare
download_config.force_download = download_mode == FORCE_REDOWNLOAD
File "/home/sasha/.local/share/virtualenvs/lib-ogGKnCK_/lib/python3.7/site-packages/nlp/builder.py", line 469, in _download_and_prepare
File "/home/sasha/.local/share/virtualenvs/lib-ogGKnCK_/lib/python3.7/site-packages/nlp/utils/info_utils.py", line 36, in verify_checksums
raise NonMatchingChecksumError(str(bad_urls))
```
Thank you in advance!
@lhoestq
| false
|
688,015,912
|
https://api.github.com/repos/huggingface/datasets/issues/538
|
https://github.com/huggingface/datasets/pull/538
| 538
|
[logging] Add centralized logging - Bump-up cache loads to warnings
|
closed
| 0
| 2020-08-28T11:42:29
| 2020-08-31T11:42:51
| 2020-08-31T11:42:51
|
thomwolf
|
[] |
Add a `nlp.logging` module to set the global logging level easily. The verbosity level also controls the tqdm bars (disabled when set higher than INFO).
You can use:
```
nlp.logging.set_verbosity(verbosity: int)
nlp.logging.set_verbosity_info()
nlp.logging.set_verbosity_warning()
nlp.logging.set_verbosity_debug()
nlp.logging.set_verbosity_error()
nlp.logging.get_verbosity() -> int
```
And use the levels:
```
nlp.logging.CRITICAL
nlp.logging.DEBUG
nlp.logging.ERROR
nlp.logging.FATAL
nlp.logging.INFO
nlp.logging.NOTSET
nlp.logging.WARN
nlp.logging.WARNING
```
| true
|
687,614,699
|
https://api.github.com/repos/huggingface/datasets/issues/537
|
https://github.com/huggingface/datasets/issues/537
| 537
|
[Dataset] RACE dataset Checksums error
|
closed
| 9
| 2020-08-27T23:58:16
| 2020-09-18T12:07:04
| 2020-09-18T12:07:04
|
abarbosa94
|
[
"dataset bug"
] |
Hi there, I just would like to use this awesome lib to perform a dataset fine-tuning on RACE dataset. I have performed the following steps:
```
dataset = nlp.load_dataset("race")
len(dataset["train"]), len(dataset["validation"])
```
But then I got the following error:
```
---------------------------------------------------------------------------
NonMatchingChecksumError Traceback (most recent call last)
<ipython-input-15-8bf7603ce0ed> in <module>
----> 1 dataset = nlp.load_dataset("race")
2 len(dataset["train"]), len(dataset["validation"])
~/miniconda3/envs/masters/lib/python3.8/site-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, features, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs)
546
547 # Download and prepare data
--> 548 builder_instance.download_and_prepare(
549 download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
550 )
~/miniconda3/envs/masters/lib/python3.8/site-packages/nlp/builder.py in download_and_prepare(self, download_config, download_mode, ignore_verifications, try_from_hf_gcs, dl_manager, **download_and_prepare_kwargs)
460 logger.info("Dataset not on Hf google storage. Downloading and preparing it from source")
461 if not downloaded_from_gcs:
--> 462 self._download_and_prepare(
463 dl_manager=dl_manager, verify_infos=verify_infos, **download_and_prepare_kwargs
464 )
~/miniconda3/envs/masters/lib/python3.8/site-packages/nlp/builder.py in _download_and_prepare(self, dl_manager, verify_infos, **prepare_split_kwargs)
519 # Checksums verification
520 if verify_infos:
--> 521 verify_checksums(
522 self.info.download_checksums, dl_manager.get_recorded_sizes_checksums(), "dataset source files"
523 )
~/miniconda3/envs/masters/lib/python3.8/site-packages/nlp/utils/info_utils.py in verify_checksums(expected_checksums, recorded_checksums, verification_name)
36 if len(bad_urls) > 0:
37 error_msg = "Checksums didn't match" + for_verification_name + ":\n"
---> 38 raise NonMatchingChecksumError(error_msg + str(bad_urls))
39 logger.info("All the checksums matched successfully" + for_verification_name)
40
NonMatchingChecksumError: Checksums didn't match for dataset source files:
['http://www.cs.cmu.edu/~glai1/data/race/RACE.tar.gz']
```
| false
|
687,378,332
|
https://api.github.com/repos/huggingface/datasets/issues/536
|
https://github.com/huggingface/datasets/pull/536
| 536
|
Fingerprint
|
closed
| 1
| 2020-08-27T16:27:09
| 2020-08-31T14:20:40
| 2020-08-31T14:20:39
|
lhoestq
|
[] |
This PR is a continuation of #513 , in which many in-place functions were introduced or updated (cast_, flatten_) etc.
However the caching didn't handle these changes. Indeed the caching took into account only the previous cache file name of the table, and not the possible in-place transforms of the table.
To fix that, I added the concept of dataset fingerprint, that is updated after each transform (in place or not), and stored inside the table metadata.
When a dataset is created, an initial fingerprint is computed. If the dataset is memory-mapped, then the fingerprint generator doesn't read the table and only looks at the filename. However if the table is in-memory, then the fingerprint generator reads the content of the table using a batched non-crypto hashing.
I added a utility class to compute hashes of arbitrary python objects in `fingerprint.py` : `Hasher`. The API is close to standard hashing tools (`.update`, `.hexdigest`). It also supports custom hashing functions depending on object types using a registry like pickle. I added a custom hashing function to hash a `pa.Table` in a batched way, and also for `nlp.DatasetInfo` to leverage its json serialization feature.
Note about this PR:
This is a draft PR because #513 needs to be merged first.
The diff that is shown is for branches fingerprint -> indices (and not master, for now)
| true
|
686,238,315
|
https://api.github.com/repos/huggingface/datasets/issues/535
|
https://github.com/huggingface/datasets/pull/535
| 535
|
Benchmarks
|
closed
| 0
| 2020-08-26T11:21:26
| 2020-08-27T08:40:00
| 2020-08-27T08:39:59
|
thomwolf
|
[] |
Adding some benchmarks with DVC/CML
To add a new tracked benchmark:
- create a new python benchmarking script in `./benchmarks/`. The script can use the utilities in `./benchmarks/utils.py` and should output a JSON file with results in `./benchmarks/results/`.
- add a new pipeline stage in [dvc.yaml](./dvc.yaml) with the name of your new benchmark.
That's it
| true
|
686,115,912
|
https://api.github.com/repos/huggingface/datasets/issues/534
|
https://github.com/huggingface/datasets/issues/534
| 534
|
`list_datasets()` is broken.
|
closed
| 3
| 2020-08-26T08:19:01
| 2020-08-27T06:31:11
| 2020-08-27T06:31:11
|
ashutosh-dwivedi-e3502
|
[] |
version = '0.4.0'
`list_datasets()` is broken. It results in the following error :
```
In [3]: nlp.list_datasets()
Out[3]: ---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
~/.virtualenvs/san-lgUCsFg_/lib/python3.8/site-packages/IPython/core/formatters.py in __call__(self, obj)
700 type_pprinters=self.type_printers,
701 deferred_pprinters=self.deferred_printers)
--> 702 printer.pretty(obj)
703 printer.flush()
704 return stream.getvalue()
~/.virtualenvs/san-lgUCsFg_/lib/python3.8/site-packages/IPython/lib/pretty.py in pretty(self, obj)
375 if cls in self.type_pprinters:
376 # printer registered in self.type_pprinters
--> 377 return self.type_pprinters[cls](obj, self, cycle)
378 else:
379 # deferred printer
~/.virtualenvs/san-lgUCsFg_/lib/python3.8/site-packages/IPython/lib/pretty.py in inner(obj, p, cycle)
553 p.text(',')
554 p.breakable()
--> 555 p.pretty(x)
556 if len(obj) == 1 and type(obj) is tuple:
557 # Special case for 1-item tuples.
~/.virtualenvs/san-lgUCsFg_/lib/python3.8/site-packages/IPython/lib/pretty.py in pretty(self, obj)
392 if cls is not object \
393 and callable(cls.__dict__.get('__repr__')):
--> 394 return _repr_pprint(obj, self, cycle)
395
396 return _default_pprint(obj, self, cycle)
~/.virtualenvs/san-lgUCsFg_/lib/python3.8/site-packages/IPython/lib/pretty.py in _repr_pprint(obj, p, cycle)
698 """A pprint that just redirects to the normal repr function."""
699 # Find newlines and replace them with p.break_()
--> 700 output = repr(obj)
701 lines = output.splitlines()
702 with p.group():
~/.virtualenvs/san-lgUCsFg_/lib/python3.8/site-packages/nlp/hf_api.py in __repr__(self)
110
111 def __repr__(self):
--> 112 single_line_description = self.description.replace("\n", "")
113 return f"nlp.ObjectInfo(id='{self.id}', description='{single_line_description}', files={self.siblings})"
114
AttributeError: 'NoneType' object has no attribute 'replace'
```
| false
|
685,585,914
|
https://api.github.com/repos/huggingface/datasets/issues/533
|
https://github.com/huggingface/datasets/pull/533
| 533
|
Fix ArrayXD for pyarrow 0.17.1 by using non fixed length list arrays
|
closed
| 0
| 2020-08-25T15:32:44
| 2020-08-26T08:02:24
| 2020-08-26T08:02:23
|
lhoestq
|
[] |
It should fix the CI problems in #513
| true
|
685,540,614
|
https://api.github.com/repos/huggingface/datasets/issues/532
|
https://github.com/huggingface/datasets/issues/532
| 532
|
File exists error when used with TPU
|
open
| 21
| 2020-08-25T14:36:38
| 2020-09-01T12:14:56
| null |
go-inoue
|
[] |
Hi,
I'm getting a "File exists" error when I use [text dataset](https://github.com/huggingface/nlp/tree/master/datasets/text) for pre-training a RoBERTa model using `transformers` (3.0.2) and `nlp`(0.4.0) on a VM with TPU (v3-8).
I modified [line 131 in the original `run_language_modeling.py`](https://github.com/huggingface/transformers/blob/master/examples/language-modeling/run_language_modeling.py#L131) as follows:
```python
# line 131: return LineByLineTextDataset(tokenizer=tokenizer, file_path=file_path, block_size=args.block_size)
dataset = load_dataset("text", data_files=file_path, split="train")
dataset = dataset.map(lambda ex: tokenizer(ex["text"], add_special_tokens=True,
truncation=True, max_length=args.block_size), batched=True)
dataset.set_format(type='torch', columns=['input_ids'])
return dataset
```
When I run this with [`xla_spawn.py`](https://github.com/huggingface/transformers/blob/master/examples/xla_spawn.py), I get the following error (it produces one message per core in TPU, which I believe is fine).
It seems the current version doesn't take into account distributed training processes as in [this example](https://github.com/huggingface/transformers/blob/a573777901e662ec2e565be312ffaeedef6effec/src/transformers/data/datasets/language_modeling.py#L35-L38)?
```
08/25/2020 13:59:41 - WARNING - nlp.builder - Using custom data configuration default
08/25/2020 13:59:43 - INFO - nlp.builder - Generating dataset text (/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d)
08/25/2020 13:59:43 - INFO - nlp.builder - Generating dataset text (/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d)
08/25/2020 13:59:43 - INFO - nlp.builder - Generating dataset text (/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d)
08/25/2020 13:59:43 - INFO - nlp.builder - Generating dataset text (/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d)
08/25/2020 13:59:43 - INFO - nlp.builder - Generating dataset text (/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d)
08/25/2020 13:59:43 - INFO - nlp.builder - Generating dataset text (/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d)
08/25/2020 13:59:43 - INFO - nlp.builder - Generating dataset text (/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d)
08/25/2020 13:59:43 - INFO - nlp.builder - Generating dataset text (/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d)
Downloading and preparing dataset text/default-b0932b2bdbb63283 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/
447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d...
Downloading and preparing dataset text/default-b0932b2bdbb63283 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/
447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d...
Downloading and preparing dataset text/default-b0932b2bdbb63283 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/
447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d...
Downloading and preparing dataset text/default-b0932b2bdbb63283 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/
447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d...
Downloading and preparing dataset text/default-b0932b2bdbb63283 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/
447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d...
Downloading and preparing dataset text/default-b0932b2bdbb63283 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/
447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d...
Exception in device=TPU:6: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
Exception in device=TPU:4: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
Exception in device=TPU:1: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
Downloading and preparing dataset text/default-b0932b2bdbb63283 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/
447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d...
Exception in device=TPU:7: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
Exception in device=TPU:3: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
Downloading and preparing dataset text/default-b0932b2bdbb63283 (download: Unknown size, generated: Unknown size, post-processed: Unknown size, total: Unknown size) to /home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/
447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d...
Exception in device=TPU:2: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
Exception in device=TPU:0: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
Traceback (most recent call last):
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/torch_xla/distributed/xla_multiprocessing.py", line 231, in _start_fn
fn(gindex, *args)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/torch_xla/distributed/xla_multiprocessing.py", line 231, in _start_fn
fn(gindex, *args)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/torch_xla/distributed/xla_multiprocessing.py", line 231, in _start_fn
fn(gindex, *args)
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 300, in _mp_fn
main()
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 300, in _mp_fn
main()
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 300, in _mp_fn
main()
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 240, in main
train_dataset = get_dataset(data_args, tokenizer=tokenizer) if training_args.do_train else None
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 240, in main
train_dataset = get_dataset(data_args, tokenizer=tokenizer) if training_args.do_train else None
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 240, in main
train_dataset = get_dataset(data_args, tokenizer=tokenizer) if training_args.do_train else None
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 134, in get_dataset
dataset = load_dataset("text", data_files=file_path, split="train")
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/load.py", line 546, in load_dataset
download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 134, in get_dataset
dataset = load_dataset("text", data_files=file_path, split="train")
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 134, in get_dataset
dataset = load_dataset("text", data_files=file_path, split="train")
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 450, in download_and_prepare
with incomplete_dir(self._cache_dir) as tmp_data_dir:
Traceback (most recent call last):
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/load.py", line 546, in load_dataset
download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/contextlib.py", line 81, in __enter__
return next(self.gen)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/load.py", line 546, in load_dataset
download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/torch_xla/distributed/xla_multiprocessing.py", line 231, in _start_fn
fn(gindex, *args)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 450, in download_and_prepare
with incomplete_dir(self._cache_dir) as tmp_data_dir:
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 422, in incomplete_dir
os.makedirs(tmp_dir)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 450, in download_and_prepare
with incomplete_dir(self._cache_dir) as tmp_data_dir:
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 300, in _mp_fn
main()
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/contextlib.py", line 81, in __enter__
return next(self.gen)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/os.py", line 220, in makedirs
mkdir(name, mode)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/contextlib.py", line 81, in __enter__
return next(self.gen)
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 240, in main
train_dataset = get_dataset(data_args, tokenizer=tokenizer) if training_args.do_train else None
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 422, in incomplete_dir
os.makedirs(tmp_dir)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/torch_xla/distributed/xla_multiprocessing.py", line 231, in _start_fn
fn(gindex, *args)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 422, in incomplete_dir
os.makedirs(tmp_dir)
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 134, in get_dataset
dataset = load_dataset("text", data_files=file_path, split="train")
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/os.py", line 220, in makedirs
mkdir(name, mode)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/load.py", line 546, in load_dataset
download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
FileExistsError: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 300, in _mp_fn
main()
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 450, in download_and_prepare
with incomplete_dir(self._cache_dir) as tmp_data_dir:
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/os.py", line 220, in makedirs
mkdir(name, mode)
FileExistsError: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 240, in main
train_dataset = get_dataset(data_args, tokenizer=tokenizer) if training_args.do_train else None
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/contextlib.py", line 81, in __enter__
return next(self.gen)
FileExistsError: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 134, in get_dataset
dataset = load_dataset("text", data_files=file_path, split="train")
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 422, in incomplete_dir
os.makedirs(tmp_dir)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/load.py", line 546, in load_dataset
download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/os.py", line 220, in makedirs
mkdir(name, mode)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 450, in download_and_prepare
with incomplete_dir(self._cache_dir) as tmp_data_dir:
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/contextlib.py", line 81, in __enter__
return next(self.gen)
FileExistsError: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 422, in incomplete_dir
os.makedirs(tmp_dir)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/os.py", line 220, in makedirs
mkdir(name, mode)
Traceback (most recent call last):
FileExistsError: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
Traceback (most recent call last):
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/torch_xla/distributed/xla_multiprocessing.py", line 231, in _start_fn
fn(gindex, *args)
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 300, in _mp_fn
main()
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 240, in main
train_dataset = get_dataset(data_args, tokenizer=tokenizer) if training_args.do_train else None
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 134, in get_dataset
dataset = load_dataset("text", data_files=file_path, split="train")
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/torch_xla/distributed/xla_multiprocessing.py", line 231, in _start_fn
fn(gindex, *args)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/load.py", line 546, in load_dataset
download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 450, in download_and_prepare
with incomplete_dir(self._cache_dir) as tmp_data_dir:
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 300, in _mp_fn
main()
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 240, in main
train_dataset = get_dataset(data_args, tokenizer=tokenizer) if training_args.do_train else None
File "/home/*****/huggingface_roberta/run_language_modeling.py", line 134, in get_dataset
dataset = load_dataset("text", data_files=file_path, split="train")
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/load.py", line 546, in load_dataset
download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 450, in download_and_prepare
with incomplete_dir(self._cache_dir) as tmp_data_dir:
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/contextlib.py", line 81, in __enter__
return next(self.gen)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/site-packages/nlp/builder.py", line 422, in incomplete_dir
os.makedirs(tmp_dir)
File "/anaconda3/envs/torch-xla-1.6/lib/python3.6/os.py", line 220, in makedirs
mkdir(name, mode)
FileExistsError: [Errno 17] File exists: '/home/*****/.cache/huggingface/datasets/text/default-b0932b2bdbb63283/0.0.0/447f2bcfa2a721a37bc8fdf23800eade1523cf07f7eada6fe661fe4d070d380d.incomplete'
```
| false
|
685,291,036
|
https://api.github.com/repos/huggingface/datasets/issues/531
|
https://github.com/huggingface/datasets/pull/531
| 531
|
add concatenate_datasets to the docs
|
closed
| 0
| 2020-08-25T08:40:05
| 2020-08-25T09:02:20
| 2020-08-25T09:02:19
|
lhoestq
|
[] | true
|
|
684,825,612
|
https://api.github.com/repos/huggingface/datasets/issues/530
|
https://github.com/huggingface/datasets/pull/530
| 530
|
use ragged tensor by default
|
closed
| 4
| 2020-08-24T17:06:15
| 2021-10-22T19:38:40
| 2020-08-24T19:22:25
|
lhoestq
|
[] |
I think it's better if it's clear whether the returned tensor is ragged or not when the type is set to tensorflow.
Previously it was a tensor (not ragged) if numpy could stack the output (which can change depending on the batch of example you take), which make things difficult to handle, as it may sometimes return a ragged tensor and sometimes not.
Therefore I reverted this behavior to always return a ragged tensor as we used to do.
| true
|
684,797,157
|
https://api.github.com/repos/huggingface/datasets/issues/529
|
https://github.com/huggingface/datasets/pull/529
| 529
|
Add MLSUM
|
closed
| 3
| 2020-08-24T16:18:35
| 2020-08-26T08:04:11
| 2020-08-26T08:04:11
|
RachelKer
|
[] |
Hello (again :) !),
So, I started a new branch because of a [rebase issue](https://github.com/huggingface/nlp/pull/463), sorry for the mess.
However, the command `pytest tests/test_dataset_common.py::LocalDatasetTest::test_load_real_dataset_mlsum` still fails because there is no default language dataset : the script throws an error as a specific config language is necessary.
I think that setting a default language would be a bad workaround for this so I kept it as it is. Putting all the train files across languages together would also be a bad idea because of the size.
Thanks for your help,
Rachel
| true
|
684,673,673
|
https://api.github.com/repos/huggingface/datasets/issues/528
|
https://github.com/huggingface/datasets/pull/528
| 528
|
fix missing variable names in docs
|
closed
| 1
| 2020-08-24T13:31:48
| 2020-08-25T09:04:04
| 2020-08-25T09:04:03
|
lhoestq
|
[] |
fix #524
| true
|
684,632,930
|
https://api.github.com/repos/huggingface/datasets/issues/527
|
https://github.com/huggingface/datasets/pull/527
| 527
|
Fix config used for slow test on real dataset
|
closed
| 0
| 2020-08-24T12:39:34
| 2020-08-25T09:20:45
| 2020-08-25T09:20:44
|
lhoestq
|
[] |
As noticed in #470, #474, #476, #504 , the slow test `test_load_real_dataset` couldn't run on datasets that require config parameters.
To fix that I replaced it with one test with the first config of BUILDER_CONFIGS `test_load_real_dataset`, and another test that runs all of the configs in BUILDER_CONFIGS `test_load_real_dataset_all_configs`
| true
|
684,615,455
|
https://api.github.com/repos/huggingface/datasets/issues/526
|
https://github.com/huggingface/datasets/pull/526
| 526
|
Returning None instead of "python" if dataset is unformatted
|
closed
| 2
| 2020-08-24T12:10:35
| 2020-08-24T12:50:43
| 2020-08-24T12:50:42
|
TevenLeScao
|
[] |
Following the discussion on Slack, this small fix ensures that calling `dataset.set_format(type=dataset.format["type"])` works properly. Slightly breaking as calling `dataset.format` when the dataset is unformatted will return `None` instead of `python`.
| true
|
683,875,483
|
https://api.github.com/repos/huggingface/datasets/issues/525
|
https://github.com/huggingface/datasets/issues/525
| 525
|
wmt download speed example
|
closed
| 8
| 2020-08-21T23:29:06
| 2022-10-04T17:45:39
| 2022-10-04T17:45:39
|
sshleifer
|
[] |
Continuing from the slack 1.0 roadmap thread w @lhoestq , I realized the slow downloads is only a thing sometimes. Here are a few examples, I suspect there are multiple issues. All commands were run from the same gcp us-central-1f machine.
```
import nlp
nlp.load_dataset('wmt16', 'de-en')
```
Downloads at 49.1 KB/S
Whereas
```
pip install gdown # download from google drive
!gdown https://drive.google.com/uc?id=1iO7um-HWoNoRKDtw27YUSgyeubn9uXqj
```
Downloads at 127 MB/s. (The file is a copy of wmt-en-de raw).
```
nlp.load_dataset('wmt16', 'ro-en')
```
goes at 27 MB/s, much faster.
if we wget the same data from s3 is the same download speed, but ¼ the file size:
```
wget https://s3.amazonaws.com/datasets.huggingface.co/translation/wmt_en_ro_packed_200_rand.tgz
```
Finally,
```
nlp.load_dataset('wmt19', 'zh-en')
```
Starts fast, but broken. (duplicate of #493 )
| false
|
683,686,359
|
https://api.github.com/repos/huggingface/datasets/issues/524
|
https://github.com/huggingface/datasets/issues/524
| 524
|
Some docs are missing parameter names
|
closed
| 1
| 2020-08-21T16:47:34
| 2020-08-25T09:04:03
| 2020-08-25T09:04:03
|
jarednielsen
|
[] |
See https://huggingface.co/nlp/master/package_reference/main_classes.html#nlp.Dataset.map. I believe this is because the parameter names are enclosed in backticks in the docstrings, maybe it's an old docstring format that doesn't work with the current Sphinx version.
| false
|
682,573,232
|
https://api.github.com/repos/huggingface/datasets/issues/523
|
https://github.com/huggingface/datasets/pull/523
| 523
|
Speed up Tokenization by optimizing cast_to_python_objects
|
closed
| 1
| 2020-08-20T09:42:02
| 2020-08-24T08:54:15
| 2020-08-24T08:54:14
|
lhoestq
|
[] |
I changed how `cast_to_python_objects` works to make it faster.
It is used to cast numpy/pytorch/tensorflow/pandas objects to python lists, and it works recursively.
To avoid iterating over possibly long lists, it first checks if the first element that is not None has to be casted.
If the first element needs to be casted, then all the elements of the list will be casted, otherwise they'll stay the same.
This trick allows to cast objects that contain tokenizers outputs without iterating over every single token for example.
Speed improvement:
```python
import transformers
import nlp
tok = transformers.BertTokenizerFast.from_pretrained("bert-base-uncased")
txt = ["a " * 512] * 1000
dataset = nlp.Dataset.from_dict({"txt": txt})
# Tokenization using .map is now faster. Previously it was taking 3.5s
%time _ = dataset.map(lambda x: tok(x["txt"]), batched=True, load_from_cache_file=False)
# 450ms
# for comparison
%time _ = tok(txt)
# 280ms
```
| true
|
682,478,833
|
https://api.github.com/repos/huggingface/datasets/issues/522
|
https://github.com/huggingface/datasets/issues/522
| 522
|
dictionnary typo in docs
|
closed
| 1
| 2020-08-20T07:11:05
| 2020-08-20T07:52:14
| 2020-08-20T07:52:13
|
yonigottesman
|
[] |
Many places dictionary is spelled dictionnary, not sure if its on purpose or not.
Fixed in this pr:
https://github.com/huggingface/nlp/pull/521
| false
|
682,477,648
|
https://api.github.com/repos/huggingface/datasets/issues/521
|
https://github.com/huggingface/datasets/pull/521
| 521
|
Fix dictionnary (dictionary) typo
|
closed
| 1
| 2020-08-20T07:09:02
| 2020-08-20T07:52:04
| 2020-08-20T07:52:04
|
yonigottesman
|
[] |
This error happens many times I'm thinking maybe its spelled like this on purpose?
| true
|
682,264,839
|
https://api.github.com/repos/huggingface/datasets/issues/520
|
https://github.com/huggingface/datasets/pull/520
| 520
|
Transform references for sacrebleu
|
closed
| 1
| 2020-08-20T00:26:55
| 2020-08-20T09:30:54
| 2020-08-20T09:30:53
|
jbragg
|
[] |
Currently it is impossible to use sacrebleu when len(predictions) != the number of references per prediction (very uncommon), due to a strange format expected by sacrebleu. If one passes in the data to `nlp.metric.compute()` in sacrebleu format, `nlp` throws an error due to mismatching lengths between predictions and references. If one uses a more standard format where predictions and references are lists of the same length, sacrebleu throws an error.
This PR transforms reference data in a more standard format into the [unusual format](https://github.com/mjpost/sacreBLEU#using-sacrebleu-from-python) expected by sacrebleu.
| true
|
682,193,882
|
https://api.github.com/repos/huggingface/datasets/issues/519
|
https://github.com/huggingface/datasets/issues/519
| 519
|
[BUG] Metrics throwing new error on master since 0.4.0
|
closed
| 2
| 2020-08-19T21:29:15
| 2022-06-02T16:41:01
| 2020-08-19T22:04:40
|
jbragg
|
[] |
The following error occurs when passing in references of type `List[List[str]]` to metrics like bleu.
Wasn't happening on 0.4.0 but happening now on master.
```
File "/usr/local/lib/python3.7/site-packages/nlp/metric.py", line 226, in compute
self.add_batch(predictions=predictions, references=references)
File "/usr/local/lib/python3.7/site-packages/nlp/metric.py", line 242, in add_batch
batch = self.info.features.encode_batch(batch)
File "/usr/local/lib/python3.7/site-packages/nlp/features.py", line 527, in encode_batch
encoded_batch[key] = [encode_nested_example(self[key], cast_to_python_objects(obj)) for obj in column]
File "/usr/local/lib/python3.7/site-packages/nlp/features.py", line 527, in <listcomp>
encoded_batch[key] = [encode_nested_example(self[key], cast_to_python_objects(obj)) for obj in column]
File "/usr/local/lib/python3.7/site-packages/nlp/features.py", line 456, in encode_nested_example
raise ValueError("Got a string but expected a list instead: '{}'".format(obj))
```
| false
|
682,131,165
|
https://api.github.com/repos/huggingface/datasets/issues/518
|
https://github.com/huggingface/datasets/pull/518
| 518
|
[METRICS, breaking] Refactor caching behavior, pickle/cloudpickle metrics and dataset, add tests on metrics
|
closed
| 2
| 2020-08-19T19:43:08
| 2020-08-24T16:01:40
| 2020-08-24T16:01:39
|
thomwolf
|
[] |
Move the acquisition of the filelock at a later stage during metrics processing so it can be pickled/cloudpickled after instantiation.
Also add some tests on pickling, concurrent but separate metric instances and concurrent and distributed metric instances.
Changes significantly the caching behavior for the metrics:
- if the metric is used in a non-distributed setup (most common case) we try to find a free cache file using UUID instead of asking for an `experiment_id` if we can't lock the cache file this allows to use several instances of the same metrics in parallel.
- if the metrics is used in a distributed setup we ask for an `experiment_id` if we can't lock the cache file (because all the nodes need to have related cache file names for the final sync.
- after the computation, we free the locks and delete all the cache files.
Breaking: Some arguments for Metrics initialization have been removed for simplicity (`version`...) and some have been renamed for consistency with the rest of the library (`in_memory` => `keep_in_memory`).
Also remove the `_has_transformers` detection in utils to avoid importing transformers everytime during loading.
| true
|
681,896,944
|
https://api.github.com/repos/huggingface/datasets/issues/517
|
https://github.com/huggingface/datasets/issues/517
| 517
|
add MLDoc dataset
|
open
| 2
| 2020-08-19T14:41:59
| 2021-08-03T05:59:33
| null |
jxmorris12
|
[
"dataset request"
] |
Hi,
I am recommending that someone add MLDoc, a multilingual news topic classification dataset.
- Here's a link to the Github: https://github.com/facebookresearch/MLDoc
- and the paper: http://www.lrec-conf.org/proceedings/lrec2018/pdf/658.pdf
Looks like the dataset contains news stories in multiple languages that can be classified into four hierarchical groups: CCAT (Corporate/Industrial), ECAT (Economics), GCAT (Government/Social) and MCAT (Markets). There are 13 languages: Dutch, French, German, Chinese, Japanese, Russian, Portuguese, Spanish, Latin American Spanish, Italian, Danish, Norwegian, and Swedish
| false
|
681,846,032
|
https://api.github.com/repos/huggingface/datasets/issues/516
|
https://github.com/huggingface/datasets/pull/516
| 516
|
[Breaking] Rename formated to formatted
|
closed
| 0
| 2020-08-19T13:35:23
| 2020-08-20T08:41:17
| 2020-08-20T08:41:16
|
lhoestq
|
[] |
`formated` is not correct but `formatted` is
| true
|
681,845,619
|
https://api.github.com/repos/huggingface/datasets/issues/515
|
https://github.com/huggingface/datasets/pull/515
| 515
|
Fix batched map for formatted dataset
|
closed
| 0
| 2020-08-19T13:34:50
| 2020-08-20T20:30:43
| 2020-08-20T20:30:42
|
lhoestq
|
[] |
If you had a dataset formatted as numpy for example, and tried to do a batched map, then it would crash because one of the elements from the inputs was missing for unchanged columns (ex: batch of length 999 instead of 1000).
The happened during the creation of the `pa.Table`, since columns had different lengths.
| true
|
681,256,348
|
https://api.github.com/repos/huggingface/datasets/issues/514
|
https://github.com/huggingface/datasets/issues/514
| 514
|
dataset.shuffle(keep_in_memory=True) is never allowed
|
closed
| 10
| 2020-08-18T18:47:40
| 2022-10-10T12:21:58
| 2022-10-10T12:21:58
|
vegarab
|
[
"good first issue",
"hacktoberfest"
] |
As of commit ef4aac2, the usage of the parameter `keep_in_memory=True` is never possible: `dataset.select(keep_in_memory=True)`
The commit added the lines
```python
# lines 994-996 in src/nlp/arrow_dataset.py
assert (
not keep_in_memory or cache_file_name is None
), "Please use either `keep_in_memory` or `cache_file_name` but not both."
```
This affects both `shuffle()` as `select()` is a sub-routine, and `map()` that has the same check.
I'd love to fix this myself, but unsure what the intention of the assert is given the rest of the logic in the function concerning `ccache_file_name` and `keep_in_memory`.
| false
|
681,215,612
|
https://api.github.com/repos/huggingface/datasets/issues/513
|
https://github.com/huggingface/datasets/pull/513
| 513
|
[speedup] Use indices mappings instead of deepcopy for all the samples reordering methods
|
closed
| 4
| 2020-08-18T17:36:02
| 2020-08-28T08:41:51
| 2020-08-28T08:41:50
|
thomwolf
|
[] |
Use an indices mapping instead of rewriting the dataset for all the samples re-ordering/selection methods (`select`, `sort`, `shuffle`, `shard`, `train_test_split`).
Added a `flatten_indices` method which copy the dataset to a new table to remove the indices mapping with tests.
All the samples re-ordering/selection methods should be a lot faster. The downside is that iterating on very large batch of the dataset might be a little slower when we have changed the order of the samples since with in these case we use `pyarrow.Table.take` instead of `pyarrow.Table.slice`. There is no free lunch but the speed of iterating over the dataset is rarely the bottleneck.
*Backward breaking change*: the `cache_file_name` argument in all the samples re-ordering/selection methods (`select`, `sort`, `shuffle`, `shard`, `train_test_split`) is now called `indices_cache_file_name` on purpose to make it explicit to the user that this caching file is used for caching the indices mapping and not the dataset itself.
| true
|
681,137,164
|
https://api.github.com/repos/huggingface/datasets/issues/512
|
https://github.com/huggingface/datasets/pull/512
| 512
|
Delete CONTRIBUTING.md
|
closed
| 2
| 2020-08-18T15:33:25
| 2020-08-18T15:48:21
| 2020-08-18T15:39:07
|
ChenZehong13
|
[] | true
|
|
681,055,553
|
https://api.github.com/repos/huggingface/datasets/issues/511
|
https://github.com/huggingface/datasets/issues/511
| 511
|
dataset.shuffle() and select() resets format. Intended?
|
closed
| 5
| 2020-08-18T13:46:01
| 2020-09-14T08:45:38
| 2020-09-14T08:45:38
|
vegarab
|
[] |
Calling `dataset.shuffle()` or `dataset.select()` on a dataset resets its format set by `dataset.set_format()`. Is this intended or an oversight?
When working on quite large datasets that require a lot of preprocessing I find it convenient to save the processed dataset to file using `torch.save("dataset.pt")`. Later loading the dataset object using `torch.load("dataset.pt")`, which conserves the defined format before saving.
I do shuffling and selecting (for controlling dataset size) after loading the data from .pt-file, as it's convenient whenever you train multiple models with varying sizes of the same dataset.
The obvious workaround for this is to set the format again after using `dataset.select()` or `dataset.shuffle()`.
_I guess this is more of a discussion on the design philosophy of the functions. Please let me know if this is not the right channel for these kinds of discussions or if they are not wanted at all!_
#### How to reproduce:
```python
import nlp
from transformers import T5Tokenizer
tokenizer = T5Tokenizer.from_pretrained("t5-base")
def create_features(batch):
context_encoding = tokenizer.batch_encode_plus(batch["context"])
return {"input_ids": context_encoding["input_ids"]}
dataset = nlp.load_dataset("cosmos_qa", split="train")
dataset = dataset.map(create_features, batched=True)
dataset.set_format(type="torch", columns=["input_ids"])
dataset[0]
# {'input_ids': tensor([ 1804, 3525, 1602, ... 0, 0])}
dataset = dataset.shuffle()
dataset[0]
# {'id': '3Q9(...)20', 'context': "Good Old War an (...) play ?', 'answer0': 'None of the above choices .', 'answer1': 'This person likes music and likes to see the show , they will see other bands play .', (...) 'input_ids': [1804, 3525, 1602, ... , 0, 0]}
```
| false
|
680,823,644
|
https://api.github.com/repos/huggingface/datasets/issues/510
|
https://github.com/huggingface/datasets/issues/510
| 510
|
Version of numpy to use the library
|
closed
| 2
| 2020-08-18T08:59:13
| 2020-08-19T18:35:56
| 2020-08-19T18:35:56
|
isspek
|
[] |
Thank you so much for your excellent work! I would like to use nlp library in my project. While importing nlp, I am receiving the following error `AttributeError: module 'numpy.random' has no attribute 'Generator'` Numpy version in my project is 1.16.0. May I learn which numpy version is used for the nlp library.
Thanks in advance.
| false
|
679,711,585
|
https://api.github.com/repos/huggingface/datasets/issues/509
|
https://github.com/huggingface/datasets/issues/509
| 509
|
Converting TensorFlow dataset example
|
closed
| 2
| 2020-08-16T08:05:20
| 2021-08-03T06:01:18
| 2021-08-03T06:01:17
|
saareliad
|
[] |
Hi,
I want to use TensorFlow datasets with this repo, I noticed you made some conversion script,
can you give a simple example of using it?
Thanks
| false
|
679,705,734
|
https://api.github.com/repos/huggingface/datasets/issues/508
|
https://github.com/huggingface/datasets/issues/508
| 508
|
TypeError: Receiver() takes no arguments
|
closed
| 5
| 2020-08-16T07:18:16
| 2020-09-01T14:53:33
| 2020-09-01T14:49:03
|
sebastiantomac
|
[] |
I am trying to load a wikipedia data set
```
import nlp
from nlp import load_dataset
dataset = load_dataset("wikipedia", "20200501.en", split="train", cache_dir=data_path, beam_runner='DirectRunner')
#dataset = load_dataset('wikipedia', '20200501.sv', cache_dir=data_path, beam_runner='DirectRunner')
```
This fails in the apache beam runner.
```
Traceback (most recent call last):
File "D:/ML/wikiembedding/gpt2_sv.py", line 36, in <module>
dataset = load_dataset("wikipedia", "20200501.en", split="train", cache_dir=my_cache_dir, beam_runner='DirectRunner')
File "C:\Users\seto\AppData\Local\Programs\Python\Python38\lib\site-packages\nlp\load.py", line 548, in load_dataset
builder_instance.download_and_prepare(
File "C:\Users\seto\AppData\Local\Programs\Python\Python38\lib\site-packages\nlp\builder.py", line 462, in download_and_prepare
self._download_and_prepare(
File "C:\Users\seto\AppData\Local\Programs\Python\Python38\lib\site-packages\nlp\builder.py", line 969, in _download_and_prepare
pipeline_results = pipeline.run()
File "C:\Users\seto\AppData\Local\Programs\Python\Python38\lib\site-packages\apache_beam\pipeline.py", line 534, in run
return self.runner.run_pipeline(self, self._options)
....
File "C:\Users\seto\AppData\Local\Programs\Python\Python38\lib\site-packages\apache_beam\runners\worker\bundle_processor.py", line 218, in process_encoded
self.output(decoded_value)
File "C:\Users\seto\AppData\Local\Programs\Python\Python38\lib\site-packages\apache_beam\runners\worker\operations.py", line 332, in output
cython.cast(Receiver, self.receivers[output_index]).receive(windowed_value)
File "C:\Users\seto\AppData\Local\Programs\Python\Python38\lib\site-packages\Cython\Shadow.py", line 167, in cast
return type(*args)
TypeError: Receiver() takes no arguments
```
This is run on a Windows 10 machine with python 3.8. I get the same error loading the swedish wikipedia dump.
| false
|
679,400,683
|
https://api.github.com/repos/huggingface/datasets/issues/507
|
https://github.com/huggingface/datasets/issues/507
| 507
|
Errors when I use
|
closed
| 1
| 2020-08-14T21:03:57
| 2020-08-14T21:39:10
| 2020-08-14T21:39:10
|
mchari
|
[] |
I tried the following example code from https://huggingface.co/deepset/roberta-base-squad2 and got errors
I am using **transformers 3.0.2** code .
from transformers.pipelines import pipeline
from transformers.modeling_auto import AutoModelForQuestionAnswering
from transformers.tokenization_auto import AutoTokenizer
model_name = "deepset/roberta-base-squad2"
nlp = pipeline('question-answering', model=model_name, tokenizer=model_name)
QA_input = {
'question': 'Why is model conversion important?',
'context': 'The option to convert models between FARM and transformers gives freedom to the user and let people easily switch between frameworks.'
}
res = nlp(QA_input)
The errors are :
res = nlp(QA_input)
File ".local/lib/python3.6/site-packages/transformers/pipelines.py", line 1316, in __call__
for s, e, score in zip(starts, ends, scores)
File ".local/lib/python3.6/site-packages/transformers/pipelines.py", line 1316, in <listcomp>
for s, e, score in zip(starts, ends, scores)
KeyError: 0
| false
|
679,164,788
|
https://api.github.com/repos/huggingface/datasets/issues/506
|
https://github.com/huggingface/datasets/pull/506
| 506
|
fix dataset.map for function without outputs
|
closed
| 0
| 2020-08-14T13:40:22
| 2020-08-17T11:24:39
| 2020-08-17T11:24:38
|
lhoestq
|
[] |
As noticed in #505 , giving a function that doesn't return anything in `.map` raises an error because of an unreferenced variable.
I fixed that and added tests.
Thanks @avloss for reporting
| true
|
678,791,400
|
https://api.github.com/repos/huggingface/datasets/issues/505
|
https://github.com/huggingface/datasets/pull/505
| 505
|
tmp_file referenced before assignment
|
closed
| 2
| 2020-08-13T23:27:33
| 2020-08-14T13:42:46
| 2020-08-14T13:42:46
|
avloss
|
[] |
Just learning about this library - so might've not set up all the flags correctly, but was getting this error about "tmp_file".
| true
|
678,756,211
|
https://api.github.com/repos/huggingface/datasets/issues/504
|
https://github.com/huggingface/datasets/pull/504
| 504
|
Added downloading to Hyperpartisan news detection
|
closed
| 2
| 2020-08-13T21:53:46
| 2020-08-27T08:18:41
| 2020-08-27T08:18:41
|
ghomasHudson
|
[] |
Following the discussion on Slack and #349, I've updated the hyperpartisan dataset to pull directly from Zenodo rather than manual install, which should make this dataset much more accessible. Many thanks to @johanneskiesel !
Currently doesn't pass `test_load_real_dataset` - I'm using `self.config.name` which is `default` in this test. Might be related to #474
| true
|
678,726,538
|
https://api.github.com/repos/huggingface/datasets/issues/503
|
https://github.com/huggingface/datasets/pull/503
| 503
|
CompGuessWhat?! 0.2.0
|
closed
| 20
| 2020-08-13T20:51:26
| 2020-10-21T06:54:29
| 2020-10-21T06:54:29
|
aleSuglia
|
[] |
We updated some metadata information associated with the dataset. In addition, we've updated the `create_dummy_data.py` script to generate data samples for the dataset.
| true
|
678,546,070
|
https://api.github.com/repos/huggingface/datasets/issues/502
|
https://github.com/huggingface/datasets/pull/502
| 502
|
Fix tokenizers caching
|
closed
| 1
| 2020-08-13T15:53:37
| 2020-08-19T13:37:19
| 2020-08-19T13:37:18
|
lhoestq
|
[] |
I've found some cases where the caching didn't work properly for tokenizers:
1. if a tokenizer has a regex pattern, then the caching would be inconsistent across sessions
2. if a tokenizer has a cache attribute that changes after some calls, the the caching would not work after cache updates
3. if a tokenizer is used inside a function, the caching of this function would result in the same cache file for different tokenizers
4. if `unique_no_split_tokens`'s attribute is not the same across sessions (after loading a tokenizer) then the caching could be inconsistent
To fix that, this is what I did:
1. register a specific `save_regex` function for pickle that makes regex dumps deterministic
2. ignore cache attribute of some tokenizers before dumping
3. enable recursive dump by default for all dumps
4. make `unique_no_split_tokens` deterministic in https://github.com/huggingface/transformers/pull/6461
I also added tests to make sure that tokenizers hashing works as expected.
In the future we should find a way to test if hashing also works across session (maybe using two CI jobs ? or by hardcoding a tokenizer's hash ?)
| true
|
677,952,893
|
https://api.github.com/repos/huggingface/datasets/issues/501
|
https://github.com/huggingface/datasets/issues/501
| 501
|
Caching doesn't work for map (non-deterministic)
|
closed
| 4
| 2020-08-12T20:20:07
| 2022-08-08T11:02:23
| 2020-08-24T16:34:35
|
wulu473
|
[] |
The caching functionality doesn't work reliably when tokenizing a dataset. Here's a small example to reproduce it.
```python
import nlp
import transformers
def main():
ds = nlp.load_dataset("reddit", split="train[:500]")
tokenizer = transformers.AutoTokenizer.from_pretrained("gpt2")
def convert_to_features(example_batch):
input_str = example_batch["body"]
encodings = tokenizer(input_str, add_special_tokens=True, truncation=True)
return encodings
ds = ds.map(convert_to_features, batched=True)
if __name__ == "__main__":
main()
```
Roughly 3/10 times, this example recomputes the tokenization.
Is this expected behaviour?
| false
|
677,841,708
|
https://api.github.com/repos/huggingface/datasets/issues/500
|
https://github.com/huggingface/datasets/pull/500
| 500
|
Use hnsw in wiki_dpr
|
closed
| 0
| 2020-08-12T16:58:07
| 2020-08-20T07:59:19
| 2020-08-20T07:59:18
|
lhoestq
|
[] |
The HNSW faiss index is much faster that regular Flat index.
| true
|
677,709,938
|
https://api.github.com/repos/huggingface/datasets/issues/499
|
https://github.com/huggingface/datasets/pull/499
| 499
|
Narrativeqa (with full text)
|
closed
| 9
| 2020-08-12T13:49:43
| 2020-12-09T11:21:02
| 2020-12-09T11:21:02
|
ghomasHudson
|
[] |
Following the uploading of the full text data in #309, I've added the full text to the narrativeqa dataset.
Few notes:
- Had some encoding issues using the default `open` so am using `open(encoding="latin-1"...` which seems to fix it. Looks fine.
- Can't get the dummy data to work. Currently putting stuff at:
```
dummy
|---- 0.0.0
|- dummy_data.zip
|-master.zip
| |- narrativeqa-master
| |- documents.csv
| |- qaps.csv
| |- third_party ......
|
| - narrativeqa_full_text.zip
| | - 001.content
| | - ....
```
Not sure what I'm messing up here (probably something obvious).
| true
|
677,597,479
|
https://api.github.com/repos/huggingface/datasets/issues/498
|
https://github.com/huggingface/datasets/pull/498
| 498
|
dont use beam fs to save info for local cache dir
|
closed
| 0
| 2020-08-12T11:00:00
| 2020-08-14T13:17:21
| 2020-08-14T13:17:20
|
lhoestq
|
[] |
If the cache dir is local, then we shouldn't use beam's filesystem to save the dataset info
Fix #490
| true
|
677,057,116
|
https://api.github.com/repos/huggingface/datasets/issues/497
|
https://github.com/huggingface/datasets/pull/497
| 497
|
skip header in PAWS-X
|
closed
| 0
| 2020-08-11T17:26:25
| 2020-08-19T09:50:02
| 2020-08-19T09:50:01
|
lhoestq
|
[] |
This should fix #485
I also updated the `dataset_infos.json` file that is used to verify the integrity of the generated splits (the number of examples was reduced by one).
Note that there are new fields in `dataset_infos.json` introduced in the latest release 0.4.0 corresponding to post processing info. I removed them in this case when I ran `nlp-cli ./datasets/xtreme --save_infos` to keep backward compatibility (versions 0.3.0 can't load these fields).
I think I'll change the logic so that `nlp-cli test` doesn't create these fields for dataset with no post processing
| true
|
677,016,998
|
https://api.github.com/repos/huggingface/datasets/issues/496
|
https://github.com/huggingface/datasets/pull/496
| 496
|
fix bad type in overflow check
|
closed
| 0
| 2020-08-11T16:24:58
| 2020-08-14T13:29:35
| 2020-08-14T13:29:34
|
lhoestq
|
[] |
When writing an arrow file and inferring the features, the overflow check could fail if the first example had a `null` field.
This is because we were not using the inferred features to do this check, and we could end up with arrays that don't match because of a type mismatch (`null` vs `string` for example).
This should fix #482
| true
|
676,959,289
|
https://api.github.com/repos/huggingface/datasets/issues/495
|
https://github.com/huggingface/datasets/pull/495
| 495
|
stack vectors in pytorch and tensorflow
|
closed
| 0
| 2020-08-11T15:12:53
| 2020-08-12T09:30:49
| 2020-08-12T09:30:48
|
lhoestq
|
[] |
When the format of a dataset is set to pytorch or tensorflow, and if the dataset has vectors in it, they were not stacked together as tensors when calling `dataset[i:i + batch_size][column]` or `dataset[column]`.
I added support for stacked tensors for both pytorch and tensorflow.
For ragged tensors, they are stacked only for tensorflow as pytorch doesn't support ragged tensors.
| true
|
676,886,955
|
https://api.github.com/repos/huggingface/datasets/issues/494
|
https://github.com/huggingface/datasets/pull/494
| 494
|
Fix numpy stacking
|
closed
| 1
| 2020-08-11T13:40:30
| 2020-08-11T14:56:50
| 2020-08-11T13:49:52
|
lhoestq
|
[] |
When getting items using a column name as a key, numpy arrays were not stacked.
I fixed that and added some tests.
There is another issue that still needs to be fixed though: when getting items using a column name as a key, pytorch tensors are not stacked (it outputs a list of tensors). This PR should help with the to fix this issue.
| true
|
676,527,351
|
https://api.github.com/repos/huggingface/datasets/issues/493
|
https://github.com/huggingface/datasets/pull/493
| 493
|
Fix wmt zh-en url
|
closed
| 1
| 2020-08-11T02:14:52
| 2020-08-11T02:22:28
| 2020-08-11T02:22:12
|
sshleifer
|
[] |
I verified that
```
wget https://stuncorpusprod.blob.core.windows.net/corpusfiles/UNv1.0.en-zh.tar.gz.00
```
runs in 2 minutes.
| true
|
676,495,064
|
https://api.github.com/repos/huggingface/datasets/issues/492
|
https://github.com/huggingface/datasets/issues/492
| 492
|
nlp.Features does not distinguish between nullable and non-nullable types in PyArrow schema
|
closed
| 7
| 2020-08-11T00:27:46
| 2020-08-26T16:17:19
| 2020-08-26T16:17:19
|
jarednielsen
|
[] |
Here's the code I'm trying to run:
```python
dset_wikipedia = nlp.load_dataset("wikipedia", "20200501.en", split="train", cache_dir=args.cache_dir)
dset_wikipedia.drop(columns=["title"])
dset_wikipedia.features.pop("title")
dset_books = nlp.load_dataset("bookcorpus", split="train", cache_dir=args.cache_dir)
dset = nlp.concatenate_datasets([dset_wikipedia, dset_books])
```
This fails because they have different schemas, despite having identical features.
```python
assert dset_wikipedia.features == dset_books.features # True
assert dset_wikipedia._data.schema == dset_books._data.schema # False
```
The Wikipedia dataset has 'text: string', while the BookCorpus dataset has 'text: string not null'. Currently I hack together a working schema match with the following line, but it would be better if this was handled in Features themselves.
```python
dset_wikipedia._data = dset_wikipedia.data.cast(dset_books._data.schema)
```
| false
|
676,486,275
|
https://api.github.com/repos/huggingface/datasets/issues/491
|
https://github.com/huggingface/datasets/issues/491
| 491
|
No 0.4.0 release on GitHub
|
closed
| 2
| 2020-08-10T23:59:57
| 2020-08-11T16:50:07
| 2020-08-11T16:50:07
|
jarednielsen
|
[] |
0.4.0 was released on PyPi, but not on GitHub. This means [the documentation](https://huggingface.co/nlp/) is still displaying from 0.3.0, and that there's no tag to easily clone the 0.4.0 version of the repo.
| false
|
676,482,242
|
https://api.github.com/repos/huggingface/datasets/issues/490
|
https://github.com/huggingface/datasets/issues/490
| 490
|
Loading preprocessed Wikipedia dataset requires apache_beam
|
closed
| 0
| 2020-08-10T23:46:50
| 2020-08-14T13:17:20
| 2020-08-14T13:17:20
|
jarednielsen
|
[] |
Running
`nlp.load_dataset("wikipedia", "20200501.en", split="train", dir="/tmp/wikipedia")`
gives an error if apache_beam is not installed, stemming from
https://github.com/huggingface/nlp/blob/38eb2413de54ee804b0be81781bd65ac4a748ced/src/nlp/builder.py#L981-L988
This succeeded without the dependency in version 0.3.0. This seems like an unnecessary dependency to process some dataset info if you're using the already-preprocessed version. Could it be removed?
| false
|
676,456,257
|
https://api.github.com/repos/huggingface/datasets/issues/489
|
https://github.com/huggingface/datasets/issues/489
| 489
|
ug
|
closed
| 2
| 2020-08-10T22:33:03
| 2020-08-10T22:55:14
| 2020-08-10T22:33:40
|
timothyjlaurent
|
[] | false
|
|
676,299,993
|
https://api.github.com/repos/huggingface/datasets/issues/488
|
https://github.com/huggingface/datasets/issues/488
| 488
|
issues with downloading datasets for wmt16 and wmt19
|
closed
| 3
| 2020-08-10T17:32:51
| 2022-10-04T17:46:59
| 2022-10-04T17:46:58
|
stas00
|
[] |
I have encountered multiple issues while trying to:
```
import nlp
dataset = nlp.load_dataset('wmt16', 'ru-en')
metric = nlp.load_metric('wmt16')
```
1. I had to do `pip install -e ".[dev]" ` on master, currently released nlp didn't work (sorry, didn't save the error) - I went back to the released version and now it worked. So it must have been some outdated dependencies that `pip install -e ".[dev]" ` fixed.
2. it was downloading at 60kbs - almost 5 hours to get the dataset. It was downloading all pairs and not just the one I asked for.
I tried the same code with `wmt19` in parallel and it took a few secs to download and it only fetched data for the requested pair. (but it failed too, see below)
3. my machine has crushed and when I retried I got:
```
Traceback (most recent call last):
File "./download.py", line 9, in <module>
dataset = nlp.load_dataset('wmt16', 'ru-en')
File "/mnt/nvme1/code/huggingface/nlp-master/src/nlp/load.py", line 549, in load_dataset
download_config=download_config, download_mode=download_mode, ignore_verifications=ignore_verifications,
File "/mnt/nvme1/code/huggingface/nlp-master/src/nlp/builder.py", line 449, in download_and_prepare
with incomplete_dir(self._cache_dir) as tmp_data_dir:
File "/home/stas/anaconda3/envs/main/lib/python3.7/contextlib.py", line 112, in __enter__
return next(self.gen)
File "/mnt/nvme1/code/huggingface/nlp-master/src/nlp/builder.py", line 422, in incomplete_dir
os.makedirs(tmp_dir)
File "/home/stas/anaconda3/envs/main/lib/python3.7/os.py", line 221, in makedirs
mkdir(name, mode)
FileExistsError: [Errno 17] File exists: '/home/stas/.cache/huggingface/datasets/wmt16/ru-en/1.0.0/4d8269cdd971ed26984a9c0e4a158e0c7afc8135fac8fb8ee43ceecf38fd422d.incomplete'
```
it can't handle resumes. but neither allows a new start. Had to delete it manually.
4. and finally when it downloaded the dataset, it then failed to fetch the metrics:
```
Traceback (most recent call last):
File "./download.py", line 15, in <module>
metric = nlp.load_metric('wmt16')
File "/mnt/nvme1/code/huggingface/nlp-master/src/nlp/load.py", line 442, in load_metric
module_path, hash = prepare_module(path, download_config=download_config, dataset=False)
File "/mnt/nvme1/code/huggingface/nlp-master/src/nlp/load.py", line 258, in prepare_module
local_path = cached_path(file_path, download_config=download_config)
File "/mnt/nvme1/code/huggingface/nlp-master/src/nlp/utils/file_utils.py", line 198, in cached_path
local_files_only=download_config.local_files_only,
File "/mnt/nvme1/code/huggingface/nlp-master/src/nlp/utils/file_utils.py", line 356, in get_from_cache
raise ConnectionError("Couldn't reach {}".format(url))
ConnectionError: Couldn't reach https://s3.amazonaws.com/datasets.huggingface.co/nlp/metrics/wmt16/wmt16.py
```
5. If I run the same code with `wmt19`, it fails too:
```
ConnectionError: Couldn't reach https://storage.googleapis.com/tfdataset-data/downloadataset/uncorpus/UNv1.0.en-ru.tar.gz
```
| false
|
676,143,029
|
https://api.github.com/repos/huggingface/datasets/issues/487
|
https://github.com/huggingface/datasets/pull/487
| 487
|
Fix elasticsearch result ids returning as strings
|
closed
| 1
| 2020-08-10T13:37:11
| 2020-08-31T10:42:46
| 2020-08-31T10:42:46
|
sai-prasanna
|
[] |
I am using the latest elasticsearch binary and master of nlp. For me elasticsearch searches failed because the resultant "id_" returned for searches are strings, but our library assumes them to be integers.
| true
|
675,649,034
|
https://api.github.com/repos/huggingface/datasets/issues/486
|
https://github.com/huggingface/datasets/issues/486
| 486
|
Bookcorpus data contains pretokenized text
|
closed
| 8
| 2020-08-09T06:53:24
| 2022-10-04T17:44:33
| 2022-10-04T17:44:33
|
orsharir
|
[] |
It seem that the bookcoprus data downloaded through the library was pretokenized with NLTK's Treebank tokenizer, which changes the text in incompatible ways to how, for instance, BERT's wordpiece tokenizer works. For example, "didn't" becomes "did" + "n't", and double quotes are changed to `` and '' for start and end quotes, respectively.
On my own projects, I just run the data through NLTK's TreebankWordDetokenizer to reverse the tokenization (as best as possible). I think it would be beneficial to apply this transformation directly on your remote cached copy of the dataset. If you choose to do so, I would also suggest to use my fork of NLTK that fixes several bugs in their detokenizer (I've opened a pull-request, but they've yet to respond): https://github.com/nltk/nltk/pull/2575
| false
|
675,595,393
|
https://api.github.com/repos/huggingface/datasets/issues/485
|
https://github.com/huggingface/datasets/issues/485
| 485
|
PAWS dataset first item is header
|
closed
| 0
| 2020-08-08T22:05:25
| 2020-08-19T09:50:01
| 2020-08-19T09:50:01
|
jxmorris12
|
[] |
```
import nlp
dataset = nlp.load_dataset('xtreme', 'PAWS-X.en')
dataset['test'][0]
```
prints the following
```
{'label': 'label', 'sentence1': 'sentence1', 'sentence2': 'sentence2'}
```
dataset['test'][0] should probably be the first item in the dataset, not just a dictionary mapping the column names to themselves. Probably just need to ignore the first row in the dataset by default or something like that.
| false
|
675,088,983
|
https://api.github.com/repos/huggingface/datasets/issues/484
|
https://github.com/huggingface/datasets/pull/484
| 484
|
update mirror for RT dataset
|
closed
| 4
| 2020-08-07T15:25:45
| 2020-08-24T13:33:37
| 2020-08-24T13:33:37
|
jxmorris12
|
[] | true
|
|
675,080,694
|
https://api.github.com/repos/huggingface/datasets/issues/483
|
https://github.com/huggingface/datasets/issues/483
| 483
|
rotten tomatoes movie review dataset taken down
|
closed
| 3
| 2020-08-07T15:12:01
| 2020-09-08T09:36:34
| 2020-09-08T09:36:33
|
jxmorris12
|
[] |
In an interesting twist of events, the individual who created the movie review seems to have left Cornell, and their webpage has been removed, along with the movie review dataset (http://www.cs.cornell.edu/people/pabo/movie-review-data/rt-polaritydata.tar.gz). It's not downloadable anymore.
| false
|
674,851,147
|
https://api.github.com/repos/huggingface/datasets/issues/482
|
https://github.com/huggingface/datasets/issues/482
| 482
|
Bugs : dataset.map() is frozen on ELI5
|
closed
| 8
| 2020-08-07T08:23:35
| 2023-04-06T09:39:59
| 2020-08-11T23:55:15
|
ratthachat
|
[] |
Hi Huggingface Team!
Thank you guys once again for this amazing repo.
I have tried to prepare ELI5 to train with T5, based on [this wonderful notebook of Suraj Patil](https://github.com/patil-suraj/exploring-T5/blob/master/T5_on_TPU.ipynb)
However, when I run `dataset.map()` on ELI5 to prepare `input_text, target_text`, `dataset.map` is **frozen** in the first hundreds examples. On the contrary, this works totally fine on SQUAD (80,000 examples). Both `nlp` version 0.3.0 and 0.4.0 cause frozen process . Also try various `pyarrow` versions from 0.16.0 / 0.17.0 / 1.0.0 also have the same frozen process.
Reproducible code can be found on [this colab notebook ](https://colab.research.google.com/drive/14wttOTv3ky74B_c0kv5WrbgQjCF2fYQk?usp=sharing), where I also show that the same mapping function works fine on SQUAD, so the problem is likely due to ELI5 somehow.
----------------------------------------
**More Info :** instead of `map`, if I run `for` loop and apply function by myself, there's no error and can finish within 10 seconds. However, `nlp dataset` is immutable (I couldn't manually assign a new key-value to `dataset `object)
I also notice that SQUAD texts are quite clean while ELI5 texts contain many special characters, not sure if this is the cause ?
| false
|
674,567,389
|
https://api.github.com/repos/huggingface/datasets/issues/481
|
https://github.com/huggingface/datasets/pull/481
| 481
|
Apply utf-8 encoding to all datasets
|
closed
| 6
| 2020-08-06T20:02:09
| 2020-08-20T08:16:08
| 2020-08-20T08:16:08
|
lewtun
|
[] |
## Description
This PR applies utf-8 encoding for all instances of `with open(...) as f` to all Python files in `datasets/`. As suggested by @thomwolf in #468 , we use regular expressions and the following function
```python
def apply_encoding_on_file_open(filepath: str):
"""Apply UTF-8 encoding for all instances where a non-binary file is opened."""
with open(filepath, 'r', encoding='utf-8') as input_file:
regexp = re.compile(r"(?!.*\b(?:encoding|rb|w|wb|w+|wb+|ab|ab+)\b)(?<=\s)(open)\((.*)\)")
input_text = input_file.read()
match = regexp.search(input_text)
if match:
output = regexp.sub(lambda m: m.group()[:-1]+', encoding="utf-8")', input_text)
with open(filepath, 'w', encoding='utf-8') as output_file:
output_file.write(output)
```
to perform the replacement.
Note:
1. I excluded all _**binary files**_ from the search since it's possible some objects are opened for which the encoding doesn't make sense. Please correct me if I'm wrong and I'll tweak the regexp accordingly
2. There were two edge cases where the regexp failed (e.g. two `open` instances on a single line), but I decided to just fix these manually in the interest of time.
3. I only applied the replacement to files in `datasets/`. Let me know if this should be extended to other places like `metrics/`
4. I have implemented a unit test that should catch missing encodings in future CI runs
Closes #468 and possibly #347
| true
|
674,245,959
|
https://api.github.com/repos/huggingface/datasets/issues/480
|
https://github.com/huggingface/datasets/pull/480
| 480
|
Column indexing hotfix
|
closed
| 2
| 2020-08-06T11:37:05
| 2023-09-24T09:49:33
| 2020-08-12T08:36:10
|
TevenLeScao
|
[] |
As observed for example in #469 , currently `__getitem__` does not convert the data to the dataset format when indexing by column. This is a hotfix that imitates functional 0.3.0. code. In the future it'd probably be nice to have a test there.
| true
|
673,905,407
|
https://api.github.com/repos/huggingface/datasets/issues/479
|
https://github.com/huggingface/datasets/pull/479
| 479
|
add METEOR metric
|
closed
| 5
| 2020-08-05T23:13:00
| 2020-08-19T13:39:09
| 2020-08-19T13:39:09
|
vegarab
|
[] |
Added the METEOR metric. Can be used like this:
```python
import nlp
meteor = nlp.load_metric('metrics/meteor')
meteor.compute(["some string", "some string"], ["some string", "some similar string"])
# {'meteor': 0.6411637931034483}
meteor.add("some string", "some string")
meteor.add('some string", "some similar string")
meteor.compute()
# {'meteor': 0.6411637931034483}
```
Uses [NLTK's implementation](https://www.nltk.org/api/nltk.translate.html#module-nltk.translate.meteor_score), [(source)](https://github.com/nltk/nltk/blob/develop/nltk/translate/meteor_score.py)
| true
|
673,178,317
|
https://api.github.com/repos/huggingface/datasets/issues/478
|
https://github.com/huggingface/datasets/issues/478
| 478
|
Export TFRecord to GCP bucket
|
closed
| 1
| 2020-08-05T01:08:32
| 2020-08-05T01:21:37
| 2020-08-05T01:21:36
|
astariul
|
[] |
Previously, I was writing TFRecords manually to GCP bucket with : `with tf.io.TFRecordWriter('gs://my_bucket/x.tfrecord')`
Since `0.4.0` is out with the `export()` function, I tried it. But it seems TFRecords cannot be directly written to GCP bucket.
`dataset.export('local.tfrecord')` works fine,
but `dataset.export('gs://my_bucket/x.tfrecord')` does not work.
There is no error message, I just can't find the file on my bucket...
---
Looking at the code, `nlp` is using `tf.data.experimental.TFRecordWriter`, while I was using `tf.io.TFRecordWriter`.
**What's the difference between those 2 ? How can I write TFRecords files directly to GCP bucket ?**
@jarednielsen @lhoestq
| false
|
673,142,143
|
https://api.github.com/repos/huggingface/datasets/issues/477
|
https://github.com/huggingface/datasets/issues/477
| 477
|
Overview.ipynb throws exceptions with nlp 0.4.0
|
closed
| 3
| 2020-08-04T23:18:15
| 2021-08-03T06:02:15
| 2021-08-03T06:02:15
|
mandy-li
|
[] |
with nlp 0.4.0, the TensorFlow example in Overview.ipynb throws the following exceptions:
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
<ipython-input-5-48907f2ad433> in <module>
----> 1 features = {x: train_tf_dataset[x].to_tensor(default_value=0, shape=[None, tokenizer.max_len]) for x in columns[:3]}
2 labels = {"output_1": train_tf_dataset["start_positions"].to_tensor(default_value=0, shape=[None, 1])}
3 labels["output_2"] = train_tf_dataset["end_positions"].to_tensor(default_value=0, shape=[None, 1])
4 tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8)
<ipython-input-5-48907f2ad433> in <dictcomp>(.0)
----> 1 features = {x: train_tf_dataset[x].to_tensor(default_value=0, shape=[None, tokenizer.max_len]) for x in columns[:3]}
2 labels = {"output_1": train_tf_dataset["start_positions"].to_tensor(default_value=0, shape=[None, 1])}
3 labels["output_2"] = train_tf_dataset["end_positions"].to_tensor(default_value=0, shape=[None, 1])
4 tfdataset = tf.data.Dataset.from_tensor_slices((features, labels)).batch(8)
AttributeError: 'numpy.ndarray' object has no attribute 'to_tensor'
| false
|
672,991,854
|
https://api.github.com/repos/huggingface/datasets/issues/476
|
https://github.com/huggingface/datasets/pull/476
| 476
|
CheckList
|
closed
| 2
| 2020-08-04T18:32:05
| 2022-10-03T09:43:37
| 2022-10-03T09:43:37
|
marcotcr
|
[
"dataset contribution"
] |
Sorry for the large pull request.
- Added checklists as datasets. I can't run `test_load_real_dataset` (see #474), but I can load the datasets successfully as shown in the example notebook
- Added a checklist wrapper
| true
|
672,884,595
|
https://api.github.com/repos/huggingface/datasets/issues/475
|
https://github.com/huggingface/datasets/pull/475
| 475
|
misc. bugs and quality of life
|
closed
| 2
| 2020-08-04T15:32:29
| 2020-08-17T21:14:08
| 2020-08-17T21:14:07
|
joeddav
|
[] |
A few misc. bugs and QOL improvements that I've come across in using the library. Let me know if you don't like any of them and I can adjust/remove them.
1. Printing datasets without a description field throws an error when formatting the `single_line_description`. This fixes that, and also adds some formatting to the repr to make it slightly more readable.
```
>>> print(list_datasets()[0])
nlp.ObjectInfo(
id='aeslc',
description='A collection of email messages of employees in the Enron Corporation.There are two features: - email_body: email body text. - subject_line: email subject text.',
files=[nlp.S3Object('aeslc.py'), nlp.S3Object('dataset_infos.json'), nlp.S3Object('dummy/1.0.0/dummy_data-zip-extracted/dummy_data/AESLC-master/enron_subject_line/dev/allen-p_inbox_29.subject'), nlp.S3Object('dummy/1.0.0/dummy_data-zip-extracted/dummy_data/AESLC-master/enron_subject_line/test/allen-p_inbox_24.subject'), nlp.S3Object('dummy/1.0.0/dummy_data-zip-extracted/dummy_data/AESLC-master/enron_subject_line/train/allen-p_inbox_20.subject'), nlp.S3Object('dummy/1.0.0/dummy_data.zip'), nlp.S3Object('urls_checksums/checksums.txt')]
)
```
2. Add id-only option to `list_datasets` and `list_metrics` to allow the user to easily print out just the names of the datasets & metrics. I often found myself annoyed that this took so many strokes to do.
```python
[dataset.id for dataset in list_datasets()] # before
list_datasets(id_only=True) # after
```
3. Fix null-seed randomization caching. When using `train_test_split` and `shuffle`, the computation was being cached even without a seed or generator being passed. The result was that calling `.shuffle` more than once on the same dataset didn't do anything without passing a distinct seed or generator. Likewise with `train_test_split`.
4. Indexing by iterables of bool. I added support for passing an iterable of type bool to `_getitem` as a numpy/pandas-like indexing method. Let me know if you think it's redundant with `filter` (I know it's not optimal memory-wise), but I think it's nice to have as a lightweight alternative to do simple things without having to create a copy of the entire dataset, e.g.
```python
dataset[dataset['label'] == 0] # numpy-like bool indexing to look at instances with labels of 0
```
5. Add an `input_column` argument to `map` and `filter`, which allows you to filter/map on a particular column rather than passing the whole dict to the function. Also adds `fn_kwargs` to be passed to the function. I think these together make mapping much cleaner in many cases such as mono-column tokenization:
```python
# before
dataset = dataset.map(lambda batch: tokenizer(batch["text"])
# after
dataset = dataset.map(tokenizer, input_column="text")
dataset = dataset.map(tokenizer, input_column="text", fn_kwargs={"truncation": True, "padding": True})
```
| true
|
672,407,330
|
https://api.github.com/repos/huggingface/datasets/issues/474
|
https://github.com/huggingface/datasets/issues/474
| 474
|
test_load_real_dataset when config has BUILDER_CONFIGS that matter
|
closed
| 2
| 2020-08-03T23:46:36
| 2020-09-07T14:53:13
| 2020-09-07T14:53:13
|
marcotcr
|
[] |
It a dataset has custom `BUILDER_CONFIGS` with non-keyword arguments (or keyword arguments with non default values), the config is not loaded during the test and causes an error.
I think the problem is that `test_load_real_dataset` calls `load_dataset` with `data_dir=temp_data_dir` ([here](https://github.com/huggingface/nlp/blob/master/tests/test_dataset_common.py#L200)). This causes [this line](https://github.com/huggingface/nlp/blob/master/src/nlp/builder.py#L201) to always be false because `config_kwargs` is not `None`. [This line](https://github.com/huggingface/nlp/blob/master/src/nlp/builder.py#L222) will be run instead, which doesn't use `BUILDER_CONFIGS`.
For an example, you can try running the test for lince:
` RUN_SLOW=1 pytest tests/test_dataset_common.py::LocalDatasetTest::test_load_real_dataset_lince`
which yields
> E TypeError: __init__() missing 3 required positional arguments: 'colnames', 'classes', and 'label_column'
| false
|
672,007,247
|
https://api.github.com/repos/huggingface/datasets/issues/473
|
https://github.com/huggingface/datasets/pull/473
| 473
|
add DoQA dataset (ACL 2020)
|
closed
| 0
| 2020-08-03T11:26:52
| 2020-09-10T17:19:11
| 2020-09-03T11:44:15
|
mariamabarham
|
[] |
add DoQA dataset (ACL 2020) http://ixa.eus/node/12931
| true
|
672,000,745
|
https://api.github.com/repos/huggingface/datasets/issues/472
|
https://github.com/huggingface/datasets/pull/472
| 472
|
add crd3 dataset
|
closed
| 1
| 2020-08-03T11:15:02
| 2020-08-03T11:22:10
| 2020-08-03T11:22:09
|
mariamabarham
|
[] |
opening new PR for CRD3 dataset (ACL2020) to fix the circle CI problems
| true
|
671,996,423
|
https://api.github.com/repos/huggingface/datasets/issues/471
|
https://github.com/huggingface/datasets/pull/471
| 471
|
add reuters21578 dataset
|
closed
| 0
| 2020-08-03T11:07:14
| 2022-08-04T08:39:11
| 2020-09-03T09:58:50
|
mariamabarham
|
[] |
new PR to add the reuters21578 dataset and fix the circle CI problems.
Fix partially:
- #353
Subsequent PR after:
- #449
| true
|
671,952,276
|
https://api.github.com/repos/huggingface/datasets/issues/470
|
https://github.com/huggingface/datasets/pull/470
| 470
|
Adding IWSLT 2017 dataset.
|
closed
| 6
| 2020-08-03T09:52:39
| 2020-09-07T12:33:30
| 2020-09-07T12:33:30
|
Narsil
|
[] |
Created a [IWSLT 2017](https://sites.google.com/site/iwsltevaluation2017/TED-tasks) dataset script for the *multilingual data*.
```
Bilingual data: {Arabic, German, French, Japanese, Korean, Chinese} <-> English
Multilingual data: German, English, Italian, Dutch, Romanian. (Any pair)
```
I'm unsure how to handle bilingual vs multilingual. Given `nlp` architecture a Config option seems to be the way to go, however, it might be a bit confusing to have different language pairs with different option. Using just language pairs is not viable as English to German exists in both.
Any opinion on how that should be done ?
EDIT: I decided to just omit de-en from multilingual as it's only a subset of the bilingual one. That way only language pairs exist.
EDIT : Could be interesting for #438
| true
|
671,876,963
|
https://api.github.com/repos/huggingface/datasets/issues/469
|
https://github.com/huggingface/datasets/issues/469
| 469
|
invalid data type 'str' at _convert_outputs in arrow_dataset.py
|
closed
| 9
| 2020-08-03T07:48:29
| 2023-07-20T15:54:17
| 2023-07-20T15:54:17
|
Murgates
|
[] |
I trying to build multi label text classifier model using Transformers lib.
I'm using Transformers NLP to load the data set, while calling trainer.train() method. It throws the following error
File "C:\***\arrow_dataset.py", line 343, in _convert_outputs
v = command(v)
TypeError: new(): invalid data type 'str'
I'm using pyarrow 1.0.0. And I have simple custom data set with Text and Integer Label.
Ex: Data
Text , Label #Column Header
I'm facing an Network issue, 1
I forgot my password, 2
Error StackTrace:
File "C:\**\transformers\trainer.py", line 492, in train
for step, inputs in enumerate(epoch_iterator):
File "C:\**\tqdm\std.py", line 1104, in __iter__
for obj in iterable:
File "C:\**\torch\utils\data\dataloader.py", line 345, in __next__
data = self._next_data()
File "C:\**\torch\utils\data\dataloader.py", line 385, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "C:\**\torch\utils\data\_utils\fetch.py", line 44, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "C:\**\torch\utils\data\_utils\fetch.py", line 44, in <listcomp>
data = [self.dataset[idx] for idx in possibly_batched_index]
File "C:\**\nlp\arrow_dataset.py", line 414, in __getitem__
output_all_columns=self._output_all_columns,
File "C:\**\nlp\arrow_dataset.py", line 403, in _getitem
outputs, format_type=format_type, format_columns=format_columns, output_all_columns=output_all_columns
File "C:\**\nlp\arrow_dataset.py", line 343, in _convert_outputs
v = command(v)
TypeError: new(): invalid data type 'str'
| false
|
671,622,441
|
https://api.github.com/repos/huggingface/datasets/issues/468
|
https://github.com/huggingface/datasets/issues/468
| 468
|
UnicodeDecodeError while loading PAN-X task of XTREME dataset
|
closed
| 5
| 2020-08-02T14:05:10
| 2020-08-20T08:16:08
| 2020-08-20T08:16:08
|
lewtun
|
[] |
Hi 🤗 team!
## Description of the problem
I'm running into a `UnicodeDecodeError` while trying to load the PAN-X subset the XTREME dataset:
```
---------------------------------------------------------------------------
UnicodeDecodeError Traceback (most recent call last)
<ipython-input-5-1d61f439b843> in <module>
----> 1 dataset = load_dataset("xtreme", "PAN-X.en", data_dir='./data')
/usr/local/lib/python3.6/dist-packages/nlp/load.py in load_dataset(path, name, version, data_dir, data_files, split, cache_dir, features, download_config, download_mode, ignore_verifications, save_infos, **config_kwargs)
528 ignore_verifications = ignore_verifications or save_infos
529 # Download/copy dataset processing script
--> 530 module_path, hash = prepare_module(path, download_config=download_config, dataset=True)
531
532 # Get dataset builder class from the processing script
/usr/local/lib/python3.6/dist-packages/nlp/load.py in prepare_module(path, download_config, dataset, force_local_path, **download_kwargs)
265
266 # Download external imports if needed
--> 267 imports = get_imports(local_path)
268 local_imports = []
269 library_imports = []
/usr/local/lib/python3.6/dist-packages/nlp/load.py in get_imports(file_path)
156 lines = []
157 with open(file_path, mode="r") as f:
--> 158 lines.extend(f.readlines())
159
160 logger.info("Checking %s for additional imports.", file_path)
/usr/lib/python3.6/encodings/ascii.py in decode(self, input, final)
24 class IncrementalDecoder(codecs.IncrementalDecoder):
25 def decode(self, input, final=False):
---> 26 return codecs.ascii_decode(input, self.errors)[0]
27
28 class StreamWriter(Codec,codecs.StreamWriter):
UnicodeDecodeError: 'ascii' codec can't decode byte 0xe2 in position 111: ordinal not in range(128)
```
## Steps to reproduce
Install from nlp's master branch
```python
pip install git+https://github.com/huggingface/nlp.git
```
then run
```python
from nlp import load_dataset
# AmazonPhotos.zip is located in data/
dataset = load_dataset("xtreme", "PAN-X.en", data_dir='./data')
```
## OS / platform details
- `nlp` version: latest from master
- Platform: Linux-4.15.0-72-generic-x86_64-with-Ubuntu-18.04-bionic
- Python version: 3.6.9
- PyTorch version (GPU?): 1.4.0 (True)
- Tensorflow version (GPU?): 2.1.0 (True)
- Using GPU in script?: True
- Using distributed or parallel set-up in script?: False
## Proposed solution
Either change [line 762](https://github.com/huggingface/nlp/blob/7ada00b1d62f94eee22a7df38c6b01e3f27194b7/datasets/xtreme/xtreme.py#L762) in `xtreme.py` to include UTF-8 encoding:
```
# old
with open(filepath) as f
# new
with open(filepath, encoding='utf-8') as f
```
or raise a warning that suggests setting the locale explicitly, e.g.
```python
import locale
locale.setlocale(locale.LC_ALL, 'C.UTF-8')
```
I have a preference for the first solution. Let me know if you agree and I'll be happy to implement the simple fix!
| false
|
671,580,010
|
https://api.github.com/repos/huggingface/datasets/issues/467
|
https://github.com/huggingface/datasets/pull/467
| 467
|
DOCS: Fix typo
|
closed
| 1
| 2020-08-02T08:59:37
| 2020-08-02T13:52:27
| 2020-08-02T09:18:54
|
bharatr21
|
[] |
Fix typo from dictionnary -> dictionary
| true
|
670,766,891
|
https://api.github.com/repos/huggingface/datasets/issues/466
|
https://github.com/huggingface/datasets/pull/466
| 466
|
[METRICS] Various improvements on metrics
|
closed
| 2
| 2020-08-01T11:03:45
| 2020-08-17T15:15:00
| 2020-08-17T15:14:59
|
thomwolf
|
[] |
- Disallow the use of positional arguments to avoid `predictions` vs `references` mistakes
- Allow to directly feed numpy/pytorch/tensorflow/pandas objects in metrics
| true
|
669,889,779
|
https://api.github.com/repos/huggingface/datasets/issues/465
|
https://github.com/huggingface/datasets/pull/465
| 465
|
Keep features after transform
|
closed
| 3
| 2020-07-31T14:43:21
| 2020-07-31T18:27:33
| 2020-07-31T18:27:32
|
lhoestq
|
[] |
When applying a transform like `map`, some features were lost (and inferred features were used).
It was the case for ClassLabel, Translation, etc.
To fix that, I did some modifications in the `ArrowWriter`:
- added the `update_features` parameter. When it's `True`, then the features specified by the user (if any) can be updated with inferred features if their type don't match. `map` transform sets `update_features=True` when writing to cache file or buffer. Features won't change by default in `map`.
- added the `with_metadata` parameter. If `True`, the `features` (after update) will be written inside the metadata of the schema in this format:
```
{
"huggingface": {"features" : <serialized Features exactly like dataset_info.json>}
}
```
Then, once a dataset is instantiated without info/features, these metadata are used to set the features of the dataset.
| true
|
669,767,381
|
https://api.github.com/repos/huggingface/datasets/issues/464
|
https://github.com/huggingface/datasets/pull/464
| 464
|
Add rename, remove and cast in-place operations
|
closed
| 0
| 2020-07-31T12:30:21
| 2020-07-31T15:50:02
| 2020-07-31T15:50:00
|
thomwolf
|
[] |
Add a bunch of in-place operation leveraging the Arrow back-end to rename and remove columns and cast to new features without using the more expensive `map` method.
These methods are added to `Dataset` as well as `DatasetDict`.
Added tests for these new methods and add the methods to the doc.
Naming follows the new pattern with a trailing underscore indicating in-place methods.
| true
|
669,735,455
|
https://api.github.com/repos/huggingface/datasets/issues/463
|
https://github.com/huggingface/datasets/pull/463
| 463
|
Add dataset/mlsum
|
closed
| 3
| 2020-07-31T11:50:52
| 2020-08-24T14:54:42
| 2020-08-24T14:54:42
|
RachelKer
|
[] |
New pull request that should correct the previous errors.
The load_real_data stills fails because it is looking for a default dataset URL that does not exists, this does not happen when loading the dataset with load_dataset
| true
|
669,715,547
|
https://api.github.com/repos/huggingface/datasets/issues/462
|
https://github.com/huggingface/datasets/pull/462
| 462
|
add DoQA (ACL 2020) dataset
|
closed
| 0
| 2020-07-31T11:25:56
| 2023-09-24T09:48:42
| 2020-08-03T11:28:27
|
mariamabarham
|
[] |
adds DoQA (ACL 2020) dataset
| true
|
669,703,508
|
https://api.github.com/repos/huggingface/datasets/issues/461
|
https://github.com/huggingface/datasets/pull/461
| 461
|
Doqa
|
closed
| 0
| 2020-07-31T11:11:12
| 2023-09-24T09:48:40
| 2020-07-31T11:13:15
|
mariamabarham
|
[] |
add DoQA (ACL 2020) dataset
| true
|
669,585,256
|
https://api.github.com/repos/huggingface/datasets/issues/460
|
https://github.com/huggingface/datasets/pull/460
| 460
|
Fix KeyboardInterrupt in map and bad indices in select
|
closed
| 2
| 2020-07-31T08:57:15
| 2020-07-31T11:32:19
| 2020-07-31T11:32:18
|
lhoestq
|
[] |
If you interrupted a map function while it was writing, the cached file was not discarded.
Therefore the next time you called map, it was loading an incomplete arrow file.
We had the same issue with select if there was a bad indice at one point.
To fix that I used temporary files that are renamed once everything is finished.
| true
|
669,545,437
|
https://api.github.com/repos/huggingface/datasets/issues/459
|
https://github.com/huggingface/datasets/pull/459
| 459
|
[Breaking] Update Dataset and DatasetDict API
|
closed
| 0
| 2020-07-31T08:11:33
| 2020-08-26T08:28:36
| 2020-08-26T08:28:35
|
thomwolf
|
[] |
This PR contains a few breaking changes so it's probably good to keep it for the next (major) release:
- rename the `flatten`, `drop` and `dictionary_encode_column` methods in `flatten_`, `drop_` and `dictionary_encode_column_` to indicate that these methods have in-place effects as discussed in #166. From now on we should keep the convention of having a trailing underscore for methods which have an in-place effet. I also adopt the conversion of not returning the (self) dataset for these methods. This is different than what PyTorch does for instance (`model.to()` is in-place but return the self model) but I feel like it's a safer approach in terms of UX.
- remove the `dataset.columns` property which returns a low-level Apache Arrow object and should not be used by users. Similarly, remove `dataset. nbytes` which we don't really want to expose in this bare-bone format.
- add a few more properties and methods to `DatasetDict`
| true
|
668,972,666
|
https://api.github.com/repos/huggingface/datasets/issues/458
|
https://github.com/huggingface/datasets/pull/458
| 458
|
Install CoVal metric from github
|
closed
| 0
| 2020-07-30T16:59:25
| 2020-07-31T13:56:33
| 2020-07-31T13:56:33
|
yjernite
|
[] |
Changed the import statements in `coval.py` to direct the user to install the original package from github if it's not already installed (the warning will only display properly after merging [PR455](https://github.com/huggingface/nlp/pull/455))
Also changed the function call to use named rather than positional arguments.
| true
|
668,898,386
|
https://api.github.com/repos/huggingface/datasets/issues/457
|
https://github.com/huggingface/datasets/pull/457
| 457
|
add set_format to DatasetDict + tests
|
closed
| 0
| 2020-07-30T15:53:20
| 2020-07-30T17:34:36
| 2020-07-30T17:34:34
|
thomwolf
|
[] |
Add the `set_format` and `formated_as` and `reset_format` to `DatasetDict`.
Add tests to these for `Dataset` and `DatasetDict`.
Fix some bugs uncovered by the tests for `pandas` formating.
| true
|
668,723,785
|
https://api.github.com/repos/huggingface/datasets/issues/456
|
https://github.com/huggingface/datasets/pull/456
| 456
|
add crd3(ACL 2020) dataset
|
closed
| 0
| 2020-07-30T13:28:35
| 2023-09-24T09:48:47
| 2020-08-03T11:28:52
|
mariamabarham
|
[] |
This PR adds the **Critical Role Dungeons and Dragons Dataset** published at ACL 2020
| true
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.